pmid stringlengths 8 8 | pmcid stringlengths 8 11 ⌀ | source stringclasses 2
values | rank int64 1 9.78k | sections unknown | tokens int64 3 46.7k |
|---|---|---|---|---|---|
31852508 | PMC6921574 | pmc | 8,462 | {
"abstract": "Objectives We report a transcriptome acquisition for the bath sponge Spongia officinalis , a non-model marine organism that hosts rich symbiotic microbial communities. To this end, a pipeline was developed to efficiently separate between bacterial expressed genes from those of eukaryotic origin. The transcriptome was produced to support the assessment of gene expression and, thus, the response of the sponge, to elevated temperatures, replicating conditions currently occurring in its native habitat. Data description We describe the assembled transcriptome along with the bioinformatic pipeline used to discriminate between signals of metazoan and prokaryotic origin. The pipeline involves standard read pre-processing steps and incorporates extra analyses to identify and filter prokaryotic reads out of the analysis. The proposed pipeline can be followed to overcome the technical RNASeq problems characteristic for symbiont-rich metazoan organisms with low or non-existent tissue differentiation, such as sponges and cnidarians. At the same time, it can be valuable towards the development of approaches for parallel transcriptomic studies of symbiotic communities and the host."
} | 296 |
35056525 | PMC8778969 | pmc | 8,464 | {
"abstract": "Ammonia from livestock manure reacts with chemical components discharged from various emission sources to produce airborne particulate matter. This study aimed to investigate a novel effective microbial agent to suppress ammonia gas emitted from manure. Both isolated L12I and 12III strains, identified as Pediococcus acidilactici (PA), were selected for their superior activity in assays performed with the evaluation criteria such as acid production, ammonia decomposition, and urease inhibition, which are key factors influencing ammonia excretion. The survivability of PA strains was confirmed by an increase in DNA abundance in the manure. PA strains lowered the pH of manure and suppressed the growth of hyper-ammonia-producing bacteria (HAB) possessing urease activity. The L12I and 12III treatment groups showed 23.58% and 38.00% emission reductions, respectively. Especially, the 12III strain was proven to be the more effective strain for reducing ammonia gas emission, with the best ability to reduce pH and inhibit HAB. The strains could have an additive effect in improving the manure quality as a nitrogen fertilizer by preserving the total nitrogen and urea content. These results suggest that PA strains can be used as unprecedented microbial agents to improve manure-derived environmental pollution and improve fertilizer quality.",
"conclusion": "4. Conclusions This study was conducted to investigate the cause of ammonia, a major factor of air pollution, emitted from livestock manure and to verify the efficacy of a novel microbial agent to reduce ammonia gas. L12I and 12III ( P. acidilactici ) were selected as effective strains to reduce pH, ammonia concentration, and urease activity, which act as emission factors. Based on the results of this study, it was verified that L12I and 12III are strains capable of acidifying manure and inhibiting the growth of HAB strains. Furthermore, it was confirmed that the selected strains with verified activity are unprecedented microbial agents with superior effects in reducing ammonia gas emitted from manure and fixing nitrogen for use as a fertilizer. Therefore, this study can be utilized as a basis for applying various control methods to reduce the emission of ammonia gas and contribute to the mitigation of air pollution derived from manure by utilizing the discovered microbial agents.",
"introduction": "1. Introduction Recently, it has been reported that the fine dust level in Korea corresponds to a high concentration risk level among countries that have signed the Convention on the Organization for Economic Co-operation and Development (OECD) [ 1 ]. Fine dust is a carcinogenic substance that causes various diseases due to oxidative penetration into human organs, such as parts of the respiratory system and skin [ 2 ]. Gaseous chemicals emitted from various sources, such as livestock, factories, and automobiles, react with each other and produce fine dust in the atmosphere, and ammonia serves as the main precursor of fine dust. Ammonia reacts with other volatile components such as oxides of nitrogen and sulfur and volatile organic compounds in the atmosphere to produce particulate fine dust, such as ammonium sulfate and ammonium nitrate [ 3 ]. In addition, ammonia itself directly irritates human eyes, the respiratory tract, and skin, and prolonged exposure can permanently damage human health [ 4 ]. According to recent reports by the European Environment Agency (EEA, EU) and the National Institute for Environmental Research (NIER, Incheon, Korea), the livestock sector accounts for a large proportion of ammonia emissions worldwide, and the field of livestock manure administration is reported to be responsible for about 90% of ammonia emission in the agriculture sector [ 5 , 6 ]. Livestock manure contains various nitrogen compounds including urea, and ammonia accumulates in the manure because of the enzymatic activity (urease and deaminase) of the microflora present in the manure [ 7 , 8 , 9 , 10 ]. The accumulated ammonia dissolves in water and increases the pH of the manure owing to the separation of hydroxide ions. The pH elevation correlates to an exponential increase in conversion of ammonia to gaseous form, thereby increasing ammonia emission [ 11 , 12 ]. Several methods that reduce the concentration and atmospheric emission of ammonia have been investigated [ 13 ]. Among these methods, microbial application has been shown to result in fewer secondary environmental pollution problems compared to physicochemical methods such as gas barriers, acidifiers, and absorbents [ 14 , 15 ]. However, many studies on biological inhibitors focus on reducing harmful gases such as ammonia through nitrogen metabolism, and studies on the cause of ammonia emission from manure and the interaction between microflora by inhibitors are insufficient [ 15 ]. Effective microorganism (EM) products are generally utilized to reduce the odor and promote fermentation in agricultural fields and comprise actinomycetes, Bacillus subtilis , lactic acid bacteria (LAB), yeasts, etc. Additionally, extensive research shows that they are known to be effective in reducing various harmful gases [ 16 , 17 ]. Bacillus subtilis and actinomycetes present in EM products are known to be effective in reducing the concentration of nitrogen present as ammonia [ 18 , 19 , 20 , 21 ]. Moreover, LAB can facilitate acid production and exert antimicrobial activity against pathogens via bacteriocin production [ 22 , 23 ]. Therefore, this study was conducted with the aim of selecting functional microorganisms that reduce pH, decompose ammonia, and inhibit urea hydrolysis, all of which can contribute to reducing ammonia emission. Furthermore, novel applications of microbial agents can be attempted to prevent the spread of contamination derived from agricultural by-products by verifying their activity in the manure.",
"discussion": "3. Results and Discussion 3.1. Screening and Identification of Potential Effective Bacterial Strains for Reducing Ammonia Emission Sixty-seven bacterial strains were isolated from commercial EM, among which 12 acid-producing bacteria were selected based on colony color change to yellow in BCP-PCA. Further, the pH changes during incubation of these strain cultures for 24 h were monitored ( Figure 2 ). As a result, eight strains were confirmed to produce acid by decreasing the pH level by 2 or more. Among the strains tested, 12III showed the best acid-producing ability. The potential strains, capable of removing ammonia–nitrogen, were screened based on analysis of the ammonia removal rate during incubation. Ammonia removal ability was tested using 67 isolated bacteria ( Figure 3 ). Consequently, all eight selected strains could remove more than 10% of the total ammonia, and L12I removed approximately 20% of the total ammonia. Moreover, a urease activity inhibition assay was conducted to select effective bacterial strains that inhibit urease activity, a major factor of ammonia release ( Figure 4 ). Approximately 20 bacterial strains inhibited urease activity by more than 50%, among which 14 strains showed 100% inhibition. 3.2. Identification of Selected Strains by Using API 50 CHL Test L12I and 12III bacterial strains that exhibited all the required properties such as pH reduction, ammonia degradation, and urease inhibition were finally selected as potentially effective microbes that could reduce ammonia emissions from manure. The selected L12I and 12III strains were both identified as Pediococcus acidilactici (PA) with a similarity of 99.9% ( Table 2 ). PA is a probiotic microorganism that is resistant to a wide range of temperatures and pH [ 40 ], and it is known to have excellent antibacterial effects due to the production of bacteriocin [ 40 ]. The selected strains are expected to be effective in inhibiting ammonia emission from manure on the basis of the screening processes and previously reported activities. 3.3. Ammonia Removal Capacity of the Isolated Bacteria in the Minimal Salt Medium Bacterial nitrogen metabolism has been modulated to reduce the emission of ammonia by nitrification, denitrification, or nitrogen fixation [ 18 , 21 , 41 ]. The ability of selected bacterial strains to utilize the nitrogen stored as ammonia was tested in the M9 minimal medium to limit possible metabolic disturbances due to medium components ( Figure 5 ). As a result, PA strains showed the capacity to remove approximately 10% of the nitrogen stored as ammonia; however, the difference in ammonia removal rate between the strains was not significant ( p > 0.05). Additionally, the ammonia emitted from the medium was not detected. PA can also produce amino acids from inorganic nitrogen sources [ 42 , 43 ]. In this respect, this result suggests that reduced amount of ammonia–nitrogen was not volatilized but instead utilized by PA strains. 3.4. DNA Abundance of Selected Strains in Manure The changes in the PA strains’ DNA abundance in manure under different culture conditions was determined to confirm that the selected bacteria are the predominant type among all the bacteria in the manure ( Figure 6 ). Since the livestock manure composting operations are performed by agitation or sedimentation [ 44 ], the survivability of the selected strains depended on the presence of oxygen. The survivability of the PA strains in manure was confirmed. Furthermore, the DNA of PA was not detected in the manure sample not treated with PA strains. The DNA abundance ratio after incubation compared to the initial of PA strains increased depending on the inoculation concentration and showed a significant increase under all culture conditions ( Figure 6 B). PA strains are thought to be active under both aerobic and anaerobic conditions [ 45 , 46 ], and have the potential to predominate among the microflora in the manure. 3.5. Changes in pH and Growth Inhibition of HAB in Manure by the Selected Strains Ammonia accumulation in the manure dissociates hydroxide ions, resulting in an increase in the pH of the manure. The pH level is closely related to the conversion of ammonia to the gaseous phase, and an increase in pH promotes the release of gaseous ammonia [ 11 , 47 ]. As such, changes in pH and ammonia concentration are essential factors affecting ammonia gas emission. According to recent research reports, iron chloride, sulfuric acid, and other acidic chemicals have been applied to reduce ammonia emissions by modulating the ammonia concentration and gaseous phase conversion [ 8 , 48 , 49 ]. In this study, it was confirmed that the selected strains (L12I, 12III) were effective in reducing the pH and inhibiting urease activity through a screening process [ 40 , 50 ]. A liquid manure medium was used for a more accurate observation of the pH level changes, and the selected bacterial strains were treated under aerobic and anaerobic conditions ( Table 3 ). As a result, the pH level of the untreated control was significantly elevated when observed at 0 h compared to the sample under aerobic conditions. In contrast, the pH level of the groups treated with selected strains showed a tendency to decrease. In the case of anaerobic culture conditions, in contrast to aerobic conditions, the pH of all groups tended to reduce after 24 h of incubation. Among them, the groups treated with PA strains showed a significant decrease in pH compared to the control ( p < 0.05). The reversal pattern between the aerobic and anaerobic control groups could be a phenomenon due to the metabolism of aerobic ammonia producing bacteria in the manure. microorganisms-10-00077-t003_Table 3 Table 3 Change of pH in manure by using selected bacterial strains in aerobic and anaerobic conditions. (L) Low-level inoculation (7.98 log CFU), (H) high-level inoculation (8.58 log CFU). Time (h) 0 24 Conditions Group Average SE Average SE Aerobic Control 7.22 0.03 7.34 * 0.04 L12I (L) 7.28 0.02 7.25 0.02 (H) 7.26 0.06 7.31 0.01 12III (L) 7.25 0.03 7.27 0.03 (H) 7.23 0.04 7.21 0.05 Anaerobic Control 7.24 0.09 6.94 a, * 0.03 L12I (L) 7.28 0.05 6.56 b, * 0.01 (H) 7.22 0.03 6.53 b, * 0.02 12III (L) 7.20 0.01 6.55 b, * 0.01 (H) 7.22 0.03 6.52 b, * 0.01 Values represent the mean ± S.E. ( n = 3). * p < 0.05 vs. 0 h indicates statistical significance. The different letters (a, b) indicate statistically significant difference between different groups at the same time and same conditions (significance level at p < 0.05). Ammonia producing bacteria via urea hydrolysis are present in livestock manure [ 9 , 10 , 51 ]. Furthermore, the DNA of C. aminophilum and P. mirabilis , representative HAB strains with urease activity [ 9 , 10 ], was present in the manure used in this study. The DNA content of C. aminophilum , an obligate anaerobe, increased by more than twofold under all culture conditions ( Figure 7 A), and the DNA of P. mirabilis , an obligate aerobe, increased only under aerobic conditions by approximately threefold ( Figure 7 B). Overall, all PA strains significantly decreased the change in DNA fold change of C. aminophilum compared to the controls under all culture conditions ( p < 0.05). Likewise, PA strains tended to inhibit the DNA increase of P. mirabilis compared to the untreated control under aerobic conditions, and especially, the 12III strain showed a significant inhibition against growth of P. mirabilis even in the low inoculation concentration group ( p < 0.05 vs. control under the aerobic conditions in Figure 7 B by using t -test). In our study, it was suggested that the inhibitory effects against HAB affected pH ( Table 3 ), which was reduced by PA treatment and was similar to the inhibition against HAB ( Figure 7 ). PA is well known for its antimicrobial-peptide-producing capacity [ 22 , 40 ] and has been verified to have positive effects on acid production and urease inhibition ( Figure 2 and Figure 4 ). Therefore, it was expected that PA strains could have a positive effect on the inhibition of ammonia emission due to their inhibitory effect on HAB proliferation through pH control and urease inhibition. Furthermore, the results suggested that decrease in ammonia emission is not only a result of the chemical reaction that occurs but is also due to the major influence of various metabolisms of the microorganisms on ammonia emission. 3.6. Ammonia Emission from Manure and Changes in the Chemical Properties of the Manure The emission pattern of ammonia from manure due to the treatment with selected strains was tested using a gas trapping device and an ammonia quantification method. Overall, the emission pattern was observed over 35 days in the experimental groups, and the pattern increased rapidly and then gradually decreased, similar to that reported for urea-derived ammonia emission from soils [ 36 , 52 ] ( Figure 8 A). Total accumulated ammonia emissions for the 35 days were 841.43 ± 38.36, 643.02 ± 31.05, and 521.71 ± 47.27 mg/kg in the control, L12I, and 12III groups, respectively ( Figure 8 B). The groups treated with L12I and 12III showed significant emission reduction effects of 23.58% and 38.00%, respectively, compared to the control group ( p < 0.05). Additionally, the chemical properties of the manure were investigated to determine how the environmental parameters of the manure changed due to the treatment with the selected strains during the discharge analyses ( Table 4 ). Among the chemical indicators, OM and pH did not show any significant differences among all groups ( p > 0.05), and the T-N of the control group decreased by 22.16% compared to the initial content of emission (at day 0). In contrast, L12I and 12III strains, which had an emission reducing effect, showed 14.93% and 10.27% T-N reduction rates, respectively; in particular, the 12III treatment group, which had the best ammonia emission reduction effect at 38.00%, also had the best nitrogen conservation capacity. In the case of urea reduction rate, the control, L12I, and 12III treated groups showed 62.18%, 45.57%, and 45.00% reduction, respectively. Consequently, it was confirmed that the change in urea content between groups showed similar patterns to the change in T-N content and ammonia emission between groups, suggesting that urea and T-N content were major factors affecting the emissions. Livestock manure, as the main raw material for organic compost, is an important nitrogen source [ 11 ]. Ammonia emissions from manure not only cause nitrogen depletion [ 11 ] but also adversely affect manure quality. The PA strains selected in this study were proven to be beneficial bacterial strains that reduced ammonia emission by preserving the nitrogen content in manure and ultimately improving the manure quality as a fertilizer. It is difficult to maintain uniformity in manure environment because the distribution of intestinal microorganisms can steadily change according to the diet and activity of the host [ 53 , 54 ]. For this reason, the determination of the environmental changes, identified in this study as decisive factors for ammonia emission, was difficult. However, in this study, it was possible to verify that their ability had an effect on nitrogen fixation of manure and reduction in ammonia emission by treating L12I and 12III, which are selected strains with growth inhibition of the HAB strain and acidification ability in manure."
} | 4,335 |
38014290 | PMC10680687 | pmc | 8,465 | {
"abstract": "Computations involved in processes such as decision-making, working memory, and motor control are thought to emerge from the dynamics governing the collective activity of neurons in large populations. But the estimation of these dynamics remains a significant challenge. Here we introduce Flow-field Inference from Neural Data using deep Recurrent networks (FINDR), an unsupervised deep learning method that can infer low-dimensional nonlinear stochastic dynamics underlying neural population activity. Using population spike train data from frontal brain regions of rats performing an auditory decision-making task, we demonstrate that FINDR outperforms existing methods in capturing the heterogeneous responses of individual neurons. We further show that FINDR can discover interpretable low-dimensional dynamics when it is trained to disentangle task-relevant and irrelevant components of the neural population activity. Importantly, the low-dimensional nature of the learned dynamics allows for explicit visualization of flow fields and attractor structures. We suggest FINDR as a powerful method for revealing the low-dimensional task-relevant dynamics of neural populations and their associated computations.",
"introduction": "1. Introduction One of the major challenges in systems neuroscience is in identifying the right level of abstraction to describe how a neural system functions, and bridging such a description to both the cellular-level implementation and behavior. In one approach, we start with the computational task that the neural system has to solve, and either hand-build ( Hopfield, 1982 ; Gerstner & van Hemmen, 1992 ; Wang, 2002 ) or train a network of model neurons ( Sussillo & Barak, 2013 ; Yang et al., 2019 ; Dubreuil et al., 2022 ; Driscoll et al., 2022 ) to solve this task. While these networks provide insights into how individual model neurons could work together to solve a particular task, these model neurons are often not directly constrained to capture the heterogeneous responses observed in real neurons. It is therefore possible that the mechanisms used by these networks to solve a task do not fully reflect the mechanisms used by real neural populations in the brain. In another approach, we start with the neural population activity measured from an animal performing a computational task, and attempt to infer latent representations, or factors, that are relevant to the task computations ( Cunningham & Yu, 2014 ). The dynamics of these representations (i.e., how they evolve over time) are thought to mediate the computations performed by neural populations ( Vyas et al., 2020 ; Duncker & Sahani, 2021 ), and unsupervised methods have been developed to infer these dynamics from neural population activity. Currently available methods make simplifying assumptions on the dynamics to facilitate inference. For example, dynamics are assumed to be autonomous ( Duncker et al., 2019 ), linear ( Macke et al., 2011 ; Gao et al., 2016 ), switching linear ( Linderman et al., 2017 ; Nassar et al., 2019 ; Zoltowski et al., 2020 ), deterministic except at specific time points ( Pandarinath et al., 2018 ; Kim et al., 2021 ; Keshtkaran et al., 2022 ), or high-dimensional ( Pandarinath et al., 2018 ; Keshtkaran et al., 2022 ). However, the assumptions on dynamics made by these inference methods may not necessarily align with the dynamics in real neural populations. Moreover, most currently available methods, applied naively, do not distinguish between task-relevant and -irrelevant dynamics. The complex response patterns observed in real neurons may in part be due to their mixed selectivity to task-related and -unrelated variables ( Rigotti et al., 2013 ), and the methods we use should separate the task-relevant and -irrelevant components in the neural population responses. To address these gaps, we propose a novel method called FINDR (Flow-field Inference from Neural Data using deep Recurrent networks) and present an overview of the method in Sections 2.1 – 2.3 . The FINDR method builds on the existing methods for latent dynamics inference in two major ways. First, leveraging the flexibility of a recently introduced class of dynamical models called neural stochastic differential equations (nSDE; Li et al. (2020) ; Kidger et al. (2021a ; b )), FINDR learns nonlinear stochastic latent dynamics. This allows FINDR to capture the heterogeneous responses of individual neurons in large populations. We demonstrate in Section 2.5 that FINDR outperforms existing methods in reconstructing the individual neural responses from frontal cortical regions involved in decision-making. Second, we take measures to improve the interpretability of the FINDR-inferred latent dynamics. We constrain the learned dynamics to lie on a low-dimensional latent space and infer how the given external inputs to the system influence the dynamics. Furthermore, we infer the task-relevant and -irrelevant dynamics separately, so that we can focus only on the dynamics of the population that are relevant to the task computations. In Section 2.4 , we show that FINDR can infer interpretable low-dimensional attractor structures from spike trains generated by synthetic neural populations that memorize continuous quantities. FINDR takes the continuous quantities used by the populations as its inputs and uses appropriate attractor structures to maintain the memory of these quantities. We also demonstrate in Section 2.5 that FINDR can express neural population activity in terms of latent representations of dimensions lower than existing methods. Because FINDR can represent neural population activity in low dimensions, even as low as two or three dimensions, we can explicitly visualize the flow field (or the velocity vector field) underlying neural population activity. In Section 2.6 , we show that when we visualize the vector field formed by the frontal cortical neural population during decision-making, we see two attractors, with each of them representing a choice alternative.",
"discussion": "3. Discussion We introduced an unsupervised deep learning method called FINDR, which infers low-dimensional latent stochastic dynamics underlying neural population activity. When FINDR is trained to spike trains simulated from a system hand-built to exhibit continuous attractors, we demonstrated that FINDR can correctly capture the low-dimensional velocity vector fields and the attractor structure. In a real neurophysiological dataset where the ground truth is not known, we demonstrated how we can increase the interpretability of the latent dynamics discovered by FINDR through separate learning of task-relevant and -irrelevant components in the neural population activity. To validate how well FINDR captures neural activities in this dataset, we demonstrated that FINDR-reconstructed PSTHs of individual neurons match the observed PSTHs across different task conditions. As a comparison, we fit SLDS, rSLDS, and autoLFADS on the same dataset and demonstrated that FINDR outperforms these methods, especially in the regime of low latent dimensions. In addition to strong performance on neural activity reconstruction, FINDR discovered interpretable latent vector fields. As an example, we showed that the rat frontal cortical neurons form attractor dynamics relevant to decision-making (see ( Luo et al., 2023 ) for scientific implications). We plan to show the applicability of FINDR on other neural datasets in the near future. While we expect FINDR to be generally applicable to a broad range of neural population data, FINDR may be less applicable to certain datasets than others. FINDR is a deep learning-based model that works well with datasets with a high number of simultaneously recorded neurons and trials. While the exact neuron and trial count that give good performance may vary depending on the dynamics in the dataset and the firing rates, generally an increase in the number of neurons should make FINDR’s estimate of each single-trial dynamical trajectory more accurate, while an increase in the number of trials should make FINDR’s estimate of the vector field more accurate, because FINDR has more latent trajectories that traverse the latent space to infer the vector field from. If we ignore the softplus rectification applied to prevent negative firing rates, FINDR uses a linear map to project the dynamics from the latent space to the neural firing rate space ( Equation (1) ). This means that in datasets where the latent dynamics live in a highly curved manifold instead of a linear subspace of the neural firing rate space, FINDR may have more difficulty fitting the data. If we replace the linear map in FINDR with a nonlinear map, the firing rate predictions could improve. However, it becomes harder to interpret the latent dynamics learned by FINDR. How to learn a latent representation that is interpretable – for example, the one that preserves the geometry of the observables – is an active area of research (e.g., Arvanitidis et al. (2018) ; Bronstein et al. (2021) ; Versteeg et al. (2023) ), and future investigations are needed to address these challenges. In conclusion, FINDR extracts low-dimensional dynamical representation from neural population activity. This representation may provide insights into the neural system, spanning multiple levels of descriptions. On one end, FINDR may add a useful constraint when building a biologically plausible network model, as the dynamics of the network should be consistent with the one discovered by FINDR and therefore real neurophysiological data. On the other end, FINDR may facilitate the development of parsimonious, algorithmic-level models of neural computation (see MMDDM in Luo et al. (2023) for an example). These features of FINDR make it a promising approach that can help bridge the gap between the neuronal-level mechanistic description and the algorithmic description of neural function."
} | 2,481 |
26824190 | null | s2 | 8,468 | {
"abstract": "Recent development and applications of calibrated, fluorescence resonance energy transfer (FRET)-based tension sensors have led to a new understanding of single molecule mechanotransduction in a number of biological systems. To expand the range of accessible forces, we systematically measured FRET versus force trajectories for 25, 40, and 50 amino acid peptide repeats derived from spider silk. Single molecule fluorescence-force spectroscopy showed that the peptides behaved as linear springs instead of the nonlinear behavior expected for a disordered polymer. Our data are consistent with a compact, rodlike structure that measures 0.26 nm per 5 amino acid repeat that can stretch by 500% while maintaining linearity, suggesting that the remarkable elasticity of spider silk proteins may in part derive from the properties of individual chains. We found the shortest peptide to have the widest range of force sensitivity: between 2 pN and 11 pN. Live cell imaging of the three tension sensor constructs inserted into vinculin showed similar force values around 2.4 pN. We also provide a lookup table for force versus intracellular FRET for all three constructs."
} | 291 |
36388997 | PMC9663882 | pmc | 8,471 | {
"abstract": "Summary From flexible electronics and multifunctional textiles to artificial tissues, polymers penetrate nearly every aspect of modern technology. High thermal conductivity of polymers is often required in their applications, where heat dissipation is crucial to maintain product reliability and functionality. However, the intrinsic thermal conductivity of bulk polymers is largely hindered by the randomly coiled and entangled chain conformation. Here, we report a copolymerization strategy that can simultaneously manipulate the intrachain and interchain hopping and increase the thermal conductivity of linear copolymerized polyimide (PI) to three times higher than that of pure PI at a low-level introduction of 2,4,5,7-tetraamino-1,8-dihydroxyanthracene-9,10-dione (10%). In addition, the large-scale copolymerized PI films display thermal stability after annealing. These remarkable results allow bulk PI to be a potential candidate for thermal management, and this copolymerization method may benefit future synthesis of interfacial thermal materials.",
"conclusion": "Conclusion In this article, we report a copolymerization method that can enhance the thermal conductivity of PI films. The integration of the 4NADA in the backbone of PI improves the intramolecular interactions and the degree of structural order, and the observed thermal conductivity is three times higher than that of bulk PI at 10% loading of 4NADA. With this common and easily achieved technology, it is now possible to synthesize large-scale thermally conductive PI that also has good electrical insulation, high mechanical strength, and high-temperature stability.",
"introduction": "Introduction Polymers are generally considered thermal insulators and have a low thermal conductivity in a narrow range of 0.1–0.3 W m −1 K −1 , which significantly limits their applications, for example, wearable sensors and electronic encapsulation. 1 , 2 , 3 To improve their thermal conductivity, conventional methods usually focus on polymer-based composites, i.e., to use fillers with high thermal conductivity, such as graphene, metal particles, and carbon nanotubes to enhance the thermal transport. 4 However, the phonon mismatch between the fillers and the polymer matrix leads to a non-negligible interfacial thermal resistance. 5 A large amount of fillers is often required to achieve the desired thermal conductivity, yet it can sacrifice the electrical insulation. 6 Meanwhile, the poor contact between the fillers and the polymer matrix along with filler aggregation may be confronted with the composites and can deteriorate the mechanical strength. 5 Therefore, the previous study to enhance filler content and decrease the thermal boundary resistance with less concentration on the thermal conductivity of the polymer matrix proves to be insufficient for composites with high thermal conductivity. 7 Instead, a pioneering study shows that a single extended polyethylene chain has an ultrahigh thermal conductivity of ∼350 W m −1 K −1 (exceeds most metals), motivating current interest in thermal manipulation in pure polymers. 1 , 8 , 9 , 10 , 11 , 12 While strong intrachain hopping can greatly enhance the thermal conductivity in stretched polymer fibers or films, the coiled chains in bulk polymers bring more scattering centers, resulting in inefficient thermal transport and, hence, low thermal conductivity. 13 , 14 , 15 , 16 For example, conjugated polymers are expected to be thermally conductive among various polymers as they possess stiff backbones and intermolecular π-π interactions, with the latter being approximately 10–100 times stronger than the van der Waals force. 17 , 18 However, they also possess low thermal conductivity (∼0.2 W m −1 K −1 ) because of the strong phonon scattering by segmental random rotations, chain entanglements, side chains, and so on. 17 , 18 Therefore, (1) improving the rigidity of main chains, (2) controlling the intermolecular interactions, and (3) inhibiting the aforementioned phonon scattering factors are the keys to achieve high thermal conductivity in conjugated polymers. Efforts have been made to tune the chain structure and interchain bonds of conjugated polymers to explore their potential as good thermal conductors. Xu et al. reported an oxidative chemical vapor deposition method that could extend the conjugated length and stabilize the planar structure in poly(3-hexylthiophene) and increase the thermal conductivity to 2.2 W m −1 K −1 . 17 More recently, researchers utilized light-triggered π-π interactions between the azobenzene groups to control the thermal conductivity in azobenzene. However, reliability and scalability in practical applications may be faced with these materials. 19 Currently, polymers with increased thermal conductivity are still in the lab stage, and it remains a great challenge to fill the gap between lab-scale proof-of-concept and commercialization. In this work, we report the enhanced thermal conductivity in linear copolymerized polyimide (PI) prepared by a common copolymerization method. We copolymerized 2,4,5,7-tetraamino-1,8-dihydroxyanthracene-9,10-dione monomers (4NADA, a sort of atypical anthraquinone-based monomer) into the backbone of PI (Kapton-type) and synthesized large-sized free-standing PI films. Taking advantage of the rigid π-conjugated bonds of 4NADA, the intramolecular thermal transport is improved. Moreover, the coplanar morphology of 4NADA can suppress chain distortions and simultaneously densely pack the chains and increase the thickness of π-stacking. In addition, the α-substituted groups in 4NADA spontaneously form intramolecular hydrogen bonds with carbonyl groups, avoiding excess phonon scattering in the copolymer. All these can facilitate thermal transport in the copolymerized PI, and the thermal conductivity reaches 0.58 ± 0.06 W m −1 K −1 at only 10% loading of 4NADA. As a typical dielectric material, an enhanced thermal conductivity may provide new opportunities for PI as a potential candidate in thermal management. 20",
"discussion": "Results and discussion Molecular design 4NADA was synthesized from 1,8-dihydroxyanthracene-9,10-dione via nitrification and subsequent reduction, following the route mentioned in our previous work. 21 The obtained diamine monomer was copolymerized with 4,4′-diaminodiphenyl ether (ODA) and 1,2,4,5-benzenetetracarboxylic dianhydride to afford a viscous polyamic acid (PAA) solution. After vacuum defoaming, the PAA solution was coated onto a glass substrate and subsequently cured by a thermal annealing process to obtain corresponding PI films. The samples are labeled as PI x , where x being the molar proportion of 4NADA among the diamines, φ 4NADA . The chemical structure of the copolymerized PI is depicted ( Figure 1 B). The 4NADA monomer shows a full domain absorption biased toward red light and, therefore, exhibits a bluish-black color. Figure 2 is a photograph of PI 10% , and the insets are images of pure PI, PI 2% , and PI 10% , where the transparency of the films decreases with the increase of φ 4NADA . It is worth noting that black PI would not hamper the intrinsic thermal properties of PI but can extend its applications in many areas, such as optical fixed attenuators and optical terminators. 21 Figure 1 Design concept for thermally conducted copolymerized polyimide (PI) (A) Left: a scheme of pure PI film, where the molecular chains (indicated by the orange curves) are highly entangled (middle panel). The main thermal pathways are the weak interchain interactions, through which the thermal transport is inefficient. Right: Chemical structure for pure PI. One can see that the benzene rings in 4,4′-diaminodiphenyl ether (ODA) are randomly rotated. (B) Left: a scheme of PI 10% . Middle: The short blue lines represent the integrated 2,4,5,7-tetraamino-1,8-dihydroxyanthracene-9,10-dione (4NADA) monomers in the PI chains. Right: Chemical structure for the copolymerized PI. The introduction of 4NADA can enhance the main chain rigidity, as 4NADA is stiffer than ODA. Besides, its coplanar structure provides more opportunities for interchain π-stacking. Thus, more efficient thermal channels are formed in the copolymerized PI (indicated by the yellow bold lines in the middle panel). n and m are the mole numbers of ODA and 4NADA. Figure 2 A representative image of PI 10% The scale bar is 1 cm. Insets: Photographs of pure PI, PI 2% , and PI 10% , respectively (from left to right). The color changes from yellow to black as the molar proportion of 4NADA increases. The scale bars are 2 cm. The design concept for the copolymerized PI is illustrated in Figure 1 . 4NADA is a rigid molecule with a coplanar structure ( Figure 1 B). Copolymerizing 4NADA into the backbone of PI can improve the stiffness of the main chains. Compared to ODA, which has a rotatable C-O-C bond between the benzene rings ( Figure 1 A), 4NADA can suppress segmental rotation. Molecular chains with fewer distortions can be formed by the substitution of 4NADA for ODA, which may offer more chances for the formation of densely-packed intermolecular stacking. In other words, stiffer, extended, and more densely-packed molecular chains ( Figure 1 B, right panel) may be developed by copolymerization, which can effectively improve intrachain and interchain thermal transport in the obtained PI films. 17 Thermal measurement and sample characterization At the low-level introduction of 4NADA, the thermal conductivity increases as a function of φ 4NADA and reaches the maximum of 0.58 ± 0.06 W m −1 K −1 at φ 4NADA = 10%, almost three times larger than that of the pure PI (0.21 ± 0.02 W m −1 K −1 ) ( Figure 3 B). Rigid polymer chains are generally better thermal conductors since their high bond-stretching energy corresponds to strong intramolecular hopping. 18 , 22 , 23 Therefore, the integrated 4NADA segments in the backbones can enhance the intrachain interactions and contribute to the thermal transport network. Figure 3 Thermal measurement results All data are represented as mean ± SEM. (A) The thermal conductivity of the samples as a function of molar proportion of 4NADA, φ 4NADA . (B) High-temperature stability of the copolymerized PI films. After annealing, the thermal conductivity of PI 10% maintains nearly the same as before. Furthermore, the coplanar structure of 4NADA constrains the atoms to be in the same atomic plane, thus greatly increasing the stiffness of segmental rotation and facilitating the interchain π-stacking. 23 We employed UV-vis spectroscopy and X-ray diffraction (XRD) pattern to characterize the π-π interactions in the films. The UV cutoff wavelength in Figure 4 A corresponds to the energy gap of electronic transfer between π-orbitals. 24 , 25 From PI to PI 10% , the red shift of it indicates stronger π-electrons transfer with increasing 4NADA. This is interpreted not only by the electron-donating ability of the substituents on the anthraquinone but also by the longer conjugation length in the copolymer. 21 , 26 Figure 4 Sample characterization (A) UV-vis transmission spectra for the pure and copolymerized PI films. The colored curves with a different cutoff wavelength from left to right indicate the PI films with increasing φ 4NADA . (B) Exemplary cross-plane XRD profiles for the pure and the copolymerized PI films, which are represented by multiple Lorentzian peaks. Figure 4 B shows the cross-plane XRD profiles of the three exemplary films. Due to the wide full width at half maximum (FWHM) and low intensity of the single peaks, the XRD pattern shows a broad shoulder for pure PI, which can be represented by three single Lorentz peaks at 19.8° (interspacing of 4.5 Å), 23.4° (interspacing of 3.8 Å), and 29.7° (interspacing of 3 Å). The coherence length ( L c ) of an ordered region is defined by the Scherrer equation 27 , 28 and equals 0.89λ/(FWHM cosθ), where λ is the wavelength of the X-ray in radian. After the introduction of 4NADA, these Lorentz peaks shift to higher diffraction angles, indicating the formation of more densely-packed chains. For PI 10%. , the peak positions are 21.1°, 25.0°, and 30.8°. Meanwhile, the last two peaks, corresponding to the ordered structures in PI films, display narrower FWHM as φ 4NADA increases, indicating the ordered region transferred to a larger one with densely-packed chains ( Table 1 ). 29 , 30 However, even for PI 10% , these peaks are dispersive. Thereby, the corresponding short L c is unlikely to form a long-range ordered structure. Although no intense peak assignable to the large crystalline region is observed, the gradually elongated L c confirms the formation of partially ordered structures with an increase of φ 4NADA . 23 Table 1 XRD analysis for the samples Sample Peak position (°) FWHM Coherence length (Å) PI 23.7 8.7 9.6 29.7 12.9 6.3 PI 2% 24.3 8.4 9.6 30.7 9.3 8.3 PI 10% 25 7.6 10.6 30.8 8.4 9.7 Data are represented as mean ± SEM. FWHM, full width at half maximum. With the introduction of 4NADA, the interchain π-stacking becomes more densely-packed, and its thickness is extended. First, ordered regions themselves have higher thermal conductivity than entangled chains. 10 Second, larger ordered regions can reduce the interfaces between the ordered and disordered regions. 31 Practically, the phonon-boundary scattering takes place if the size of the sample is smaller than the phonon mean free path, causing a size-dependent or thickness-dependent thermal conductivity. 32 , 33 A larger-sized sample usually has higher thermal conductivity, in which phonons process less boundary scattering. 31 , 32 , 33 In this case, regarding the entire film that contains both the ordered and disordered regions, larger thicknesses of π-stacking can reduce the ordered-disordered interfaces and, therefore, contribute to the enhanced thermal conductivity. Moreover, π-stacking has significant implication on ordering and stretching the polymer chains and, thus, indirectly enhances thermal transport along the chain via the intrachain covalent bonding. Additionally, as denoted by the nuclear magnetic resonance results in our previous study, the side groups of 4NADA manifest different nucleophilic activities due to the formation of intramolecular hydrogen bonds, leading to the linear conformation of the copolymer. 21 In this case, no extra phonon scattering would be introduced by side groups in the copolymerization process. 34 However, with the addition of 4NADA, the thermal conductivity slowly decreases. The reduced thermal conductivity is rationalized by the stacking of polymer chains at the high molar proportion of 4NADA. The possible increasing of the free volume in the random copolymer with φ 4NADA may increase the difficulty in the formation of intermolecular hydrogen bond and hence lead to a nonuniform PI film, demonstrated by the reduced glass transition temperature ( T g ) at φ 4NADA of 15% ( Figure S2 ). 21 Another justification is the breaking strain of the films, which decreases as φ 4NADA increases ( Figure 4 B) and becomes too fragile to measure when φ 4NADA exceeds 15%. As a result, the thermal conductivity drops with further addition of 4NADA and to 0.41 ± 0.04 W m −1 K −1 at φ 4NADA of 20%. Our previous study suggests several thermal physical properties (other than thermal conductivity) of the copolymerized films. 21 From φ 4NADA = 0 to 10%, the glass transition temperature ( T g ) slightly increases from the initial 421°C to 434°C ( Figure S2 ), and the temperature of 5% weight loss ( T d 5% ) remains nearly constant, both indicate good thermal stability properties. Such improvement of T g is influenced by the coplanar structure and rigid bonding of 4NADA, as the backbone of copolymerized PI became more rigid and more difficult to rotate with the increasing 4NADA content. 21 , 35 To prove the stability of thermal conductivity, we placed the T-bridge sensor carrying a PI 10% film into a chamber and annealed it at T = 200°C for 1 hour. After this process, the whole device was cooled down to room temperature, and then the thermal conductivity of PI 10% was measured again. We find that after annealing, the change of the thermal conductivity is within the experimental errors ( Figure 3 B). This is comprehensible as 200°C is far below their T g , the copolymerized films are still in the glassy state, and the chains would not be softened due to the temperature rise. Although we did not test the thermal conductivity at a temperature higher than 200°C, the thermal conductivity stability at T = 200°C can satisfy the needs of thermal management in many cases, like chip cooling. Physical variables contributing to the enhanced thermal conductivity In the previous paragraphs, we discussed the molecular interactions (both intrachain and interchain) and heat-carrier scattering effects that can impact the thermal conductivity. How the chain structure and conformation affect the physical variables contributing to the thermal conductivity has been proposed and is quite important to know. 7 Due to the lack of periodic lattice in amorphous polymers, atomic vibrations described by the phonon picture are always a settlement. Allen et al. proposed that the vibration modes in amorphous polymers are categorized as propagons, diffusons, and locons, where diffusons, neither fully localized nor propagating, are the main heat carriers. 36 However, recent works show that disordered materials can have a phonon-like mode which can propagate hundreds of nanometers, supporting the phonon picture. 37 , 38 Moreover, the PI studied in this work is not fully amorphous but with ordered structures, which can be reflected by XRD patterns and the changes in FWMH and L c , due to the extended and more densely-packed molecular chains. Therefore, the concept of “mean free path” is still usable to describe the average length for vibration modes here. 2 To provide a straightforward picture and for convenience, the kinetic theory is used here. According to the kinetic theory, the thermal conductivity is described as κ ∼ cvl , where c , v , and l are, respectively, the specific heat, phonon group velocity, and phonon mean free path. To resolve this, we measured the density and Young’s modulus of the films. The density of the copolymerized films remains nearly constant as x varies ( Figure 5 A). The specific heat is measured by using a differential scanning calorimeter, and the results are shown in Figure 5 B, which decreases from and by 22% for PI 10% . Young’s moduli ( E ) of the films are determined by tensile tests. E ∼ ρv 2 , where ρ and v are density and average sound velocity, respectively, increases from 1.5 ± 0.2 to 2.2 ± 0.2 GPa with φ 4NADA increasing from 0 to 10% ( Figure 5 C). 21 Thus, the average sound velocity is increased by 20% for PI 10% . Therefore, we conclude that the enhanced thermal conductivity of copolymerized PI is mainly due to the improved phonon mean free path, consistent with the picture of extended and more densely-packed molecular chains. Figure 5 Physical variables contributing to the enhanced thermal conductivity Data are represented as mean ± SEM. (A) Density of the pure and the copolymerized PI films. (B) Specific heat of the samples. (C) Young’s modulus and breaking strain of the pure and the copolymerized PI films. In conventional bulk polymers, the phonon localization and phonon scattering lead to a relatively low phonon mean free path. By introducing 4NADA in the main chains, the structure order has been improved, and a longer phonon mean free path must underpin the higher thermal conductivity than the bulk value. Conclusion In this article, we report a copolymerization method that can enhance the thermal conductivity of PI films. The integration of the 4NADA in the backbone of PI improves the intramolecular interactions and the degree of structural order, and the observed thermal conductivity is three times higher than that of bulk PI at 10% loading of 4NADA. With this common and easily achieved technology, it is now possible to synthesize large-scale thermally conductive PI that also has good electrical insulation, high mechanical strength, and high-temperature stability. Limitation of this study In this study, we propose a copolymerization method enabling the fabrication of large-sized free-standing PI films. In comparison to conventional bulk polymers, the copolymerized PI films have an enhanced thermal conductivity owing to the simultaneously improved intramolecular and intramolecular interactions. The thermal conductivity reaches its maximum at 10% loading of 4NADA and decreases with the addition of the rigid monomer. Here, we only use Kapton-type PI for copolymerization. Other configurations of PI are not synthesized and tested."
} | 5,205 |
28165005 | PMC5303878 | pmc | 8,472 | {
"abstract": "By nature of their small size, dense growth and frequent need for extracellular metabolism, microbes face persistent public goods dilemmas. Genetic assortment is the only general solution stabilizing cooperation, but all known mechanisms structuring microbial populations depend on the availability of free space, an often unrealistic constraint. Here we describe a class of self-organization that operates within densely packed bacterial populations. Through mathematical modelling and experiments with Vibrio cholerae, we show how killing adjacent competitors via the Type VI secretion system (T6SS) precipitates phase separation via the ‘Model A' universality class of order-disorder transition mediated by killing. We mathematically demonstrate that T6SS-mediated killing should favour the evolution of public goods cooperation, and empirically support this prediction using a phylogenetic comparative analysis. This work illustrates the twin role played by the T6SS, dealing death to local competitors while simultaneously creating conditions potentially favouring the evolution of cooperation with kin.",
"discussion": "Discussion Phase separation is well-known to drive pattern formation in biology 51 52 , but has mainly been investigated using either Turing activator-inhibitor feedbacks 53 54 , or positive density-dependent movement, described by the Cahn-Hilliard equation 51 55 56 57 . In this paper, we describe a third general mechanism of self-organized pattern formation: targeted killing of non-kin competitors. This drives a ‘Model A' phase separation; the kinetics of this coarsening process—described by the Allen-Cahn equation—only depend on a few cellular details. While we restrict our analysis in this paper to the T6SS, the role of antagonistic interactions in structuring biological communities it is probably far more general, applying to diffusible compounds that kill adjacent non-kin in both micro-organisms (for example, antibiotics) and macro-organisms (for example, allelopathy in plants 58 and animals 59 ). However, while ‘Model A' coarsening is universal, the realization of such dynamics in a densely packed, immobile, athermal system is likely unique to biology. Physically, this system bears similarities to active matter 18 51 56 57 ; phase separation has also been observed in these far from equilibrium active systems, wherein constituents expend energy to move. Phase separation in these systems typically occurs due to differences in mobility as a function of density; constituents move slowly through crowded regions, and quickly through low density regions. Mobility-induced phase separation has been observed (or predicted) in systems as varied as swimming bacteria 60 , self-propelled colloids 61 62 , mussels 51 , granular rods 63 , active filaments 64 65 , rotating particles 66 , among other systems 52 . In the current system, activity is derived from reproduction and killing events at high density rather than constituent mobility 67 , leading to a ‘Model A' transition. Model A coarsening captures the behaviour of a broad array of phase transitions that lack conservation. This transition was originally developed to model magnetization in ferromagnetic materials via the Ising model. Ferromagnetic spins have minimum energy when they align; they do so via Glauber spin flips, leading to a change in the overall magnetization. The physical universality of this transition may be reflected in the strong correlation between secretome and T6SS effectors and apparatuses seen in Fig. 4 . The microscopic details of the system do not strongly affect coarsening, so long as densely packed cells are equipped with T6SS. In recent years, there has been a growing appreciation that many microbial behaviours requiring extracellular metabolism are susceptible to social exploitation. Here we show how simple cell-cell aggression can, as a consequence, create a structured population favourable to cooperation. Clearly, many factors contribute to the structure and function of microbial communities 1 4 15 16 19 20 21 22 42 47 . However, because T6SSs are common (found in ∼25% of Gram-negative bacteria 68 ), and microbes often live in dense communities, phase-separation driven by contact-mediated killing may have a fundamental role in defining the genetic composition and ecosystem-level functionality of microbial communities worldwide."
} | 1,093 |
34906223 | PMC8670125 | pmc | 8,474 | {
"abstract": "Background Diatoms are well known for high photosynthetic efficiency and rapid growth rate, which are not only important oceanic primary producer, but also ideal feedstock for microalgae industrialization. Their high success is mainly due to the rapid response of photosynthesis to inorganic carbon fluctuations. Thus, an in-depth understanding of the photosynthetic carbon fixation mechanism of diatoms will be of great help to microalgae-based applications. This work directed toward the analysis of whether C4 photosynthetic pathway functions in the model marine diatom Phaeodactylum tricornutum , which possesses biophysical CO 2 -concentrating mechanism (CCM) as well as metabolic enzymes potentially involved in C4 photosynthetic pathway. Results For P. tricornutum , differential proteome, enzyme activities and transcript abundance of carbon metabolism-related genes especially biophysical and biochemical CCM-related genes in response to different concentrations of CO 2 were tracked in this study. The upregulated protein abundance of a carbonic anhydrases and a bicarbonate transporter suggested biophysical CCM activated under low CO 2 (LC). The upregulation of a number of key C4-related enzymes in enzymatic activity, transcript and protein abundance under LC indicated the induction of a mitochondria-mediated CCM in P. tricornutum . Moreover, protein abundance of a number of glycolysis, tricarboxylic acid cycle, photorespiration and ornithine–urea cycle related proteins upregulated under LC, while numbers of proteins involved in the Calvin cycle and pentose phosphate pathway were downregulated. Under high CO 2 (HC), protein abundance of most central carbon metabolism and photosynthesis-related proteins were upregulated. Conclusions The proteomic and biochemical responses to different concentrations of CO 2 suggested multiple carbon metabolism strategies exist in P. tricornutum . Namely, LC might induce a mitochondrial-mediated C4-like CCM and the improvement of glycolysis, tricarboxylic acid cycle, photorespiration and ornithine–urea cycle activity contribute to the energy supply and carbon and nitrogen recapture in P. tricornutum to cope with the CO 2 limitation, while P. tricornutum responds to the HC environment by improving photosynthesis and central carbon metabolism activity. These findings can not only provide evidences for revealing the global picture of biophysical and biochemical CCM in P. tricornutum , but also provide target genes for further microalgal strain modification to improve carbon fixation and biomass yield in algal-based industry. Supplementary Information The online version contains supplementary material available at 10.1186/s13068-021-02088-5.",
"discussion": "Discussion Improvement of glycolysis, TCA cycle, photorespiration and OUC activity contribute to operating a LC-induced mitochondria-mediated C4-like CCM in P. tricornutum to cope with the CO 2 limitation Based on the higher activity of mitochondria-localized PEPCK, PEPC2, NAD-MDH, NAD-ME and PYC1 under LC conditions, we proposed a LC-induced mitochondria-mediated C4-like CCM in P. tricornutum (Fig. 7 ), whereby P. tricornutum regenerates OAA through direct carboxylation of PYR by PYC1 rather than firstly converting PYR into PEP via PPDK, which is different from these NA(D)P-dependent ME and PEPCK-dependent C4 pathway in higher plants [ 26 ], and also different from the PEPCK-dependent C4-like pathway reported in the diatom T. weissflogi i [ 7 ]. Kustka et al. [ 4 ] previously proposed that diatom Thalassiosira pseudonana operates a “closed-loop biochemical model” in response to LC conditions, in which the generation and the subsequent decarboxylation of the C4 acid (OAA) were brought out by plastid-localized PEPC2 and PYC, respectively, and the regeneration of PEP from PYR was in a glycine decarboxylase- dependent manner instead of the PPDK-mediated manner. It should be noted that the biochemical CCM in P. tricornutum has many similarities with the biochemical model in T. pseudonana . For example, neither ME nor PEPCK, mediated decarboxylation for C4 pathway, are located in chloroplast, and both proteins are localized in mitochondria; PPDK not with a role in maintaining the operation of C4 pathway; and the presences of chloroplast localized PEPC and PYC may ensure the regeneration and decarboxylation of OAA in C4 pathway. However, the decarboxylation of OAA by PYC requires a specific reaction condition, namely, at a neutral pH with an ATP / ADP ratio of 2.5 and an OAA concentration > 1 mM. Although the transcriptional abundance of the chloroplast-localized PYC2 in the present study was significantly increased, it remains uncertain whether PYC2 in P. tricornutum plays the same role as that in T. pseudonana due to the lack of data on the exact cellular environment that requires for decarboxylation of OAA by PYC2. In this study, with the operation of the mitochondria-mediated C4-like pathway, the concentration of CO 2 generated from the decarboxylation of OAA and MAL via PEPCK and NAD-ME, respectively, might be elevated in mitochondria. Although it is known that diatom plastids are surrounded by four membranes, there are very close physical interactions observed between plastids and mitochondria, which may make energetic interactions between the two organelles possible [ 27 ], facilitating the entry of mitochondria-generated CO 2 into plasmids via passive diffusion or active transport (after conversing to HCO 3 − ), followed by fixation of this molecule in the proximity of RuBisCo. For the operation of the mitochondria-mediated C4-like pathway in LC-cultured P. tricornutum , the carbon skeleton and energy supply are essential. The lower expression of photosynthesis, Calvin cycle and PPP-related proteins in P. tricornutum under LC conditions indicated that these pathways make no contribution to the operation of this mitochondria-mediated C4-like pathway. The glycolysis and TCA cycle are not only the sources of carbon skeletons, but also the main suppliers of ATP (energy) needed for cell metabolism. Since carbon concentration is an energy-consuming process, we hypothesized that the LC-induced upregulation of glycolysis and TCA cycle may be the ATP supplier for the operation of the C4-like CCM in mitochondria. This might be the strategy by which P. tricornutum copes with carbon limitations under LC conditions. Additionally, with LC cultivation, several characteristic related proteins in photorespiration, such as GOX, AGAT and glycine decarboxylation-related proteins was upregulated, indicated the activity of photorespiration was enhanced in P. tricornutum . The enhancement of photorespiratory carbon cycle was also found in LC measured Chlamydomonas reinhardtii [ 28 , 29 ], and Nannochloropsis oceanica [ 30 ], and T. pseudonana [ 4 ] based on omics analysis. As we know, photorespiration is an important pathway to recapture carbon potentially lost due to the oxygenation reaction of Rubisco [ 31 ]. However, with the activation of photorespiration, the release of ammonium ions and CO 2 in the mitochondria would increase. The former has to be recaptured as excessive accumulation of ammonia can cause damage to algal cells [ 24 ]. The OUC has been suggested an effective way to recycle ammonia and CO 2 and thus to avoid the diffusive loss of nitrogen and carbon. However, none of the OUC proteins quantified were upregulated in T. pseudonana under LC conditions, while the upregulated cytosolic carbamoyl phosphate synthetase (CPS) suggested a possible recuperation of ammonia through glutamine synthetase (GLNA) and CPS activity [ 4 ]. For C. reinhardtii , in which the OUC is absent, the upregulation of both the cytosolic GLNA1 and the chloroplastic GLNA2 gene expression suggests that ammonia generated by photorespiration is mainly recaptured by GLNA [ 28 , 29 ]. Unlike T. pseudonana and C. reinhardtii , N. oceanica recapture ammonia through the upregulated ornithine shuttle associated with OUC rather than GLNA that was not upregulated under LC conditions [ 30 ]. In this study, we found that the way in which P. tricornutum refixed ammonia is largely similar to that of in N. oceanica , as the abundance of proteins involved in the OUC (e.g., mitochondrial CPSIII) increased to different extents while the mitochondrial GLNA was downregulated under LC conditions. As noted, the enhancement of photorespiration and OUC activity might be responsible for the efficient recycling of mitochondrial ammonia and CO 2 for chloroplastic carbon fixation and with a role in the completion of the proposed mitochondria-mediated C4-like pathway. P. tricornutum responds to the HC environment by improving photosynthesis and central carbon metabolism activity For photosynthetic microalgae, the content of light-harvesting pigments is closely related to photosynthetic activity. In Chlorella pyrenoidosa [ 32 ] and Trebouxia sp. [ 33 ], the increase in chlorophyll content promoted the enhancement of photosynthetic efficiency when the environmental CO 2 concentration was elevated. Increased photosynthetic performance of P. tricornutum simultaneously with increasing light harvesting pigment (chlorophyll a + c and fucoxanthin) content under HC conditions was also found in our previous study [ 18 ], consistent with the significant upregulation of the pigment synthesis and photosynthesis-related proteins in HC-cultivated P. tricornutum shown in Fig. 1 and Additional file 5 : Fig. S2. The improvement in photosynthetic activity under HC conditions suggested that the efficiency of light energy conversion was enhanced and that the provisioning of ATP and NADPH from the light reaction used for CO 2 fixation in the Calvin cycle increased. In general, the Calvin cycle is considered to be the source of intermediates for protein and nucleic acid biosynthesis, which is closely related to algal growth. Glycolysis and the TCA cycle are generally considered to provide ATP for intracellular metabolism, while the PPP, in particular, is the main source of reductants in the cytoplasm, playing an important role in the elongation of the fatty acid chain and the synthesis of lipids by supplying NADPH. With HC cultivation, not only the pigments, the photosynthesis performance, and the algal growth of P. tricornutum increased [ 18 ], but also the intracellular biosynthesis of total lipid (including the TAGs) and insoluble sugar content enhanced. As expected, the expression of proteins related to central carbon metabolism (involving Calvin cycle, PPP, TCA cycle and glycolysis) was upregulated to different extents under HC cultivation (Fig. 2 ), which were consistent with the transcriptomic results of P. tricornutum cultured at different carbon concentrations; that is, elevated CO 2 promotes upregulation of the transcriptional abundance of carbon metabolism-related genes [ 34 ]. We therefore hypothesized that the upregulation of photosynthesis and central carbon metabolism activity guarantees the higher growth rate and yields of intracellular fatty acids, lipids, insoluble sugar and other metabolites or storage materials under HC cultivation, which might be the strategy adopted by P. tricornutum in response to high concentrations of CO 2 . The HC-induced upregulation of photosynthesis and central carbon metabolism were distinct from LC cultures, indicating tightly regulated carbon metabolism in P. tricornutum . These HC- or LC-induced upregulated carbon metabolism-related genes, such as PEPCK, ME, PEPC, PYC, GAPDH, and G6PDH, are promising targets for functional exploitation for algal strain improvement. Seo et al. [ 35 ] found that overexpression of PEPC promotes the growth of P. tricornutum , and Yang et al. [ 11 ] proposed that silencing PEPCK could promote the accumulation of TAG in P. tricornutum without affecting its growth. These are quite consistent with high growth rate and high lipid content in HC-cultured P. tricornutum , in which PEPC upregulated 3.91-fold and PEPCK significantly downregulated in protein expression. Additionally, rising interest has recently been focused on the genetic modification of central carbon metabolism, especially the key rate-limiting enzymes involved in the PPP for microalgal strain improvement to develop algal-based biodiesel production [ 19 , 36 – 38 ]. Besides, elevated CO 2 has also been used for microalgal domestication to obtain high biomass and biodiesel or to produce other high-value metabolites. For example, increasing the CO 2 level can enhance the growth rate of Skeletonema costatum [ 39 ] and Chaetoceros spp. [ 40 ] and promote the synthesis and accumulation of fatty acids and TAG in Nannochloropsis oculate [ 41 ], Chlorella vulgaris [ 42 ], Chlorella sorokinia [ 43 ] , etc. It follows that the rational engineering of key genes related to carbon metabolism for algal strains development in combination with high CO 2 cultivation may provide an efficient and economical route for biodiesel production from microalgae."
} | 3,274 |
34671368 | PMC8521030 | pmc | 8,475 | {
"abstract": "Numerous studies have confirmed that arbuscular mycorrhizal fungi (AMF) can promote plant nitrogen and phosphorus absorption, and prime systemic plant defense to plant pathogenic microbes. Despite that, the information on the interaction between AMF and plant pathogenic microbes is limited, especially the influence of plant pathogenic microbes on the effect of AMF promoting plant growth. In this study, 650 independent paired-wise observations from 136 published papers were collected and used to calculate the different effect of AMF with plant pathogenic microbes (DAPP) in promoting plant growth through meta-analysis. The results showed that AMF had a higher effect size on plant growth with pathogenic microbes comparing to without pathogenic microbes, including the significant effects in shoot and total fresh biomass, and shoot, root, and total dry biomass. The results of the selection models revealed that the most important factor determining the DAPP on plant dry biomass was the harm level of plant pathogenic microbes on the plant dry biomass, which was negatively correlated. Furthermore, the change of AMF root length colonization (RLC) was the sub-important factor, which was positively correlated with the DAPP. Taken together, these results have implications for understanding the potential and application of AMF in agroecosystems.",
"conclusion": "Conclusion This intensive meta-analysis of the interaction between AMF and plant pathogenic microbes significantly advances the understanding of plant pathogenic microbes on the functions of AMF plant growth promotion. The dependence of plant growth on AMF was negatively correlated with the harm level of plant pathogenic microbes on plant and positively correlated with the RLC change ratio and had no significant relationships with other biotic factors through a model selection method. All these results help us to understand the beneficial potential of AMF and to find a more efficient AMF species with regard to resistance to plant pathogen and applications in agroecosystems.",
"introduction": "Introduction Feeding an increasing global human population while maintaining the sustainability of the farmland is the most important challenge in the twenty-first century ( Godfray et al., 2010 ; Tilman et al., 2011 ). The primary limiting factor in this challenge is the poor availability of soil nitrogen (N) and phosphorus (P) needed to increase crops yields ( Bennett et al., 2013 ). Consequently, arbuscular mycorrhizal fungi (AMF) have a great potation for more efficient agriculture ( Rodriguez and Sanders, 2014 ), because AMF are the one of the key mechanisms of enhancing the acquisition of N and P by crops ( Smith and Read, 2008 ; Hodge and Fitter, 2010 ; Zhang et al., 2019 ) and also can improve mineral acquisition in plants ( Clark and Zeto, 2000 ; Lehmann and Rillig, 2015 ). AMF form mutualistic associations with the roots of over 80% of land plant species ( Smith and Read, 2008 ) and provide N and P to host plant in return for lipids and/or sugars ( Bago et al., 2000 ; Jiang et al., 2017 ). However, plant pathogenic microbes are wildly distributed in agriculture system and are the cause of least to 15% crop yield losses globally ( Savary et al., 2019 ). In comparison with the function of absorption on N and P, there are few researches focusing on the effect of AMF on plant pathogenic microbes ( Dugassa et al., 1996 ; Veresoglou and Rillig, 2012 ), especially on the interaction between them and how plant pathogenic microbes may influence the plant growth promotion function of AMF. Therefore, systematic research on the interaction between AMF and pathogenic microbes, particularly the influence of plant pathogenic microbes on function of AMF, can enrich our poor understanding of the application of AMF in pathogen protection. Normally, the presence of AMF can lessen the harm of plant pathogenic microbes on plant growth ( Newsham et al., 1995 ; Dugassa et al., 1996 ; Veresoglou and Rillig, 2012 ). Whether this effect is caused by the role of AMF in promotion plant growth, or by a specific microbe–microbe function is still unclear. Pozo et al. (2009) and Campos-Soriano et al. (2012) have found that AMF-associated plants will employ a more efficient defense response against plant pathogenic microbes. This mechanism indicates that AMF may play a direct role in the defense reaction to the plant pathogenic microbes, rather than solely aiding in promoting plant growth. Thus, we hypothesize that AMF may play a more efficient role in promoting plant growth in environments with plant pathogenic microbes in comparison with without pathogens. Furthermore, we also hypothesize that an increased AMF effect due to plant pathogenic microbes would be correlated with the harm level of the pathogens and the effect size of AMF on plant growth, because a greater harm caused by plant pathogenic microbes on plant growth may stimulate a higher defense reaction from plant colonized by AMF through producing more defensive compounds ( Gianinazzi-Pearson et al., 1996 ; Pozo et al., 2009 ; Campos-Soriano et al., 2012 ). In the meanwhile, logically the difference of the effect level of AMF on plant growth with pathogenic microbes may be also determined by the effect size of AMF expressed without pathogenic microbes. Therefore, it is necessary to investigate the key factors of this potential interaction. For both AMF and plant pathogen, their influences on plant growth were normally determined by the abiotic and biotic factors, for example, nutrient condition, the species of host plant, species of AMF, and species of pathogenic microbes ( Berg and Koskella, 2018 ; Jiang et al., 2018 ; Mommer et al., 2018 ; Qin et al., 2020 ). But if and how these factors further influence the interaction between AMF and plant pathogenic microbes on the plant growth is also unknown. In order to reveal a more quantitative understanding of the influence of plant pathogenic microbes on the function of AMF on plant growth, we conducted a global meta-analysis on published articles reporting the influence of AMF on plant biomass with and without plant pathogenic microbes. Furthermore, all the influences on the performance of AMF with plant pathogenic microbes in influencing plant growth including both abiotic and biotic factors were investigated. To do so, a database of the basic abiotic and biotic factors, changes of AMF root length colonization (RLC), and influence of AMF and plant pathogenic microbes on plant growth was constructed from collected data. Using these database, we aimed to address the following two questions: (1) What is the function difference of AMF on plant growth between with plant pathogenic microbes and without? (2) What is the main factor determining this function difference?",
"discussion": "Results and Discussion In total, 138 published papers met our criteria, and 650 independent paired-wise observations were included in this study ( Supplementary Data File S1 ). The result on the effect of AMF and plant pathogenic microbe on plant biomass showed that the inoculation of AMF significantly increased the plant fresh and dry biomass, and plant pathogenic microbe decreased the plant fresh and dry biomass ( Figure 1 ). These results fitted the previous results and confirmed that AMF had positive effect and plant pathogenic microbes had negative effect on plant growth ( Mendes et al., 2013 ; Treseder, 2013 ; Zhang et al., 2019 ). We found that the RLC of AMF was reduced by the pathogenic microbes from the paired T test for combination of fresh and dry biomass for shoot, root, and total ( Figure 2 ). This result revealed that host plant reduced the carbon offering to the AMF when colonized additionally with plant pathogenic microbes ( Dar and Reshi, 2017 ; Singh and Giri, 2017 ). Figure 1 The effect of AMF and pathogenic microbes on plant fresh and dry biomass. Values near the error bar were numbers of observations included in the analysis. Figure 2 Boxplot for root length colonization (RLC; sqrt-transformed) of AMF inoculation and AMF+pathogen inoculation treatments for each sub-database. Paired T test is used to calculate the significant difference: *** p <0.001. N is the number of observations used for paired T test. Our paired-wise comparison revealed that under the plant pathogenic condition AMF had higher effects on plant biomass, which was significant for the plant shoot (DAPP=0.138, CI: 0.015–0.262) and total (DAPP=0.151, CI: 0.025–0.277) fresh biomass, and shoot (DAPP=0.149, CI: 0.071–0.228), root (DAPP=0.133, CI: 0.044–0.223), and total (DAPP=0.163, CI: 0.096–0.231) dry biomass, and was nearly significant for root fresh biomass (DAPP=0.087, CI: −0.064–0.238; Figure 3 ). Taken together these results revealed that plant pathogenic microbes promoted the effect of AMF on plant growth, except for root fresh biomass. The effect size of AMF on plant growth is also regarded as the plant dependence on AMF; in another word, our results revealed that plant pathogenic microbes enhance the dependence of host plant on AMF. If only considering the influence of AMF on nutritional benefits for the host plant, the effect of AMF on a plant with pathogenic microbes should logically not be higher than that without pathogenic microbes. A likely explanation to our results is that AMF not only promote plant growth when plant pathogens are present, but also help to protect the plant from those pathogenic microbes ( Saldajeno et al., 2008 ; Sikes, 2010 ) and compete for the same fatty acid resource with pathogenic microbes ( Jiang et al., 2017 ), which presents an indirect mechanism to increase the effect of AMF on plant growth by reducing the harm of plant pathogenic microbes. Figure 3 The difference of the effect size of AMF on the plant biomass under plant pathogenic microbes and no-pathogenic condition (DAPP). Values near the error bar were numbers of observations included in the analysis. The model selection showed that the effect of pathogenic microbe on plant dry biomass was the most important moderator to determine the DAPP ( Figure 4 ; Table 1 ), which was all negatively correlated with DAPP for shoot, root, and total dry biomass ( Table 1 ; Figure 5 ). The ∆RLC was the sub-important moderator with a positive effect in influencing the DAPP of all the types of biomass ( Figures 4 , 5 ; Table 1 ). These results revealed that the most important factor to determining DAPP is harm level of plant pathogenic microbes on plant growth, which was negatively correlated. This directly demonstrated that the worse harm to host plant from pathogenic microbes, the greater role in systemic defense AMF would play. This could be related to defense priming; the increased harm of pathogenic microbes would further stimulate defense mechanisms in host plants using AMF ( Gianinazzi-Pearson et al., 1996 ; Pozo et al., 2009 ; Campos-Soriano et al., 2012 ). Additionally, this might potentially make the competition of pathogenic microbes with AMF for fatty acid fiercer under the colonization of AMF, which would also reduce the infection of plant microbes in plants and ostensibly increase the effect of AMF on plant growth ( Jiang et al., 2017 ). These results also showed that the presence of AMF could mitigate harmful effects of plant pathogenic microbes and highlighted that the appropriate regimes to maintain AMF diversity and abundance in arable soil are necessary to maintain the health of crops with regard to plant pathogenic microbes. However, our results were calculated from pot experiments, which are simpler systems than agricultural ecosystems. More studies are needed in the future to investigate the actual dependence of diseased crops on AMF through measuring, for example, RLC, AMF abundance, or biomass. We found a better performance for AMF comparing to the normal condition (without pathogenic microbes), despite the reduction of RLC of AMF caused by plant pathogenic microbes ( Figure 2 ). According to the role of the colonization extent (RLC) determining the nutrient exchanging ratio between AMF and host plant and revealing the host plant dependence level on AMF ( Treseder, 2013 ), the lower the reduction in RLC (or even increase) by plant pathogenic microbes, the higher the performance of AMF with pathogenic microbes. This explains the positive correlation between ∆RLC and DAPP. Furthermore, this result also demonstrated that the AMF species and/or host plant, which suffered less RLC reduction by plant pathogenic microbes, has a higher pathogen resistance and should be applied to control the pathogenic microbes. While, we did not analyze the potential impacts of other moderators on the DAPP separately in this meta-analysis, even of which would be significant, our model selection method avoided exaggeration the effect of a single moderator as well as overlapping effects among different moderators in the impact on the DAPP ( Terrer et al., 2016 ; Crawford et al., 2019 ). Thus, our results are likely closer to the true biological factors influencing the DAPP without overrepresentation of falsely “significant” moderators due to methodological artefacts. Figure 4 Variable importance of moderators for the effect size difference of AMF on the plant dry biomass under plant pathogenic microbes and no-pathogenic conditions (DAPP). The importance values are the sum of the weights for the models in which the variable appears. The averaged models included the top 128 candidate models. Moderators with an importance of 0.8 or greater are considered for the significance tests. Table 1 The details of significant importance of moderators for the effect size difference of arbuscular mycorrhizal fungi (AMF) on the plant biomass under plant pathogenic microbes and no-pathogenic conditions. Moderator Importance Estimate 95% CI Lower Upper Shoot dry biomass \n RR of pathogen 1.000 −0.717 −0.838 −0.595 ∆RLC 1.000 0.499 0.292 0.706 Root dry biomass \n RR of pathogen 1.000 −0.757 −0.868 −0.646 ∆RLC 1.000 0.366 0.200 0.533 Total dry biomass \n RR of pathogen 1.000 −0.717 −0.838 −0.595 ∆RLC 1.000 0.499 0.292 0.706 Figure 5 Linear relationships between the effect size difference of AMF on the plant dry biomass under plant pathogenic microbes and no-pathogenic conditions (DAPP) with ∆RLC and effect of plant pathogenic on plant biomass. Significance test for the linear relationship was based on a mixed-effects model with a REML method and values of p ≤0.05 were significant. The point size is proportional to the weight of each observation in model. N means the number of observations."
} | 3,646 |
34443819 | PMC8399360 | pmc | 8,477 | {
"abstract": "Superhydrophobic surfaces attract a lot of attention due to many potential applications including anti-icing, anti-corrosion, self-cleaning or drag-reduction surfaces. Despite a list of attractive applications of superhydrophobic surfaces and demonstrated capability of lasers to produce them, the speed of laser micro and nanostructuring is still low with respect to many industry standards. Up-to-now, most promising multi-beam solutions can improve processing speed a hundred to a thousand times. However, productive and efficient utilization of a new generation of kW-class ultrashort pulsed lasers for precise nanostructuring requires a much higher number of beams. In this work, we introduce a unique combination of high-energy pulsed ultrashort laser system delivering up to 20 mJ at 1030 nm in 1.7 ps and novel Diffractive Laser-Induced Texturing element (DLITe) capable of producing 201 × 201 sub-beams of 5 µm in diameter on a square area of 1 mm 2 . Simultaneous nanostructuring with 40,401 sub-beams resulted in a matrix of microcraters covered by nanogratings and ripples with periodicity below 470 nm and 720 nm, respectively. The processed area demonstrated hydrophobic to superhydrophobic properties with a maximum contact angle of 153°.",
"conclusion": "4. Conclusions Fabrication of functional hydrophobic to superhydrophobic nanostructured surface has been demonstrated on stainless steel. Two different types of nanostructures have been produced with respect to periodicity and input beam polarization. For a low number of pulses around 5, nanogratings have been fabricated with periodicity below half of the input wavelength and with parallel orientation to the beam polarization. In addition, the periodicity of nanograting can be tailored by fluence. For a number of pulses above 10, ripples can be observed with a periodicity close to laser wavelength and perpendicular orientation to the beam polarization. The wettability of the surface can be tailored by following the structure geometry. Ripple structures demonstrated significantly better water repellence compared to shallow nanograting, reaching a contact angle up to 153 ± 3°. The combination of a high-energy pulsed picosecond laser system and the beam-splitting and focusing element allows simultaneous production of 40,401 nanostructured spots with a diameter below 5 µm ordered in a square-shaped matrix covering an area of 1 mm × 1 mm. By moving the stage, a processing rate of more than 8 million microcraters containing periodic nanograting per second can be achieved, significantly improving the throughput compared to the single beam approach and showing great potential for rapid large-scale surface functionalization.",
"introduction": "1. Introduction The inspiration and design of functional surfaces may often come from nature by replicating examples including lotus or rice leaf, mosquito eye, shark skin or cicada wings [ 1 , 2 ]. Among many natural examples, superhydrophobic biological systems attract a lot of attention for applications including anti-icing, anti-corrosion, self-cleaning or drag-reduction surfaces [ 3 ], lab-on-chip devices [ 4 ], control of cellular attachment and bacterial adsorption [ 5 ]. The wettability of a given surface is characterized by the static contact angle (CA) which is the visually measurable angle that a liquid makes with a solid. The surface is hydrophilic for a CA below 90°, hydrophobic above this value. The surface can be called superhydrophobic when CA exceeds 150°. In general, the wetting conditions are described by the two factors: the chemical composition and the surface roughness, which both affect the surface free energy (SFE)—the work that would be necessary to increase the surface area of a solid phase. If the surface is ideally flat, the wettability is determined by the chemical nature of the different phases. In this case, the maximum CA is limited to ~120° [ 6 , 7 ]. To improve wettability further, roughness has to be introduced on the surface. The Wenzel and Cassie–Baxter theories explain wettability through surface roughness [ 8 , 9 ]. According to the Wenzel model describing a fully wetted surface, the originally hydrophilic surface will become more hydrophilic and originally hydrophobic more hydrophobic with the increase in surface roughness [ 10 ]. On the other hand, the Cassie–Baxter equation describes a rough surface, which is able to trap air between surface features when a droplet is deposited on the surface. In this case, water repellence can be enhanced for both, originally hydrophilic and hydrophobic surfaces. A large variety of fabrication methods have been developed for fabrication of functional surfaces including chemical vapour deposition [ 11 ], lithography [ 12 ], chemical etching [ 13 ], plasma treatments [ 14 ] or laser surface texturing [ 15 ]. However, most of these techniques are too slow to be implemented in an industrial environment or they require chemicals and thus are not environmentally friendly. Among these, laser surface texturing offers a flexible, fast and environmentally friendly method for high-quality fabrication of desired micro and nano geometries with high precision and on a large variety of materials. The high-precision laser fabrication of functional structures with dimensions below a few micrometres may often require maximum pulse energy and power to be close to the ablation threshold. In addition, close-to-threshold irradiation with ultrashort pulses may lead to the development of regular nanoscale structures known as Laser-induced Periodic Surface Structures (LIPSS), ripples or nanogratings, which may further enhance surface functionality [ 16 , 17 , 18 ]. As a result, only a small portion of available laser power is used during processing, making the potential industrial application expensive and less ecological. One of the most promising solutions for industrial utilization of lasers for micromachining is multi-beam processing. In this solution, the incident beam is split into a matrix of sub-beams for simultaneous fabrication inside the optimal processing window [ 19 ]. This can be achieved by several techniques including Direct Laser Interference Patterning (DLIP) [ 20 , 21 ] and beam splitting by Diffractive Optical Element (DOE) [ 22 ] or Spatial Light Modulator (SLM) [ 23 ]. Employing up-to-date laser and optical technology these multi-beam approaches can reach high efficiency with high-energy pulsed laser systems by splitting the beam into hundreds of sub-beams or interference maxima positions. For example, Lang et al. [ 24 ] reached the world record in DLIP structuring of polymers with the throughput of 0.9 m 2 /min utilizing an elliptical interference area of 15,000 µm × 50 µm featuring ~676 interference maxima positions. In addition, more than 1000 individual microcraters were fabricated simultaneously by four beam interference patterning used by authors in previous work [ 25 ]. However, the common drawback for DLIP is interference maxima positions follow Gaussian distribution of the input laser beam. Therefore, slowing down patterning over a larger area due to required overlap of interference areas to achieve homogeneous pattern distribution [ 26 ]. More straightforward solution is the use of diffractive elements which may be applied directly in the optical setup. Moreover, advanced DOE design enables to split the beam into the square-shaped matrix of sub-beams with equal intensity. Thus, eliminating overlap for stitching over larger areas. Kuang et al. [ 27 ] demonstrated fast parallel microstructuring with 1 mJ femtosecond laser system utilizing 40 sub-beams in a rectangular matrix. Efficient micromachining with high power-ultrashort laser pulses and 144 sub-beams was also reported by Gillner et al. [ 28 ]. More than 784 sub-beams were applied for micromachining of invar in the previous work of authors [ 19 ]. However, to efficiently utilize the new generation of ultrashort laser systems with power above 0.5 kW and millijoule pulse energy [ 29 ] significantly higher number of sub-beams would be required for high-quality micro and nanostructuring together with a new type of DOEs. This work introduces a novel method for rapid and cost-efficient production of superhydrophobic surfaces by utilizing a compact protype of processing optical system responsible for beam splitting and focusing. The optical system includes a new type of DOE capable of splitting the incident high-quality laser beam into the rectangular matrix of 201 × 201 sub-beams. By applying this optical system, a square-shaped area of 1 mm 2 can be efficiently nanostructured during a few laser pulses by 40,401 sub-beams, which is, to the best of our knowledge, the highest number of sub-beams utilized for simultaneous mask-less laser nanostructuring.",
"discussion": "3. Results and Discussion The combination of advanced beamsplitting element and tightly focused sub-beams requires precise alignment of all components with respect to the laser beam. Especially crucial is the distance between the splitting and focusing system and the sample surface (z-distance). A slight change in a z-distance for 10 µm already results in significant changes in pattern homogeneity. Figure 2 depicts the calculated change in a pattern homogeneity from the ideal shape ( Figure 2 a) to the defocused pattern for 10 µm ( Figure 2 b) and 20 µm ( Figure 2 c) and compares the calculation with experiment ( Figure 2 d–f). Figure 3 shows a detail of the best-quality pattern achieved by scanning the z-distance with a step of 1 µm. The pattern still shows some inhomogeneities which might be improved in the future by applying a precise automated focusing system with sub-micron resolution. In the following step, the formation of nanostructures was investigated with respect to a different number of pulses (N) in a range of 5–50 and pulse energies of 1 mJ, 1.5 mJ and 2 mJ, which corresponds to peak fluences of 0.33 J/cm 2 , 0.49 J/cm 2 and 0.65 J/cm 2 in each spot of the matrix. These experiments revealed a formation of periodic nanostructures formed inside each individual spot, as depicted in Figure 4 . As can be observed in Figure 4 , different kinds of nanostructures can be produced within individual microholes. A significant change in morphology is observed especially with the growing number of pulses. For a single pulse and fluence up to 0.49 J/cm 2 , the surface starts to be modified and slightly disrupted increasing the initial roughness ( Figure 4 a,g). In the case of a higher fluence of 0.65 J/cm 2 , the surface appears to be smoothed by rapid melting and solidification of the thin surface layer ( Figure 4 m). With the increased number of pulses to 5, small periodic nanograting can be observed inside each spot ( Figure 4 b,h,n). The higher is the fluence, the greater is the periodicity of this nanograting and the more melt can be observed inside each spot. The nanograting periodicity was measured as 220 ± 15 nm for 0.33 J/cm 2 , 302 ± 21 nm for 0.49 J/cm 2 and 470 ± 40 nm for 0.65 J/cm 2 . By increasing the number of pulses to 10, this periodicity increased to 229 ± 31 nm for 0.33 J/cm 2 and 381 ± 48 nm for 0.49 J/cm 2 . In the case of the highest fluence of 0.49 J/cm 2 , the small nanograting is already disrupted ( Figure 4 o). Except for the nanograting, a formation of ripples with the periodicity of 720 ± 24 nm and perpendicular orientation to the laser polarization starts to be observed ( Figure 4 c,i,o). The reason for this behaviour might be found in heat accumulation during consecutive laser pulses, when the transient temperature of the material may reach the melting point [ 32 , 33 ]. In addition, Marangoni shear-generated convection could lead to hydrothermal waves and eventually to ripple formation [ 34 ]. As the number of pulses increase, ripples start to dominate and from 50 pulses there is no nanograting observed ( Figure 4 f,l,r). High fluence and a high number of pulses also results in a lot of accumulated heat and melting around the main periodic structures ( Figure 4 l,r). In the case of 0.65 J/cm 2 and 50 pulses, the beginning of microdrilling process can be observed ( Figure 4 r). For the wettability analysis selected structures were produced on the area of 6 mm × 6 mm including structures depicted in Figure 4 b,e,h,k,n,q which were produced by 5 and 20 pulses with a peak fluence of 0.33 J/cm 2 , 0.49 J/cm 2 and 0.65 J/cm 2 . Immediately after laser treatment, samples show hydrophilic properties due to the formation of high-temperature metal oxides which have a high affinity towards the water molecules leading to hydrophilic behaviour [ 30 ]. Ageing in ambient conditions for several days is conventionally used to transfer samples from hydrophilic to hydrophobic state [ 35 ]. To speed up the transition samples were stored in a low-pressure environment containing hydrocarbons due to an oil-rotary vacuum pump [ 36 ]. As explained in our previous work [ 37 ], the dominant presence of non-polar hydrocarbons together with the absence of water molecules and convenient surface geometry is responsible for the superhydrophobic behaviour of the laser-treated sample stored in a low-pressure environment. As shown in Figure 5 , samples stored for 6 h in vacuum conditions exhibit hydrophobic to superhydrophobic behaviour. As can be observed in Figure 5 , contact angle evolution is dependent on applied fluence and the number of pulses. Generally, higher fluence results in a larger spot, and thus, a higher percentage of the surface is textured with a fixed spot separation distance of 5 µm, which results in a higher contact angle. The type of structure also plays a significant role. Only a slight increase in contact angle from 81 ± 2° to 97 ± 4° is observed for shallow nanogratings fabricated by five consecutive laser pulses. On these surfaces, the droplets are pinned down with roll-off angles always above 20°. On the other hand, ripple structures exhibited a contact angle of 153 ± 3° with a roll-off angle of 7 ± 2.5° which can be considered as superhydrophobic surface. The reason for different wettability of these two types of surfaces may be related to the depth of the surface features [ 30 ]. In the case of nanogratings, the depth is always below 80 nm which might have a minimal effect on a heavy 10 µL water droplet that may penetrate these shallow surface features. In contrast, ripple structures exhibit depth between 230–280 nm. Moreover, a higher number of pulses required for the ripple formation also results in a microcrater ablation with a diameter close to 4.9 µm and depth of ~1 µm for the superhydrophobic structure in Figure 4 q. Hence the highest contact angle is reached for the deepest hierarchical structures composed of micro and nanoscale structures, which are often an optimal solution for a stable Cassie–Baxter state [ 9 ]. In addition to wettability, structural colour was observed on the processed sample surface as can be observed in the inset in Figure 5 . The advantageous square-shaped patterned area allows stitching each pattern with 0% overlap and thus significantly improve the throughput compared to the single beam approach. Within 5 ms for 5 consecutive laser pulses (repetition rate of 1 kHz), the simultaneous fabrication with 40,401 sub-beams allows production of 8,080,200 microcraters containing periodic nanograting ( Figure 4 h) per one second. In addition, this throughput may be easily up-scaled by increasing the repetition rate. Thus, in the future, this technique may demonstrate a great potential for rapid large-scale nano and micropatterning for the production of functional surfaces in industrial applications."
} | 3,919 |
27986083 | PMC5159968 | pmc | 8,479 | {
"abstract": "Existing workflows for the analysis of multi-omic microbiome datasets are lab-specific and often result in sub-optimal data usage. Here we present IMP, a reproducible and modular pipeline for the integrated and reference-independent analysis of coupled metagenomic and metatranscriptomic data. IMP incorporates robust read preprocessing, iterative co-assembly, analyses of microbial community structure and function, automated binning, as well as genomic signature-based visualizations. The IMP-based data integration strategy enhances data usage, output volume, and output quality as demonstrated using relevant use-cases. Finally, IMP is encapsulated within a user-friendly implementation using Python and Docker. IMP is available at http://r3lab.uni.lu/web/imp/ (MIT license). Electronic supplementary material The online version of this article (doi:10.1186/s13059-016-1116-8) contains supplementary material, which is available to authorized users.",
"conclusion": "Conclusions IMP represents the first self-contained and standardized pipeline developed to leverage the advantages associated with integrating MG and MT data for large-scale analyses of microbial community structure and function in situ [ 4 , 6 ]. IMP performs all the necessary large-scale bioinformatic analyses, including preprocessing, assembly, binning (automated), and analyses within an automated, reproducible, and user-friendly pipeline. In addition, we demonstrate that IMP vastly enhances data usage to produce high-volume and high-quality output. Finally, the combination of open development and reproducibility should promote the general paradigm of reproducible research within the microbiome research community.",
"discussion": "Discussion The microbiome analysis workflow of IMP is unique in that it allows the integrated analysis of MG and MT data. To the best of our knowledge, IMP represents the only pipeline that spans the preprocessing of NGS reads to the binning of the assembled contigs, in addition to being the first automated pipeline for reproducible reference-independent metagenomic and metatranscriptomic data analysis. Although existing pipelines such as MetAMOS or MOCAT may be applied to perform co-assemblies of MG and MT data [ 44 ], these tools do not include specific steps for the two data types in their pre- and post-assembly procedures, which is important given the disparate nature of these datasets. The use of Docker promotes reproducibility and sharing, thereby allowing researchers to precisely replicate the IMP workflow with relative ease and with minimal impact on overall performance of the employed bioinformatic tools [ 29 , 46 – 48 ]. Furthermore, static websites will be created and associated with every new version of IMP (Docker image), such that users will be able to download and launch specific versions of the pipeline to reproduce the work of others. Thereby, IMP enables standardized comparative studies between datasets from different labs, studies, and environments. The open source nature of IMP encourages a community-driven effort to contribute to and further improve the pipeline. Snakemake allows the seamless integration of Python code and shell (bash) commands and the use of make scripting style, which are arguably some of the most widely used bioinformatic scripting languages. Snakemake also supports parallel processing and the ability to interoperate with various tools and/or web services [ 49 , 51 ]. Thus, users will be able to customize and enhance the features of the IMP according to their analysis requirements with minimal training/learning. Quality control of NGS data prior to de novo assemblies has been shown to increase the quality of downstream assembly and analyses (predicted genes) [ 63 ]. In addition to standard preprocessing procedures (i.e., removal of low quality reads, trimming of adapter sequences and removal), IMP incorporates additional tailored and customizable filtering procedures which account for the different sample and/or omic data types. For instance, the removal of host-derived sequences in the context of human microbiomes is required for protecting the privacy of study subjects. The MT-specific in silico rRNA removal procedure yielded varying fractions of rRNA reads between the different MT datasets despite the previous depletion of rRNA (section “Tailored preprocessing and filtering of MG and MT data”), indicating that improvements in wet-lab protocols are necessary. Given that rRNA sequences are known to be highly similar, they are removed in IMP in order to mitigate any possible misassemblies resulting from such reads and/or regions [ 65 , 66 ]. In summary, IMP is designed to perform stringent and standardized preprocessing of MG and MT data in a data-specific way, thereby enabling efficient data usage and resulting in high-quality output. It is common practice that MG and MT reads are mapped against a reference (e.g., genes, genomes, and/or MG assemblies) [ 28 , 29 , 40 ] prior to subsequent data interpretation. However, these standard practices lead to suboptimal usage of the original data. IMP enhances overall data usage through its specifically tailored iterative co-assembly procedure, which involves four measures to achieve better data usage and yield overall larger volumes of output (i.e., a larger number of contigs ≥1 kb and predicted unique and complete genes). First, the iterative assembly procedure leads to increases in data usage and output volume in each additional iterative assembly step (section “Data usage: iterative assembly”). The exclusion of mappable reads in each iteration of the assembly serves as a means of partitioning the data, thereby reducing the complexity of the data and overall, resulting in a higher cumulative volume of output [ 60 , 63 , 67 ]. Second, the initial assembly of MT-based contigs enhances the overall assembly, as transcribed regions are covered much more deeply and evenly in MT data, resulting in better assemblies for these regions [ 43 ]. The MT-based contigs represent high-quality scaffolds for the subsequent co-assembly with MG data. Third, the co-assembly of MG and MT data allows the integration of these two data types while resulting in a larger number of contigs and predicted complete genes against which, in turn, a substantially higher fraction of reads can be mapped (section “Data usage: multi-omic iterative co-assembly”). Furthermore, the analyses of the human fecal microbiome datasets (HF1–5) demonstrate that the numbers of MG reads mapping to the IMP-based iterative co-assemblies for each sample are comparable to the numbers of reads mapping to the comprehensive IGC reference database (Table 2 ). Previously, only fractions of 74–81% of metagenomic reads mapping to the IGC have been reported [ 35 ]. However, such numbers have yet to be reported for MT data, in which case we observe lower mapping rates to the IGC reference database (35.5–70.5%) compared to IMP-based assemblies (Additional file 2 : Table S3). This may be attributed to the fact that the IGC reference database was generated from MG-based assemblies only, thus creating a bias [ 35 ]. Moreover, an excess of 90% of MG and MT reads from the human fecal datasets (HF1–5) are mappable to either the IGC reference database and/or IMP-based iterative co-assemblies, emphasizing that a combined reference-based and IMP-based integrated-omics approach vastly improves data usage (Table 2 ). Although large fractions of MG and/or MT reads can be mapped to the IGC, a significant advantage of using a de novo reference-independent approach lies within the fact that reads can be linked to genes within their respective genomic context and microbial populations of origin. Exploiting the maximal amount of information is especially relevant for microbial communities with small sample sizes and which lack comprehensive references such as the IGC reference database. Fourth, the assembly refinement step via a contig-level assembly with cap3 improves the quality of the assemblies by reducing redundancy and increasing contiguity by collapsing and merging contigs (section “Assembly quality: multi-omic iterative co-assembly”). Consequently, our results support the described notion that the sequential use of multi- k mer-based de Bruijn graph assemblers, such as IDBA-UD and MEGAHIT, with overlap-layout-consensus assemblers, such as cap3, result in improved MG assemblies [ 38 , 62 ] but importantly also extend this to MG and MT co-assemblies. When compared to commonly used assembly strategies, the IMP-based iterative co-assemblies consisted of a larger output volume while maintaining a relatively high quality of the generated contigs. High-quality assemblies yield higher quality taxonomic information and gene annotations while longer contigs (≥1 kb) are a prerequisite for unsupervised population-level genome reconstruction [ 14 , 19 , 56 ] and subsequent multi-omics data integration [ 39 , 43 , 44 ]. Throughout all the different comparative analyses which we performed, IMP performed more consistently across all the different datasets when compared to existing methods, thereby emphasizing the overall stability and broad range of applicability of the method (section “Assembly quality: multi-omic iterative co-assembly”). Integrated analyses of MG and MT data with IMP provide the opportunity for analyses that are not possible based on MG data alone, such as the detection of RNA viruses (section “Identification of RNA viruses”) and the identification of transcriptionally active populations (section “Identification of populations with apparent high transcriptional activity”). The predicted/annotated genes may be used for further analyses and integration of additional omic datasets, most notably metaproteomic data [ 39 , 43 , 44 ]. Furthermore, the higher number of complete genes improves the downstream functional analysis, because the read counts per gene will be much more accurate when having full length transcript sequences and will increase the probability to identify peptides. More specifically, the large number of predicted genes may enhance the usage of generated metaproteomic data, allowing more peptides, and thus proteins, to be identified."
} | 2,554 |
26035711 | PMC4452783 | pmc | 8,480 | {
"abstract": "Clostridium phytofermentans was isolated from forest soil and is distinguished by its capacity to directly ferment plant cell wall polysaccharides into ethanol as the primary product, suggesting that it possesses unusual catabolic pathways. The objective of the present study was to understand the molecular mechanisms of biomass conversion to ethanol in a single organism, Clostridium phytofermentans , by analyzing its complete genome and transcriptome during growth on plant carbohydrates. The saccharolytic versatility of C . phytofermentans is reflected in a diversity of genes encoding ATP-binding cassette sugar transporters and glycoside hydrolases, many of which may have been acquired through horizontal gene transfer. These genes are frequently organized as operons that may be controlled individually by the many transcriptional regulators identified in the genome. Preferential ethanol production may be due to high levels of expression of multiple ethanol dehydrogenases and additional pathways maximizing ethanol yield. The genome also encodes three different proteinaceous bacterial microcompartments with the capacity to compartmentalize pathways that divert fermentation intermediates to various products. These characteristics make C . phytofermentans an attractive resource for improving the efficiency and speed of biomass conversion to biofuels.",
"conclusion": "Conclusions Analysis of the C . phytofermentans genome revealed a diverse array of genes for metabolism of lignocellulosic biomass and production of alcohols and hydrogen that constitute a unique repertoire among sequenced clostridial genomes with relevance for the biofuels industry. Our analysis of the genome revealed the genomic basis for the generalist behavior of this microbe. The diverse CAZy likely enable complete hydrolysis of cellulosic and hemicellulosic substrates. Unexpectedly, we found no evidence of cellulosomes in C . phytofermentans , suggesting that C . phytofermentans has evolved alternative strategies to optimize degradation and uptake of plant cell wall components [ 16 ]. The absence of a cellulosome simplifies strategies for engineering levels of individual enzymes to improve the conversion of plant biomass to fermentable sugars. Many active sugar transporters are in close proximity to polysaccharide hydrolases, likely cooperating for efficient simultaneous degradation and uptake of carbohydrate growth substrates. Further investigation will be required to determine the substrate-specificity of these transporters. Genomic analysis and transcriptional profiling also suggest that high levels of ethanol production by C . phytofermentans may be due to a combination of factors. These include: 1) increasing the energetic yield of glycolysis by utilizing pyrophosphate-dependent enzymes; 2) high levels of expression of the enzymes involved in ethanol production coupled with the ability to utilize both NADH and NADPH for ethanol biosynthesis; 3) the presence of multiple pathways for the dissipation of excess reducing equivalents; and 4) the presence of sodium-dependent energy generating pathways. Experimental studies will be required to determine if these hypotheses are indeed valid. Examination of the genomes of several well-studied cellulolytic and solventogenic clostridia indicated that very few of the central metabolic enzymes and complexes discussed above are unique to C . phytofermentans . It may therefore be the specific combination of enzymes and their transcriptional regulation that makes C . phytofermentans metabolism unique. Efficient direct conversion of biomass to bioproducts using a microbial catalyst such as C . phytofermentans requires an increased understanding of cell growth dynamics, rate-limiting steps of biomass conversion, enzyme production and regulation. These genome-based experiments and analyses provide a blueprint for identifying bottlenecks and guiding strategies to generate novel productive strains for specific uses.",
"introduction": "Introduction Plant biomass is one of the most abundant renewable energy sources on Earth and a largely underutilized feedstock for biofuels [ 1 ]. Production of biofuels from the lignocellulose fraction of plant biomass differs from production from grains in two fundamental aspects: (1) different types of saccharolytic enzymes are required to break down lignocellulose into soluble carbohydrates; and (2) fermentation of pentose sugars, in addition to hexoses, is required to harvest the majority of energy stored in lignocellulose [ 2 ]. At present, the cost of producing saccharolytic enzymes and the complexity of the hydrolysis and fermentation processes limit the use of plant biomass as a competitive alternative to gasoline and pose key challenges in the development of a global biomass industry for manufacturing a wide range of products from agricultural and forestry wastes [ 3 ]. One potential solution is the use of microbes that produce lignocellulose-decomposing enzymes and simultaneously ferment the resulting hexose and pentose carbohydrates to products such as ethanol. Merging these processes in a single microbe could substantially reduce the costs of lignocellulosic biofuel production [ 4 ]. Such microbes, primarily members of the Clostridiales , are found in natural anoxic environments where vast quantities of cellulose and other plant cell wall components are decomposed. Species of Clostridium have a rich tradition in the development of biofuels. Clostridium acetobutylicum is a long-standing commercially valuable bacterium that has been used to produce acetone, butanol and ethanol from starch [ 5 ]. Processes based on acetone-butanol-ethanol fermentation were industry standards until the late 1940's, when low oil prices favored processes based on hydrocarbon cracking and petroleum distillation techniques. C . acetobutylicum and its relative Clostridium beijerinckii have recently regained market interest for use in the production of butanol as a gasoline and diesel fuel replacement. Microbial fermentation of cellulose has been studied extensively in Clostridium cellulolyticum and Clostridium thermocellum [ 6 – 9 ]. Carbon metabolism during growth on cellulose and cellobiose in C . cellulolyticum has been investigated using carbon isotope labeling and metabolic flux analysis [ 8 ]. To degrade cellulose, C . cellulolyticum and C . thermocellum produce extracellular enzymatic complexes (cellulosomes) that permit bacterial adhesion to insoluble substrates and promote the hydrolysis of cellulose [ 6 , 9 , 10 ]. C . acetobutylicum laboratory strains do not grow on cellulose although they contain genes for cellulosome synthesis [ 7 ] and secrete a small cellulosome [ 11 ]. Products of cellulose degradation, such as cellobiose, are transported across the cell membrane and enter into the Embden-Meyerhof-Parnas pathway. C . acetobutylicum laboratory strains do not grow on cellulose although they contain genes for cellulosome synthesis [ 7 ] and secrete a small cellulosome [ 11 ]. We isolated a new species, Clostridium phytofermentans (strain ISDg ATCC 700394) from forest soil near the Quabbin Reservoir in Massachusetts, U.S.A. that directly ferments all major components of plant biomass, including cellulose, hemicellulose, pectin and starch to yield ethanol as the primary product of fermentation [ 12 ]. The combination of carbohydrate substrate versatility and high ethanol yield in a single organism distinguishes C . phytofermentans from other described species and suggests that it possesses unusual catabolic pathways. Extensive metabolism of the complex sugars within lignocellulose (without high ethanol yield) appears to be a trait found in several hyperthermophiles [ 13 ] but C . phytofermentans stands out as one of few, if not the only mesophile with this capacity. Thus, C . phytofermentans offers opportunities to understand the molecular mechanisms of plant biomass conversion to biofuels in a single organism. These attributes have led to the adoption of C . phytofermentans as a study system by groups in the US, Japan and Europe [ 14 – 25 ]. Here we investigate the unique properties of C . phytofermentans through analyses of its complete genome sequence and transcriptional profiling during growth on key components of plant biomass.",
"discussion": "Results and Discussion \n C . phytofermentans is distinct from other well-studied solventogenic and cellulolytic species found within clostridial Clusters I (Clostridiaceae), III (Ruminococcaceae), and X (Thermoanaerobacteraceae) ( Fig 1 ). A member of Cluster XIV (Lachnospiraceae), C . phytofermentans is closely related to human commensals that have been sequenced as part of the International Human Microbiome Consortium [ 53 ], and to bacteria isolated from rice paddy soils, earthworm intestines and other anaerobic, carbon rich environments ( Fig 1 ). As a genetically tractable [ 15 ] member of this under-explored group, and the first with a publicly available genome sequence, C . phytofermentans is an important point of reference for comparative genomic analyses. 10.1371/journal.pone.0118285.g001 Fig 1 Neighbor-joining tree of C . phytofermentans and related taxa within the class Clostridia based on 16S rRNA gene sequences. Taxa with sequenced genomes are marked with an asterisk. Cluster numbers correspond to the cluster system of Collins et al. [ 68 ]. Bootstrap values were determined for 1,000 replicates. Features of the C . phytofermentans genome \n C . phytofermentans has a single circular 4.8 Mbp chromosome, no plasmids and a G+C content of 35%. The genome encodes 3,926 CDS, 27% of which lack a predicted function ( Table 1 ). Genes encoding eight rRNA clusters were found in proximity to the origin of replication and 61 tRNAs were detected ( Table 1 ). Two putative prophage regions were found ( S1 File ). Genes for processes typical of clostridia, such as sporulation ( S2 File ), motility and chemotaxis, are present. Identification of sporulation-related genes is typically based on sequence homology to those of Bacillus subtilis– the model organism for studying the sporulation cycle [ 54 – 56 ]. The genome of C . phytofermentans contains a homolog of the master regulator of sporulation of B . subtilis , SpoOA (Cphy_2497, 55% amino acid identity to SpoOA of B . subtilis , Table A in S2 File ). Although the majority of the genes in the sporulation cascade of B . subtilis downstream of the master regulator SpoOA are present in C . phytofermentans , the ones upstream (the sensory histidine kinase and phosphorelay system) are not. C . phytofermentans is motile, moving by means of one or a few sub-terminal flagella [ 12 ]. Genes predicted to be involved in flagellar biosynthesis are found in two clusters (Cphy_0303–0316 and Cphy_2687–2720). Chemotaxis genes are found within one of these clusters (Cphy_2687–2691). C . phytofermentans also has three distinct genetic loci coding for proteinaceous bacterial microcompartments [ 57 , 58 ] ( S3 File ), two of which are predicted to be involved in choline, ethanolamine and 1,2-propanediol metabolism and one whose function cannot be inferred from sequence analysis. 10.1371/journal.pone.0118285.t001 Table 1 General features of the genome of C . phytofermentans . Parameter Value Size (bp) 4,847,594 G+C content (%) 35 Protein coding genes No. similar to known proteins (%) 2,870 (73.1) No. similar to proteins of unknown function \n a \n (%) 170 (4.3) No. of conserved hypotheticals \n b \n (%) 265 (6.7) No. of hypotheticals \n c \n (%) 621 (15.8) Total 3,926 Average ORF size (bp) 1,009 Coding (%) 81 No. of rRNA clusters 8 No. of tRNA genes 61 \n a Unknown function indicates significant sequence similarity to a named protein to which no specific function is currently attributed. \n b Conserved hypothetical proteins share significant sequence similarity to a translation of an open reading frame (ORF) in another organism for which no experimental evidence of protein expression is not available. \n c Hypothetical proteins with no significant similarity to any other sequenced gene. Genes encoding carbohydrate-active enzymes \n C . phytofermentans is capable of breaking down the recalcitrant, insoluble components of plant cell walls including cellulose, hemicellulose, pectin and starch [ 12 ] as well as switchgrass, corn stover and pulp wastes that have been minimally processed without thermo-chemical pretreatment ( Fig 2 ). Numerous carbohydrate-active enzymes (CAZy) predicted to be involved in the degradation of various plant cell wall components are encoded throughout the C . phytofermentans genome, including glycoside hydrolases (GH), polysaccharide lyases (PL) and carbohydrate esterases (CE). The diversity of GH families in the C . phytofermentans genome is unparalleled among sequenced clostridial genomes ( Fig 3 ). A total of 116 GHs distributed among 44 families are encoded in the genome of C . phytofermentans including but not limited to endo- and exo-cellulases, hemicellulases, chitinases, pectinases, amylases, and lichenases ( Fig 3 and S4 File ). Only the GH content of a distant relative in Cluster I, Clostridium cellulovorans , is comparable, with 113 GH domains distributed among 37 families ( Fig 3 ). A closer relative of C . phytofermentans , Butyrivibrio proteoclasticus ( Fig 3 ), has a comparable number of GH domains (113), but less diversity with only 25 families and no exo-cellulase (GH48). 10.1371/journal.pone.0118285.g002 Fig 2 Fermentation products on different growth substrates. \n (A) Fermentation products during growth on 2% (w/v) cellobiose. Data are an average of two samples; error bars represent range. (B) Ethanol produced on a variety of substrates expressed as the molar percentage of non-gaseous products. All substrates were present at a concentration of 1% (w/v) except where otherwise indicated. The particle size of insoluble substrates was reduced by grinding; the substrates were not otherwise pre-treated. Fermentation products were measured after obvious growth ceased (3–5 days) at 30°C. In most cases, substrate conversion was incomplete. 10.1371/journal.pone.0118285.g003 Fig 3 Comparative analysis of AraC transcriptional regulators, glycoside hydrolases (GH), and ABC transporters among selected sequenced clostridial genomes. \n (A) A conceptual illustration of how GH (blue), ABC transporters (purple) and AraC regulators (red) may work together. (B) Number of AraC transcriptional regulators per genome. (C) Number of GH domains per genome. Organisms having both GH48 and GH9 are marked with two asterisks, and organisms having GH9 alone are marked with one asterisk. (D) Number of putative ABC transporters per genome. To gain insight into the origin of the GHs of C . phytofermentans , we identified the closest relatives of the GHs of C . phytofermentans in the GenBank database using BLASTP and compared their distribution to that of the closest relatives of all of the protein-coding genes within the C . phytofermentans genome. The latter analysis was performed to calibrate how much similarity to other bacteria would be expected on average. In total, approximately 40% of the GHs of C . phytofermentans were most similar to GHs present in species outside the class Clostridia ( Fig 4 ), whereas only 20% of all the genes in the C . phytofermentans genome were most similar to genes from outside the Clostridia. The higher than expected proportion of GHs with distant relatives is statistically significant (Pearson's Chi-squared test, X-squared = 77.8583, df = 9, p-value = 4.299e-13) ( Fig 4 ). This result suggests that horizontal gene transfer from diverse origins rather than vertical divergence from an ancestral genome played a key role in the assembly of the unique set GHs present in C . phytofermentans . 10.1371/journal.pone.0118285.g004 Fig 4 Comparison of the distribution of the closest relatives of all C . phytofermentans open reading frames among sequenced bacterial genomes (left) to that of closest relatives of its glycoside hydrolases (right). In some bacteria, notably C . cellulolyticum and C . thermocellum , lignocellulose-degrading enzymes are attached to complex extracellular structures called cellulosomes that are believed to be critical for efficient plant cell wall breakdown. However, there is no genomic evidence for the production of cellulosomes by C . phytofermentans ( S4 File ). In fact, two critical cellulases of C . phytofermentans , the GH9 family endocellulase and GH48 family exocellulase are more similar to the soluble cellulases of C . thermocellum than to cellulosomal cellulases [ 15 ]. The majority of the GHs of C . phytofermentans are multimodular. Carbohydrate-binding modules (CBMs) are found within 17% of the GHs of C . phytofermentans , including the critical endocellulase (GH9) [ 15 ]. In the absence of a cellulosome, these CBM domains may enable GHs to adhere to plant cell wall substrates, facilitating degradation of the heterogeneous, highly cross-linked lignocellulose polysaccharides. Among the 31 GH enzymes predicted to be extracellular, 16 contain domains involved in anchoring proteins to the cell surface, including transmembrane helices and/or cell-wall binding domains, suggesting that these enzymes are cell-associated. Despite the absence of cellulosomal assembly domains, the striking multimodular nature of cellulosomal proteins, in which multiple domains from diverse families of GH, CE, PL and carbohydrate-binding modules (CBM) are found within individual proteins, is preserved in C . phytofermentans (Table A in S4 File ). C . phytofermentans has 19 multimodular GH proteins, representing about 17% of all putative GH genes (Table A in S4 File ). In fact, the largest protein in the proteome is the multimodular glycoside hydrolase family 10 protein Cphy_3862, with 2457 amino acids and a predicted molecular weight of 266 kD [ 16 ]. This protein contains consecutive GH10, CE15, and CBM domains. In non-cellulolytic bacteria, the corresponding GH domains are found mainly in single-domain polypeptides, which are cytosolic and act on smaller, soluble carbohydrate substrates [ 59 ]. Thus, the multi-modular organization that seems to be characteristic of enzymes from cellulolytic species, may reflect their involvement in the extracellular processing of heterogeneous insoluble substrates, such as plant cell walls [ 59 ]. Biofilm formation may also play an important role in the orchestration of the degradation of the plant cell wall polysaccharides. Cells might adhere to each other via a variety of different domains such as pfam07705 (CARDB, cell adhesion domain in bacteria) and pfam01391 (Collagen, Collagen triple helix repeat), both of which are found in the C . phytofermentans genome. Genes potentially involved in carbohydrate transport Further examination of the genome revealed 148 genes encoding subunits of ATP-binding cassette (ABC) transporters, more than found in other clostridia ( Fig 3 ). These genes are typically organized in operons consisting of two permeases and one solute-binding component. The majority of the ABC transporter-encoding operons lack an ATPase, suggesting that these transporter complexes may interact with a multitasking ATPase. Cphy_3611 is similar to MsmX of Bacillus subtilis , which is proposed to be an ATPase for several oligosaccharide transporters [ 60 ]. These findings suggest that C . phytofermentans is capable of active uptake of a diverse array of metabolites, including multiple oligosaccharides and simple sugars. The presence of GH genes adjacent to 50% of the transporter loci, suggests that carbohydrate degradation and uptake are frequently coupled. C . phytofermentans may feed cytoplasmic oligosaccharides into glycolysis via cellobiose/cellodextrin phosphorylases as occurs in other cellulolytic bacteria [ 8 ]. Import of oligosaccharides followed by internal hydrolysis via phosphorolysis minimizes ATP consumption [ 3 ]. Genes potentially involved in the regulation of carbohydrate metabolism To orchestrate the regulation of diverse metabolic pathways in response to changing growth substrates, C . phytofermentans has numerous transcriptional regulators, including 70 AraC ( Fig 3 ) and 23 PurR family members. AraC regulators typically activate transcription of genes involved in carbon metabolism, stress responses and pathogenesis [ 61 ], whereas PurR regulators act as repressors [ 62 ]. The abundance of these regulators suggests a complex regulatory network allowing rapid adaptation to varying substrate availability. Among the ABC-transporter genes found clustered with GHs, 50% are adjacent to AraC and 25% to PurR regulator genes. Analysis of gene expression during growth on a variety of simple and complex carbohydrates We designed a custom Affymetrix GeneChip to identify genes expressed in C . phytofermentans during growth on monosaccharides that are common in plant cell walls (glucose, galactose, xylose, arabinose, mannose), purified polysaccharides (cellobiose, cellulose, xylan and pectin) and fibrous plant biomass ( Brachypodium distachyon ) ( S4 File and S6 File). These microarray studies suggest that C . phytofermentans regulates the stoichiometry of the plant degradative and assimilatory machinery in response to growth substrate. When C . phytofermentans was cultured with glucose, genes involved in biomass degradation (e.g. cellulase and xylanase) were essentially off ( Fig 5 ), and the most abundant transcript was a putative ABC monosaccharide transporter ( S6 File ). During growth on xylose and xylan, transcripts for enzymes involved in pentose interconversion (xylose isomerase (Cphy_0200, and Cphy_1219) and xylulokinase (Cphy_3419) were among the most highly expressed ( S6 File ). When C . phytofermentans was grown with cellulose as substrate, the GH9 cellulase gene was among the most abundant transcripts ( Fig 5 , cellulase_Cphy_3368). This cellulase gene has been shown by gene inactivation to be essential for growth on cellulose in C . phytofermentans [ 15 ]. On nearly all substrates tested, we observed specific sets of co-regulated groups of genes, often consisting of GHs, an ABC transporter and a transcriptional regulator (Tables C and D in S4 File and S5 File). The putative multitasking ABC transporter ATPase subunit Cphy_3611, was expressed during growth on all substrates (transcript abundance within the 50 th percentile) ( S5 File ). To orchestrate the regulation of these genes, a number of transcriptional regulators, typically physically close to the transporters and CAZy, are highly expressed on a given substrate (Tables C and D in S4 File ). Thus, microarray experiments facilitated identification of enzymes involved in the breakdown and transport of specific carbohydrates. In addition, expression profiling with defined substrates was useful for deciphering data from more complex fibrous substrates. When plant biomass was used as growth substrates, GH expression profiles were similar to each other and to profiles with cellulose as substrate, with the exception that a putative xylanase (Cphy_2105) and mannanase (Cphy_1071) were more highly expressed on the plant biomass than on cellulose or xylan ( S4 File ). Thus, gene expression analysis proved to be a useful strategy for deciphering the functions of diverse enzymes involved in lignocellulose degradation ( Fig 5 , S4 File). 10.1371/journal.pone.0118285.g005 Fig 5 Illustration of the variation in transcription level of selected genes on various substrates. Transcript rank abundance curves during growth on (A) glucose, (B) hemicellulose, (C) cellulose and (D) Brachypodium. ADH_Cphy_1029 refers to a putative alcohol dehydrogenase. Cellulase_Cphy_3368 denotes the putative cellulose. Xylanase_Cphy_2105 denotes the putative xylanase. Genomics and transcriptomics investigation of C . phytofermentans central metabolism Perhaps the most industrially relevant property of C . phytofermentans is that it produces ethanol as the major fermentation product during growth on a wide variety of substrates including, simple sugars, cellulose and minimally processed plant biomass [ 12 ] ( Fig 2 ). The fact that C . phytofermentans produces predominantly ethanol, suggests that it can maintain its redox balance without forming equivalent levels of lactate and/or formate and that it can generate sufficient energy for growth in the absence of high levels of acetate synthesis, which yields ATP via substrate-level phosphorylation. To gain insight into the basis for high levels of ethanol production, we used a combination of transcriptional profiling and comparative genomic analysis to identify a subset of genes that were both highly expressed on all growth substrates and predicted to be involved in ethanol production, energy conservation, and/or redox balance. The results of this analysis are the basis of a simplified model of the core physiology of C . phytofermentans ( Fig 6 , S7 File ) and indicate that high levels of ethanol production may be due to a combination of factors. Firstly, pyruvate appears to be funneled to ethanol. The levels of the transcripts of the enzymes within the ethanol biosynthesis pathway are extremely high, and exceed those of all enzymes involved in the synthesis of alternate carbon fermentation products ( Fig 5 , Table A in S7 File ). In particular, two alcohol dehydrogenases (ADH), Cphy_3925 and Cphy_1029, were constitutively transcribed at levels that rivaled or exceeded those of many ribosomal protein genes (average transcript abundances within the 98th percentile, Fig 5 and Table A in S7 File ). Examination of these genes revealed that both NADH and NADPH are likely to contribute to ethanol production, another factor that may increase ethanol production. Secondly, reduced ferredoxin generated during conversion of pyruvate to the ethanol precursor, acetyl-CoA, by pyruvate ferredoxin oxidoreductase may contribute to ethanol production both directly, through reduction of NAD and NADP and indirectly, by participating in energy conservation. Two constitutively highly expressed protein complexes are likely to play a role in enabling reduced ferredoxin to contribute to ethanol production: NfnAB, an NADH-dependent reduced ferredoxin:NADP oxidoreductase [ 63 ] and Rnf, a sodium-translocating NADH:ferredoxin oxidoreductase [ 64 ]. C . phytofermentans may be able to exploit the sodium gradient produced by Rnf for energy production by way of the highly expressed sodium-translocating F 1 F o -ATPase (Cphy_3735–42). Thus, Rnf may contribute to favorable energetics of ethanol production and reduce dependence on ATP generation via acetate production. Hydrogenases may also play an important role in ferredoxin metabolism in C . phytofermentans . In other clostridia, hydrogenases dissipate excess ferredoxin-reducing equivalents [ 65 , 66 ]. C . phytofermentans generates free hydrogen as a product of the fermentation of cellulose and cellobiose [ 67 ], and three cytoplasmic [FeFe]-hydrogenase-encoding clusters, one encoding a putative ferredoxin-dependent hydrogenase and two encoding NAD-dependent hydrogenases were constitutively highly expressed. The simultaneous expression of a ferredoxin-oxidizing hydrogenase with NAD-dependent hydrogenases, which could catalyze hydrogen-dependent NAD reduction and feed ferredoxin reducing equivalents into ethanol production, may prevent excess hydrogen accumulation thus enabling C . phytofermentans to maintain a high rate of ferredoxin turnover. Finally, C . phytofermentans may further reduce the requirement for acetate production by utilizing pyrophosphate–dependent glycolytic enzymes, which can substantially increase the ATP yield of glycolysis. 10.1371/journal.pone.0118285.g006 Fig 6 Model of C . phytofermentans central metabolism including proposed pathways involved in high ethanol yield."
} | 7,033 |
35056506 | PMC8779289 | pmc | 8,481 | {
"abstract": "The success of mine site restoration programs in arid and semi-arid areas poses a significant challenge and requires the use of high-quality seedlings capable of tolerating heavy metal stresses. The effect of ectomycorrhizal fungi on different physiological traits was investigated in Pinus halepensis seedlings grown in soil contaminated with heavy metals (Pb-Zn-Cd). Ectomycorrhizal (M) and non-ectomycorrhizal (NM) seedlings were subjected to heavy metals stress (C: contaminated, NC: control or non-contaminated) soils conditions for 12 months. Gas exchange, chlorophyll fluorescence, water relations parameters derived from pressure–volume curves and electrolyte leakage were evaluated at 4, 8 and 12 months. Ectomycorrhizal symbiosis promoted stronger resistance to heavy metals and improved gas exchange parameters and water-use efficiency compared to the non-ectomycorrhizal seedlings. The decrease in leaf osmotic potentials (Ψ π 100 : osmotic potential at saturation and Ψ π 0 : osmotic potential with loss of turgor) was higher for M-C seedling than NM-C ones, indicating that the ectomycorrhizal symbiosis promotes cellular osmotic adjustment and protects leaf membrane cell against leakage induced by Pb, Zn and Cd. Our results suggest that the use of ectomycorrhizal symbiosis is among the promising practices to improve the morphophysiological quality of seedlings produced in forest nurseries, their performance and their tolerance to multi-heavy metal stresses.",
"conclusion": "5. Conclusion and Research Needs The present study emphasizes the importance of ectomycorrhizae in enhancing physiological processes in P. halepensis seedlings that are subjected to heavy metal contaminated soil, and supports the use of a mycorrhizoremediation approach in reforestation and rehabilitation of heavy metal contaminated sites. Our results, combined with our previous results [ 47 ], suggest that the use of ectomycorrhizal fungi is among the promising practices that would improve the morphophysiological quality of tree seedlings produced in forest nurseries, their performance and their tolerance to multiple heavy metal stresses (Pb, Zn and Cd). Therefore, further study is warranted regarding the combined effects of heavy metals and ectomycorrhizae on the mineral status of the seedlings using vector analysis of foliar nutrients and biomass.",
"introduction": "1. Introduction Mining activities that are conducted in a Mediterranean climate can exert their greatest effects on the environment through water pollution, contamination and alteration of agricultural soil due to the spread of heavy metals [ 1 , 2 ]. In North Africa, land losses due to increasingly heavy metal-polluted soils are the highest in the world [ 1 ]. Therefore, rehabilitation strategies targeting mine-degraded areas have been launched to combat land degradation and have been recognized as one of several sustainable development goals [ 1 ]. However, phytoremediation of abandoned mine lands by installing forest plantation presents a great challenge [ 3 ]. Indeed, severe environmental stresses and their interactions can negatively affect the survival of forest trees [ 4 , 5 , 6 ], the sustainability of ecosystems and the success of reforestation programs on abandoned mine sites in the context of climate change. North Africa (Morocco, Algeria and Tunisia) is recognized as one of the Mediterranean region’s most vulnerable to climate change, given that it is characterized by a significant decrease in precipitation and a significant increase in temperature, which reinforce both the pressures and the phenomena that are associated with ecosystem degradation [ 5 , 7 ]. In North Africa, various efforts have been made to modernize the seedling production chain and forest nurseries that produce seedlings of high morphophysiological quality, which are capable of surviving, growing and tolerating various environmental stresses in reforestation sites [ 8 , 9 , 10 ]. The results of this modernization project have shown that growth of ectomycorrhizal seedlings, which are produced in containers in modern forest nurseries, was much higher than that of seedlings that were produced in polybags in traditional nurseries [ 9 ]. Yet, these restoration projects did not focus upon the rehabilitation of mining sites using adapted species and ectomycorrhizal seedlings that were produced in modern forest nurseries. The development of mycorrhizoremediation technology involving the use ectomycorrhizal seedlings is seen as a way for enhancing their abilities to tolerate various multi-metal stresses [ 11 , 12 , 13 ]. Other studies have reported the efficiency of ectomycorrhizal seedlings in overcoming the detrimental effects of abiotic stresses, such as drought [ 14 , 15 ] and salinity [ 16 , 17 ]. In contrast, numerous studies have demonstrated the negative effects of heavy metal contamination on the survival and growth of a wide range of non-mycorrhizal plant species [ 18 , 19 , 20 , 21 , 22 , 23 ]. On one hand, an excess of heavy metals in soils limits the efficiency of water- and mineral nutrient-use [ 18 , 19 , 24 ]. In addition, heavy metals substantially reduce leaf gas exchange parameters (net photosynthesis, transpiration, stomatal conductance, etc.), thereby causing lower water flow from the soil to the leaves [ 18 , 25 ], which may cause water stress in seedlings [ 24 ]. On the other hand, CO 2 assimilation and leaf transpiration are decreased due to reduced stomatal opening [ 26 , 27 ], which may be induced by direct interaction of metal ion toxicity with guard cells [ 19 ]. Furthermore, long-term exposure to high levels of toxic heavy metals is often followed by water deficits [ 19 , 28 ]. This leads to the appearance of physiological responses consistent with those found under drought stress, including reductions in root water uptake, leaf turgor and stomatal conductance [ 19 , 29 , 30 ]. Heavy metals, such as Pb, Zn and Cd, reduce cell membrane permeability [ 18 , 31 ], chlorophyll concentrations [ 32 , 33 ] and photosystem II activity [ 25 ], which in turn can limit photosynthesis leading to metabolic disruptions [ 27 , 34 , 35 ]. To improve the survival, growth and physiological processes of tree seedlings that are intended for reforestation and mine site restoration programs, the use of seedlings with high morphophysiological quality that are produced in modern forest nurseries is necessary [ 8 , 9 , 36 ]. In addition to the choice of local forest species that are already adapted to the interactions of different environmental stresses, improving the root system using compatible host-ectomycorrhizal fungi that are resistant to environmental stresses would improve the survival, growth and physiology of the seedlings after their installation in mining and reforestation sites. Under drought conditions, as is the case in arid and semi-arid areas of North Africa, it was shown that ectomycorrhizal seedlings with fungal genotypes that produce mycelial strands or rhizomorphs capable of transporting large amounts of water [ 37 ] can substantially improve physiological processes (gas exchange and water relations parameters, mineral nutrition and hydraulic conductivity of the roots, among others). Consequently, the drought tolerance of the seedlings is improved [ 9 , 14 , 38 , 39 , 40 ]. In contrast, it is recognized that in the presence of heavy metals, ectomycorrhizal fungi improve survival, growth and various physiological processes (transpiration, gas exchange parameters: stomatal conductance, transpiration, net photosynthesis and osmotic adjustment, among others) of forest seedlings under controlled conditions and on mining sites [ 11 , 41 , 42 ]. Despite the publication of several reviews on the water relations of different species in response to heavy metal stresses [ 18 , 19 , 24 , 43 ], no information has been made available, to our knowledge, regarding ectomycorrhizal effects on water relations parameters that are derived from pressure–volume curves [ 44 , 45 , 46 ] in response to heavy metal stresses (Pb, Zn and Cd). These water relations variables for ectomycorrhizal tree seedlings, include osmotic potential at saturation (Ψ π 100 ), osmotic potential with loss of turgor (Ψ π 0 ), relative water content at loss of turgor (RWC 0 ), water content of the symplasm (SWC), the modulus of elasticity (ε max ) and osmotic adjustment (OA). The determination of these variables that are specific to water relations is fundamental to better quantifying and understanding the effects of ectomycorrhizal fungi on the physiology of cells and tissues of forest seedlings, which allow them to survive, grow and maintain their physiological functions (net photosynthesis, transpiration, etc.) in response to multi-heavy metal stresses. In addition, the results of this study would help advance operational practices, thereby further improving the morphophysiological quality of seedlings that are produced in forest nurseries, together with their performance in mining sites that are located in arid and semi-arid regions. This research continues our recent work [ 47 ], which showed that ectomycorrhizal fungi ( Rhizopogon sp.) improved growth and mineral nutrient contents of Pinus halepensis mill seedlings that were subjected to heavy metal stresses. The presence of ectomycorrhizal fungi also reduced translocation factors for Zn and Cd, and bioaccumulation factors for Pb and Cd. This study was designed to test the hypothesis that ectomycorrhizal fungi can improve the water relations and the gas exchange parameters of P. halepensis seedlings grown in response to multi-heavy metal stresses. The objectives of this study were: (i) to compare gas exchange variables, water use-efficiency (WUE), electrolyte leakage and chlorophyll fluorescence under multiple heavy metal stresses (Pb, Zn and Cd) in ectomycorrhizal and non-ectomycorrhizal Pinus halepensis seedlings; (ii) to determine water relations parameters that were derived from pressure–volume curves (Ψ π 100 , Ψ π 0 , RWC 0 , SWC and ε max ) of ectomycorrhizal and non-ectomycorrhizal seedlings that were grown in the absence and presence of multi-metal stresses; and (iii) to examine whether osmotic adjustment occurs as a result of long-term exposure of ectomycorrhizal seedlings to high levels of heavy metal toxicity. This evaluation of gas exchange and water relations parameters will help to understand the physiological processes relevant to the performance of plants in response to multi-heavy metal stresses.",
"discussion": "4. Discussion The use of ectomycorrhizal P. halepensis seedlings that were infected with Rhizopogon sp. in the presence of soil contaminated with heavy metals significantly improved gas exchange and water relations ( Figure 2 , Figure 3 and Figure 4 ). After 12 months of growth, the mycorrhizae also conferred on Aleppo pine seedlings an osmotic adjustment and a high elasticity allowing the plants to survive, grow and maintain various physiological processes ( Figure 2 , Figure 3 , Figure 4 and Figure 5 ) despite the presence of extremely high concentrations of heavy metals in the soil (Pb, Zn and Cd). Our previous results [ 47 ] showed that after 12 months of growth in contaminated soil (NM-C) with heavy metals (Pb, Zn, and Cd), shoot and root dry masses of P. halepensis seedlings were reduced compared to the control (NM-NC), while no differences were observed for M-C compared to the control (NM-NC). Prolonged exposure to heavy metals primarily affects seedling growth and significantly decreases nutrient uptake, water-use efficiency, photosynthetic activity and cell membrane integrity [ 19 , 22 , 25 , 29 ]. In contrast, other studies showed that mycorrhizal fungi alleviate metal toxicity by improving physiological mechanisms and adaptation of host plants [ 23 , 61 ]. This observation is consistent with the current research indicating a significant increase in photosynthetic rate among ectomycorrhizal seedlings (M-C) to a level similar to seedlings under controlled conditions (NM-NC) ( Figure 2 ). The increase in M-C was two times greater than non-ectomycorrhizal seedlings (NM-C) after 8 and 12 months of exposure to heavy metal contaminated soil ( Figure 2 ). Other results have demonstrated an adverse effect of heavy metals on gas exchange parameters [ 25 , 62 , 63 ]. This was evaluated through analysis of maximum quantum yield of PSII photochemistry (F v /F m ), which is considered to be a critical parameter of plant photosynthetic performance and a useful tool for evaluating plant tolerance to heavy metal toxicity. Our results revealed a significant F v /F m reduction under Pb, Zn and Cd toxicity, particularly in non-ectomycorrhizal seedlings (NM-C), while M-C seedlings always exhibited higher values ( Figure 3 ). For non-ectomycorrhizal seedlings, the decline in F v /F m ratio suggests that photodamage and photoinhibition occurred to PSII [ 64 , 65 ]. This effect significantly decreases net photosynthesis as was observed in non-ectomycorrhizal seedlings ( Figure 2 ). Under controlled conditions, net photosynthesis was positively correlated with internal CO 2 concentration [ 66 ]. Yet, an increase in internal CO 2 (C i ) concentrations may be associated with low net uptake of CO 2 . Therefore, photosynthetic limitation was posed by internal conductance to CO 2 movement in seedlings that were exposed to contaminated soil (NM-C and M-C) after 8 months ( Figure 2 ). It has been shown that Cd stress is able to increase CO 2 concentrations in the intercellular spaces of the mesophyll [ 67 ]. High internal CO 2 levels can be explained by the decreased capacity of chloroplasts to assimilate CO 2 [ 68 ]. In contrast, ectomycorrhizae enabled the seedlings (M-C) to withstand metal toxicity and to maintain C i concentrations that were similar to the control after 12 months of growth ( Figure 2 ). Ectomycorrhizae also increased transpiration (E) rates in seedlings (M-C) after 8 and 12 months of exposure to excess of Pb, Zn and Cd ( Figure 2 ). Similarly, Han et al. [ 42 ] demonstrated an increase in photosynthetic and transpiration rates in ectomycorrhizal hybrid poplar ( Populus alba × tremula var. glandulosa ) seedlings that had been subjected to high concentrations of Cd. Mycorrhizal plants often exhibit higher stomatal conductance (g s ) than non-mycorrhizal ones [ 69 ]. Our results showed that stomatal conductance of ectomycorrhizal seedlings (M-C) is not different to the control (NM-NC), but it is always significantly greater than the non-ectomycorrhizal seedlings (NM-C) throughout the growing period (12 months) ( Figure 2 ). Heavy metals negatively affect absorption of mineral nutrients due, in part, to stomatal closure, which leads to the decrease in CO 2 uptake and photosynthesis [ 33 , 70 , 71 , 72 ]. Other major positive effects of ectomycorrhizae on water relations and gas exchange parameters are associated with (i) increasing soil root exploration by extending the extraradical phase (hyphae and mycelial strands) of ectomycorrhizal fungi that are capable of penetrating soil micropores, which are not accessible to the roots, to extract water [ 37 , 39 , 40 , 73 , 74 , 75 , 76 ]. (ii) The formation of mycelial strands or rhizomorphs by certain ectomycorrhizal fungi, as is the case in our study ( Rhizopogon sp.), which are capable of transporting large quantities of water [ 37 , 74 ], and the extraradical mycelium of ectomycorrhizal fungi, by itself (without the roots), make it possible to satisfy the moisture requirements of the seedling in order to maintain photosynthesis and transpiration [ 14 , 38 ]. (iii) Furthermore, water stress is reduced at the soil-root interface [ 39 ], together with (iv) improvement of soil structure and soil water reserves depending upon the density and expanse of the fungal hyphae in the soil [ 75 , 77 , 78 ]. Our previous results have shown that ectomycorrhizae significantly increase the dry root masses of P. halepensis seedlings [ 47 ]. The development of root systems in ectomycorrhizal plants was linked to higher rates of CO 2 uptake and greater WUE in stone pine ( Pinus pinea L.) [ 79 ]. It should be noted that ectomycorrhizal seedlings (M-C) had the most rapid and efficient recovery from the harmful effects of heavy metals compared to the non-ectomycorrhizal ones (NM-C) ( Figure 2 and Figure 3 ). The recovery time was associated to the rapid growth of plants, which causes a reduction in internal metal concentrations due to a dilution effect [ 28 ]. Our previous results showed that ectomycorrhizal P. halepensis seedlings exhibit higher growth and lower Pb, Zn and Cd uptake than non-ectomycorrhizal seedlings [ 47 ], which demonstrates the substantial benefits of ectomycorrhizal fungi to their hosts under multi-metal stress. Stomatal closure disrupts the flow of sap and, consequently, the water relations of the seedlings [ 80 ]. Therefore, the importance is demonstrated for osmotic adjustment (OA) as an adaptation mechanism [ 81 ]. This is consistent with our findings showing a substantial increase in OA for ectomycorrhizal (M-C) seedlings that was two-fold higher than non-ectomycorrhizal ones (NM-C) at the end of the experiment ( Figure 4 f). The increase in osmotic adjustment may be associated with decreases in osmotic potentials (Ψ π 100 and Ψ π 0 ) ( Figure 4 a,b), suggesting that mycorrhizae can give plants increased resistance to various heavy metals. A greater decline in Ψ π 100 and Ψ π 0 for ectomycorrhizal seedlings (M-C), together with greater osmotic adjustment ( Figure 4 ), may grant them the ability to absorb water from contaminated soil, even when the former is hardly available due to effects of contamination [ 18 ]. As a result, M-C seedlings can maintain their turgor for a long time before reaching the loss of turgor point [ 46 ], which is consistent with lower values of RWC 0 and increased elasticity of cell membranes ( Figure 4 ). This allows gas exchange (stomata conductance and net photosynthesis) to be maintained for longer periods [ 52 ], as was observed in this study. The positive effects of these significant improvements in physiological processes were reflected on several growth parameters [ 47 ]. In contrast, Muhsin and Zwiazek [ 82 ] revealed that ectomycorrhizae increase apoplastic water transport and root hydraulic conductivity in American elm ( Ulmus americana L.) seedlings, and suggested that this is related to decreased resistance to the flow of water from the apoplast through ectomycorrhizal hyphae. The increased root water flow in mycorrhizal seedlings may be related to the nutritional and metabolic effects of mycorrhizae on the activity of root water channels [ 82 ]. It has been demonstrated that ectomycorrhizal fungi improve water transport and water relations of seedlings [ 37 , 38 , 39 ] through the improvement in the mineral nutrition of plants. Phosphorus and potassium, especially, are two key elements that are involved in active adjustments [ 73 ]. Several ectomycorrhizal fungi improve the osmotic adjustment of their host seedling’s cells. This improvement came from the synthesis of organic acids and accumulation of mineral nutrients [ 83 , 84 ]. For example, high concentrations and contents of nitrogen, phosphorus and calcium increase the rate of net photosynthesis [ 48 , 76 , 85 ], which could increase water-use efficiency compared to non-ectomycorrhizal ones (NM-C), as was noted in the present study ( Figure 3 ). Indeed, our recent findings revealed that ectomycorrhizal association of P. halepensis seedlings improved mineral nutrition, particularly in terms of nitrogen and calcium [ 47 ]. This last mineral nutrient (Ca) stimulates photosynthesis, cell division, cell wall rigidity and absorption of major nutrients (N, P and K) [ 86 , 87 , 88 ]. Furthermore, osmotic adjustments are known to prevent oxidative damage [ 26 , 27 , 89 ]. This effect becomes more important in the presence of mycorrhizae [ 90 ]. This may explain the low electrolyte leakage rates that were sustained by ectomycorrhizal contaminated (M-C) seedlings ( Figure 5 b,c). Reduction in electrolyte leakage may be associated with the cell conservation of water absorption and transport structures within the seedlings [ 27 ]. Membrane stability of ectomycorrhizal seedlings may be further linked to the decrease in the modulus of elasticity (ε max ) ( Figure 4 e), which provides more flexibility and stability to the membranes [ 46 ]. This flexibility provides the plant with the possibility of undergoing significant variation in the water content of the apoplast without affecting the dynamic structure of the cell walls [ 91 ]. This is confirmed by the significant decrease in the symplastic water content (SWC) in ectomycorrhizal (M-C) seedlings, compared to the control (NM-NC) and the non-ectomycorrhizal ones (NM-C) ( Figure 4 c). As such, the influence of metal contamination on membrane stability was not associated with heavy metal exposure time, unlike the other variables ( Table 1 ). This response is probably due to the protection that is offered by the ectomycorrhizal symbiosis, which confers greater resistance to stress, thereby lowering the occurrence of stress symptoms, such as membrane stability maintenance and decreased osmo-protectant production [ 92 ]."
} | 5,340 |
37141097 | PMC10242440 | pmc | 8,483 | {
"abstract": "Summary Mouse models are key tools for investigating host-microbiome interactions. However, shotgun metagenomics can only profile a limited fraction of the mouse gut microbiome. Here, we employ a metagenomic profiling method, MetaPhlAn 4, which exploits a large catalog of metagenome-assembled genomes (including 22,718 metagenome-assembled genomes from mice) to improve the profiling of the mouse gut microbiome. We combine 622 samples from eight public datasets and an additional cohort of 97 mouse microbiomes, and we assess the potential of MetaPhlAn 4 to better identify diet-related changes in the host microbiome using a meta-analysis approach. We find multiple, strong, and reproducible diet-related microbial biomarkers, largely increasing those identifiable by other available methods relying only on reference information. The strongest drivers of the diet-induced changes are uncharacterized and previously undetected taxa, confirming the importance of adopting metagenomic methods integrating metagenomic assemblies for comprehensive profiling.",
"introduction": "Introduction Evolutionary, anatomical, and physiological proximity to humans make the mouse a successful model organism for biomedical research. Ease of breeding, validated disease models, and fast proliferation, as well as the possibility to perform multi-generation experiments and diet-related interventions, established mice as the main preclinical model for the study of the human gut microbiome. 1 , 2 , 3 , 4 , 5 , 6 In mice, microbiome experiments can be conducted while controlling for several variables such as genetics, nutritional or pharmacological exposures, and other experiment-confounder factors. 1 , 2 , 3 , 5 , 6 , 7 , 8 , 9 , 10 , 11 , 12 , 13 , 14 , 15 , 16 , 17 , 18 , 19 However, because the composition of the microbiome of laboratory and wild mice is different from that of humans, 3 , 4 the mouse microbiome structure and diversity is far from being comprehensively addressed, with consequent limitations for fundamental and translational research in mice. Analyses of the microbiome features characterizing diet, disease, and phenotype-related changes in mice have been extensively performed using 16S rRNA gene amplicon sequencing, 5 , 14 , 15 , 20 , 21 , 22 , 23 which, despite the reduced costs, can be considered limited in its phylogenetic, taxonomic, and functional resolution. The high-resolution shotgun metagenomic approach, which is now the standard in human microbiome studies, 24 is still much less employed in mouse studies for the lack of reference genomes covering the majority of the members of the mouse microbiome. 25 Efforts at cataloging de novo the diversity of the mouse microbiome by systematic bioinformatic assembly of mice metagenomes have been undertaken, 25 , 26 , 27 , 28 , 29 but it remains challenging to efficiently exploit them for new studies and in support of reference-based taxonomic profiling. In this work, we show how the mouse gut microbiome and its links with nutritional patterns can be investigated accurately and comprehensively via shotgun metagenomics by leveraging a computational taxonomic profiling approach called MetaPhlAn 4, 30 which integrates massive assemblies in its database. MetaPhlAn 4 considers over 22,718 metagenome-assembled genomes (MAGs) retrieved from 1,906 mouse-derived fecal, cecal, and ileal metagenomes with the species-level genome bin (SGB) strategy. 31 By applying the approach on a large and heterogeneous catalog of mouse microbiomes, we show that accounting for metagenomically defined species is necessary in the context of nutritional studies in mouse models. We also found that the microbial species not detectable by mapping against genomes from isolates account for the greatest proportion of the diet-associated microbiome changes.",
"discussion": "Discussion Here, we taxonomically profiled the gut microbiome of laboratory mice at an enhanced resolution by the integration of a massive number of MAGs (22,718 mice-derived MAGs) in the reference database of the MetaPhlAn 4 marker-based approach. To investigate whether this increased resolution can lead to the discovery of relevant associations between the mouse microbiome and host conditions, we focused on the analysis of the microbiome links with diet. We collected, manually curated, and profiled a set of nine mouse microbiome datasets (one of which was made available by this work), all characterized by the presence of multiple diet regimes differing in the percentage of fat. Machine learning and meta-analyses on the cohorts profiled with the MAG-enhanced database revealed cross-cohort associations with diet that are stronger than what was previously available and that were mostly driven by uncharacterized microbial species (uSGBs). Our results highlight the need for inclusions of genomes from uncultured microorganisms in the process of taxonomic profiling of mice microbiome data and the key role that species available only through metagenomic analyses may play for host-microbiome interaction specifically in laboratory mice. Importantly, we showed that MetaPhlAn 4 is able to efficiently integrate uSGB profiling in the metagenomic analysis and thus largely improve the analysis of microbiome in mouse models. Limitations of the study Our study and tools can be the basis for more nuanced study of nutritional effects on the microbiome and host-microbiome interactions in preclinical models. Improved study designs could, for example, account for the differences in saturated vs. polyunsaturated fat intake. 52 In our study, we could not correct our analysis by the weight of the mouse at baseline or consider the polysaccharide nutritional content, 13 although these aspects were shown to be only minor confounders with respect to the diet-induced obesity development. 53 , 54 , 55 As the diet-microbiome-host links remain intricate and mouse models can be useful in studying them, it will also be crucial to extend the ability to profile uncharacterized aspects of the microbiome to microbial transcripts, metabolites, and proteins, and thus future work should be focused on integrative computational methods to profile microbiome mouse models with meta-omic approaches. 56"
} | 1,552 |
34168966 | PMC8209186 | pmc | 8,485 | {
"abstract": "Highlights • Widely applicable UV mutagenesis method to improve transgene expression in the microalga Chlamydomonas . • New CC-1690*** chassis strain with shear stress tolerance, mating-competence and high transgene expression capacity. • Proof-of-concept cadaverine production from CO 2 with CC-1690***.",
"conclusion": "4 Conclusion A new Chlamydomonas reinhardtii cell wall-containing strain derived from the robust wildtype CC-1690 was generated by combining UV mutagenesis and subsequent candidate selection based on an increased Zeocin antibiotic tolerance. This novel strain (UV strain) will be further on designated CC - 1690 *** (spoken CC-1690 triple star) because it possess three key traits for industrial production processes. Besides possessing a cell wall-mediated high shear force resistance, it efficiently expresses nuclear transgenes, which are both prerequistes for industrial bioproduction schemes. As am third trait, CC - 1690 *** can be crossed to other strains, thus enabling fast trait combination and outcrossing of antibiotic resistance cassettes.",
"introduction": "1 Introduction Green biotechnology could play an essential role in the reduction of CO 2 emissions. Within these concepts, metabolically-engineered microalgae could become renewable and environment-friendly sources of diverse carbon-based compounds [ 1 ] and fuels [ 2 ]. Among biotechnologically suitable microorganisms, microalgae are especially promising, since they are photoautotrophic and readily cultivateable in cheap water-based media, using sunlight and carbon dioxide [ 3 ]. Metabolic engineering and synthetic biology approaches can substantially expand the range of products obtainable from microalgae [ 4 ] and increase their photoautotrophic productivity [ 5 ]. Chlamydomonas reinhardtii represents a well-established model organism for basic research in the field of photosynthesis research [ 6 ] and also microalgal biotechnology, including metabolic engineering [ [7] , [8] , [9] ], since for many fast-growing microlagal species suitable for large-scale cultivation, a molecular toolbox is not available [ 10 ]. In 2009 Neupert et al. created Chlamydomonas cell lines capable of efficiently expressing transgenes, by applying UV mutagenesis and subsequent selection of high transgene expressors based on their improved resitance towards the antibiotic emetine [ 11 ]. Since then, the UVM4/11 expression strains have successfully been used for the high-level expression of various nuclear transgenes [ 7 , 9 , 12 , 13 ]. Although, these strains efficiently express transgenes, UVM4/11 cell lines have two main disadvantages complicating their application in biotechnology. Firstly, the lack of a cell wall renders these strains susceptible to shear stress and thus unsuitable for cultivation at a larger scale, which is normally accompanied by intense mixing based on stirring or gassing with high flow rates [ 14 ]. Secondly, their mating inability prevents crossing experiments as a fast and efficient way of removing antibiotic selection markers or adding further genetic traits. Recently, phototrophic production of the diamine cadaverine (1,5-pentanediamine) via decarboxylation of l -lysine, catalyzed by the enzyme l -Lysine decarboxylase ( CadA , EC 4.1.1.18), was reported for C. reinhardtii , using UVM4-derivative strains [ 9 ]. 1,5-pentanediamine is used as a building block for the synthesis of (bio-) polyamides, which have excellent material properties and find application as medicinal plastics, fibers for textiles, or films and coatings [ 15 , 16 ]. In order to transfer sustainable cadaverine production into a robust C. reinhardtii phototrophic production chassis, we applied UV-mutagenesis to the cell wall-containing and nitrate-assimilating wildtype CC-1690 [ 17 ]. Here, we demonstrate that an iteration of UV mutagenesis and selection of improved transgene expressors, based on their enhanced resistance towards the antibiotic Zeocin, can be used to equip robust C. reinhardtii wildtype strains with the ability to express transgenes at a high level, which is a prerequisite for their application in sustainable production schemes.",
"discussion": "3 Results and discussion 3.1 UV-mutagenesis and Zeocin selection yield in a Chlamydomonas strain with increased reporter expression To add the capablility of expressing transgenes at a high level to the robust C.reinhardtii wildtype strain CC-1690 [ 17 ], the following strategy was applied ( Fig. 1 ). First, strain CC-1690 was transformed with a Sh Ble containing plasmid [ 20 ] conferring resitance against the antibiotic Zeocin [ 27 , 28 ] and a transformant displaying only a low resistance level was isolated. This strain was transformed with a nuclear expression construct for the expression of a Clover green fluorescent protein reporter alone or a red fluorescent (mRuby2) reporter protein fused to the l -lsyine decarboxylase CadA. To obtain mutant strains expressing transgenes at a higher level by inactivating or circumventing the transgene silencing mechanisms inherent to Chlamydomonas [ [29] , [30] , [31] ], UV-mutagenesis was applied. UV mutagenized strains were then tested for an elevated resitance towards Zeocin. The Zeocin resistance level is proportional to the amount of expressed Zeocin-binding protein, since it forms a 1:1 complex with the antibiotic, which prevents DNA cleavage [ 32 ]. Zeocin resitance can be thus used as a proxy for the transgene expression capacity of UV mutants. UV mutants showing an elevated antibiotic resistance level were then also transformed with the fluorescent Clover reporter and transformants screened for high reporter expression levels. After obtaining a strain with retained robust phototrophic growth and high transgene expression capacity, this strain was transformed with a lysine decarboxylase (CadA)-mRuby2 reporter fusion construct as a proof-of-concept application for phototrophic cadaverine production. Fig. 1 Workflow chart depicting the different steps which led to the isolation of novel CC-1690 strains with desired properties. Created with BioRender.com. Fig. 1 To isolate the starting strain with a low Zeocin resitance level, 95 Zeocin-resitant transformants were subjected to growth assays on agar plates containing Zeocin concentrations in the range of 5−200 μg/mL of the antibiotic. Among all transformants, 11 showed little growth on 100 μg/mL Zeocin and could not grow at all on 150 μg/mL of the antibiotic (Fig. S1). Among them, strain 40p showed the lowest resitance level and was selected as the starting strain. To this starting strain, which is further on designated “low resistance strain”, we applied UV-mutagenesis to obtain mutants with an altered transgene expression capacity. Among the isolated UV mutants seven could survive on up to 1200 μg/mL Zeocin and we focussed on one of these mutants (Fig. S2; 7a; CC-1690*** ), designated “CC-1690***” in further experiments. Along with UVM4, CC-1690*** was subjected to growth analyses under phototrophic conditions in a flat panel bioreactor ( Fig. 2 A) with vigorous bubbling (gas flow rate of 2.5 vvm, 1.7 % (v/v) CO 2 ) to test the shear force resitance of both strains. The final biomass yield noted for the cell wall-containing strain CC-1690*** was about two-fold higher compared to the cell-wall deficient [ 11 ] UVM4 strain (1.46 ± 0.03 g L −1 for CC1690*** vs. 0.70 ± 0.07 g L −1 for UVM4 at day 5). Fig. 2 UV mutagenesis and Zeocin selection yields in a strain with improved reporter expression. A ) Analysis of phototrophic biomass accumulation for strain CC-1690*** and UVM4 grown in 400 mL flat panel photobioreactors with HS medium. The mean value and standard deviations (error bars) of two biological replicates is shown. B ) Scheme showing the tandem DNA construct, which was used for transformation of CC-1690 derivative strains (low resistance strain and UV strain). Nuclear expression of the fluorescent Clover gene and the paromomycin resitance cassette (AphVIII – ParoR) are driven by HSP70A and RBCS2 promoters (red and black arrows, respectively), RBCS2 3‘UTR and PSAD 5` UTR regulatory elements. RBCS2 intron 1 (yellow) was incorporated into the RBCS2 promoter region and RBCS2 intron 2 (orange) was incorporated into the Clover fluorescence gene. Strep-tag II is a synthetic peptide / epitope tag for immunodetection. C ) Fluorescence in Clover-expressing transformants derived from either the low resistance strain (white box) or the UV-mutagenized low resistance strain ( CC-1690*** ; dark grey box). For analysis, 25 out of 133 mutants from low resistance strain and 25 out of 294 mutants from UV strain have been used. Fluorescence was normalized to chlorophyll absorption (OD 680nm ) and is given relative to wild type values (background fluorescence set to 1). An asterisk indicates a P value lower than 0.05 according to Student’s t -test. Fig. 2 It is well-known, that cell wall-reduced C. reinhardtii strains cannot resist the high shear forces resulting from vigorous mixing in bioreactors [ 33 ]. The much better performance of CC-1690*** compared to UVM4 should therefore result from the fact that CC1690*** contains a wildtype-like robust cell wall, whereas UVM4 does not. CC-1690*** and the low resistance strain were transformed with a construct for the expression of a Clover green fluorescent protein ( Fig. 2 B). For each transformation 25 distinct transformants were analysed in regard to their Clover expression based on fluorescence emission ( Figs. 2 C and S3). The Clover fluorescence emitted by UV strain-derived transformants was significantly ( P < 0.05) higher than in those derived from the low resitance strain. 3.2 The novel UV strain displays a UVM4-like capacity to express nuclear transgenes The UV strain, its progenitor the low resitance strain and UVM4 were then transformed with a nuclear expression construct ( Fig. 3 A) encoding a fusion of the lysin decarboxylase CadA [ 9 , 15 , 16 ] and the fluorescent reporter mRuby2 [ 34 ]. Fig. 3 Expression of a mRuby2-CadA fusion protein in the UV strain. A ) The DNA construct, which was used for transformation in CC-1690 derivative strains (wild type, low resistance strain, CC-1690*** ) and UVM4. The hygromycin resistance gene (AphVII – HygroR) was placed under control of a PSAD promoter and FDX 3‘UTR. The lysine decarboxylase gene was regulated by HSP70A , RBCS2 and FDX1 3‘UTR regulatory elements and the open reading frame of CadA fused to the ORF of the mRuby 2 fluorescence reporter. Yellow vertical bars indicate RBCS2 intron 1, which was inserted into the RBCS2 promoter region as well as lysine decarboxylase and mRuby2 genes. HA indicates the human influenza hemagglutinin (HA) epitope tag. B ) Distribution of mRuby2 fluorescence in strains CC-1690 before and after UV mutagenesis in comparison to UVM4. All strains were transformed with the plasmid shown in panel A). For analysis, 12 out of 600 mutants from low resistance strain, 11 out of 200 mutants from CC-1690*** strain and 11 out of 288 mutants from UVM4 were used. mRuby2 fluorescence in CC-1690 lines was normalized first to OD680 and then to wild type background fluorescence. In the case of UVM4 mutants the mRuby2 fluorescence was normalized first to OD 680nm and then to UVM4 background fluorescence. Fig. 3 After transformation, transformants were randomly picked, obtaining a broad spectrum of transgene expression levels ( Fig. 3 B), which is due to positioning effects, as a result from random integration of expression constructs into the nuclear genome of C. reinhardtii [ 35 ]. Fluorescence levels in transformants derived from CC-1690* ** (11 out of 200 transformants analysed) and UVM4 (11 out of 288 transformants) were significanty (P < 0.05) higher when compared to those derived from the low resistance strain ( Fig. 3 B). Although, the median fluorescence and the upper quartile range was higher in UVM4-derived transformants, fluorescence differences between the two transformant populations were insignificant according to a Student`s t -test (p < 0.01, two-tailed hypothesis). Therefore the UV strain displays a high transgene expression capacity, comparable to that of strain UVM4, a strain which along with UVM11 was for many years the only option, if high level transgene expression in C. reinhardtii was required. The nuclear genomes of both strains, UVM4 and UVM11 carry mutations in a gene encoding the Sir2-type histone deacetylase SRTA [ 36 ], which were shown to be causative for the high transgene expression phenotype. Sequencing of the SRTA gene in CC-1690 *** confirmed that UV mutagenesis did not cause mutations in its coding sequence, pointing at mutations in other genes as the cause for higher transgene expression in this strain. Future experiments will aim at identifying the genotype responsible for the phenotype of CC-1690 ***. 3.3 The novel UV strain efficiently produces the diamine cadvarine under photoautotrophic conditions Among the analysed UV strain-derived transformants UV_6, UV_10 and UV_3 displayed the highest mean mRuby2 fluorescence ( Fig. 4 A) and were subjected to further analyses ( Fig. 4 B) together with UV_1 and UV_2, as strains with an intermediate fluorescence. UV_6, UV_3 and UV_2 showed the highest cadaverin titers in the range of 3.5–3.7 mg L −1 after 7 days of cultivation in mixotrophic TAP medium ( Fig. 4 B) and were further analysed regarding their cadaverine production capacity under phototrophic conditions ( Fig. 4 C) and using the novel 6xP medium [ 9 ], developed for high cell density cultivation of C. reinhardtii . Under these conditions UV_6 displayed the highest biomass (20.49 g L −1 biomass dry weight) and cadaverine accumulation (22.5 mg L −1 ) after 14 days of phototrophic cultivation. Fig. 4 Selection of strain UV_6 for further analyses. A ) Analysis of the mean mRuby2 fluorescence in transformants derived from strain CC-1690*** (grey and coloured bars) or UVM4 (black bars). Fluorescence readings were normalized to chlorophyll absorption determined at 680 nm. Values are given relative to the normalized fluorescence recorded for the wildtype (set to 1). Error bars indicate the standard error derived from three biological replicates (n = 3). B ) Cadaverine titers reached in different UV strain-derived transformants after 7 days of cultivation in TAP medium containing acetate for mixotrophic growth. C ) Biomass dry weight (left y-axis, lines) and cadaverine titers (right y-axis, bars) determined for strains UV_2 (red), UV_10 (blue) and UV_6 (green) after 14 days of cultivation in 6xP medium for high cell density cultivation under photoautotrophic conditions. Fig. 4 In order to compare the cadaverine production capacity of UV_6 to that present in the best UVM4-derived strain (UVM4_3; Fig. 4 A), cultivations in photoautotrophic HSM medium were performed ( Fig. 5 ), which, in contrast to 6xP medium, contains ammonium instead of nitrate as the nitrogen source [ 9 , 37 ]. Fig. 5 Phototrophic cadaverine production in strains UV_6 and UVM4_3. A ) Biomass dry weight determined for strains UV_6 (green) and UVM_3 (black) after 14 days of cultivation in phototrophic HSM medium. Error bars indicate the standard error derived from three biological replicates (n = 3). B ) Cadaverine accumulation in UV_6 (green) and UVM_3 (black) during 14 days of cultivation in phototrophic HSM medium. Error bars indicate the standard error derived from three biological replicates (n = 3). C ) Immunodetection of the CadA-mRuby2 fusion in UV_6 and UVM4_3 using an antibody raised against HA-tag and whole cell extracts derived from samples taken at day 2 of the cultivation shown in A) and B). Three distinct biological replicates (#1-#3) were analysed for each strain and protein loading assessed by a Coomassie Brilliant Blue (CBB) stain of whole cell extracts. Fig. 5 Strain UV_6 accumulated about 20 % more biomass than strain UVM4_3 (1.63 g L −1 in UV_6 vs. 1.38 g L −1 ) at day five, where biomass accumulation peaked for both strains ( Fig. 5 A). Cadaverine titers ( Fig. 5 B) showed a comparable increase in both strains, reaching a similar level (23.2 ± 1.6 mg L −1 in UVM4_3 vs. 22.3. ± 2.8 mg L −1 in UV_6) at day 11, before cadaverine further accumulated until day 14 (34.2 ± 3.1 mg L −1 in UVM4_3 vs. 21.4 ± 6.7 mg L −1 in UV_6). Immunodetection of the CadA-reporter fusion protein after 3 days of cultivation ( Fig. 5 C) showed that in UVM4_3 the lysine decarboxylase enzyme accumulated to higher levels, explaining the higher cadaverine titer in UVM4_3 at day 3. The higher biomass accumulation in UV_6 at day 11 seems to compensate for the lower transgene expression, resulting in comparable cadaverine titers. Based on the observed traits of the UV strain including high transgene expression capacity ( Figs. 2 C, 3 B, 4 A and 5 C) and robust growth ( Figs. 2 A and 5 A), it was tested if this strain could be crossed with other C. reinhardtii strains of the opposite mating type ( Fig. 6 ). Fig. 6 CC-1690*** is mating-competent. A ) Microscopic images taken in two distinct mating experiments with the novel strain CC-1690*** (mt + ) and CC-1691 (mt − ). Zygote-forming dikaryons possessing four flagella are highlighted by white boxes. B ) Magnification of two representative zygotes derived from CC-1690*** . Fig. 6 3.4 The novel UV strain is mating-competent As expected from the presence of long flagella and the motility of the UV strain, dikaryons (white boxes) possessing four flagella could be identified in microscopic images taken after mixing UV strain (derived from CC1690 mt + ) with strain CC1691 (mt − ). It is well established that almost every dikaryon (>99.9 %; [ 38 ]) becomes a zygote, and indeed these dykaryons turned into zygotes ( Fig. 6 B), showing the typical thick cell wall [ 39 ]. The cell-wall deficiency and immotility of UVM4/11 precluded mating experiments [ 11 , 36 ] representing a major drawback for their use a bioproduction chassis strain. Only recently, UVM11 could be crossed with wildtype strain CC-124 and the progeny retained the capacity to express transgenes at a high level [ 36 ]. The mating-competence of the UV strain will facilitate future map-based cloning or whole genome re-sequencing experiments [ 40 ] to identify the causative mutation."
} | 4,594 |
24710692 | null | s2 | 8,486 | {
"abstract": "Should iron and copper be added to the environment to stimulate the natural bioremediation of marine oil spills? The key enzymes that catalyze the oxidation of alkanes require either iron or copper, and the concentration of these ions in seawater is vanishingly low. Nevertheless, the dependence of alkane oxidation activity on external metal concentrations remains unclear. This perspective will summarize what is known about the co-regulation of alkane oxidation and metal acquisition and pose a series of critical questions to which, for the most part, we do not yet have answers. The paucity of answers points to the need for additional studies to illuminate the cellular biology connecting microbial growth on alkanes to the acquisition of metal ions."
} | 189 |
36968555 | PMC9995162 | pmc | 8,487 | {
"abstract": "Abstract Biomanufacturing in the form of industrial sugar fermentation is moving beyond pharmaceuticals and biofuels into chemicals, materials, and food ingredients. As the production scale of these increasingly consumer‐facing applications expands over the next decades, considerations regarding the environmental impact of the renewable biomass feedstocks used to extract fermentable sugars will become more important. Sugars derived from first‐generation biomass in the form of, for example, corn and sugarcane are easily accessible and support high‐yield fermentation processes, but are associated with the environmental impacts of industrial agriculture, land use, and competition with other applications in food and feed. Fermentable sugars can also be extracted from second‐ and third‐generation feedstocks in the form of lignocellulose and macroalgae, respectively, potentially overcoming some of these concerns. Doing so, however, comes with various challenges, including the need for more extensive pretreatment processes and the fermentation of mixed and unconventional sugars. In this review, we provide a broad overview of these three generations of biomass feedstocks, outlining their challenges and prospects for fuelling the industrial fermentation industry throughout the 21st century.",
"introduction": "1 INTRODUCTION In the last several years, products made via the industrial fermentation of sugars from renewable biomass feedstocks by genetically engineered microbes have moved beyond their traditional role in biofuels and medicine to consumer‐facing applications in food and materials (Figure 1 ) [ 1 , 2 , 3 ]. This growth will likely continue as both synthetic biology capacities and consumer demand for fossil‐free and animal‐free products expand, such that over the next decades, products made via fermentation and engineered biology will increasingly permeate society [ 4 ]. FIGURE 1 Overview of the sugar‐based biomanufacturing process, showing how sugars from a variety of sources may be fermented by genetically engineered microorganisms to create products for different industries. Examples of the product categories shown include bioethanol, compostable plastics derived from, for example, 1–4 butanediol, textiles made from recombinant proteins, recombinant heme as an ingredient of plant‐based meats, and recombinant insulin. Heme and textile images were reproduced with permission from Impossible Foods and Spiber Inc., respectively Currently, a substantial part of the industrial fermentation industry that makes these products is fuelled by sugars from first‐generation feedstocks, that is, the edible parts of sucrose‐ and starch‐containing biomass crops like sugarcane and maize (corn) [ 1 , 5 , 6 , 7 ]. Products made from these feedstocks often display more favourable environmental impact profiles compared to fossil‐based and animal‐based alternatives [ 8 ], but concerns regarding the impacts of the intensive industrial agriculture practices required to grow these feedstocks at a large scale have been voiced in the context of corn bioethanol and may be echoed in other parts of the industry as production scales increase [ 9 , 10 ]. Related to these concerns is the fact that the expansion of the bioeconomy does not occur in isolation, but in the context of a growing global population [ 11 ] whose per capita demand for crops has historically been a function of per capita real income—largely due to higher meat consumption and the associated need for animal feed [ 12 ]. Although biotechnology could potentially offset some of the pressure of this demand by improving crop yields or enabling plant‐based and cell‐based meat alternatives [ 13 , 14 ], any business‐as‐usual growth scenario could see significant increases in biomass demand for food and feed in addition to that of a larger bioeconomy. To put the potential growth of the bioeconomy and its impact in a practical perspective, consider the size of annual global textile fibre production, estimated at around 100 million tonnes per year in 2017, and projected to grow to 120 million tonnes in 2025 [ 15 ]. Several biomanufacturers are producing bio‐based fibres via the fermentation of sugars derived from first‐generation feedstocks (e.g. protein‐based fibres [ 1 ] and bio‐based nylon [ 16 ]). Assuming a hypothetical 10% yield of biomanufactured fibre on corn starch‐derived glucose, scaling up to 10% of 2017's global fibre production would require around 22 million acres of corn, which is 1.7 times the number of harvested corn acres in Iowa [ 17 ], the world's most productive corn region [ 18 ], and on the same order of corn acres that were used for bioethanol production in the United States in 2018/2019 (∼28 million acres) [ 19 ]. Although the fermentation‐based textile fibre industry is far from reaching this scale and improvements in yield would reduce biomass requirements relative to a given amount of product, various other industries are being targeted by fermentation as well. Together, their demand for biomass may eventually start adding up to magnitudes on the order of the above and beyond. For biomanufacturers relying on first‐generation biomass, this means that it becomes increasingly important to evaluate the environmental impacts of their current feedstock strategy, to mitigate negative externalities as much as possible, and to eventually develop the capacity to use sugars from alternative sources if warranted by future trends in biomass supply and demand. Regarding the latter, alternative sugar sources that do not compete with food resources exist in various forms, out of which lignocellulosic biomass (second generation) and brown macroalgae (third generation) have been among the most prominent and promising [ 20 , 21 ]. Despite the earlier‐stage nature of these feedstock and mixed results with early lignocellulosic biofuel facilities [ 22 ], progress in the processing of these feedstocks and their fermentation into target compounds has been ongoing [ 23 , 24 , 25 ]. In this perspective, we provide a broad overview of these three generations of biomass feedstocks, outlining the challenges and prospects for sustainably and efficiently using them as the bioeconomy expands.",
"discussion": "5 DISCUSSION Although biomass‐derived sugar fermentation may be the dominant paradigm in industrial biomanufacturing at present, it is not the only one. Specifically, single carbon (C1) substrates—including methane, methanol, CO 2 , and formate—are increasingly acknowledged as promising feedstocks for the biomanufacturing industry, largely due to their natural abundance, availability as industrial by‐products, and capacity to be generated electrochemically with excess electricity [ 104 , 105 , 106 ]. As with the xylose or alginate content of second‐ and third‐generation biomass feedstocks, biomanufacturers that wish to use C1 feedstocks face the choice of either developing natural C1‐utilizing organisms for industrial production or engineering the effective use of C1 compounds into existing industrial microbes. Several companies are taking the former approach [ 107 , 108 ], but doing so for a wider range of products will require extensive development of new hosts, an effort that is currently constrained by a lack of industrial familiarity and synthetic biology tools applicable to such hosts. As an alternative approach, recent advances in genome engineering and laboratory evolution have enabled the creation of synthetic C1‐utilizing microbes, illustrated by the first eukaryote ( Pichia pastoris ) and prokaryote ( E. coli ) able to synthetically fix CO 2 in 2019 [ 109 , 110 ], as well as the first E. coli cells able to grow on methanol and formate in 2020 [ 111 , 112 ]. At present, the growth rate of these synthetic C1‐utilizing microorganisms is still well below their growth rate on sugar; for example, the doubling time of wild‐type E. coli on sugar feedstocks is typically around 20 min, whereas methylotrophic E. coli growing on methanol and autotrophic E. coli growing on CO 2 as the sole carbon sources have doubling times of 8.5 h and 18 ± 4 h, respectively [ 106 ]. Although there is thus much work to be done to make C1‐based production for a wider range of products a reality, throughout the 21 st century, these efforts could create opportunities to increasingly shift biomanufacturing feedstocks from biomass‐derived sugars to single carbons. In the meantime, the extent to which the biomanufacturing industry might transition from first‐generation biomass feedstocks to second‐ and third‐generation ones depends on both basic economic and technical considerations as well as global trends in biomass supply and demand. As an illustration of the latter, the Nova Institute evaluated a number of scenarios for global biomass supply and demand projections until 2050 [ 113 ]. In their ‘bio‐based’ demand scenario, biomass demand for the bioeconomy would rise sharply (including an increase in the share of biomass needed to cover the demand of the chemical and plastic industry from 10% to 40% by 2050), and would only be met by supply if both extensive intensification as well as an expansion of agricultural cropland area with 360 million additional ha would take place. In a context of accelerating climate change, biodiversity preservation goals, and the sensitivity of the public when it comes to using biomass for non‐food purposes, such growth—if realized—may stimulate a transition to second‐ and third‐generation feedstocks to minimize these trade‐offs. Among the impacts of first‐generation biomass utilization for bio‐based products, much has been published on land use change in particular. Associated impact estimates vary widely, ranging from large increases in indirect land usage and related biodiversity and carbon losses as a result of diverting corn to ethanol production [ 33 , 114 , 115 ], to a study showing no evidence of ‘lands converting to agriculture because of biofuel’ [ 116 ]. Since 1) it seems a priori unlikely that a large increase in demand would not result in a significant additional land requirement to meet this demand unless compensated by yield increases and/or demand decreases elsewhere, and 2) future biomass demand for food and feed is set to grow in accordance with increases in population and prosperity, we suspect the actual land use change impact of a larger bio‐economy will be non‐negligible. In this context, the yield of a target product on sugar will be a key metric for biomanufacturers to optimize towards its theoretical maximum. A doubling of a compound's yield on sugar, for example, translates to a halving of the amount of feedstock required to produce a similar amount of product. Given how yields on sugar for bio‐based products are initially typically much lower than the theoretical maximum yield [ 117 ], there is likely much to be gained here by applying the tools of synthetic biology to iteratively improve on this important metric. It should be noted that although second‐ and third‐generation feedstocks may largely avoid the land change issue, they do come with their own inherent trade‐offs. Agricultural waste residues like sugarcane bagasse or corn stover, for example, currently fulfil respective roles as fuel in sugar mills and as fertilizer [ 118 , 119 ]. Using this biomass instead to generate fermentable sugars may, therefore, result in an increase of synthetic fertilizer application and fossil fuel usage, offsetting lignocellulose's environmental benefits for various impact metrics [ 35 ]. For third‐generation feedstock processing, the trade‐offs are currently less known and will depend on cultivation location and harvesting methods, but may affect marine ecosystems in various ways. Beyond environmental concerns, the practical feasibility of using any particular feedstock is largely an economic matter. Although the procurement of second‐ and third‐generation biomass is relatively cheap compared to purchasing first‐generation sugars ― $24 to $121 per ton for lignocellulosic biomass depending upon the crop, yield, region, and method of analysis [ 120 ] and $112 per ton for brown macroalgae in the form of Laminaria sp. [ 21 ] versus $440 per ton for sucrose [ 121 ]―the efforts required for feedstock aggregation, processing into fermentable sugars, and reductions in yield due to the often more dilute and impure nature of the resulting sugars, all add up to a final cost that will likely be significantly higher than first‐generation sugars costs. The practical reality of this has been shown before in the biofuels industry, in which numerous commercial efforts at lignocellulosic biofuels were attempted, but many have subsequently been shelved due to economic intractabilities [ 122 ]. Ethanol, however, averaged a price of less than $2 per gallon from 2016 to 2021 [ 123 ], which means that feedstock costs can quickly make up a substantial fraction of the final biofuel price. The fact that some efforts do in fact appear successful [ 124 ] suggests that the economics can be feasible with the right feedstock aggregation strategy, process design, and fermentation process. For higher‐value products like speciality chemicals, food ingredients, and fermentation‐derived alternatives to animal materials, the feedstock price component will be a smaller fraction of the price, suggesting more flexibility and robustness in terms of the maximum affordable feedstock price. Techno‐economic evaluations can quantify such economic considerations for various feedstock choices and do so in the context of environmental impact trade‐offs when combined with an LCA [ 125 ]. Such evaluations will likely be an important first step for biomanufacturers making a stepwise transition from first‐ to second‐ and third‐generation biomass feedstocks when the incentives to do so emerge as a result of, for example, consumer and investor demand. Ultimately, sustainable biomanufacturing involves a complex set of trade‐offs between different social and environmental impact metrics and their magnitude. At present, the trade‐offs inherent in making bio‐based products via the fermentation of first‐generation feedstocks are often (though not always) favourable compared to fossil‐ or animal‐based alternatives, especially from a climate change perspective [ 8 ]. As the bio‐economy expands throughout the next decades though, trade‐offs related to land usage and the conflicting use‐cases for first‐generation biomass may become more dominant, potentially spurring at least a partial transition to the second or third‐generation feedstocks outlined here. Practically speaking, we, therefore, believe it not only prudent to put into place sustainable agriculture practices for first‐generation biomass cultivation, but also to make strategic investments in the technologies needed to efficiently procure and ferment sugars from alternative biomass sources such as the ones described here."
} | 3,738 |
30966494 | PMC6415215 | pmc | 8,488 | {
"abstract": "Silk-based materials are water-sensitive and show different physical properties at different humidities and under wet/dry conditions. To overcome the water sensitivity of silk-based materials, we developed a silk composite material with a fluoropolymer. Blending and coating the silk protein-based materials, such as films and textiles, with the fluoropolymer enhanced the surface hydrophobicity, water vapor barrier properties, and size stability during shrinkage tests. This material design with a protein biopolymer and a fluoropolymer is expected to broaden the applicability of protein-based materials.",
"conclusion": "4. Conclusions In this study, we developed a silk composite material with a fluoropolymer to overcome the water sensitivity of silk-based materials. Blending and coating silk protein-based materials, such as films and textiles, with a fluoropolymer enhanced the surface hydrophobicity, water vapor barrier properties, and size stability during washing/drying cycles. However, the blending of the fluoropolymer with silk proteins cannot stabilize the silk materials perfectly. In the case of the silk films coated with the fluoropolymer, the water and biodegradation resistances of the silk materials were improved. Thus, fluoropolymer treatment is expected to broaden the applicability of silk materials as well as protein-based materials. The other technique to modify the surface property of silk materials is plasma treatment. However, plasma treatment sometimes digests silk molecules at the surface of the silk materials, resulting in more water-sensitive surface. The fluoropolymer coating also has a disadvantage, namely, the loss of the original texture of the silk materials. To resolve this issue, we plan to design and synthesize other fluoropolymers to realize a fluoro-coating layer with a thickness of molecular level. The combination of protein-based materials and fluoropolymers will open a door for many applications in biomaterial, structural material and apparel material fields.",
"introduction": "1. Introduction Spider and silkworm silk proteins (fibroins) have been investigated widely because of their toughness, light weight, biodegradability and biocompatibility [ 1 , 2 , 3 , 4 ]. In addition to silk fibers and textiles, silk protein is processable and has been used to generate various biopolymer-based materials, including nanoparticles, porous materials, films, sponges, and hydrogels, for use in biomedical applications such as tissue engineering, drug/gene delivery and regenerative medicine [ 5 , 6 , 7 ]. In nature, silk fiber is used as a structural material in spider webs, spider draglines and silkworm cocoons [ 8 , 9 ]. However, silk protein has not yet been used as a structural material on a bulk scale. One of the limitations and drawbacks to the practical use of silk as a structural material is its water sensitivity, namely, silk protein-based materials are highly sensitive to water. Water molecules in silk-based materials are categorized roughly into two types, namely, bound water and free water. Bound water molecules strongly interact with silk molecules and show different characteristics to those of bulk water; in contrast, free water means unbound water molecules that behave similarly to bulk water [ 10 , 11 , 12 , 13 ]. The influences of bound water on the biological and physical properties of silk molecules have been widely reported by several groups. Asakura and coworkers reported that the hydration of Bombyx mori silk molecules induces the stabilization of silk I forms based on solid-state NMR analysis [ 14 ]. The secondary structure and dynamics of swollen B. mori silk molecules were also characterized by 13 C- and 1 H-pulsed NMR [ 15 , 16 ]. Asakura et al. have also reported that hydration does not affect the crystalline fraction of B. mori silk fibers [ 17 ]. It was reported that bound water influences the glass transition temperature ( T g ) of B. mori silk molecules, that is, the T g decreases with an increase in the water content of the silk films [ 18 , 19 , 20 ]. In addition to B. mori silk, the storage modulus and loss tangent of Nephila edulis spider dragline [ 21 , 22 ] and the elastic modulus values of Antheraea pernyi silkworm silk [ 23 ] and Argiope trifasciata spider silk [ 24 ] are reported to be affected by bound water. Bound water is considered to disrupt the hydrogen bonds between silk molecules in amorphous phase, and hence to enhance the mobility of the silk molecules, as well as influence the glass transition behavior [ 18 , 19 , 20 , 21 , 25 ]. Considering the biological properties of silk-based materials, the state of the water molecules in a silk hydrogel controls the effect of the hydrogel on cell viability, namely, human cell lines and cell-adhesion proteins in the extracellular matrix preferentially expand and adhere on silk molecules hydrated with more bound water [ 13 ]. Macroscopic studies on the effect of water and relative humidity (RH) on silk materials have been reported by several groups. Cebe and coworkers reported that hot-water vapor annealing induces crystallization in silk films [ 26 ]. In our previous studies, we demonstrated the effects of the water content in silk films and fibers on crystallization, bio- and thermal degradation [ 27 , 28 ]. By using thermal gravimetric analysis and differential scanning calorimetry (DSC), the silk samples prepared at different RHs were analyzed in terms of the effects of the water content on thermal degradation, crystallization and transition of B. mori silk materials. The hydration state and RH affected the mechanical properties of silk fibers [ 23 ]. At a relatively high RH, approximately 97%, the toughness and degree of crystallinity of silk films increase dramatically, indicating that the appropriate hydration of silk molecules induces crystallization and plasticization simultaneously [ 28 ]. The RH from 20% to 60% resulted in tough and strong silk materials by using various types of silk hydrogels. Dehydration did not negatively impact the biodegradability of the silk resins and hydrogels [ 27 ]. Thus, the thermal stability, mechanical properties and other attributes of silk materials are regulated by their water content and crystallinity. To exploit silk and silk-based materials as practical structural materials for human use, stabilization against water molecules is necessary. As introduced above, silk-based materials are water-sensitive and show different physical properties at different humidities. However, this instability of silk under wet conditions is detrimental to its use as a structural material. To overcome the water sensitivity of silk-based materials, in this study we developed a silk composite material with a fluoropolymer, which is famous for its hydrophobicity and waterproofness. Blend films of silk proteins and fluoropolymer showed enhanced surface hydrophobicity and vapor barrier properties. The coating of fluoropolymer on silk textiles was resistant to washing and shrinkage treatments. This material design, with a protein biopolymer and a fluoropolymer, will broaden the applicability of protein-based materials.",
"discussion": "3. Results and Discussion 3.1. Basic Properties of Silk/Fluoropolymer Blend Films Blend films of silk and fluoropolymer were prepared at different blend ratios (wt %) ( Figure 1 ). All the films were transparent, but the films with lower silk contents such as silk/fluoropolymer (FP) = 1/9 were too elastic to be handled. The mechanical properties of the silk/FP blend films were characterized in terms of their stress–strain curves ( Figure 2 ). With an increase in the fluoropolymer content, the blend films became more elastic and stretchable. The surface morphologies of the film samples were observed by SEM ( Figure 3 a). We could recognize the domain structures at the surfaces. In the enlarged images of the 7/3 and 5/5 silk/FP blend films, the fluoropolymer domains were obvious. In the cases of the 9/1 and 1/9 silk/FP blend films, minor domains were obscure but detectable. After stretching the film samples, we imaged the film surfaces close to the fracture edges ( Figure 3 b). The round domains were stretched and became more crack-like, indicating that silk/FP domains have a substantial impact on mechanical properties. Furthermore, silk and the fluoropolymer were not perfectly miscible at blend ratios between 9/1 and 1/9. To clarify the blending state, namely, to determine the miscibility of silk and fluoropolymer, the blend films were characterized by DSC. The peaks originating from the removal of water, and the T g of silk were assigned as shown in Figure 4 . On the other hand, the fluoropolymer did not show clear peaks or transitions. In terms of the shift in the silk T g , silk domains were present at silk/FP ratios from 9/1 to 1/9, and their T g values were not significantly influenced by the fluoropolymer. This result indicates that the silk and fluoropolymer were not miscible in the blend films, which was confirmed by the SEM micrographs shown in Figure 3 . The hydrophobicities of the blend films were characterized based on their advancing contact angles ( Figure 5 ). Based on the contact angles of water droplets on the blend films, we clarified the trend in the hydrophobicity of the blend films as a function of fluoropolymer content ( Figure 6 ). As expected, the contact angle increased with an increase in fluoropolymer content. However, surprisingly, the blend film (silk/FP = 1/9) showed the highest contact angle, approximately 120°, which might be because the perfluoro groups of the fluoropolymer were aligned on the surface by the presence of the silk molecules. In addition to the hydrophobicities of the blend films, we evaluated their water vapor barrier properties ( Figure 7 ). The water vapor barrier properties are related to not only the surface morphologies, but also the interior structures of the blend films. The water vapor permeability decreased with an increase in fluoropolymer content. The silk/FP blend film (1/9) showed a water vapor permeability of approximately 170 g/m 2 /day. Although nylon has a similar polyamide structure to that of proteins, in that both contain amide bonds; the water vapor permeability of nylon 6 and nylon 6, 10 are 47 and 22 g/m 2 /day, respectively. For practical uses of silk materials, we expect that the water vapor permeability of silk blend materials will be approximately 100 g/m 2 /day. However, the blend film did not show such a low water vapor permeability in this study. 3.2. Fluoropolymer Coating with Silk Materials To enhance the water resistance of silk materials, we evaluated the fluoropolymer coating of the silk materials. B. mori silk textile and silk film were coated with the fluoropolymer. The coating of the fluoropolymer on the silk films was approximately 50 µm thick, based on the cross-section SEM observation of the fluoropolymer-coated silk films. The fluoropolymer-coated silk films show a contact angle of approximately 90°. The water vapor permeability of the fluoropolymer-coated silk films was approximately 110 g/m 2 /day. These properties of the fluoropolymer-coated silk films indicate the potential advantages of a fluoropolymer coating on silk materials. To show the excellent water resistance of the fluoropolymer-coated silk-based materials, shrinkage tests, which were similar to the washing and drying process used for clothing, were performed with the silk textiles. A shrinkage test was performed on the silk textile, but not on the silk film, as the shrinkage test is designed for clothes. We coated B. mori silk textile with fluoropolymer and subjected it to the shrinkage test ( Figure 8 ). The coated textile showed great size stability during the washing/drying cycles ( Figure 8 a,c). In contrast, the silk textile without the fluoropolymer coating did not show size stability ( Figure 8 b). After the third drying stage, the sizes had decreased by approximately 5% ( Figure 8 d). Thus, the fluoropolymer coating can enhance the water resistance of silk materials in practical applications. 3.3. Biodegradability of Silk, Fluoropolymer-Coated Silk, and Silk/Fluoropolymer Blend To study the effect of the fluoropolymer on the biodegradability of silk materials, a BOD test was performed using the silk film, fluoropolymer-coated silk film, silk textile, fluoropolymer-coated silk textile, and silk/FP (1/9) blend film in active sludge at 25 °C ( Figure 9 ). All the silk materials showed increases in BOD with increases in degradation time, except for the fluoropolymer-coated silk film. The fluoropolymer-coated silk film was not significantly degraded after being subjected to active sludge for 14 days, indicating that the fluoropolymer coating prevents biological degradation of silk films. In contrast, the fluoropolymer coating on the silk textile did not prevent biological degradation. Before and during the biodegradation tests, the sample morphologies noticeably changed ( Figure 10 ). The silk film without the fluoropolymer coating became white during the biodegradation test, indicating that the surface of the silk film degraded and became rough. The fluoropolymer-coated silk film also became white and opaque, even though the surface was coated with fluoropolymer. The silk textile samples with and without the fluoropolymer coating gradually changed, especially at the edges of the samples. The silk/FP (1/9) blend film became white and opaque. To visualize the details of the surface morphologies, as well as to clarify the difference in the degradability of silk film and textiles, we observed the silk samples by SEM before and after the biodegradation test ( Figure 11 ). The surface of the mock silk film without any coating was degraded and roughed by the biodegradation treatment, while the FP-coated silk film was not noticeably degraded. Based on the BOD tests ( Figure 9 ), silk textile samples, even those with a fluoropolymer coating, were degraded in active sludge. The SEM images of the fluoropolymer-coated silk textile showed degradation at the surface (FP-coated silk textile after BOD test shown in Figure 11 ). Thus, the simple dipping of the silk textile in the fluoropolymer solution was not sufficient to perfectly coat the silk textile and protect it against biological degradation. The silk/FP blend film showed a rough surface morphology after the BOD test, indicating that the blend film was also degraded in the active sludge. To confirm the biodegradation-resistance of the fluoropolymer-coated silk films, we further characterized their mechanical properties before and after the BOD test. After the treatment with active sludge for 14 days, the silk films were too brittle for the mechanical tests. Therefore, we could not measure the stress-strain curves of the silk films after the BOD test. The fluoropolymer-coated silk films were not dramatically changed by the biodegradation test in terms of their mechanical properties ( Figure 12 a). On the other hand, the silk/FP (1/9) blend film was weakened by biodegradation ( Figure 12 b). Thus, the combination of the BOD test and mechanical analysis of the silk materials confirmed that fluoropolymer coating can prevent biodegradation of silk films, even in active sludge."
} | 3,822 |
25242287 | null | s2 | 8,490 | {
"abstract": "Gram-negative bacteria use N-acyl L-homoserine lactone (AHL) quorum-sensing (QS) signals to regulate the expression of myriad phenotypes. Non-native AHL analogs can strongly attenuate QS receptor activity and thereby QS signaling; however, we currently lack a molecular understanding of the mechanisms by which most of these compounds elicit their agonistic or antagonistic profiles. In this study, we investigated the origins of striking activity profile switches (i.e., receptor activator to inhibitor, and vice versa) observed upon alteration of the lactone head group in certain AHL analogs. Reporter gene assays of mutant versions of the Pseudomonas aeruginosa QS receptor LasR revealed that interactions between the ligands and Trp60, Tyr56, and Ser129 govern whether these ligands behave as LasR activators or inhibitors. Using this knowledge, we propose a model for the modulation of LasR by AHL analogs-encompassing a subtly different interaction with the binding pocket to a global change in LasR conformation."
} | 255 |
36984700 | PMC10054235 | pmc | 8,491 | {
"abstract": "Superhydrophobic poly(vinylidene fluoride) (PVDF) membranes were obtained by a surface treatment consisting of oxygen plasma activation followed by functionalisation with a mixture of silica precursor (SiP) (tetraethyl-orthosilicate [TEOS] or 3-(triethoxysilyl)-propylamine [APTES]) and a fluoroalkylsilane (1H,1H,2H,2H-perfluorooctyltriethoxysilane), and were benchmarked with coated membranes without plasma activation. The modifications acted mainly on the surface, and the bulk properties remained stable. From a statistical design of experiments on surface hydrophobicity, the type of SiP was the most relevant factor, achieving the highest water contact angles (WCA) with the use of APTES, with a maximum WCA higher than 155° for membranes activated at a plasma power discharge of 15 W during 15 min, without membrane degradation. Morphological changes were observed on the membrane surfaces treated under these plasma conditions, showing a pillar-like structure with higher surface porosity. In long-term stability tests under moderate water flux conditions, the WCA of coated membranes which were not activated by oxygen plasma decreased to approximately 120° after the first 24 h (similar to the pristine membrane), whilst the WCA of plasma-treated membranes was maintained around 130° after 160 h. Thus, plasma pre-treatment led to membranes with a superhydrophobic performance and kept a higher hydrophobicity after long-term operations.",
"conclusion": "4. Conclusions Superhydrophobic PVDF membranes were prepared by means of surface treatment based on an initial oxygen plasma activation followed by functionalisation with a fluoroalkylsilane and silica precursors. The outcomes of the statistical design of the experiments showed that the type of silica precursor had the highest effect on the water contact angle of the modified membrane. Particularly, the use of APTES as silica precursors always led to membranes with higher water contact angles than those obtained by using TEOS, which was ascribed to the asymmetric structure of APTES. Other factors, such as the power and, especially, the time of the plasma discharge, showed less significant effects on the water contact angle. From those results, optimal values for the oxygen plasma treatment were established at a power of 15 W and a time of 15 min, obtaining contact angles higher than 155° and avoiding membrane degradation. In fact, the modifications acted primarily on the surface, and the bulk properties remained stable and maintained the thermal features of PVDF membranes after the treatments. Moreover, at those plasma conditions, the membrane surface showed a pillar-like morphology, with higher porosity and roughness, which favoured the hydrophobicity. During long-term operation tests under a continuous liquid flux, the contact angle decreased with the time of use, which was attributed mainly to the removal of fluoroalkylsilane molecules from the surface since a lower fluorine content was observed after the long-term tests, and also to the reduction of the surface porosity due to a surface plastic deformation. However, the functionalisation layer of the oxygen plasma-activated membranes showed high stability, in comparison with the coated membranes without any activation step. After 160 h of operation, the plasma-treated membranes showed a water contact angle of approximately 130°, which was still higher than that of the non-modified membranes (119°). Further research should focus on the improvement of the anchoring of the modifying agents onto the PVDF substrate and the feasibility and scalability of the functionalisation methodologies for their implementation at an industrial scale.",
"introduction": "1. Introduction The interest in membrane technology is continuously increasing due to the high efficiency in separation processes, compactness of the membrane modules and lower energy consumption compared with conventional separation units that present problems related to flooding, foaming and emulsions [ 1 , 2 ]. In this regard, recent efforts have especially focused on the development of membrane distillation for the water desalination of seawater, brackish water and brines [ 3 , 4 , 5 ], gas separation for carbon dioxide (CO 2 ) removal from flue gases and biogas [ 6 , 7 ], and pervaporation systems for the biobutanol and dissolved methane (CH 4 ) recovery from biological effluents and wastewaters [ 8 , 9 ]. However, membranes tend to suffer from wetting and fouling, especially in those applications that involve highly contaminated or quite complex liquid feeds, such as industrial brines or anaerobic effluents [ 10 , 11 , 12 , 13 ]. Wetting and fouling phenomena lead to an additional mass transfer resistance located inside and on the surface of the membrane, respectively, reducing the separation efficiency and involving additional cleaning processes [ 14 ]. Therefore, the useful lifetime of the membrane can decline considerably, and the operational cost would rise [ 15 , 16 , 17 ]. Up to now, these issues are not completely resolved, hindering the large-scale application of membrane technology in areas such as the desalination of seawater or methane recovery from anaerobic effluents [ 18 , 19 ]. Different techniques for tailoring polymeric membrane characteristics have been successfully implemented to improve the wetting resistance of the membranes [ 20 , 21 ]. In this regard, superhydrophobic membranes with a water contact angle (WCA) higher than or equal to 150° [ 22 ] have been reported to significantly mitigate the wetting, which was attributed to the low contact area between the liquid phase and the membrane [ 23 , 24 ]. Additionally, superhydrophobicity has been related to the improvement in the self-cleaning properties of the membrane surface [ 12 , 19 ]. Thus, membrane surface functionalisation has been applied to increase its hydrophobicity at the surface level, whilst the bulk properties remain unchanged [ 25 ] by means of the addition of new hydrophobic functional groups, such as siloxanes and fluoroalkyls [ 26 , 27 ]. Surface modification techniques can be classified into two main categories: physical and chemical treatments [ 27 ]. Physical treatments involve a physical interaction between the modifying agents and the membrane, and the initial composition of the membrane remains unchanged. Among physical treatments, surface coatings have been widely studied to confer hydrophobicity to surfaces [ 28 , 29 ] and, more recently, different lithography and nanotexturing techniques have been reported to increase the fouling resistance and self-cleaning behaviour [ 30 , 31 , 32 , 33 ]. However, physical treatments such as surface coating could be unstable over long-term operations [ 34 ]. In contrast, chemical treatments involve the grafting of the modifying agents on the membrane surface by means of chemical bonding, such as covalent, ionic and hydrogen bonds, thus achieving a stronger adhesion force [ 35 ]. An activation step prior to the grafting of the modifying agents is frequently needed in membranes that present a high inertness and chemical stability, such as polyvinylidene fluoride (PVDF) [ 3 , 27 ]. This activation is usually based on the addition of oxygen-containing functional groups such as hydroxyls (-O-H), peroxides (-O-O-) and carbonyls (-C=O) that act as active sites for the subsequent grafting [ 36 , 37 , 38 ]. This activation has been successfully carried out by an alkali treatment with sodium hydroxide (NaOH) or lithium hydroxide (LiOH), plasma treatment and high-energy radiation, among other processes [ 3 , 39 , 40 ]. Plasma treatment is considered an environmentally friendly, versatile, reproducible, easily scalable and inexpensive method for activating and texturing polymer surfaces [ 16 , 41 , 42 , 43 , 44 , 45 ]. This technique consists of a high-energy discharge that ionises the gas near the electrodes and produces a complex gas mixture of excited ions and electrons, atoms and molecule fragments and free radicals [ 39 , 46 , 47 ]. The gas plasma is also a source of radiation that can break chemical bonds of the material [ 48 ]. The effect of the plasma discharge in the treated membrane relies on the type and conditions of the supplied gas, pressure, the power of the discharge, the duration of the treatment and the configuration of the chamber and electrodes [ 44 , 49 , 50 , 51 ]. Thus, a chemical and/or physical modification can be induced on the membrane surface depending on the plasma conditions [ 47 , 49 ] since the ion bombardment and interaction with the different reactive species contained in the plasma can produce sputtering of the membrane material (etching), substitution reactions, atom abstraction, removal of volatile substances and/or scission of polymer chains [ 4 , 46 , 52 , 53 ]. Additionally, the use of oxidative gases, such as oxygen (O 2 ), CO 2 and water (H 2 O), for plasma treatment creates a more reactive environment capable of adding oxygen-containing functional groups onto the membrane surface. Plasma treatment has been reported as a useful approach for tailoring the chemical composition and/or surface morphology of different polymeric membranes and surfaces, such as polydimethylsiloxane (PDMS) [ 41 , 54 , 55 ], polypropylene (PP) [ 36 , 56 ], polytetrafluoroethylene (PTFE) [ 49 ], polyethylene (PE) [ 57 ], polycarbonate (PC) [ 33 , 58 ], polyacrylonitrile (PAN) [ 35 , 59 ] and polyethersulfone (PES) [ 35 ]. In contrast, PVDF has been extensively studied and commercialised as a membrane material, owing to its outstanding features such as high mechanical, chemical and thermal resistance, inertness, ease of processing and relatively low cost of the raw materials [ 25 , 27 , 60 , 61 ]. Hence, PVDF has been treated to improve the hydrophobicity for membrane distillation and CO 2 absorption [ 24 , 39 , 62 ] or to induce hydrophilicity for filtration processes and enhance the fouling resistance [ 42 , 61 , 63 ]. The effects of membrane modifications with plasma treatments have mostly been evaluated based on the membrane properties at microscopic (morphology, chemical composition) and macroscopic (hydrophobicity) levels after the modification procedure, and the performance of the modified membranes is often benchmarked against the pristine membrane. Only a few studies have evaluated the effects of long-term operation on the chemical properties, morphology and stability of hydrophobicity of the modified membranes [ 36 ]. For example, the work by Jiménez-Robles et al. [ 64 ] showed that the modification of PVDF membranes with alkali activation and functionalisation with fluoroalkylsilanes (FAS) kept a higher surface hydrophobicity and avoided the water breakthrough that non-modified PVDF suffered after approximately 800 h treating an anaerobic effluent for the dissolved CH 4 recovery. Therefore, it is an area of interest requiring further studies into the effects of the combination of plasma activation and surface grafting on the long-term stability of membrane properties. In this context, the aim of this work was to evaluate the effect of the oxygen plasma treatment on the surface modification of a commercial PVDF membrane. An evaluation of different silica precursors for further grafting of a FAS in order to produce superhydrophobic membranes was also carried out. First, a statistical experimental design was conducted considering the power and time of the plasma treatment and the type of silica precursors to maximise the surface hydrophobicity of the membrane by measuring the static water contact angle. Second, membrane stability tests over long-term operation were conducted in a flat-sheet membrane module with a constant flux of deionised water. The stability of the membranes was evaluated and benchmarked against the pristine PVDF in terms of hydrophobicity, thermal properties, morphology and chemical composition.",
"discussion": "3. Results and Discussion 3.1. Effect of the Oxygen Plasma Activation and Organofluorosilanisation on Membrane Hydrophobicity A statistical experimental design evaluating the main parameters that affect the membrane surface hydrophobicity was conducted to maximise the response variable, i.e., the WCA of the plasma-activated and functionalised membranes. It is relevant to remark that after the previous plasma activation, with no further functionalisation, the droplets deposited onto the membranes were quickly absorbed, resulting in superhydrophilic surfaces with a WCA lower than 10°, as reported by other authors [ 38 ]. This hydrophilic behaviour was mainly attributed to the generation of oxygen functional groups (hydroxyls, peroxides and carbonyls) [ 38 ]. After the preliminary experiments, and based on the literature [ 50 ], the power and time of the plasma treatment and the type of SiP were found to be the main factors to be optimised. The effects of these factors and their interactions can be observed in the Pareto diagram of the design shown in Figure 1 . The critical standardised effect (2.015) was calculated from the Student’s T distribution (t α/2 ,ν) with a significance level (α) of 0.05 and 16 degrees of freedom (ν), associated with the error of the design of the experiments. As observed in the Pareto diagram, the SiP presented the highest standardised effect (8.830), indicating that the WCA was mainly influenced using TEOS or APTES. Likewise, the power of the plasma treatment significantly affected the WCA in the tested conditions with a standardised effect value of 3.723. However, the plasma time showed a standardised effect of 2.045, which was similar to the critical value (2.015), indicating a low effect on the response, at least with the evaluated operational conditions. The interaction between the factors did not appear to affect the WCA of the modified membranes since lower values of the standardised effects were obtained respective to the critical value. Hence, the effects of the interactions in the WCA of the treated membranes could be neglected, thereby easing future scale-up technology. The effects of the individual factors on the response are depicted in the main effects plot shown in Figure 2 . A WCA higher than that of the pristine PVDF (119.4 ± 1.7°) was obtained for all the tested plasma-treated membranes, with an overall mean value of 153.3°, showing that the modification procedure with the oxygen plasma activation proposed in this work was suitable to obtain superhydrophobic PVDF membranes. As observed by other authors [ 19 , 72 ], droplets placed on the membranes for the WCA measurements easily rolled off the surface in most of plasma-treated membranes, which is an essential condition for superhydrophobic and self-cleaning surfaces. As can be observed in the main effects plot ( Figure 2 ), the WCA of the PVDF membranes achieved values of around 152.5° with the lowest plasma power (3 W) and maintained similar values for power less than or equal to 10 W. This could indicate that the membrane was saturated with oxygen active sites at only 3 W of power, limiting the grafting of the SiP and FAS. During plasma treatment, C-H and C-F bonds from PVDF chains are broken by the effect of the ion bombardment and radicals and electron interactions [ 44 , 51 , 53 , 67 ], generating volatile substances such as carbon monoxide (CO), CO 2 or hydrogen fluoride (HF) [ 4 , 47 , 48 ] that are removed from the membrane, inducing active sites on the surface. The high reactivity of some removed substances inside the plasma chamber could lead to reabsorption on those active sites on the membrane [ 16 ]. Thus, the rate of atom removal and reabsorption on the membrane seemed to be in equilibrium under the low plasma power values of 3 to 10 W at the time points evaluated. However, the WCA of the modified membranes continuously increased with plasma power values higher than or equal to 10 W, indicating a greater performance of the plasma treatment at high power. This improvement was mainly attributed to the etching effect on the membrane surface, inducing an increase in the roughness. This phenomenon will be discussed in further sections. The time of plasma treatment presented the lowest effect in the WCA since a slight increase in the WCA (less than 2°) was observed when increasing the plasma time from 3 to 17 min. Kim et al. [ 73 ] reported that the changes in the WCA of PVDF membranes after the oxygen plasma treatment at 10 W occurred in the first 60 s of the discharge, after which the WCA remained constant. It is worth mentioning that PVDF samples treated at a plasma power and time higher than 17 W and 17 min, respectively, became brittle and easily destroyed during their manipulation in the membrane modification and/or analysis. Hence, modified membranes could neither be evaluated nor obtained at higher plasma power and time, limiting the working upper limit of these factors for the treatment of PVDF with oxygen plasma. This is in agreement with previously reported results, which have shown an increase in membrane rigidity after the plasma treatment [ 74 ]. Regarding the main effect of the categorical factor ( Figure 2 ), the type of SiP presented the highest effect, as previously indicated by the Pareto diagram. The highest WCA values of the plasma-treated membranes were obtained with APTES, with an overall mean value of 155°, compared to those membranes modified with TEOS, with a significantly lower overall mean value of 151°. The use of TEOS as SiP led to the seeding of siloxane chains (-Si-O-Si-O-) onto the membrane surface in the active sites generated during the plasma treatment. The reactions involved in the grafting process with TEOS are detailed elsewhere [ 64 ]. In contrast, the use of APTES can lead to the formation of siloxane chains with additional alkyl chains [ 39 ] that are naturally hydrophobic [ 40 , 50 ] and come from the aminopropyl group present in the APTES molecule. Moreover, the presence of an amine group can involve additional grafting reactions in the oxygen-rich active sites on the plasma-treated surfaces. Such reactions include amide formation and even breakage or scission of the PVDF backbones [ 22 , 26 , 36 , 75 ]. Hence, the incorporation of an additional reactive group, such as the primary amine of the APTES, can lead to the formation of a more complex and uneven structure on the membrane surface, compared to the more symmetric siloxane structure originated from TEOS. Likewise, subsequent grafting of the FAS could have taken place in a heterogeneous way when APTES was used, which positively increased the WCA. An ANOVA and surface response analysis based on a linear multiple regression were conducted to create a model that fit the experimental results and to determine the values of the factors that maximised the WCA of the plasma-treated membranes. The response surface for each categorical factor value (TEOS and APTES) is shown in Figure 3 . From the ANOVA of the model, the F-values of the quadratic and two-way interaction terms were low (less than 2), indicating that they could be neglected compared to the lineal terms with a higher F-value of 32, which agreed with the previous main effects analysis. In addition, the ANOVA showed no evidence of lack-of-fit ( p -value of 0.58), indicating that the model can adequately predict the WCA of the modified membranes under the operational conditions tested, as observed when comparing the response surface with the experimental data (open symbols in Figure 3 ). The maximum WCA values predicted by the model were 155.2 ± 2.7° and 157.3 ± 2.1° with TEOS and APTES, respectively, at the highest plasma power and time of 17 W and 17 min, respectively. However, these maximum WCAs did not show significant differences compared with those WCAs predicted at a lower plasma power and time of 15 W and 15 min, respectively, and the experimental WCAs obtained at these conditions were 155.5 ± 1.5° and 157.0 ± 0.9° for TEOS and APTES, respectively. Thus, the conditions for the plasma treatment were established at 15 W and 15 min in order to achieve high superhydrophobic membranes for their further characterisation. The oxygen plasma-activated and functionalised PVDF membranes at these conditions and TEOS or APTES as SiP are labelled as PO 2 -TEOS and PO 2 -APTES, respectively. The infrared spectra of these membranes and the pristine PVDF are shown in Supplementary Material (S2) , even though no significant differences were observed due to the penetration depth of the FTIR-ATR analysis [ 21 ]. Only a few works evaluating a membrane modification procedure with plasma activation followed by grafting of FASs can be found in the literature. In a similar approach to this work, Liu et al. [ 19 ] activated a PVDF membrane with oxygen plasma at 50 W for 1 min, followed by grafting of the FAS 1H,1H,2H,2H-perfluorodecyltriethoxysilane, achieving a WCA of 162.0 ± 2.3° in membrane surfaces with a pillar surface structure. The higher WCA reported in that work could be ascribed to the higher fluorine chain length of the modifying agent used [ 39 ] and the high initial WCA of the pristine membrane (155.3 ± 1.7°) [ 19 ]. Sairiam et al. [ 39 ] also evaluated the previous FAS for the modification of PVDF membranes with a helium plasma activation at 80 W for 180 s, and they reported an increase in the WCA from 68.9 ± 0.9° to 145.6 ± 3.1°. 3.2. Membrane Stability Tests in Long-Term Operation Long-term performance is a design requirement of high relevance for polymer-based dispositive [ 76 ], but research works evaluating the effect of long-term operation on the stability of modified membranes are still very scarce and mainly focused on the stability in the separation performance [ 39 ]. In this section, the effect of long-term operation on the stability of different membrane properties was evaluated. First, the results regarding the stability of the membrane bulk properties after long-term operation are shown, followed by the stability evaluation of the surface properties during the operation. For comparison purposes, the outcomes obtained with the different treated membranes were benchmarked with the pristine PVDF membrane (p-PVDF). 3.2.1. Stability of the Bulk Properties after Operation Thermal analysis is a quality tool to characterise the performance of polymers at the design stage and after operation [ 77 ]. The results from thermogravimetric analysis under different atmospheres are shown in Figure 4 in the form of derivative thermogravimetric (DTG) curves for the pristine, coated and plasma-treated PVDF membranes before and after the long-term stability tests. Under an oxidative environment, all the membranes before the stability test showed a similar trend until the complete decomposition of the samples was observed in the range of 450 to 600 °C ( Figure 4 a). Under an inert environment ( Figure 4 c), the decomposition profile was similar to the outcomes under an oxidative atmosphere, showing main decomposition peaks at temperatures around 475 °C. However, a char was observed at the end of the TGA analysis for all the PVDF membranes representing 20 to 30% of the initial weight of the sample, similar to the pristine PVDF analysed by other authors [ 59 ]. These outcomes indicate that the chemical structure of the pristine PVDF was not affected by plasma activation and functionalisation. The DTG thermograms of the membranes after long-term operation ( Figure 4 b,d) showed the same trends that the non-used membranes had with no significant variation in the decomposition temperatures, indicating that long-term operation did not affect the thermal behaviour and stability of the bulk membrane. DSC measurements were conducted to detect potential modifications in the amorphous/crystalline configuration and thermal performance of the membranes after the modification procedures and operation. The DSC thermograms of the pristine, coated and modified PVDF membranes under a nitrogen atmosphere are shown in Figure 5 . The melting temperature and enthalpy, together with the crystallinity degree before and after the long-term stability tests, are shown in Table 3 . Similar values were observed for all membranes, regardless of the membrane treatment and operation, which highlights the stability of the bulk PVDF toward surface modification technologies, as stated by other authors [ 35 , 39 , 43 , 48 ]. The characteristic melting temperature was approximately between 160 and 180 °C, with an endothermic peak at around 163 °C. These outcomes are consistent with the literature [ 38 , 67 ]. The melting enthalpies of the samples were between 53 and 59 J g −1 , and maximum crystallinity with a value of 61% was observed for the pristine PVDF. The treated membranes showed slightly lower crystallinity degrees with values between 54 and 58%. Other authors have also reported similar crystallinity degrees of around 54% for PVDF membranes treated with argon and oxygen plasma at 100 W for a period ranging from 200 to 600 s [ 38 ]. Similar findings and trends were observed on the membranes after the stability tests ( Figure 5 b and Table 3 ). In general terms, neither membrane treatment nor long-term operation induced any significant modification of the thermal behaviour of the membranes. The works of Correia et al. [ 38 , 67 ] also reported no significant effect of argon and oxygen plasmas on the crystallinity of PVDF membranes. 3.2.2. Stability of the Surface Hydrophobicity in Operation The stability of PVDF membranes during long-term operation was evaluated by monitoring its WCA as a useful indicator of changes or alterations occurring on the membrane surface [ 28 , 64 , 65 , 66 ]. The results of the variation in WCA with the time of use of the coated PVDF membranes (Coat-TEOS and Coat-APTES) and plasma-treated PVDF membranes at the optimal conditions (PO 2 -TEOS and PO 2 -APTES) are shown in Figure 6 a,b, respectively. The initial WCA of Coat-TEOS and Coat-APTES were 144.8 ± 3.2° and 149.8 ± 2.7°, respectively. These values were lower than their respective plasma-treated membranes (PO 2 -TEOS and PO 2 -APTES), suggesting a lower presence of FAS and SiP on the membrane surface due to the absence of a previous activation. Kaur et al. [ 25 ] studied the grafting of methacrylic acid on non-treated and plasma-treated PVDF membranes, and they reported that grafting was facilitated by the formation of radicals and peroxide groups in argon plasma-treated membranes. As can be seen in Figure 6 a, the WCA of both coated membranes decreased pronouncedly during the first 24 h, reaching similar values to those of the p-PVDF. Then, the WCA of the coated membranes remained constant after 24 h, as did the p-PVDF. Hence, it can be concluded that the effect of the coating with FAS/TEOS and FAS/APTES solutions was completely lost from the surface-coated membranes before 24 h. This fact suggests that only a physical deposition of FAS and SiP was involved without further chemical reactions, creating an unstable surface layer that would be removed by dragging effects under a constant liquid flux over the membrane [ 34 , 64 ]. In contrast, although the WCA of the plasma-treated membranes PO 2 -TEOS and PO 2 -APTES decreased from the initial values of 155.5 ± 1.5° and 157.0 ± 0.9° to 127.0 ± 3.4° and 129.9 ± 3.8° at approximately 75 h, respectively, the WCA values stayed almost constant and slightly higher than those of the p-PVDF after 75 h ( Figure 6 b). These results suggested that the effect of the functionalisation was partially lost, likely due to removal of grafted molecules by the dragging effect, as reported in a previous work with modified membranes activated with an alkali solution [ 64 ]. However, WCA of the plasma-treated membranes after the stability tests was slightly higher than that of p-PVDF, suggesting a remnant of FAS and/or SiP on the surface. Hence, stronger chemical bonding and interactions could be inferred between the SiP, FAS and activated PVDF than those shown by the coated membranes. Comparing the effect of the SiP on the membrane stability, both PO 2 -APTES and PO 2 -TEOS membranes experienced a similar decrease rate with the time of use. However, PO 2 -APTES always showed slightly higher WCAs, mainly attributed to the additional hydrophobic aminopropyl segments from the APTES grafted on the membrane, as previously discussed. Research works evaluating the effect of long-term operation on the stability of modified membranes are still very scarce, with most of them focusing on the stability of the separation performance. For example, our previous study [ 64 ] evaluated a functionalisation protocol using an alkali-based activation and showed that the WCA of the modified PVDF decreased to values lower than the pristine membrane in the first 50 h of operation at a liquid flow rate of 27 L h −1 . In this regard, the plasma pretreatment seemed to provide a more stable hydrophobic surface. Moreover, Sairiam et al. [ 38 ] performed long-term tests for CO 2 absorption with helium plasma-treated and functionalised PVDF membranes, and they reported a stable gas flux during the 15 days, unlike the unmodified membrane and the membrane activated with an alkali solution. Liang et al. [ 60 ] evaluated the stability of PVDF membranes modified with Ar plasma activation and grafting with methacrylic acid by immersing the membranes in solutions with different pH for only 15 min, and they reported a similar WCA before and after the stability tests. Moreover, Gryta [ 36 ] reported that PP membranes treated with helium (He) plasma for enhancing hydrophilicity showed a greater and more stable performance than non-treated membranes during 300 h of treating actual seawater for water purification, in which a higher water flux and lower permeate conductivity have been reported. 3.3. Structure and Chemical Composition of the Modified Membranes Microscopy using FESEM and EDX analyses was conducted in order to determine the morphology of the surface, cross-section and chemical composition of the pristine, coated and plasma-treated membranes, both before and after the stability tests. The FESEM images of the surface and cross-section of the different analysed membranes before the stability tests are shown in Figure 7 . The surface morphology of the Coat-TEOS and Coat-APTES membranes ( Figure 7 b1,c1) was similar to that of the p-PVDF ( Figure 7 a1), and a surface coating layer covering the membrane surface was not observed. In addition, the measured surface porosities of the pristine and coated membranes were similar, with values around 10% ( Table 4 ). The surface morphology of the plasma-treated membranes ( Figure 7 d1,e1) was similar to that of the pristine PVDF when a low plasma power and time of 5 W and 5 min were applied, respectively, independently of using TEOS or APTES as SiP. In contrast, at a higher plasma power and time of 15 W and 15 min, respectively, the surface became a pillar-like structure, resulting in a rougher and more open surface structure with larger pore sizes ( Figure 7 d1,e1), and the surface porosity increased to approximately 15 % ( Table 4 ) for both PO 2 -TEOS and PO 2 -APTES. This increase in the surface roughness also contributed to the increase of the WCAs, leading to superhydrophobic membranes in which the droplets easily rolled off, suggesting a Cassie state, which is desirable for wetting and fouling resistant surfaces [ 33 , 55 ]. This severe change at the surface was explained by the etching effect of plasma treatment, especially at a plasma time higher than or equal to 15 min [ 46 ]. Different authors have reported similar observations regarding surface morphology changes after plasma treatment. Lin et al. [ 43 ] reported an increase of the pores on the surface caused by etching during a plasma treatment with methane at a power higher than 50 W and time less than 5 min, and they suggested that the formation and breakage rates of C-F bonds were equal for plasma time periods higher than 5 min. Liang et al. [ 61 ] also reported an increase of the surface pore size and porosity of PVDF membranes after argon plasma treatment at 18 W and time less than or equal to 120 s, attributed to the etching effect. Yang et al. [ 62 ] obtained a more open structure on PVDF membranes after plasma treatment with carbon tetrafluoride at 150 W and 15 min. Xu et al. [ 60 ] observed an increase in the surface porosity from 13 to 34% after an argon plasma treatment at 30 W and 120 s, and this value declined to 12% after the grafting of an organosilane similar to the FAS used in this work. Jeong et al. [ 4 ] observed that the PVDF surface became a pillar-like structure after an oxygen and carbon tetrafluoride plasma treatment at 62 W for plasma time periods higher than 30 min. The FESEM images of the membrane cross-section before the stability tests were taken and shown in Figure 7 . They focused on the target surface at high magnification. In Supplementary Material (S3) , images were taken at lower magnification to show the entire membrane. The cross-section of the upper layer of p-PVDF ( Figure 7 a2) changed after the coating treatment ( Figure 7 b2,c2), showing a more porous structure. Only in the case of the plasma-treated membranes under soft conditions of 5 W and 5 min ( Figure 7 d2,e2), a dense-like cross-section was observed at the upper surface (≤7 µm), likely owing to the grafting of SiP and FAS, especially when TEOS was used, as observed in our previous study [ 64 ]. In contrast, the upper layers of the plasma-treated membranes at 15 W and 15 min showed a highly porous structure and rougher surface ( Figure 7 d2,e2), in which a surface profile with ridges and valleys was observed, as reported by other authors [ 38 ]. Regarding the thickness of the PVDF membranes, it remained unchanged in values ranging from 120 to 130 µm after the coating treatment and plasma activation and functionalisation ( Supplementary Material (S3) ), as reported by other authors [ 24 ]. After the stability tests, all the membranes experienced a reduction in surface porosity to a value of approximately 5% ( Table 4 ), which was attributed to plastic deformation observed at the surface level ( Figure 8 ), which also contributed to the WCA decrease during the stability test. However, no significant changes were observed on the cross-section with membrane thickness values of approximately 120 µm ( Figure 8 ). This indicated a high membrane bulk stability and mechanical resistance under the tested operational conditions. The surface chemical composition of the pristine, coated and plasma-treated membranes was determined by the EDX, and the atomic ratios of F, O, Si and N with respect to C (% atomic /% atomic ) are shown in Table 5 . The F/C ratio of p-PVDF was 0.92, quite far from the theoretically value of 1.00 for pure PVDF, indicating a raw PVDF membrane with the presence of other organic carbon compounds. After the coating of PVDF, in both Coat-TEOS and Coat-APTES membranes, the F/C ratio slightly increased to 0.95 due to the deposition of FAS. The presence of O, Si and N indicated the deposition of SiPs. A similar deposition grade of FAS and SiP could be inferred independently of the SiP applied since similar atomic ratios were obtained. The highest F/C ratio of 0.99 was obtained for the PO 2 -APTES membrane, indicating a higher grafting grade of FAS with respect to the coated membranes and PO 2 -TEOS. This can help explain the greater hydrophobicity of the membranes treated with oxygen plasma and APTES. Both PO 2 -TEOS and PO 2 -APTES membranes presented slightly higher O/C, Si/C and N/C ratios than Coat-TEOS and Coat-APTES, respectively, indicating a higher presence of SiP molecules because of the generation of active sites on the membrane during the plasma treatment. The higher O/C ratios in the plasma-treated membranes with respect to the coated membranes could also be attributed to the increase in the oxygen content during the plasma treatment [ 48 , 53 , 67 ]. After the stability tests, the F/C ratio decline was more noticeable in the plasma-treated and functionalised membranes ( Table 5 ), indicating a loss of fluorine in accordance with the decrease of the WCA with the time of use. The F/C ratio decreased from 0.95 to 0.80 and 0.99 to 0.85 in the PO 2 -TEOS and PO 2 -APTES membranes, respectively. These outcomes showed a similar decrease of around 15% in the F/C ratio, suggesting that a similar amount of FAS moieties was removed. In addition, the O/C, Si/C and N/C ratios stayed the same or slightly higher after the stability test, which could indicate that the condensation and grafting of SiP led to a stronger interaction/chemical bonding with the membrane than those involved with FAS, hindering the dragging of SiP by the liquid flux. Furthermore, the higher amount of fluorine and SiP on the PO 2 -APTES after the stability tests could explain its higher WCA with respect to the PO 2 -TEOS at the end of the test. The F/C ratio of the coated membranes decreased slightly after the stability test and kept at similar values to that of the pristine membrane, indicating a loss of the coating layer and showing similar WCAs. In addition, the F/C ratio of the plasma-treated membranes was lower than those of the pristine and coated membranes despite their higher WCAs after the stability tests. This result suggests that the higher surface roughness and pore size had a predominant effect on the WCA. From the previous results, new modification protocols should be evaluated in future studies in this direction, with special focus on the stability of grafting compounds on the modified membranes, preventing the removal or dragging of the modifying agents under moderate liquid fluxes at long-term operations. This prolonged useful lifetime of the functionalisation layer would lead to a reduction in the operational costs during large-scale applications."
} | 9,484 |
27759096 | PMC5069546 | pmc | 8,493 | {
"abstract": "Living organisms produce finely tuned biomineral architectures with the aid of biomineral-associated proteins. The functional amino acid residues in these proteins have been previously identified using in vitro and in silico experimentation in different biomineralization systems. However, the investigation in living organisms is limited owing to the difficulty in establishing appropriate genetic techniques. Mms6 protein, isolated from the surface of magnetite crystals synthesized in magnetotactic bacteria, was shown to play a key role in the regulation of crystal morphology. In this study, we have demonstrated a defect in the specific region or substituted acidic amino acid residues in the Mms6 protein for observing their effect on magnetite biomineralization in vivo . Analysis of the gene deletion mutants and transformants of Magnetospirillum magneticum AMB-1 expressing partially truncated Mms6 protein revealed that deletions in the N-terminal or C-terminal regions disrupted proper protein localization to the magnetite surface, resulting in a change in the crystal morphology. Moreover, single amino acid substitutions at Asp123, Glu124, or Glu125 in the C-terminal region of Mms6 clearly indicated that these amino acid residues had a direct impact on magnetite crystal morphology. Thus, these consecutive acidic amino acid residues were found to be core residues regulating magnetite crystal morphology.",
"discussion": "Discussion The magnetite biomineralization process in magnetotactic bacteria is comprised of multiple steps, including vesicle formation 27 39 , assembly of the vesicle into a chain structure along the filament protein 40 41 , sorting of protein 42 , iron transport 43 , redox control in the vesicles 44 45 46 , and crystal formation 29 30 47 . Mms6 protein is a key protein for the crystal formation, and most probably localizes to the magnetosome membrane before or during the crystal nucleation 20 29 31 . Mms6 protein with eliminated GL region, which is predicted to be a transmembrane region, was abolished from the magnetosome membrane. The subcellular localization analysis indicated that elimination of this putative transmembrane region resulted in the change in localization of the Mms6 protein. Alternatively, the Mms6 variants are thought to be digested by endogenous proteases prior to the localization of the protein to the proper position in the magnetosome membrane 48 because the His-tag fused proteins were also absent in the cytoplasm and cell membrane fractions. Moreover, elimination of the C-terminal region influenced the localization of proteins onto the surface of magnetite crystals. This suggests that direct association of Mms6 to the crystal surface is also important for the proper localization of this protein. The interactions between the protein and crystal surface are believed to be important for biomineral formation 49 . The interaction of biomineral protein with the specific crystal face inhibits the crystal growth and is known to control the biomineral morphology 8 . Investigations into the possible mechanisms of interaction between several biomineral proteins and crystal surfaces have been proposed. Osteopontin is specifically adsorbed onto {100} faces of calcium oxalate monohydrate as a result of electrostatic interactions with its acidic amino acid and phosphate residues 8 . Statherin has a hydroxyapatite recognition region comprising of 15 amino acids in the N-terminal region, which causes the formation of a helical structure 50 . Four amino acid residues in this helical structure are positioned to associate with the hydroxyapatite surface 51 . Analysis of the X-ray crystal structure of osteocalcin revealed that the location of 5 acidic amino acids in the α-helix of this protein corresponds to the placement of calcium ions on the surface of hydroxyapatite 10 . Owing to this structural feature, osteocalcin can recognize a specific crystal surface of hydroxyapatite. These studies attributed the interactions between protein and mineral surfaces to negatively charged residues matching the crystal lattice. The key residues crucial for the function of Mms6 protein were identified by establishing expression vectors for single amino acid-substituted Mms6 and transforming the same into the Δ mms6 strain. The substitution of 3 consecutive acidic amino acids, Asp123, Glu124, and Glu125, impaired the function of Mms6 protein as a morphological regulator of magnetite crystal in magnetotactic bacteria. In vitro analyses of mutated Mms6 suggested that the C-terminal region affects the iron-binding ability and stability of the self-assembled protein structure 34 . The self-assembled C-terminal domains of Mms6 may form a macromolecular interface with appropriate spacing among the acidic amino acid residues, enabling its interaction with the specific crystal surface of magnetite 41 . The deletion of the mms6 gene in M. magneticum AMB-1 resulted in the expression of {110} and high-index faces in the crystals 29 30 . As these faces are uncommon in the crystals from the wild-type strain, Mms6 protein was suggested to be involved in facilitating the formation of {110} face 30 . These suggested that the 3 acidic amino acid residues may directly associate with the {110} crystal face. The crystals synthesized by the Δ mms6 strain were smaller than those produced by the wild-type strain, indicating that Mms6 protein is involved in crystal growth 29 30 . A previous study analyzing the crystallization of hydroxyapatite on a collagen fibril using poly-l-aspartic acid showed that poly-l-aspartic acid induced crystallization; this attributed the local super-saturation of calcium ions surrounding the fibril to the interactions between poly-l-aspartic acid and collagen 52 . Similarly, the amino acid residues of Mms6 protein may also be involved in the development of local super-saturation of iron ions that facilitate the growth of a specific crystal face. In conclusion, both the C-terminal acidic amino acid region and N-terminal hydrophobic region including GL repetitive sequence are crucial for protein conformation and localization on the surface of magnetite crystals. The continuous acidic amino acids, Asp123, Glu124, and Glu125, in the C-terminal region, are key residues affecting the function of Mms6. Consequently, these three consecutive acidic amino acids play an essential role in the growth of crystals and morphological regulation of bacterial magnetite in M. magneticum AMB-1. Although other residues, such as basic amino acids, are important for biomineral-protein interaction 51 , our results correspond to the results of the in vitro and structural analyses of biomineral-protein interactions caused by acidic amino acids. Because acidic proteins commonly exist in biominerals, further experimental analyses may expand our understanding of in vivo biomineralization."
} | 1,725 |
22483202 | null | s2 | 8,494 | {
"abstract": "Just as synthetic organic chemistry once revolutionized the ability of chemists to build molecules (including those that did not exist in nature) following a basic set of design rules, cell-free synthetic biology is beginning to provide an improved toolbox and faster process for not only harnessing but also expanding the chemistry of life. At the interface between chemistry and biology, research in cell-free synthetic systems is proceeding in two different directions: using synthetic biology for synthetic chemistry and using synthetic chemistry to reprogram or mimic biology. In the coming years, the impact of advances inspired by these approaches will make possible the synthesis of nonbiological polymers having new backbone compositions, new chemical properties, new structures, and new functions."
} | 201 |
34673920 | PMC8589332 | pmc | 8,496 | {
"abstract": "Optimising the function of a protein of length N amino acids by directed evolution involves navigating a ‘search space’ of possible sequences of some 20 N . Optimising the expression levels of P proteins that materially affect host performance, each of which might also take 20 (logarithmically spaced) values, implies a similar search space of 20 P . In this combinatorial sense, then, the problems of directed protein evolution and of host engineering are broadly equivalent. In practice, however, they have different means for avoiding the inevitable difficulties of implementation. The spare capacity exhibited in metabolic networks implies that host engineering may admit substantial increases in flux to targets of interest. Thus, we rehearse the relevant issues for those wishing to understand and exploit those modern genome-wide host engineering tools and thinking that have been designed and developed to optimise fluxes towards desirable products in biotechnological processes, with a focus on microbial systems. The aim throughput is ‘making such biology predictable’. Strategies have been aimed at both transcription and translation, especially for regulatory processes that can affect multiple targets. However, because there is a limit on how much protein a cell can produce, increasing k cat in selected targets may be a better strategy than increasing protein expression levels for optimal host engineering.",
"conclusion": "Concluding remarks This has been a purposely high-level overview of some of the possibilities in host engineering predicated on genome-wide analyses. Our main aim has been to draw attention to these developments, and to some of the means by which readers who are only loosely acquainted with them can incorporate these methods into their own work. Take-home messages include\n Host engineering, like directed protein evolution, is a combinatorial search problem. Every enzyme potentially has an optimal expression level for every process. This is not normally its maximal level, since the maximum amount of protein a cell can produce is fixed, including for a given growth rate; protein synthesis is largely a zero-sum game. Changes in the individual concentrations of most enzymes at their operating point necessarily have little effect on fluxes. Some areas of transcription and translation effect a more global control and thus can have greater effects and hence serve as better targets for host engineering. k cat is a much better target for host and protein engineering than is V max. Modern methods of modelling, including deep learning, are beginning to provide the ability to assess desirable changes in silico , as a prelude to developing a fully predictive biology. The success of these messages will be judged by the rapidity with which the strategies they contain are adopted.",
"introduction": "Introduction Much of microbial biotechnology consists conceptually of two main optimisation problems [ 1 ]: (i) deciding which proteins whose levels should be changed, and (ii) by which amounts. The former is ostensibly somewhat simpler, e.g. when a specific enzyme is the target for overproduction, since the assumption is then that the aim is simply the maximal production of the active target (whether intracellularly or in a secreted form). Where the overproduction of a small molecule is the target the optimal levels of individual metabolic network enzyme proteins depend on their specific kinetic properties and the consequent distribution of flux control (e.g. [ 2–8 ]). Since both circumstances ultimately seek to maximise the flux to the product of interest, we shall discuss them both, albeit mostly at a high level. Recognising that many pathways are poorly expressed in their natural hosts we shall be somewhat organism-agnostic [ 9 , 10 ], (though we largely ignore cell-free systems) since we are more interested in the principles (whether microscopic [ 11 ] or macroscopic [ 12 ]) than the minutiae. The possible number of discrete manipulations one can perform on a given system is referred to as the ‘search space’. The overriding issue is that the number of changes one might make scales exponentially with the number of those considered, and is simply astronomical; the trick is to navigate the search space intelligently [ 13 ]. Modern methods, especially those recognising the potential of synthetic biology and host engineering to make ‘anything’ (e.g. [ 14–23 ]), are improving both computational [ 24 ] and experimental approaches. The main means of making such navigation more effective is by seeking to recognise those areas that are most ‘important’ or ‘difficult’ for the problem of interest, and focusing on them; this is generally true of combinatorial search problems (and to illustrate this, a nice example is given by the means by which the Eternity puzzle https://en.wikipedia.org/wiki/Eternity_puzzle was solved).\n\nMethods for genome-wide introduction of mutations Many of the genetic variations that improve the performance of microbial cell factories are not currently possible to design rationally, despite the large degree of genetic knowledge around many platform strains [ 378 ]. This is in large part due to the high degree of epistasis and the combinatorial problems discussed in detail above. While advances in AI are rapidly changing this (see above), improvements in microbial cell factories are presently still in many cases being found by wet laboratory techniques ( Table 2 ) that introduce more-or-less random mutations across the genome and then select for strains with desired properties. These strains can be used directly, or with the plummeting costs of next-generation sequencing, beneficial mutations can be identified revealing new mechanisms and targets for further rational design. A further advantage of random mutagenesis relevant to some applications is that strains generated through random mutagenesis are considered ‘GMO free’, which allows one to avoid legal regulations that have been set up around some kinds of so-called genetically modified organisms [ 445 , 446 ]. Table 2. Example applications of techniques to introduce genome-wide mutations Technique Species Purpose Notes References UV \n Kluyveromyces marxianus \n Improved ethanol production Used an automated platform incorporating UV mutagenesis. [ 480 ] \n Yarrowia lipolytica \n Improved oil production [ 481 ] Chemical mutagenesis \n Chlorella vulgaris \n Light tolerance [ 482 ] \n Brettanomyces bruxellensis \n Reduced production of 4-ethylphenol, an undesirable by-product in wine fermentation [ 483 ] \n Yarrowia lipolytica \n Increased lipid production [ 484 ] \n Lipomyces starkeyi \n Increased production of triacylglycerol [ 485 ] Atmospheric and room temperature Plasma mutagenesis \n Zymomonas mobilis \n Acetic acid tolerance [ 486 ] \n Spirulina platensis \n Astaxanthin production [ 487 ] \n Escherichia coli \n L-lysine production Incorporated a biosensor for cell sorting [ 488 ] \n Actinosynnema pretiosum \n Production of the antibiotic Ansamitocin Used in combination with genome shuffling [ 489 ] \n Streptomyces mobaraensis \n Production of the enzyme transglutaminase epWGA \n S. cerevisiae \n Ethanol tolerance [ 451 ] \n Lactobacillus pentosus \n Lactic acid production [ 490 ] \n Zymomonas mobilis \n Furfural tolerance [ 491 ] \n E. coli \n Butanol tolerance. [ 492 ] Serialised ALE \n Saccharomyces cerevisiae \n β-caryophyllene production [ 493 ] \n Corynebacterium glutamicum \n Glutarate production [ 494 ] \n E. coli \n Ionic liquid tolerance [ 454 ] Continuous ALE \n Methylobacterium extorquens \n Methanol tolerance [ 495 ] \n E. coli \n Conversion to generate all its biomass from CO 2 [ 496 ] GREACE \n E. coli \n Lysine production [ 478 ] \n E. coli \n Butanol tolerance [ 476 ] \n E. coli \n Cadmium resistance [ 497 ] \n S. cerevisiae \n Acetic acid tolerance, reduced acetaldehyde production [ 479 ]"
} | 1,968 |
24547877 | PMC4260171 | pmc | 8,499 | {
"abstract": "To predict effects of climate change and possible feedbacks, it is crucial to understand the mechanisms behind CO 2 responses of biogeochemically relevant phytoplankton species. Previous experiments on the abundant N 2 fixers Trichodesmium demonstrated strong CO 2 responses, which were attributed to an energy reallocation between its carbon (C) and nitrogen (N) acquisition. Pursuing this hypothesis, we manipulated the cellular energy budget by growing Trichodesmium erythraeum IMS101 under different CO 2 partial pressure (pCO 2 ) levels (180, 380, 980 and 1400 µatm) and N sources (N 2 and NO 3 − ). Subsequently, biomass production and the main energy-generating processes (photosynthesis and respiration) and energy-consuming processes (N 2 fixation and C acquisition) were measured. While oxygen fluxes and chlorophyll fluorescence indicated that energy generation and its diurnal cycle was neither affected by pCO 2 nor N source, cells differed in production rates and composition. Elevated pCO 2 increased N 2 fixation and organic C and N contents. The degree of stimulation was higher for nitrogenase activity than for cell contents, indicating a pCO 2 effect on the transfer efficiency from N 2 to biomass. pCO 2 -dependent changes in the diurnal cycle of N 2 fixation correlated well with C affinities, confirming the interactions between N and C acquisition. Regarding effects of the N source, production rates were enhanced in NO 3 − grown cells, which we attribute to the higher N retention and lower ATP demand compared with N 2 fixation. pCO 2 effects on C affinity were less pronounced in NO 3 − users than N 2 fixers. Our study illustrates the necessity to understand energy budgets and fluxes under different environmental conditions for explaining indirect effects of rising pCO 2 .",
"conclusion": "Conclusions Despite the change in energy demand imposed by the different pCO 2 levels and N sources, Trichodesmium showed no alteration in energy producing pathways. Yet, elevated pCO 2 increased cellular POC and PON contents in both N treatments. In N 2 fixers, also nitrogenase activity was strongly enhanced with pCO 2 . Concurrently, CCM activity was downregulated, reducing the use of ATP in active HCO 3 − uptake and allowing its allocation to N 2 fixation. The increase in N 2 fixation was, however, not reflected in PON production, possibly due to an increase in N loss with increasing pCO 2 . In NO 3 − users, the lower N-normalized ATP demand for PON production (Table 3 ) and the better N retention allowed for higher production rates of POC as well as PON compared with N 2 fixers. A calculation of the theoretical energy demands of the measured POC and PON production rates (Table 3 ) revealed that most of the ATP saved from the switch to NO 3 − use (approximately 80%) was invested into increasing the production rates of POC and PON, resulting in almost unaltered ATP demand in our cultures (0.5 ATP residue, Table 3 ). The concomitant increase in the demand of reducing equivalents may have prevented a full implementation of ATP savings into the production of particulate organic matter (POM). The effects of pCO 2 on CCM activity were smaller in NO 3 − users than in N 2 fixers, highlighting the dependence of energy reallocation on the stoichiometric demands in energy equivalents: As NO 3 − assimilation requires only little ATP and is limited by electrons ( Fig. 6 ), any spare ATP arising from downregulation of the CCM would not have the same stimulatory effect as in N 2 fixers. Interestingly, the diurnal pattern in O 2 fluxes usually attributed to protection of nitrogenase was maintained also in NO 3 − grown cells. Further studies are necessary to unravel the effects of different environmental conditions on cellular energy budgets, focusing on energization of the CCM as well as the intricate effects of the NDH-1 4 complex on C use efficiency and energy balance.",
"introduction": "Introduction The release of anthropogenic carbon (C) has caused atmospheric CO 2 partial pressure (pCO 2 ) to increase from 280 to 390 µatm since pre-industrial times and pCO 2 levels are expected to rise further to 750 µatm or even beyond 1000 µatm by the end of this century ( IPCC 2007 , Raupach et al. 2007 ). As CO 2 is taken up by the ocean, seawater CO 2 concentrations increase and pH levels decrease, a phenomenon termed ocean acidification ( Caldeira and Wickett 2003 ). These changes in carbonate chemistry are expected to have diverse effects on marine phytoplankton ( Rost et al. 2008 , Riebesell and Tortell 2011 ). By fixing CO 2 into organic matter, phytoplankton acts as a C sink and plays a potential role as a negative feedback mechanism to atmospheric pCO 2 increase ( Raven and Falkowski 1999 , De La Rocha and Passow 2007 ). In marine ecosystems, phytoplankton productivity is often limited by availability of nitrogen (N). Fixation of atmospheric N 2 by diazotrophic cyanobacteria thus plays a crucial role for primary productivity, particularly in oligotrophic regions of the world ocean. With global change, the marine N cycle is subject to an array of perturbations. On the one hand, increasing deposition of anthropogenic N leads to eutrophication in coastal regions ( Duce et al. 2008 ). On the other hand, the expansion of oxygen minimum zones favors N loss processes such as denitrification and anammox ( Lam and Kuypers 2011 ). Additionally, ocean acidification is expected to decrease marine nitrification rates ( Beman et al. 2011 ), and global warming intensifies stratification and therewith lowers nutrient input into the upper mixed layer ( Doney 2006 ). As the latter processes are likely to decrease the overall NO 3 − availability in the surface ocean, marine N 2 fixation may become more important, helping to restore the global N budget. The cyanobacterium Trichodesmium is considered one of the most important marine N 2 fixers with an estimated contribution of up to 50% to global marine N 2 fixation ( Mahaffey et al. 2005 ). Previous studies found this diazotroph to be exceptionally sensitive to rising pCO 2 . Laboratory experiments exposing cultures to pCO 2 levels projected for the end of this century showed significant increases in the production of particulate organic C and particulate organic nitrogen (POC and PON) as well as N 2 fixation rates ( Barcelos é Ramos et al. 2007 , Hutchins et al. 2007 , 2013, Kranz et al. 2009 , Levitan et al. 2007 ); the magnitude of these effects yet differed strongly between investigations. In several follow-up studies, CO 2 effects on Trichodesmium were found to be strongly modulated by other environmental factors such as iron ( Shi et al. 2012 ) and light ( Kranz et al. 2010 , Levitan et al. 2010 , Garcia et al. 2011 ), the latter highlighting the importance of energy in the modulation of CO 2 effects. Cyanobacteria have to invest a considerable share of energy into the accumulation of inorganic carbon (C i ) by carbon concentrating mechanisms (CCMs) owing to a competing reaction with O 2 and a particularly low CO 2 affinity of their ribulose-1,5-bisphosphate carboxylase/oxygenase (RubisCO) ( Badger et al. 1998 ). The CCM of Trichodesmium involves a distinct assembly of RubisCO and carbonic anhydrase (CA) within carboxysomes, as well as two C i acquisition systems ( Badger et al. 2006 , Price et al. 2008 ). HCO 3 − is taken up via a Na + -dependent HCO 3 − transporter (BicA) whereas diffusive uptake of CO 2 is facilitated by the so-called NDH-1 4 complex, converting CO 2 to HCO 3 − . Next to C i acquisition, another important energy sink in Trichodesmium is N 2 fixation ( Kranz et al. 2011 ). As CCM activity was found to be downregulated at high pCO 2 levels, while N 2 fixation rates were simultaneously increased in this species, a reallocation of energy between C and N 2 fixing pathways has been suggested to fuel the increase in production at high pCO 2 ( Kranz et al. 2010 ). Similarly to RubisCO, nitrogenase is characterized by a high sensitivity toward O 2 ( Falkowski 1997 ). In consequence, while the fixation of N 2 is an extremely energy demanding reaction in itself (Eqn 1 ), diazotrophs face additional costs, which are related to the protection of nitrogenase from photosynthetically evolved O 2 ( Großkopf and LaRoche 2012 ). To separate O 2 evolution from N 2 fixation, Trichodesmium has a tightly regulated diurnal cycle of N 2 fixation and photosynthesis ( Berman-Frank et al. 2001 ), involving daily synthesis and degradation of nitrogenase ( Capone et al. 1990 , Sandh et al. 2009 ) and alternation of photosynthetic activity states ( Küpper et al. 2004 ). Moreover, nitrogenase is expressed only in subsets of cells within filaments, the diazocytes ( Lin et al. 1998 , Berman-Frank et al. 2001 ). As no trans-cellular transport mechanisms for N compounds have been found in Trichodesmium , diazocytes have to release N for use by their neighboring cells ( Mulholland and Capone 2000 ). Uptake mechanisms for N sources other than N 2 are thus indispensable for this species. Laboratory studies have shown that Trichodesmium can use NO 3 − and NH 4 + as well as organic N compounds (glutamine, glutamate or urea; Mulholland et al. 1999 ), all of them requiring different amounts and types of energy equivalents. NO 3 − is taken up in cyanobacteria by high-affinity ATP-dependent transporters and subsequently reduced to NH 4 + in a two-step ferredoxin-dependent reaction catalyzed by nitrate reductase and nitrite reductase ( Flores et al. 2005 , Wang et al. 2000 ) (Eqn 2). (1) (2) In Trichodesmium , N 2 fixation was shown to be inhibited in cultures grown in NO 3 − -containing media ( Ohki et al. 1991 , Fu and Bell 2003 , Holl and Montoya 2005 , Sandh et al. 2011 ). As the uptake and reduction of NO 3 − requires little ATP (Eqn 2), it can be expected that NO 3 − addition to culture media will alter the energy budget of the cells in comparison to N 2 fixing conditions. In this study, Trichodesmium erythraeum IMS101 was acclimated in a matrix of four different pCO 2 levels (ranging from 180 to 1400 µatm) and two different N sources (N 2 and NO 3 − ). In addition to acclimation effects on the level of growth and composition (C, N and pigments), physiological key processes (N 2 fixation, O 2 fluxes and electron transport) were analyzed to improve our understanding of the plasticity in energy and resource allocation under the different energetic requirements imposed by changing environmental conditions.",
"discussion": "Discussion To investigate CO 2 effects on Trichodesmium under altered energy requirements, cultures were grown over a range of different pCO 2 levels under N 2 fixing conditions as well as with NO 3 − , the latter providing a N source with a significantly lower demand in ATP but higher electron requirements (Eqns 1 and 2 ). We also tested NH 4 + as an alternative N source, which would have altered the energy requirements most strongly, lowering the ATP as well as the electron demand compared with N 2 fixation. However, measurements revealed NH 4 + to be toxic to Trichodesmium in concentrations as low as 10 µ M (data not shown), which equaled the average daily N consumption in our cultures, and therefore argued against the applicability in dilute batch incubations. Additionally, concentrations could not be kept stable because of pH-dependent out-gassing of NH 3 (data not shown), rendering it impossible to perform pCO 2 manipulations without simultaneously affecting the N availability. The addition of NO 3 − , on the other hand, had no negative effects on Trichodesmium and was not influenced by pH. As a consequence, we chose NO 3 − to impose a change in the energy status of cells. The change in N usage upon NO 3 − addition was demonstrated by direct measurements of nitrogenase activity as well as by the change in 15 N composition of PON. NO 3 − assimilation resulted in corresponding changes in TA, which were compensated by additions of HCl in equimolar amounts to keep the carbonate system comparable between N 2 fixing and NO 3 − using cultures. Growth rates and Fv/Fm indicate that cells were not stressed in any of the treatments. Please note that the light level applied in the acclimations (150 µmol photons m −2 s −1 ) was below saturation ( Fig. 4 ), imposing a general energy constraint in the cell. Acclimation effects of different pCO 2 levels and N sources The increase of POC as well as PON with pCO 2 ( Fig. 1 ) is in accordance with previous results ( Kranz et al. 2009 , 2010 ). Respective production rates, however, stayed relatively constant due to the concomitant decrease in growth (Table 2 , Fig. 1 ). In other words, cells contained less biomass and divided more quickly at low and medium pCO 2 , while at high pCO 2 , cell quotas were higher and cells divided more slowly. Among the previous studies on T . erythraeum IMS101 testing the effect of pCO 2 levels up to 750 or 1000 µatm, some showed an increase in growth rate with pCO 2 (Barcelos é Ramos et al. 2007 , Levitan et al. 2007 , Kranz et al. 2010 , Garcia et al. 2011 ), while others did not find significant differences ( Hutchins et al. 2007 , Kranz et al. 2009 ). Only one study has tested CO 2 levels comparable to our highest CO 2 treatment, finding that positive effects on growth leveled off between 760 and 1500 µatm in T . erythraeum GBRTRL101 ( Hutchins et al. 2007 ). In a recent study investigating pCO 2 effects under low iron availability representative for oligotrophic oceans, growth rates of Trichodesmium were shown to decrease with pCO 2 (380 vs 750 µatm; Shi et al. 2012 ). The stimulation in PON production in NO 3 − grown cells may be directly attributed to the lower energy requirement for N assimilation (Table 2 , Eqns 1 and 2 ) as well as the fact that filaments are not subject to N loss during transfer from diazocytes to non-diazotrophic cells when grown on NO 3 − . The effect on POC production, however, cannot be directly linked to N assimilation and suggests a more global effect of NO 3 − on the cells' metabolism such as reallocation of energy from N to C assimilation. The cost reduction associated with the switch from N 2 to NO 3 − assimilation is also reflected in the lower POC:PON ratios under these conditions (Table 2 ). Changes in POC:PON ratios have been found in response to nutrient limitation in different phytoplankton ( Sterner and Elser 2002 ). Even though none of our treatments was N limited, the observed changes in POC:PON ratios may simply reflect the higher N assimilation costs in N 2 fixers. O 2 fluxes and electron transport To better understand how the observed effects of pCO 2 and N source on POC and PON were fuelled, we investigated photosynthesis as a measure of energy generation. Concerning treatment effects, neither pCO 2 nor the N source had a significant effect on net O 2 evolution ( Fig. 3 ). However, net O 2 evolution can be uncoupled from energy generation by high rates of O 2 uptake or cyclic electron transport ( Heber 2002 ). Thus, gross and net O 2 fluxes as well as chlorophyll fluorescence need to be considered to obtain a more complete picture of energy generating processes. In all cultures, irrespective of pCO 2 or N source, about one third of the gross O 2 evolved was consumed by dark respiration and light-dependent O 2 uptake ( Fig. 4 ), the latter being indicative for either the classical Mehler reaction ( Mehler 1951 ) or the equivalent reduction of O 2 by flavoproteins ( Helman et al. 2003 ). High rates of O 2 uptake by dark respiration and Mehler reaction have been suggested to protect nitrogenase from O 2 inhibition in Trichodesmium ( Kana 1993 , Carpenter and Roenneberg 1995 , Berman-Frank et al. 2001 , Milligan et al. 2007 ). Rates of Mehler reaction equaled only about 10% of gross O 2 evolution at acclimation light intensity, yet rates increased when light intensities exceeded acclimation levels ( Fig. 4 ). This light dependency could either indicate a role for Mehler reaction in photoprotection and/or reflect the enhanced need for nitrogenase protection at high gross O 2 evolution rates. Moreover, the fact that light-dependent O 2 uptake was not significantly affected by the N source seems surprising, considering the proposed role of Mehler reaction in nitrogenase protection in Trichodesmium ( Milligan et al. 2007 ). Chlorophyll fluorescence showed a light response typical for cyanobacteria, with dark-adapted fluorescence being controlled by respiratory electron flow that introduces electrons into the plastoquinone (PQ) pool (reviewed by Campbell et al. 1998 ). At low light intensities, electron flux through PSI is induced, oxidizing the PQ pool and thereby increasing Fv′/Fm′ and decreasing Q A re-oxidation time ( τ ) ( Fig. 4 ). When light intensities increase beyond acclimation light, input of excitation energy can become higher than the cells' capacity of ferredoxin re-oxidation, making cells vulnerable to photodamage. However, being adapted to high and variable light regimes, Trichodesmium employs effective photoprotective mechanisms ( Breitbarth et al. 2008 , Andresen et al. 2009 ). First of all, state transitions lead to a re-arrangement of phycobilisomes toward PSI, decreasing the PSII functional absorption cross section ( σ ) and therewith Fv′/Fm′ ( Fig. 4 ). Second, the enhanced rates of Mehler reaction dissipate electrons at high light ( Fig. 4 ). The effectiveness of these photoprotective mechanisms is reflected in a decreasing Q A re-oxidation time ( τ ) whilst gross O 2 evolution increases with light ( Fig. 4 ). To cover the high ATP demand of N 2 fixation (Table 3 ; Eqn 1 ), Trichodesmium depends on high rates of cyclic electron transport and Mehler reaction, increasing the ATP:NADPH ratio beyond that provided by linear photosynthetic electron transport. High rates of cyclic electron transport have been proposed to result in chemical reduction of the PQ pool, increasing re-oxidation time of Q A ( Berman-Frank et al. 2001 ). Assuming that cells adjust their energy generation closely to their needs, we expected the treatment-dependent differences in energy demand to be reflected in chlorophyll fluorescence. Contrary to our assumption, none of the fluorescence parameters measured was affected by pCO 2 or N source with the exception of a small pCO 2 effect on functional absorption cross section of PSII ( σ , Fig. 5 ). Table 3 Theoretical ATP and electron (e − ) costs of cellular processes and costs calculated for the observed POC and PON production rates under two different N sources (N 2 and NO 3 − ). Theoretical demands of C fixation, CCM and POC production were normalized to mol N using the average POC:PON ratio measured in the experiment. CCM costs are based on 80% HCO 3 − use and a transport cost of 0.5 mol ATP per mol HCO 3 − , assuming 50% CO 2 leakage. Costs of NO 3 − assimilation include 1 mol ATP for uptake. Loss of fixed nitrogen (e.g. NH 4 + ) is not accounted for. Please note that numbers given do not include costs for synthesis of enzymes and transporters, which would significantly increase the estimates for fixation of carbon as well as nitrogen ( Brown et al. 2008 ). POM (particulate organic matter) is the sum of POC and PON. Process Unit ATP e − ATP:NADPH + H + Reference C fixation mol (mol N) −1 14 19 1.5 Allen 2002 CCM mol (mol N) −1 4 0 Hopkinson et al. 2011 POC production mol (mol N) −1 18 19 1.9 N 2 assimilation to NH 4 + mol (mol N) −1 8 4 4.0 Flores and Herrero 1994 NO 3 − assimilation to NH 4 + mol (mol N) −1 1 8 0.3 Flores et al. 2005 NH 4 + assimilation to glutamate mol (mol N) −1 1 2 1.0 Flores et al. 2005 PON production N 2 fixer mol (mol N) −1 9 6 3.0 PON production NO 3 − user mol (mol N) −1 2 10 0.4 POC production measured in N 2 fixer µmol (µg chl a) −1 day −1 5.9 6.2 1.9 POC production measured in NO 3 − user µmol (µg chl a) −1 day −1 7.5 7.9 1.9 POC production difference NO 3 − vs N 2 µmol (µg chl a) −1 day −1 1.6 1.7 PON production measured in N 2 fixer µmol (µg chl a) −1 day −1 2.9 1.9 3.0 PON production measured in NO 3 − user µmol (µg chl a) −1 day −1 0.8 4.2 0.4 PON production difference NO 3 − vs N 2 µmol (µg chl a) −1 day −1 −2.0 2.3 Total POM production in N 2 fixer µmol (µg chl a) −1 day −1 8.8 8.2 2.2 Total POM production in NO 3 − user µmol (µg chl a) −1 day −1 8.3 12.1 1.4 POM production difference NO 3 − vs N 2 µmol (µg chl a) −1 day −1 −0.5 3.9 Regarding the diurnal cycle, there was a characteristic downregulation of maximal net photosynthesis as well as Fv/Fm during midday, which has been shown previously in Trichodesmium as part of the cells' mechanisms to reduce O 2 concentrations during the phase of highest N 2 fixation ( Berman-Frank et al. 2001 ). In the morning, highly efficient electron transport was indicated by high Fv/Fm and a large PSII functional absorption cross section ( Fig. 5 ), which is in line with the high gross O 2 evolution (data not shown). Dark respiration, as indicated by τ dark and O 2 flux measurements, was lowest in the morning, while rates of Mehler reaction were at their maximum. Later during the day, rates of photosynthetic electron transport decreased, reflected by lower Fv/Fm, functional absorption cross section of PSII ( σ ), O 2 evolution as well as Mehler reaction, while dark respiration increased. Interestingly, the diurnal cycle of O 2 evolution and uptake as well as electron transport was maintained also in NO 3 − grown cultures. Studies on the diurnal cycle of nitrogenase protein abundance in Trichodesmium showed that nitrogenase is synthesized de novo every day ( Zehr et al. 1996 ), resulting in a significant energy demand for protein synthesis ( Brown et al. 2008 ). Nitrogenase was found to be synthesized, yet not activated by post-translational modification, in cells grown even at high levels of NO 3 − ( Ohki et al. 1991 ). These findings suggest that although nitrogenase was not active, NO 3 − grown cells in our study may still have invested a considerable amount of energy for synthesis of nitrogenase. This would cause similar energy requirements as well as protection of nitrogenase from O 2 also in NO 3 − grown cells (i.e. O 2 consumption by dark respiration and Mehler reaction as well as downregulation of photosynthesis during midday), explaining the lack of N effects on chlorophyll fluorescence and O 2 fluxes observed in our study. There is, however, also data suggesting significantly lower expression levels of nitrogenase subunits NifK and NifH in NO 3 − grown cells ( Sandh et al. 2011 ). In summary, the lack of a clear pCO 2 or N effect on photosynthesis, dark respiration or Mehler reaction confirms that there was no difference in energy generation (ATP and reducing equivalents). The observed treatment effects on contents and production of POC and PON can thus not be explained by differences in the overall energy availability, indicating potential changes down-stream of the electron transport chain. To identify alterations in the energy consuming processes we therefore measured rates of N 2 fixation and C acquisition. N 2 fixation In agreement with previous results ( Kranz et al. 2010 ), a characteristic change in the diurnal pattern of N 2 fixation was observed at elevated pCO 2 , with the phase of high nitrogenase activity being prolonged toward the evening ( Fig. 2 ). Although integrated daily N 2 fixation rates increased by as much as 60% between 380 and 1400 µatm pCO 2 , PON production was not significantly affected by the different pCO 2 levels. The ARA used for estimating N 2 fixation rates gives a measure of the maximal nitrogenase enzyme activity under the respective assay conditions (approximating gross N 2 fixation) while PON production rates reflect how much N is ultimately incorporated into the cells (approximating net N 2 fixation). While there are indications that a considerable share of fixed N is lost before incorporation into PON ( Mulholland and Capone 2000 , Mulholland 2007 ), significant uncertainties remain with respect to the absolute values due to methodological issues ( Mulholland and Capone 2001 and references therein). It also has to be noted that, in contrast to acetylene reduction during ARA, actual N 2 fixation is dependent on ammonium consumption by downstream metabolism (e.g. Herrero et al. 2001 ). However, interpretation of trends within results of each of the methods should be valid. In accordance with our findings on CO 2 sensitivity, previous studies found ARA-based estimates of N 2 fixation to increase more strongly with pCO 2 than estimates of PON production based on cell quotas or 15 N fixation ( Kranz et al. 2010 , Garcia et al. 2011 ). In the natural environment, N release by Trichodesmium has been suggested to provide an important N source for a range of associated organisms ( Mulholland and Capone 2000 , Mulholland 2007 ), which may be enhanced under elevated pCO 2 according to our data. The high assimilation costs and unavoidable N loss in N 2 fixers impose higher energy requirements compared with NO 3 − users, especially under elevated pCO 2 . As all treatments, however, showed the same energy generation, we expect changes in other energy sinks. Inorganic C acquisition Acquisition of inorganic C constitutes a major energy sink in Trichodesmium due to the high CCM activities required to compensate for the poor CO 2 affinity of its RubisCO ( Kranz et al. 2009 ). Similarly to O 2 and electron fluxes as well as N 2 fixation, also the affinity for inorganic C was subject to a strong diurnal cycle ( Fig. 3 ), which was previously described by Kranz et al. (2009) . The high affinity for inorganic C in the mornings observed in all treatments is in line with the high rates of photosynthesis discussed above. The overall lower affinities at high pCO 2 , especially during the second half of the day, suggest significantly lower operational costs for the CCM which, in turn, allow for the enhanced N 2 fixation observed ( Figs 2 and 3 ). These CO 2 -dependent changes in affinities and the anti-correlation with N 2 fixation are in agreement with previous results ( Kranz et al. 2010 ). The fact that pCO 2 effects are larger in N 2 fixers than in NO 3 − using cells can be attributed to the higher overall energy requirements of N 2 fixation as well as differences in the stoichiometry of ATP and electron demand ( Fig. 6 ): Provided that the downregulation of CCM activity mainly saves ATP, this surplus energy can be readily used in N 2 fixers to cover the high ATP demand of nitrogenase. In contrast, NO 3 − usage requires only little ATP (for uptake) and is, instead, likely to be limited by the supply of reducing equivalents. Consequently, a downregulation of the CCM in NO 3 − users would not have the same stimulatory effect on PON production as in N 2 fixers. Fig 6 Schematic diagram of the distribution of energy equivalents for PON production under different N sources (N 2 and NO 3 − ). Due to the different requirements of N 2 and NO 3 − assimilation with respect to ATP and electron (e − ) stoichiometry, N 2 fixation is prone to limitation by ATP while NO 3 − assimilation tends to be limited by e − supply. Please note that the ultimate outcome in terms of PON production in the different N treatments is strongly dependent on the ratio of ATP per e − available, which is, in turn, controlled by the ratio of (pseudo-) cyclic to linear e − transport and the use of energy equivalents by other cellular processes. NaR, nitrate reductase; NiR, nitrite reductase. Energy requirements of the CCM are generally dependent on the C sources and uptake mechanisms. CCM operation in Trichodesmium is considered to predominantly consume ATP, as the main C source for this species is HCO 3 − (approximately 80%; Kranz et al. 2009 , 2010 ), which is taken up via a transporter fuelled indirectly by ATP (BicA; Price et al. 2008 ). Such HCO 3 − transporters are dependent on a Na + gradient across the plasma membrane and presumably consume 0.5 mol ATP per mol HCO 3 − (Espie and Kandasamy 1994, Hopkinson et al. 2011 ). Furthermore, the so-called NDH-1 4 complex converts CO 2 to HCO 3 − , thereby driving uptake of CO 2 as well as an internal recycling to prevent CO 2 leakage ( Price et al. 2002 , 2008 ). The reaction is involved in the electron transport chain, receiving electrons from NADPH or ferredoxin that are subsequently transferred to PQ. Intriguingly, NDH-1 4 activity leads to a release of protons into the thylakoid lumen, which in turn increases the pH gradient used for ATP synthesis. The observation that this complex seems to be especially active at high pCO 2 ( Kranz et al. 2010 ) is in line with the increased ATP demand by enhanced N 2 fixation under these conditions ( Fig. 2 ). It has to be noted that the operational costs for BicA and NDH-1 4 are still under debate. Provided that the two CCM components have opposing effects on cellular ATP levels, it is crucial to investigate their differential regulation in response to different environmental conditions."
} | 7,335 |
26156085 | null | s2 | 8,500 | {
"abstract": "We present a nanofluidic device for targeted manipulations in the quarternary structure of single DNA molecules. We demonstrate the folding and unfolding of hairpin-shaped regions, similar to chromatin loops. These loops are stable for minutes at nanochannel junctions. We demonstrate continuous scanning of two DNA segments that occupy a common nanovolume. We present a model governing the stability of loop folds and discuss how the system achieves specific DNA configurations without operator intervention."
} | 127 |
34950156 | PMC8688851 | pmc | 8,501 | {
"abstract": "The interdependence of multiple traits allows plants to perform multiple functions. Acquiring an accurate representation of the interdependence of plant traits could advance our understanding of the adaptative strategies of plants. However, few studies focus on complex relationships among multiple traits. Here, we proposed use of leaf trait networks (LTNs) to capture the complex relationships among traits, allowing us to visualize all relationships and quantify how they differ through network parameters. We established LTNs using six leaf economic traits. It showed that significant differences in LTNs of different life forms and growth forms. The trait relationships of broad-leaved trees were tighter than conifers; thus, broad-leaved trees could be more efficient than conifers. The trait relationships of shrubs were tighter than trees because shrubs require multiple traits to co-operate efficiently to perform multiple functions for thriving in limited resources. Furthermore, leaf nitrogen concentration and life span had the highest centrality in LTNs; consequently, the environmental selection of these two traits might impact the whole phenotype. In conclusion, LTNs are useful tools for identifying key traits and quantifying the interdependence of multiple traits.",
"conclusion": "Conclusion Leaf trait networks provide an effective approach to explore how plants respond to the environment, with many promising applications.",
"introduction": "Introduction Plant functional traits are defined as morpho-physio-phenological traits that indirectly impact fitness via their effects on growth, reproduction, and survival ( Violle et al., 2007 ; He et al., 2018 ; Liu et al., 2021 ). Functional traits are not independent of each other, and their relationships are often represented by positive and negative correlations and allometry, resulting from the different biomechanical and physiological requirements of plants ( Freschet et al., 2015 ). Bivariate trait relationships mainly arise from three different reasons ( Sack et al., 2013 ) (1) direct mechanistic (i.e., physiological structure function relationships), where for instance the size or number of a given structure determines the physiological output of a process; (2) optimal design, in which each trait independently contributes structurally to an overarching function ( Sack and Holbrook, 2006 ); and (3) concerted convergence, in which each trait contributes independently to advantage in a given environment ( Givnish et al., 2005 ). In particular, bivariate trait relationships have been tested from the species to the community level, and from the local to global scale ( Bruelheide et al., 2018 ). However, many functional traits interact with each other and jointly optimize functioning, allowing plants to apply multiple strategies for environmental adaptation, resource competition, and development ( Figure 1A ; Diaz et al., 2016 ; La Riva et al., 2016 ; Bruelheide et al., 2018 ; Liu et al., 2019 ). Thus, focusing on the interdependence of multiple traits, rather than bivariate trait relationships, could provide more realistic insights into how plants adapt to their environment. FIGURE 1 Theoretical basi s and method used to calculate leaf trait networks (LTNs). Multiple leaf traits jointly interact with each other to adapt to the environment or to optimize leaf functions. Integrative LTNs could help capture highly complex relationships among different traits and explore the underlying strategies of plants (A,B) . Considering that plants can adjust their relationships through strength and distance, actual LTNs are shown in panel (C) . LTNs can be represented as a set of nodes (circles) connected by edges (lines) (D) . There are many ways to show the interdependence of multiple traits. Correlation matrices and heatmaps are often used to describe trait–trait interrelationships ( Reich et al., 1999 ); however, both these approaches are limited to quantifying bivariate trait relationships. Structural equation models can be used to study the interrelationships of multiple traits ( Vile et al., 2006 ); however, such models are often used to analyze directed relationships (e.g., cause and effect). Yet, most traits are equal in status. Principal component analysis has been used to analyze multiple traits and group them as independent units by clustering. Within this framework, ecologists strive to reduce the number of linkages among multiple traits to a few axes of variation, also termed “spectra” or “leading dimensions” ( Westoby et al., 2002 ; Diaz et al., 2004 ; Laughlin, 2014 ). Furthermore, leaf economic traits and hydraulic traits are decoupled in tropical-subtropical forests ( Li et al., 2015 ). However, dimensionality reduction and clustering methods only qualitatively (not quantitatively) describe the interrelationships of traits within groups and among groups. Network analysis seems to be an effective solution to this problem. A few studies have applied network visualization to show how traits are correlated ( Poorter et al., 2013 , 2014 ; Mason and Donovan, 2015 ; Schneider et al., 2017 ). In addition to providing a tool for network visualization, trait networks also could capture variation in the interdependency of traits, allowing the important traits of plants to be distinguished using network parameters ( Messier et al., 2017 ; Kleyer et al., 2019 ; He et al., 2020 ). Recently, Kleyer et al. (2019) used network parameters to explore the relationships among plant traits. The authors showed that stem mass and stem-specific length are “hub” traits, meaning that they are correlated with most traits. He et al. (2020) pointed out that plant trait network is an effective method to explore the complex relationship between multiple plant traits, and made some prospects for the application of trait network. Network analysis has rigorous network parameters ( Table 1 ), with this approach potentially providing a higher resolution of the complex relationships among multiple leaf traits. TABLE 1 Key parameters of leaf trait networks (LTNs). Parameters Definition Ecological significance Overall parameters Edge density ( ED ) ED was the ratio of the sum of actual weighted edges to the maximum possible weighted edges. A network with higher ED may allow for the more efficient acquisition and mobilization of resources (as all traits are connected with all other traits) \n \n Diameter ( D ) D refers to the maximum shortest path length between any two connected traits in the network. Higher D represent higher independence among any plant traits \n \n Average path length ( AL ) AL was the mean shortest path between all traits in the network. Higher AL represent stronger interdependence among plant traits \n \n Individual parameters Degree ( k ) k was defined as the sum of edges that connect the focal node traits to other nodes, and the number of connections and the strength of relationships influence the degree of trait. Traits with higher k favor the efficient use and acquisition of resources within and across plant tissues \n \n Closeness ( C ) C was defined as the reciprocal of the mean shortest path between a focal node trait and all other nodes in LTNs. Traits with higher C refer to the traits closely related to other traits in the network \n \n Betweenness ( B ) B was defined by the number of shortest paths going through a focal node trait. Traits with higher B values could serve as a broker in the network The worldwide leaf economics spectrum consists of leaf chemical, structural, and physiological traits. This spectrum was used to show that fast-growing species have higher photosynthetic rates and nitrogen concentrations than slow-growing species, which have higher leaf mass per area and higher leaf longevity ( Wright et al., 2004 ; Wright and Suttongrier, 2012 ). Fives models were used to explain these patterns (two based on structural allocation, two on venation networks, and one on resource allocation to cell walls and cell contents). Each model yielded different explanations for the correlation between these functional traits ( Benjamin et al., 2015 ). Using graph theoretic methods and structural equation modeling, Shipley et al. (2006) showed that the trade-off strategy of plants may lead to a certain quantitative relationship among leaf mass per area (LMA), photosynthetic assimilation rate (A mass ), and leaf lifespan (LL). The origins of the bivariate trait relationships between leaf economic traits are controversial; however, their bivariate trait relationships have been recorded at multiple scales ( Wright et al., 2005 ; Wright and Suttongrier, 2012 ). Because all leaf economic traits are closely correlated with photosynthesis and productivity, it is necessary to quantify variation in the interdependency of leaf economic traits. Mediating the trade-off between cost and benefit, leaf economic traits interactively and jointly optimize photosynthesis; consequently, their complex relationships are expected to form a huge network ( Figure 1B ). This network is represented as a set of nodes (traits) connected by edges (bivariate trait relationships). The width and length of edges are important for network analysis ( Figure 1D ), and given those bivariate trait relationships between leaf economic traits are observed from local to global scales, weighted trait networks must be established ( Figure 1C ). This approach facilitates the accurate expression and measurement of the interdependency of leaf economic traits. Specifically, here we aimed: (1) construct leaf trait networks (LTNs) using leaf economic traits and reveal their complex relationships; (2) explore differences in the interdependence of multiple traits among different growth forms and life forms; and (3) identify the key traits among six leaf economic traits.",
"discussion": "Discussion Leaf Trait Networks Provide an Effective Approach for Exploring Complex Relationships Among Leaf Economic Traits The interdependency of multiple traits is the basis of multiple functions. Due to environmental pressure and plant trade-off strategy, there will be a certain quantitative relationship between traits with different functions. Many relationships were integrated into leaf organs and plant level to form a complex and orderly trade-off relationship network of economic spectrum traits. Throughout the relevant studies of leaf economic spectrum ( Wright et al., 2004 , 2005 ; Shipley et al., 2006 ; Osnas et al., 2013 ), it is always inseparable from the exploration of the pattern of leaf functional traits and the trade-off relationship between them, in other words, the complex and stable “economic” strategies and relationships between functional traits, It is the basis and starting point of leaf economics spectrum research, and functional traits are the nodes of these relationship networks. Therefore, the network analysis was used to explore the complex relationships among plant traits here. There are some minor, but fundamental differences between LTNs and traditional network analysis. For example, the microbial network is used to explore how soil microorganisms coexist, whereas leaf traits are permanent and cannot be removed from plants ( Wang et al., 2018 ). In transportation networks, the distance between airports/railway stations is real and measurable, whereas the distance between leaf traits is difficult to quantify ( Wang et al., 2011 ). Studies conducted within the last 10 years established the basis of the plant trait network, completing the visualization of interdependent relationships among multiple traits ( Poorter et al., 2013 , 2014 ; Sack et al., 2013 ; Mason and Donovan, 2015 ; Messier et al., 2017 ; Kleyer et al., 2019 ; He et al., 2020 ). However, the concept of using the parameters of a network in an LTN to quantify variation in the interdependency of multiple traits is novel and effective. Other fields of research have already developed methods to construct networks and evaluate associated parameters. For instance, a special website for network structure and parameter analysis has been developed in the field of molecular ecology ( Deng et al., 2012 ). However, the edges of LTNs differ from those of the microbial network and transportation network. Thus, we weighted the network with the absolute value of the correlation coefficient. There is general consensus that the more similar the traits, the closer the distance between them ( Kleyer et al., 2019 ). However, the calculation of the distance between traits is a challenge. Some scholars used unweighted approaches to construct an unweighted network that connects traits ( Flores-Moreno et al., 2019 ). Studies within the last 5 years used the reciprocal of the correlation coefficient as a proxy for the distance between multiple traits ( Kleyer et al., 2019 ); Here, we explored using the Euclidean distance under the principal component of traits as the distance between traits. Within this framework, we calculated the relevant network parameters. We recommend that this approach may be widely implemented in future studies investigating plant trait networks. Many studies compared and analyzed the indication intensity of two types of standardized traits on plant function, the correlation between traits and the relationship with environmental factors ( Osnas et al., 2013 ; Westoby et al., 2013 ). In this study, variation in network parameters and node parameters was used to quantify differences between mass-based LTNs and area-based LTNs. Mass-based LTNs had a higher edge density , diameter , and shorter average path length than area-based LTNs. Therefore, leaf traits appear to be more strongly coordinated on a mass basis than on an area basis. This result supported those of previous studies ( Wright et al., 2004 ). It might be attributed to the LMA-LL spectrum being related to mass-based nutrient concentrations ( Wright et al., 2004 ). Leaf Lifespan and Leaf Nitrogen Are the Key Traits in Leaf Trait Networks The environmental selection of functional traits with high centrality in LTNs may affect the whole phenotype. Consequently, it is necessary to identify the “key traits” in complex relationships among multiple traits. The “key traits” in LTNs might play important roles in regulating critical functioning or might be involved in regulating key functions, strongly influencing higher-level properties (e. g., fitness) ( Koschützki and Schreiber, 2008 ). In this study, we showed that the degree of LL was the highest in mass-based LTNs; thus, LL was the “hub trait.” The cost of constructing carbon and carbon gain in leaves is directly related to LL ( Reich et al., 1992 , 1999 ), with LL likely being determined by LMA mechanistically. In other words, a higher LMA facilitates a higher LL owing to the higher carbon mass per area, or LL may also be associated with A mass . For instance, at the whole-plant scale, non-optimal resource use might arise if high-performance leaves lived long enough to experience self-shading from canopy growth ( Ackerly and Bazzaz, 1995 ). Thus, the environmental selection of LL might strongly limit the variability of other leaf economic traits. We also found that N mass had the highest closeness and betweenness ; thus, N mass acts like a bridge in LTNs, linking other leaf economic traits. This phenomenon might be attributed to nitrogen being allocated to cell walls and Rubisco ( Onoda et al., 2017 ), mediating the trade-off between the structure and physiology of leaves, to some extent. Variation in Leaf Trait Networks Among Growth Forms and Life Forms Variation in the network parameters of LTNs could quantify the interdependence of multiple traits. Compared with the LTNs of trees, the LTNs of shrubs had higher edge density , shorter diameter , and shorter average path length . All these network parameters suggest that the interdependence of leaf economic traits was higher in shrubs than in trees. A higher interdependence among traits might allow for the efficient acquisition and mobilization of resources ( Flores-Moreno et al., 2019 ). Many studies have pointed out that plants with low resource availability likely face stronger selection and, thus, tend to have tighter trait correlations and trade-offs ( Liu et al., 2019 ). For example, leaf economic traits and leaf hydraulic traits are decoupled in humid regions ( Li et al., 2015 ), but are coupled in arid regions ( Yin et al., 2018 ). Compared with trees, the availability of light resources could be limited for understory shrubs, and the availability of water resources could be limited for shrubs of open habitats. Consequently, shrubs adopt a cost-effective strategy that allows leaf economic traits to strongly correlate with each other, facilitating efficient functioning. The interdependence among leaf economic traits of broadleaved trees was higher than that of conifer trees. Therefore, compared with coniferous trees, the higher photosynthetic rate of broad-leaved trees may be the result of stronger interdependence of individual traits and multiple traits. Our results (except for leaf vein traits ( Brodribb and Feild, 2010 )) might provide novel evidence explaining why angiosperms, rather than gymnosperms, dominate the plant world. Future Directions and Challenges for Leaf Trait Networks The network analysis is used to explore complex relationships among global leaf economic traits, and LTNs are established. The network parameters could help us identify key traits and quantify the interdependence of multiple traits. However, several hurdles challenge network optimization. Theoretically, the construction of plant trait network needs more matching trait data from different organs and different plant species. First, measuring and collecting many trait data from different organs is an important premise. Plant trait database TRY 3 is one of the largest databases in the world, which can provide a strong database for the construction of trait networks ( Kattge et al., 2011 , 2020 ; Borgy et al., 2017 ). However, compared with the measured data, many interpolation data may cause large errors. For the construction of PTNs, it is best to use the same method for the measurement of traits as much as possible, even on the same plant ( Westoby et al., 2002 ). Such networks could then be used to obtain an accurate representation of the ability of plants to adapt to various environments in the future. Second, this study only presents a typical example; however, more parameters must be studied and more quantitative methods with ecological significance must be developed. Ultimately, the concept of trait networks could be applied to explore how trait networks of plants: (1) vary across climate zones and different regions; (2) vary along successional gradients, and (3) respond to disturbance and global climate change."
} | 4,720 |
34455069 | null | s2 | 8,502 | {
"abstract": "Despite their simple body plan, stony corals (order Scleractinia, phylum Cnidaria) can produce massive and complex exoskeletal structures in shallow, tropical and subtropical regions of Earth's oceans. The species-specific macromorphologies of their aragonite skeletons suggest a highly coordinated biomineralization process that is rooted in their genomes, and which has persisted across major climatic shifts over the past 400 + million years. The mechanisms by which stony corals produce their skeletons has been the subject of interest for at least the last 160 years, and the pace of understanding the process has increased dramatically in the past decade since the sequencing of the first coral genome in 2011. In this review, we detail what is known to date about the genetic basis of the stony coral biomineralization process, with a focus on advances in the last several years as well as ways that physical and chemical tools can be combined with genetics, and then propose next steps forward for the coming decade."
} | 256 |
33506576 | PMC8048489 | pmc | 8,504 | {
"abstract": "Abstract Disturbance is a key factor shaping ecological communities, but little is understood about how the effects of disturbance processes accumulate over time. When disturbance regimes change, historical processes may influence future community structure, for example, by altering invasibility compared to communities with stable regimes. Here, we use an annual plant model to investigate how the history of disturbance alters invasion success. In particular, we show how two communities can have different outcomes from species introduction, solely due to past differences in disturbance regimes that generated different biotic legacies. We demonstrate that historical differences can enhance or suppress the persistence of introduced species, and that biotic legacies generated by stable disturbance history decay over time, though legacies can persist for unexpectedly long durations. This establishes a formal theoretical foundation for disturbance legacies having profound effects on communities, and highlights the value of further research on the biotic legacies of disturbance.",
"introduction": "INTRODUCTION How disturbance shapes communities has remained a central focus in ecological research for decades (Picket & White 1985 ; Walker, 2012 ), and interactions between disturbance properties (Box 1 ), species traits, and competition have all been investigated theoretically (Levin & Paine 1974 ; Lubchenco 1978 ; Connell 1978 ; Tilman 1994 ; Roxburgh et al . 2004 ; Shea et al . 2004 ; Turner 2010 ; Walker 2012 ; Newman 2019 ). For example, the complex nonlinear interactions between aspects of disturbance and species’ traits are known to have strong effects on diversity (e.g. Miller et al . 2011 ; Liao et al . 2016 ). Box 1 Ontology of disturbance in ecology Disturbance is a general concept in ecology that encompasses many related phenomena, including specific destructive events as well as the recurring process of multiple such events. There are many definitions of disturbance in the literature, and we draw from them to clarify the interrelated concepts and terms associated with disturbance. Though there is inherent ambiguity in this broad topic, carefully parsing key aspects and properties of disturbance can improve clarity when studying how disturbance shapes ecosystems. Disturbance event A particular event at a specified place and time that disrupts ecosystem structure or the physical environment (following Pickett & White 1985 ). This disruption can occur at any level of organization: the ecosystem structure includes the communities, populations, and individuals, and the physical environment includes e.g. abiotic resources, substrate availability, and weather. Disturbance type The physical nature of a disturbance event. Different disturbances such as fires and floods may have similarities in their effects, but each type of disturbance has specific impacts on individuals. The type of the disturbance is closely related to its cause. Disturbance regime The distribution of disturbance events occurring over a given space and time. Empirically, a past collection of events has a joint statistical distribution of the quantified aspects of the disturbance events. Theoretically, regimes can be prescribed several ways, e.g. by defining a probability distribution of events, or a periodic or stochastic process. Dynamically generated disturbances can also be simulated. See Krebs et al . ( 2010 ) for detailed discussion of the term “fire regime.” Aspects of disturbance The distinct and conceptually independent properties of disturbance when considered as a complex, multifaceted concept (Roxburgh et al . 2004 ; Fraterrigo & Russak 2008 ). Both disturbance events and regimes have associated aspects. Events can have: intensity , the vigor of the disturbing force (e.g. Fraterrigo & Rusak 2008 ); duration , the temporal span of effects (e.g. Lake 2000 ); extent , the spatial span of effects (e.g. Miller et al . 2012a ); and timing , the temporal placement measured with respect to season or life cycle (e.g. Miller et al . 2012b ). Regimes have aspects that events cannot, such as autocorrelation , the statistical serial relation of events with themselves (e.g. Liao et al . 2016 for spatial autocorrelation, Garrison et al . 2012 for temporal). Some aspects of regimes have analogs in events. For example, frequency of a regime is the expected number of events per time period, and is analogous to the aspect of time since last disturbance associated with an event. Pace of change , the rate of onset of disturbance effects for an event. E.g. two floods could have the same extent and duration, but trigger different effects and responses depending on their speed of occurrence. For a regime, pace of change indicates that some aspect of the disturbance regime changes over time. Press and pulse disturbance A dichotomy distinguishing discrete and specific disturbance events (pulse) from disturbance that occurs uniformly and generally over time (press). This punctual vs. continual aspect can be treated as a distinction in duration of events. This allows the distinction between repeated acute disturbance and chronic disturbance (McCormick et al . 2015 ). Disturbance cause, effect, response Disturbance shapes ecosystems through a chain of impacts. The cause is the material impetus, the effect is its primary action, and the response is how species react. E.g. a fire could be caused by lightning, the effect is destruction of biomass, and the responses may include resprouting or increased germination. Disturbance history The record of disturbance experienced at a given location over a period of time. This can be theoretically represented by a procedure, or determined empirically as a set of events. This term relates to other concepts of the ecological past, e.g. “ecological memory” and “antecedent effects” (Ogle et al . 2015 ). Disturbance refugia Locations that are disturbed less intensely or less frequently than other areas within the surrounding landscape. These can act as important reservoirs of biotic and abiotic legacies (from Krawchuk et al . 2020 ). Biotic legacy The accumulated biological effects of a disturbance process acting on a community over time. “Material” legacies are the physical effects (e.g. biomass, species abundances, seed banks) and the “informational” legacies are the result of selection and filtering acting on populations subject to disturbance (following Franklin et al . 2000 ; Johnstone et al . 2016 ). Here, the persistent seedbank shaped by a stable regime over time constitutes the biotic legacy of historical disturbance. Abiotic legacy The accumulated material effects of a process acting over time that are not biological in nature. Examples include erosion, changes to stream beds, soil leaching, sediment deposition, etc. Recently, interest in how the effects of disturbance regimes accumulate over time (Seabloom et al . 2020 ) and how changes in disturbance regimes influence invasion success has grown (Theoharides & Dukes 2007 ; Moles et al . 2012 ; Johnstone et al . 2016 ). Over time, disturbance regimes build a biotic legacy (Box 1 ) reflective of that environmental variation through cumulative effects; the history and its legacy may then influence present ecological processes. Importantly, when regimes change, there is a potential for mis‐match between a community shaped by one regime but experiencing another. This may alter niche opportunities, and historical legacies of disturbance may therefore affect both current and future dynamics as the impacts of changes to disturbance regimes unfold. History of disturbance may affect how new species integrate into communities in many ways, and we hypothesise that the persistence or extirpation of introduced species can be better understood if we explicitly account for the history of disturbance and changes in disturbance regimes. Here, we develop a theoretical approach to this emerging research field, and address our basic hypotheses using a model of an annual plant community of two resident species and one invader. While previous work has shown how abiotic legacies can influence communities (Perring et al . 2016 ), our interest is the potential for disturbance history to create biotic legacies, and to understand how these legacies influence invasion success in the absence of abiotic legacies. Although these concepts have been observed repeatedly – and increasingly – in empirical systems (Newman, 2019 ), there is relatively little theory developed on the topic. We ask two fundamental questions about changing disturbance regimes and species introductions in a model annual plant community. First (Q1): do historical changes in disturbance regimes alter present invasion success, compared to introductions taking place in a community operating continually under a stable disturbance regime? Because biotic legacies will vary across communities with different histories, we hypothesise that invaders may be more likely to succeed in a community subject to a recent change in disturbance regime compared to a community that has been operating under a single stable regime for a long time. It is not clear how long effects of historical disturbance regimes may last. For example, recent disturbance events may influence the current abundances of species, but legacies of disturbance may interact with populations over longer timeframes. Thus, our second question is (Q2) : how long does a given historical period of disturbance continue to affect community structure? The annual life cycle is rapid, and populations can change dramatically within a few years. Thus, we hypothesise that the effects of disturbance history will fade relatively rapidly after changing to a new regime. Our findings demonstrate that overlooking disturbance history can lead to erroneous conclusions because two communities comprising the same species with the same disturbance regime can have different invasion outcomes due to differences in disturbance history. We also demonstrate that the effects of disturbance history may endure for remarkably long times. Our results provide a theoretical understanding to inform research examining disturbance and invasive species, and because they are based on exhaustively sampling all frequency‐intensity disturbance regime combinations, we capture all possible disturbance‐mediated dynamics that can be described within the present model context for a given community. Our prior work has shown that insights from this model can also explain features of microbial microcosm systems (Hall et al . 2012 ), and there is potential that our results will be more broadly applicable (Miller et al . 2011 ). We also use our findings to generate empirically testable hypotheses, for example, an increase in disturbance frequency or intensity can enable a species to successfully invade, whereas introduction under a regime with the higher value and a stable history would fail.",
"discussion": "DISCUSSION Disturbance regimes around the world are changing due to rapid global change (Vitousek et al., 1994 ; Vitousek et al., \n 1996 ; Hooper et al . 2012 ; Essl et al . 2015a ). Moreover lags between anthropogenic changes and ecosystem effects are ubiquitous and worsen with time (Essl et al . 2015b ; Komatsu et al . 2019 ). Our work suggests that historical regimes can influence future community invasibility, and that effects of past changes to disturbance regimes can linger due to long‐lasting biotic legacies. Importantly, we show that two systems with the same current disturbance regime may nevertheless have very different fates, solely due to differences in history. Thus, we raise awareness that analyzing community assembly or invasion based on stable processes whose distributions do not change over time may disregard important information about the past and how it shapes future dynamics. We have shown that changing the history of disturbance can alter invasion success (Q1, Fig. 3c ). Since the only difference between Regime Change and Introduction experiments is the disturbance history, this demonstrates that history of disturbance has clear and direct effects on community invasibility through creation of biotic legacies. The Simultaneous experiment displays a rich nonlinear response of diversity to disturbance (Figs. 2a , S10A), and regime changes and invasion result in extremely complex patterns of community change (Figs. 2d , S2D). While we cannot predict performance of a specific system, we now understand the types of behavior that may arise, and demonstrate how historical processes can accumulate to shape ecological communities. In addition to changing invasion success and community structure, we have shown that the effects of disturbance history decay over time (Q2). Fig. 5I shows that 4500 years after a change to a historical regime, the effects of that biotic legacy can still register in community composition, for both positive and negative shifts in disturbance frequency and intensity. Though effects can decay relatively quickly, in some cases they can persist for millennia, and there seems to be no characteristic time scale for this decay. This is surprising, given that the effect of a historical period on the species composition can be seen after a time period more than double that of the original historical regime period. In longer‐lived organisms, there is precedent for even longer legacy effects of history, for example, the climate of 21 000 years ago was found to have a strong influence on patterns of tree species richness in Europe (Svenning & Skov 2007 ). Thus, our results illustrate that transient behavior is often more ecologically meaningful than asymptotic states (e.g. Hastings 2001 ; Hastings 2018 ; Fig. S13 ). Using sub‐models of the present model, we have previously established stability of coexistence and mechanisms thereof (Miller 2011 ), and demonstrated high richness of persisting species (Roxburgh 2004 ). However, for the timescale of a century after shifts in disturbance, species persistence is of primary interest, and stability results may be misleading (e.g. Fukami & Nakajima 2011 , Yamamichi et al . 2014 , Hastings 2018 ). Observational studies have suggested the importance of disturbance history, and our work demonstrates a theoretical basis for strong effects of historical changes to disturbance regimes on community invasibility and community composition. We designed our experiments to rule out other possible explanations: all effects we describe must be due to the effects of disturbance history, and the biotic legacies left by the past processes. The historical aspects we consider are closely related to ‘antecedent effects’ and ‘memory’ of ecological systems, which have also been a growing topic of research, showing promising ability to better understand the future of ecosystems by considering the past as well as the present (Ogle et al . 2015 ). Thus, although it is commonly assumed that ecological processes can be well‐understood using stationary processes and current status, evidence is building that this assumption may be a poor one. Our annual plant model is well‐suited as a tool to investigate our current and future questions, because the persistence of the seedbank allows for cumulative effects over time to be stored, and there is a rich record of theoretical work on annual plant systems (e.g. Ellner 1985 ; Roxburgh et al . 2004 ) that can inform our inquiry. Our results highlighting the importance of legacies also provide avenues for future research by integrating additional processes shown to be potentially important in determining how disturbance affects diversity. For example, disturbance only acts on seed yield in the present model. When disturbance alters other factors such as germination or mortality in this model, different mechanisms of diversity maintenance are engaged (Miller et al . 2012b ). Thus, regime changes may alter invasion success and community structure via additional mechanisms. The regimes studied here are described entirely by two constant parameters ( I, F ). However, variation in intensity, I , and frequency, F , can alter community dynamics (Miller et al . 2019 ) and potentially interact with the formation of legacies. Likewise, we have studied abrupt change in disturbance regime, but gradual changes to disturbance regimes will be common in the coming decades of global change, and species may respond differently to gradual changes, compared to rapid shifts. Histories featuring regime change with an increase in disturbance tend to favor invasion in our model, though not always. In fact, when the ruderal species R is treated as an invader (Fig. S5 ), only minimal effects of disturbance history on invasion success are evident. More research will be necessary to understand the complex interactions between species’ traits and the importance of history, and these issues can be investigated within the present framework. Other factors may affect the influence of disturbance history in annual plant systems; for example, productivity also interacts with disturbance to shape communities (Kondoh 2001 ), and this would enable yet another way for biotic legacies to accumulate. Additionally, regimes can be patchy in space, with local disturbance refugia acting as important sources of biotic and abiotic legacy (Box 1 , Krachuk et al . 2020 ), and allowing for these would likely increase the effects of disturbance history. Moreover we expect that the interplay between ecological filtering and biotic legacies will allow for a variety of additional complex behaviours resulting from regime change and species introduction, because disturbance can act as a selective agent on a pool of potential resident species (e.g. Diaz et al . 1998 ; Grime 2006 ; Gompert et al . 2014 ). Interestingly, invasive species can often form their own legacies, such that effects of the invasion can persist even after the invader is removed (Corbin & D’Antonio 2012 ). Our results on historical effects on invasions may apply to many different plant and non‐plant systems. Although our quantitative results are specific to a certain case, every three‐species community we have simulated has demonstrated effects of historical regimes change on invasion success for some invading species (Figs S5, S6, S9–14), indicating these effects are widespread and not dependent on specific model configurations. We have empirically tested qualitative predictions of our annual plant model on impacts of disturbance on diversity in laboratory microbial experiments (Hall et al . 2012 ), and these results can also help explain results of related studies in microcosms (Kassen et al . 2004 ; Benmayor et al . 2008 ). Another system that is characterised by strong influence of disturbance and can be described by a similar model is the community of macroinvertebrates at hydrothermal vents (Miller et al . 2018 ), which are expected to experience rapid changes to disturbance regimes due to deep‐sea mining (Dunn et al . 2018 ). Our model is also structurally similar to those used to study insect populations that reproduce on an annual cycle and compete through shared parasitoids (Holt & Lawton, 1993 ), and this hints at the potential importance of biotic legacies in insect invasions as well. Finally, our techniques can be used to study effects of historical resource pulses (Yang et al . 2008 ) because when frequency is high, the disturbance regime can be interpreted as describing rare resource pulses that lead to increases in seed yield. Our results show that historical changes to disturbance regimes can alter future community dynamics. They also generate rich areas for future theoretical (described above) and empirical research. Tests of hypotheses in laboratory and field systems will be required to fully assess how our findings apply to real‐world communities. Disturbance is known to play multiple roles in the establishment of exotic species (McIntyre & Lavorel 1994 ; Lake & Leishman 2004 ; Buckley et al . 2007 ; Lockwood et al . 2007 ) and biotic legacies of historical disturbance regimes are anticipated to affect productivity and carbon cycling (Volkova et al . 2018 ), diversity (Lunt & Spooner, 2005 ), nutrient cycling (Johnstone et al . 2010 ), and resilience (e.g. Johnstone et al . 2016 ). We propose that including history of disturbance and resultant biotic legacies in analyses will aid in developing better predictions of species invasions, community dynamics and ecosystem function. Research at this forefront will increase our basic ecological understanding, and also inform applications to conservation biology, invasion ecology, restoration efforts and management. As our global ecosystems experience accelerating change, an understanding of these issues will allow us to build a foundation on which applied work can stand, so that each new problem does not need to be addressed de novo . As such, this research will then help to ameliorate and mitigate the ecological and economic damage that global change causes."
} | 5,303 |
30555443 | PMC6281684 | pmc | 8,505 | {
"abstract": "Electron exchange reactions between microbial cells and solid materials, referred to as extracellular electron transfer (EET), have attracted attention in the fields of microbial physiology, microbial ecology, and biotechnology. Studies of model species of iron-reducing, or equivalently, current-generating bacteria such as Geobacter spp. and Shewanella spp. have revealed that redox-active proteins, especially outer membrane c -type cytochromes (OMCs), play a pivotal role in the EET process. Recent (meta)genomic analyses have revealed that diverse microorganisms that have not been demonstrated to have EET ability also harbor OMC-like proteins, indicating that EET via OMCs could be more widely preserved in microorganisms than originally thought. A methanotrophic bacterium Methylococcus capsulatus (Bath) was reported to harbor multiple OMC genes whose expression is elevated by Cu starvation. However, the physiological role of these genes is unknown. Therefore, in this study, we explored whether M. capsulatus (Bath) displays EET abilities via OMCs. In electrochemical analysis, M. capsulatus (Bath) generated anodic current only when electron donors such as formate were available, and could reduce insoluble iron oxides in the presence of electron donor compounds. Furthermore, the current-generating and iron-reducing activities of M. capsulatus (Bath) cells that were cultured in a Cu-deficient medium, which promotes high levels of OMC expression, were higher than those cultured in a Cu-supplemented medium. Anodic current production by the Cu-deficient cells was significantly suppressed by disruption of MCA0421, a highly expressed OMC gene, and by treatment with carbon monoxide (CO) gas (an inhibitor of c -type cytochromes). Our results provide evidence of EET in M. capsulatus (Bath) and demonstrate the pivotal role of OMCs in this process. This study raises the possibility that EET to solid compounds is a novel survival strategy of methanotrophic bacteria.",
"introduction": "Introduction Extracellular electron transfer (EET) is the process by which some microorganisms exchange intracellular electrons with an extracellular electron donor/acceptor, including naturally occurring metal compounds and artificial electrodes, across the cell membrane ( Lovley, 2008 ; Kato, 2015 ). Microorganisms harboring EET abilities have received considerable attention as the biocatalysts of bioelectrochemical systems, such as microbial fuel cells and microbial electrosynthesis ( Lovley and Nevin, 2013 ; Igarashi and Kato, 2017 ; Sasaki et al., 2018 ). Intensive study of the molecular mechanisms of EET in model microorganisms such as Geobacter sulfurreducens , Shewanella oneidensis , and Acidithiobacillus ferrooxidans , has revealed that redox-active proteins, especially outer membrane c -type cytochromes (OMCs), play an important role in EET ( Weber et al., 2006 ; Castelle et al., 2008 ; Shi et al., 2016 ). Recent studies have demonstrated that phylogenetically and physiologically diverse microorganisms, including sulfate-reducing bacteria, acetogenic bacteria, filamentous Chloroflexi bacteria, and methanogenic archaea, exhibit EET abilities ( Bond and Lovley, 2002 ; Nevin et al., 2010 ; Yamada et al., 2014 ; Kato et al., 2015 ; Deng et al., 2018 ; Kawaichi et al., 2018 ). Furthermore, metagenomics analysis of environmental samples revealed that various uncultured microorganisms, including anaerobic methane-oxidizing archaea, harbor OMC-like genes ( McGlynn et al., 2015 ). Considering that there has been no report that OMCs are used for metabolic reactions other than EET, we can assume that EET via OMCs could be more widely conserved in microorganisms than originally thought. Aerobic methane-oxidizing bacteria (methanotrophs) utilize methane as their sole carbon and energy source, and play a significant role in global climate by mitigating emissions of methane into the atmosphere. Methanotrophs have also received considerable attention as biocatalysts for conversion of methane into valuable chemicals ( Strong et al., 2016 ; Cantera et al., 2018 ; Hwang et al., 2018 ). Methylococcus capsulatus (Bath) is a model methanotroph whose physiology, biochemistry, and genetics have been extensively investigated ( Kelly et al., 2005 ; Hakemian and Rosenzweig, 2007 ). Availability of Cu markedly affects the gene expression and physiology of M. capsulatus (Bath) by a process called the “copper switch” ( Semrau et al., 2018 ). M. capsulatus (Bath) uses two different methane monooxygenases (MMOs) according to the availability of Cu: a Cu-containing particulate enzyme (pMMO) is highly expressed in the presence of Cu, whereas the soluble counterpart (sMMO) predominantly functions in the absence of Cu ( Semrau et al., 2018 ). In addition to the MMO enzymes, some OMC-like genes in M. capsulatus (Bath) are highly expressed in the absence of Cu ( Karlsen et al., 2005 , 2008 ). Furthermore, some of the OMC genes form a gene cluster with a gene homologous to the β-barrel outer membrane protein ( mtrB ) of S. oneidensis , which has been observed to complex with OMCs in the outer membrane ( Larsen and Karlsen, 2016 ). Based on these findings, we considered it likely that M. capsulatus (Bath) possesses EET capability via OMCs; however, no experimental evidence has been provided for the existence of EET capability in methanotrophs, and the physiological functions of their OMCs have not been clarified. In this study, the EET ability of M. capsulatus (Bath) was verified by assessing current-producing and Fe(III)-reducing activities of cells cultured in the presence or absence of Cu. The contribution of OMCs to the EET ability was also verified by investigating the effects of deletion of an OMC gene and supplementation of carbon monoxide (CO) (an inhibitor of cytochromes) on current-producing activities.",
"discussion": "Results and Discussion Effects of Cu Availability on OMC Expression To verify the EET ability of M. capsulatus (Bath), culture conditions that induce different OMC expression levels were examined. Cells were cultured in inorganic basal medium in the presence or absence of 20 μM CuCl 2 until the late logarithmic phase. The expression level of the MCA0421 gene, which encodes a highly expressed OMC ( Karlsen et al., 2005 , 2008 ), was measured by qRT-PCR (Figure 1 ). Genes for two types of MMOs, mmoX and pmoA , whose expressions are up- and down-regulated by Cu starvation, respectively ( Semrau et al., 2018 ), were analyzed as controls. Consistent with the previous studies, mmoX was up-regulated (41 fold, p < 0.05) in the Cu-deficient condition, whereas pmoA was down-regulated (0.32 fold, p < 0.05). MCA0421 gene expression was significantly higher (58 fold, p < 0.05) in the absence of Cu than in the presence of Cu. Therefore, M. capsulatus (Bath) cultured in Cu-deficient and -supplemented conditions were utilized in the following experiments as models of cells highly or poorly expressing OMCs, respectively. FIGURE 1 Effects of Cu availability on the expression of genes for soluble methane monooxygenase (sMMO) ( mmoX ), particulate methane monooxygenase (pMMO) ( pmoA ), and an outer membrane c -type cytochrome (OMC) (MCA0421). Methylococcus capsulatus (Bath) was cultured in the presence and absence of Cu (+Cu and –Cu, respectively), and subjected to quantitative reverse transcription (qRT)-PCR analysis. The expression values of each gene were normalized with that of a house-keeping gene (membrane protein B). The vertical axis represents the expression fold changes of –Cu vs. +Cu. Data are presented as means of triplicate experiments, and error bars represent standard deviation. Ferrihydrite Reduction by M. capsulatus (Bath) To determine the EET capabilities of M. capsulatus (Bath), the reduction of insoluble iron oxides (ferrihydrite) by M. capsulatus (Bath) cell suspensions was assessed. In the presence of M. capsulatus (Bath) cells and an electron donor compound (10 mM formate), ferrihydrite particles that were reddish brown in color turned black (Figure 2A ), indicating occurrence of Fe(III) reduction. M. capsulatus (Bath) cells cultured in Cu-deficient medium (referred to as -Cu cells, which highly expressed OMCs) produced Fe(II) via ferrihydrite reduction in the presence of formate (Figure 2B ). Ferrihydrite reduction by the -Cu cells also occurred in the absence of formate, whereas the reduction activities were much lower. Organic compounds accumulated in the cells during pre-culture may serve as electron donors for Fe(III) reduction. M. capsulatus (Bath) cells cultured in Cu-supplemented medium (referred to as +Cu cells, which poorly expressed OMCs) also reduced ferrihydrite, but the activity was much lower than that by the -Cu cells. Fe(III) reduction by M. capsulatus (Bath) was also observed when methanol was used as the electron donor instead of formate (Supplementary Figure S1 ). The same trend was observed for the effect of Cu deficient pre-culture, while the Fe(III) reduction activity with methanol was lower than that with formate. These results indicate that M. capsulatus (Bath) has EET capability and that their OMCs are involved in the EET process. We also assessed the growth of M. capsulatus (Bath) under the Fe(III)-reducing conditions; however, no growth was observed in the absence of oxygen even when methane or formate was added as an electron donor (data not shown), indicating that EET reactions are not a sufficient means of energy acquisition to support growth of M. capsulatus (Bath). FIGURE 2 Reduction of ferrihydrite by M. capsulatus (Bath). (A) Changes in color of ferrihydrite particles after 15-d incubation in the absence (“No cells”) or presence of M. capsulatus (Bath) cells. Cells were prepared under Cu-deficient (–Cu) or Cu-supplemented (+Cu) conditions. (B) Production of Fe(II) via reduction of ferrihydrite by the –Cu and +Cu cells of M. capsulatus (Bath) in the presence or absence of formate. Data are presented as means of triplicate experiments, and error bars represent standard deviation. Anodic Current Production by M. capsulatus (Bath) To quantitatively evaluate the EET activities of M. capsulatus (Bath), anodic current production (i.e., electron transfer from microbial cells to an anode) was measured in an electrochemical cell (See Materials and Methods). M. capsulatus (Bath) cells prepared from the Cu-deficient culture produced anodic current (maximum current density, 118 ± 20 nA cm -2 ) in the presence of formate (Figures 3A,C ). In contrast, the current density was quite low in the control cultures omitting either bacterial cells or formate (<8 nA cm -2 ). M. capsulatus (Bath) cells prepared from the Cu-supplemented culture also exhibited anodic current production, but the maximum current density (22 ± 2 nA cm -2 ) was significantly lower than that produced by the -Cu cells ( p < 0.05, Figure 3C ). These results are consistent with the Fe(III) reduction assays, and confirm that M. capsulatus (Bath) has EET abilities via OMCs. FIGURE 3 Anodic current production by wild-type (WT) and OMC-deletion mutant (ΔMCA0421) cells of M. capsulatus (Bath). (A) Anodic current production by WT cells in the presence or absence of formate. (B) Anodic current production by WT and ΔMCA0421 cells. (A,B) \n M. capsulatus (Bath) cells were pre-cultured under Cu-deficient (–Cu) or Cu-supplemented (+Cu) conditions. Data plots are representative of at least three independent experiments. (C) Comparison of the maximum current densities calculated from the data of A and B. Data are presented as means of triplicate experiments, and error bars represent standard deviation. Effects of Deletion of an OMC Gene on Anodic Current Production To explore the involvement of OMCs in the EET activities, a mutant strain of M. capsulatus (Bath) that was deficient in an OMC gene was constructed and its current-generating activity was measured. The MCA0421 gene, which encodes the OMC with the highest expression level under Cu-limited conditions in previous reports ( Karlsen et al., 2005 , 2008 ), was selected as the target of disruption. The mutant strain (ΔMCA0421) was constructed by the double crossover method using mutated pheS as the counter-selectable marker (see Materials and Methods; Ishikawa et al., 2018 ). The ΔMCA0421 cells produced anodic currents, and the effects of Cu availability in the pre-cultures were similar to that observed for wild-type (WT) M. capsulatus (Bath) (Figure 3B ). However, the current produced by ΔMCA0421 cells prepared under the Cu-deficient conditions (maximum current density, 86 ± 8 nA cm -2 ) was significantly smaller than that produced by WT cells ( p < 0.05, Figure 3C ). These results indicate that the major OMC gene, MCA0421, at least partly involves in the EET process of M. capsulatus (Bath). Similar to the results in our study, a decrease but not abolition of EET activities was also observed following deletion of a single OMC gene that is a major contributor to EET activities in other bacterial strains such as S. oneidensis and G. sulfurreducens ( Holmes et al., 2006 ; Newton et al., 2009 ). These previous reports propose that multiple OMCs whose expression levels are low in WT are involved in the EET reactions of mutants lacking the major OMCs. Among more than 50 c -type cytochrome genes encoded by the genome of M. capsulatus (Bath) ( Ward et al., 2004 ), 15 have been confirmed by proteome analysis to be OMCs, in that they are expressed and localized on the cell surface ( Larsen and Karlsen, 2016 ). Among the 15 putative OMCs, 10 have been proved to be up-regulated under Cu-deficient conditions ( Larsen and Karlsen, 2016 ). Therefore, we assume that multiple OMCs, including MCA0421, play pivotal roles in the EET reactions of M. capsulatus (Bath). Further investigations are required to clarify the exact roles of multiple OMCs in M. capsulatus (Bath). Effects of Inhibition of Cytochromes on Anodic Current Production To further elucidate the involvement of OMCs in EET, the effect of an inhibitor of cytochromes on current production by M. capsulatus (Bath) was investigated. CO inhibits electron transfer activity of cytochromes by specifically and tightly coordinating to the cytochrome heme groups, and is utilized as the inhibitor of OMC-dependent EET activities in current-generating bacteria ( Shibanuma et al., 2011 ; Ishii et al., 2015 ). Upon the treatment of M. capsulatus (Bath) cells with CO, the anodic current sharply decreased (Figure 4 ). In contrast, the current production was not affected when the cells were treated with Ar. These results strongly support the notion that OMCs play pivotal roles in current production by M. capsulatus (Bath). FIGURE 4 Effects of carbon monoxide (CO) gas on the anodic current production by M. capsulatus (Bath). The electrochemical measurements were performed under N 2 atmosphere, and injection of CO or Ar gas started at the time indicated by the arrowhead. The data plots are representative of at least three independent experiments. Physiological and Ecological Implications This study provides the first experimental evidence of EET capability in methanotrophs. However, the physiological and ecological implications remain unclear. Generally, microorganisms with EET abilities acquire the energy required for growth by using solid compounds such as iron oxides and electrodes as the sole electron acceptor. However, we observed that M. capsulatus (Bath) did not grow under Fe(III)-reducing conditions. Furthermore, the EET activity of M. capsulatus (Bath) was considerably lower than that of known current-producing and/or iron-reducing bacteria. For example, the current producing activity of M. capsulatus (Bath) observed in this study (c.a., 100 nA/cm 2 ) is one order of magnitude lower than that of S. oneidensis (c.a., several μA/cm 2 ) under the same experimental settings ( Liu et al., 2010 ). From these observations, we assume that M. capsulatus (Bath) does not employ EET to utilize extracellular solid compounds as the electron acceptor instead of oxygen. Furthermore, M. capsulatus (Bath) does not have the advantage of using electron acceptors other than oxygen, because it requires molecular oxygen as a substrate for the MMO reactions. Hence, we hypothesize that M. capsulatus (Bath) employs EET to maintain cellular redox homeostasis. It has been reported that M. capsulatus (Bath) compensates for insufficient reducing equivalents by oxidizing molecular hydrogen to supplement methanotrophic growth, but does not grow if molecular hydrogen is the sole electron donor ( Hanczár et al., 2002 ). Similarly, M. capsulatus (Bath) might utilize insoluble solid compounds as an additional electron source (or sink) when the cellular redox state becomes oxidative (or reductive). This hypothesis is consistent with our finding that M. capsulatus (Bath) cells over-expressed OMCs under Cu-deficient conditions. M. capsulatus (Bath) preferentially utilize sMMO instead of pMMO as the methane-oxidizing enzyme under Cu-deficient conditions ( Semrau et al., 2018 ). The methane oxidation reaction requires reducing equivalents for activation of methane ( Keltjens et al., 2014 ). Electrons are supplied to pMMO via reduced quinone or c -cytochrome, whereas sMMO requires electron carriers with higher energy such as NAD(P)H. Hence, we propose that reducing power tends to be insufficient under Cu-deficient conditions, and that this induces expression of OMCs to enable the cells to acquire reducing power via EET. Further investigations are required to clarify the physiological and ecological significance of EET in methanotrophic bacteria."
} | 4,451 |
33584008 | null | s2 | 8,506 | {
"abstract": "Growing algae in darkness for biodiesel production eliminates the challenges of evaporation and light penetration reported for open ponds and the costs and fouling that plague photobioreactors. The current study demonstrated that "
} | 57 |
30062001 | PMC6050626 | pmc | 8,507 | {
"abstract": "Adhesive curli nanofibers, bacterial biofilms' major protein component, were utilized to mediate the growth of MOFs on various polymeric substrates.",
"conclusion": "Conclusions We demonstrated a new technique platform that leverages adhesive CNFs as universal coatings on diverse polymeric substrates or objects with complex geometries to mediate controlled growth of MOF materials. As a result, significantly improved MOF surface coverage was achieved on CNF-coated substrates as compared to that of bare substrates. Further extending this strategy led to the successful growth of ZIF-8 on PFA tubing, 3D printed PR pyramid scaffolds, and PET fabrics. By pre-patterning the PTFE surface with CNFs, a corresponding ZIF-8 pattern replica was obtained with good fidelity and high contrast. Transferring this technology to membrane fabrication afforded a 760 nm ZIF-8 TFC membrane which showed a C 3 H 6 /C 3 H 8 mixed-gas separation factor up to 10, C 3 H 6 permeance up to 1110 GPU and operational stability up to 7 days. We believe that this method will serve as a general route for mediating the interfaces between MOFs and polymers, thus paving the way towards the development of more sophisticated MOF membranes and devices.",
"introduction": "Introduction As a class of extremely diverse materials, metal–organic frameworks (MOFs) have shown great promise in numerous applications. 1 – 5 While certain applications such as gas storage, catalysis, sorption-based separation simply call for MOF powder or pellets, many others demand the growth of MOFs on various substrates including metals, oxides, and polymers. 6 – 8 These substrates serve as mechanical supports to spatially arrange and organize MOF particles in a controlled fashion thus realizing the potential deployment of MOF materials as separation membranes, 9 filters 10 and sensors. 11 Therefore, engineering energy-favorable interfaces that facilitate controlled growth and robust immobilization of MOFs on diverse substrates is of profound technological significance. The heterogeneous nucleation of MOFs preferentially occurs on surfaces with high surface energy. Therefore, the presence of polar functional groups such as –COOH, –OH, and –NH 2 on substrate surfaces is normally essential for nucleation to occur. 12 – 14 For metal and oxide substrates, a complete toolbox can be found in the literature to introduce chemical functionality. 15 – 19 Hydrophobic polymers, however, are considerably more challenging for chemical modification due to the lack of functional moieties as well as their low tolerance to high temperature and aggressive organic solvents. To surmount this challenge, several methods have been developed. These include microwave facilitated synthesis, 20 direct solvothermal synthesis, 21 contra-diffusion, 22 , 23 interfacial polymerization, 8 chemical modification of polymer surfaces, 24 hot-pressing, 25 seeded growth 26 , 27 and polydopamine 28 , 29 or protein 30 mediated growth. While each method has its own advantages, these approaches either require high temperature, the use of aggressive organic solvents, multiple-step fabrication or are only limited to special types of substrates or geometry. Here, we reported a new technical platform that harvests bacterial adhesive curli nanofibers (CNFs) as universal coatings to mediate spatially controlled growth of MOF crystals on diverse substrates or objects with complex geometries in a scalable fashion. Using ZIF-8 31 , 32 as a model system, we demonstrated that significantly improved MOF surface coverage was achieved on all 12 polymeric substrates tested based on this simple approach. Furthermore, by pre-patterning a polytetrafluoroethylene (PTFE) plate with CNFs, MOF growth can be spatially controlled affording MOF patterns with excellent fidelity and contrast. Finally, a thin film composite (TFC) membrane with a sub-micrometer seamless ZIF-8 selective layer grown on a microporous polyvinylidene fluoride (PVDF) support was successfully fabricated which exhibited high C 3 H 6 permeance, reasonable C 3 H 6 /C 3 H 8 separation selectivity and high operation stability. Our technical platform was essentially inspired by E. coli biofilms, which exhibit strong adhesion towards diverse biotic or abiotic interfaces and shield the living bacteria against antibiotics and chemicals with robust extracellular substances. 33 In particular, curli nanofiber networks, major protein components of E. coli biofilms, contribute to the adhesion and integrity of biofilms. 34 CNFs are formed via extracellular self-assembly of CsgA proteins, which themselves are composed of five repeating strand-loop-strand motifs ( Scheme 1a ). Each repeating unit is composed of conserved glycine, glutamine and asparagine residues. The glutamine and asparagine residues are predicted to form a hydrogen bonding network that contributes to the self-assembly and extreme stability of these fibers. 34 Moreover, the hydrophobic amino acids such as alanine, proline and valine endow CNFs with strong adhesion towards hydrophobic surfaces. 35 In addition, residual polar functional groups such as asparagine, serine, glutamine, tyrosine, aspartic acid and histidine, rich in –COOH, –OH, –NH 2 , and imidazole groups, may provide potential nucleation sites for MOF crystallization ( Scheme 1b ). Recently, CNFs have been harvested as organic templates for anchoring various nanomaterials and macromolecules including quantum dots, metal nanoparticles and enzymes. 36 – 39 Finally, similar to other amyloid structures, CNFs are well known for their strong mechanical properties and high chemical/thermal stability. 40 Based on these attributes, we rationalize that adhesive CNFs might serve as universal coatings to mediate the nucleation and growth of MOF materials on diverse substrates. Scheme 1 (a) Structural hierarchy of a typical E. coli biofilm. (b) Three major functions of CsgA proteins and their responsible amino acids. (c) Schematic illustration of typical experimental procedures of adhesive CNF mediated growth of ZIF-8 on polymeric substrates. Briefly, an aqueous solution containing fresh CsgA monomers was first cast onto the substrate. CsgA proteins then spontaneously self-assembled into CNFs affording CNFs-substrates. After immersing CNFs-substrates into ZIF-8 growth solutions, ZIF-8 crystallites started to appear on CNFs. Extended growth period led to a uniform ZIF-8 layer with high surface coverage.",
"discussion": "Results and discussion As a general procedure to apply curli nanofiber coatings, the targeting substrate was immersed in a CsgA protein buffer solution for a given amount of time to allow the self-assembly and deposition of CNFs on the substrate surface. Using this method, we first deposited CNFs onto a transmission electron microscopy (TEM) grid. The TEM image clearly showed that the self-assembly of CsgA protein led to the formation of curli nanofiber networks ( Fig. 1a ). Next, ZIF-8 was selected as a model system to study the feasibility of using CNFs as nucleation centers for MOF growth. After 10 min growth in a methanolic solution of Zn(NO 3 ) 2 ·6H 2 O and 2-methylimidazole (HMIM) on a CNF-coated TEM grid at room temperature, sub-100 nm ZIF-8 particles emerged and interconnected into a fibrous network suggesting that the nucleation process of ZIF-8 was guided by CNFs ( Fig. 1b ). To monitor the early growth process of ZIF-8, atomic force microscopy (AFM) images were taken after 10, 30 and 60 min growth of ZIF-8 on a CNF-coated silicon substrate. Fig. S1a † clearly shows the formation of a fibrous network composed of ZIF-8 nanoparticles after 10 min reaction. At 30 min, more ZIF-8 particles started to emerge (Fig. S1b † ). Meanwhile, the fibrous morphology became less evident. Further extending the growth period to 60 min led to the formation of larger ZIF-8 particles and disappearance of ZIF-8 nanoparticles (Fig. S1c † ) due to Ostwald ripening. Both TEM and AFM results suggest that the nucleation of ZIF-8 did occur preferentially on CNFs. Fig. 1 TEM images of CNFs (a) before and (b) after ZIF-8 mineralization. (c) The photograph of a CNFs-PTFE plate stained by Congo red (inset: the photograph of a bare PTFE plate stained by Congo red). AFM images of CNFs deposited on (d) PSU and (g) PTFE. SEM images of ZIF-8 particles grown on (e) bare PSU, (h) bare PTFE, (f) CNFs-PSU and (i) CNFs-PTFE. Insets in (f) and (i) are the cross-section SEM images of the corresponding samples. Next, polytetrafluoroethylene (PTFE), widely recognized as a superb non-stick substrate, was selected as the first representative example to evaluate the adhesion capability of CNFs. After immersing PTFE in freshly made CsgA/buffer solution for 24 h, a red dye, Congo red, was applied to specifically stain CNFs. Fig. 1c shows that the surface of CNF-coated PTFE (denoted as CNFs-PTFE) displays a uniform red color whereas the bare PTFE is unable to be stained (inset in Fig. 1c ). This indicates that the PTFE surface can be completely and uniformly covered by CNFs on a macroscopic scale. X-ray photoelectron spectroscopy (XPS) was performed to analyze the surface composition of CNFs-PTFE. A peak at ∼400 eV (Fig. S2b † ) corresponding to the 1s orbital of amide nitrogen was observed confirming the presence of CsgA proteins on the PTFE substrate. AFM images of CNF-coated PTFE and polysulfone (PSU) revealed interconnected networks of CNFs ( Fig. 1d and g ) in agreement with TEM observation ( Fig. 1a ). We next turned to evaluate the effectiveness of CNFs in assisting the nucleation and growth of ZIF-8 on PTFE and PSU. CNFs-PTFE and CNFs-PSU were again submerged in methanolic solutions, supplemented with Zn(NO 3 ) 2 ·6H 2 O, HMIM and 1-methylimidazole at room temperature for 24 h. Scanning electron microscopy (SEM) images revealed that ZIF-8 microcrystals with an average size of ∼3 μm were uniformly grown on both substrates with a surface coverage of 87 ± 3% and 99 ± 1% respectively ( Fig. 1f, i and 2c ). In contrast, only 9 ± 6% and 9 ± 7% of surfaces were found to be covered by ZIF-8 for bare PSU and PTFE substrates under the same growth conditions ( Fig. 1e, h and 2c ). Cross-section SEM images showed that the ZIF-8 coatings were composed of a single layer of ZIF-8 crystallites with good continuity (insets in Fig. 1f and i ). Powder X-ray diffraction (PXRD) patterns confirmed that the particles grown on PTFE and CNFs-PTFE were indeed ZIF-8 ( Fig. 2a ). However, the diffraction intensity of ZIF-8 crystals on bare PTFE was significantly weaker than that of the crystals grown on CNFs-PTFE apparently due to less surface coverage, as indicated by the SEM results ( Fig. 1e and h ). Fig. 2 (a) PXRD patterns of different PTFE samples before and after ZIF-8 growth. (b) ZIF-8 surface coverage on bare PTFE and CNFs-PTFE substrates after different chemical pre-treatments. (c) ZIF-8 coverage on non-coated (orange columns) and CNF-coated (green columns) polymeric substrates. The purple line shows the water contact angle values of corresponding substrates (right y axis). To evaluate the universality of this approach, we further applied this method to nine additional commercially available polymeric substrates: polycarbonate (PC), polyethylene terephthalate (PET), polyvinyl chloride (PVC), polyoxymethylene (POM), poly( p -phenylene oxide) (PPO), poly(methyl methacrylate) (PMMA), microporous PVDF, polystyrene (PS), polypropylene (PP), and one as-synthesized polyimide, 4,4′-(hexafluoroisopropylidene)diphthalic anhydride (6FDA)–2,4,6-trimethyl-1,3-phenylenediamine (DAM). After forming curli nanofiber coatings, a decrease in the water contact angle was observed for all 12 substrates (Fig. S16 † ), as a quick indication of the successful surface modification by CNFs. Fig. 2c shows the statistical analysis of ZIF-8 coverage on all 12 substrates with and without curli nanofiber coatings. For non-coated substrates, six relatively more hydrophilic ones (highlighted in the blue region, Fig. 2c ) were relatively easy for ZIF-8 to grow achieving medium ZIF-8 coverage from 41% to 65%. One exception was PSU which only achieves 9 ± 6% coverage by ZIF-8. The more hydrophobic substrates, however, only had 0.9% to 11% ZIF-8 coverage with PS performing slightly better (38% coverage). In contrast, for CNF-coated substrates, significantly improved ZIF-8 coverage in the range from 63% to 99% was observed in all cases. Microporous PVDF, as the most extreme case, exhibited a boost of ZIF-8 coverage from 0.9 ± 0.8% to 93 ± 1%, a striking two orders of magnitude improvement. Moreover, ZIF-8 grown on CNF-coated surfaces also exhibited more even particle distribution than on bare substrates, as revealed by the significantly smaller error bars ( Fig. 2c ). One prominent feature of CNFs is their high chemical and thermal stability towards harsh conditions owing to the strong hydrogen bonding networks. To assess the robustness and stability of curli nanofiber coatings on substrates in the presence of organic solvents, strong acids and strong bases, we treated five CNFs-PTFE samples with dichloromethane (DCM), dimethylformamide (DMF), dimethyl sulfone (DMSO), HCl solution (pH = 2), and NaOH solution (pH = 12) for 48 h, respectively, before ZIF-8 growth. As a result, ZIF-8 coverage of 66 ± 2%, 60 ± 2%, 88 ± 2%, 69 ± 3% and 82 ± 2% was obtained for each sample, respectively ( Fig. 2b ). Despite some variation, these coverage values are still considerably higher than that of the bare PTFE substrate, demonstrating the robustness of this methodology towards aggressive chemical environments. To further evaluate the mechanical stability of CNF mediated ZIF-8 coatings, adhesive tape peel tests were performed on three ZIF-8 coated PS substrates. After one cycle of tape peel test using a standard high-tack tape (VHB, 3M, with an adhesion to steel value of 2600 N m –1 ), micron-sized ZIF-8 particles were completely removed from the CNF-coated PS surface as shown in Fig. S18a and b. † However, sub-100 nm ZIF-8 particles grown on CNF-coated PS were significantly more resistant to mechanical peeling as most of the area on PS remained covered by ZIF-8 (Fig. S18c and d † ) after one peeling cycle. In contrast, ZIF-8 particles gown on bare PS surfaces were completely removed by the adhesive tape (Fig. S18e and f † ) suggesting that the mechanical stability of ZIF-8 coatings benefits tremendously from the adhesion of CNFs. However, such an effect becomes less evident when the particle size increases. Next, we investigated whether this method is suitable for ZIF-8 growth on complex-shaped objects or patterned surfaces with spatial control. To facilitate the visualization of the macroscopic distribution of ZIF-8 particles on different subjects, a red dye, new coccine, was added to the growth solution to incorporate into ZIF-8 crystals during growth so that the crystals displayed a vivid red color. We first tested ZIF-8 growth with three identical pieces of perfluoroalkoxy alkane (PFA) tubing, with tube surfaces fully (tube #2, Fig. 3b ), partially (tube #3, Fig. 3c ) or not (tube #1, Fig. 3a ) coated with CNFs. After an identical growth period, both the interior and exterior of tube #2 were coated with a layer of ZIF-8 crystals ( Fig. 3b ). On tube #3, ZIF-8 was only found on the inner wall, on which CNFs were selectively deposited ( Fig. 3c ). In contrast, ZIF-8 crystals failed to grow on the surface of tube #1 ( Fig. 3a ), suggesting that the CNF modification is critical for the mineralization of ZIF-8 to occur on PFA surfaces. When a 3D printed photosensitive resin (PR) pyramid scaffold was applied for ZIF-8 growth, complete coverage of ZIF-8 particles was found from the exterior to the very interior of the object despite its complex geometry ( Fig. 3f–h ). When this practice was applied on a piece of woven PET fabric, ZIF-8 particles were again observed on every single fiber across the whole fabric ( Fig. 3j and k ). Note that without curli nanofiber coatings, only very low ZIF-8 coverage was achieved (Fig. S19 and S20 † ). By applying a mask to restrict the deposition of CNFs only to the exposed area on PTFE, we successfully fabricated a “ShanghaiTech” university logo comprising ZIF-8 particles with good fidelity and sharp contrast ( Fig. 3l and m ). Collectively, these results suggested that our approach for ZIF-8 growth could be widely applicable to a variety of polymeric substrates with diverse surface chemistry and geometry. Fig. 3 The photographs of ZIF-8 grown on a piece of PFA tubing (a) without pre-treatment, (b) with curli nanofiber coatings and (c) with curli nanofiber coatings only on the inner wall. (d) The SEM image of ZIF-8 particles on the inner wall of the tubing. The photographs of a CNF-coated 3D printed PR pyramid scaffold (e) before and (f and g) after ZIF-8 growth. (h) The SEM image of one of the edge areas on the pyramid. The photographs of a piece of CNF-coated woven PET fabric (i) before and (j) after ZIF-8 growth. (k) The SEM image of the PET fiber coated by ZIF-8 particles. (l) The photograph of ZIF-8 grown on a PTFE plate pre-patterned by CNFs on the surface. (m) The SEM image of the dot on the letter i. To further illustrate its potential application, we exploited this method for the fabrication of MOF-based TFC membranes for propylene/propane separation. Industrial propylene/propane separation using cryogenic distillation is an extremely energy intensive process. Membrane-based separation is regarded as an attractive alternative to surmount this challenge. 41 Recently, ZIF-8 membranes showed exceptionally high performance for propane/propylene separation. 15 , 42 Moreover, MOF membranes, in general, are more resilient to plasticization owing to their structural rigidity. Therefore, they are expected to have higher operation stability than their polymeric counterparts. However, one major roadblock for the industrial deployment of pure MOF membranes is the lack of a reliable and scalable fabrication method for growing a thin yet defect-free layer of MOF on porous polymeric supports. 43 – 45 One unique aspect of curli nanofiber-mediated MOF growth is that while it provides a macroscopically uniform nucleation layer for MOF growth, the loose fibrous network also creates openings for the passage of gases hence maintaining the flux. As a proof-of-concept, we successfully fabricated ZIF-8@CNFs-PVDF TFC membranes through a three-step process. First, ZIF-8 microcrystals were uniformly grown on a curli nanofiber-mediated PVDF microporous membrane with a surface coverage of ∼93% (Fig. S12 † ). Next, a layer-by-layer technique developed by Tanaka et al. 46 was applied to seal the remaining gaps (Fig. S21a, † high-resolution SEM image). Finally, a PDMS gutter layer was applied on top of ZIF-8 to further seal off the pinholes (Fig. S21b, † high-resolution SEM image). The cross-section SEM image showed that a 760 ± 80 nm ZIF-8 layer was uniformly and seamlessly grown on the microporous PVDF substrate covered by an ∼1 μm PDMS skin layer ( Fig. 4a ). No apparent pinholes were observed on the ZIF-8 layer (Fig. S21b † ). A mixed-gas transport experiment was performed with an equimolar binary C 3 H 6 /C 3 H 8 feed using a constant pressure set-up (Fig. S22 † ). Pressure dependent permeation data showed that under low transmembrane pressure (TMP, 3.6 psi), C 3 H 6 and C 3 H 8 permeances of 1110 and 116 GPU were recorded (1 GPU = 3.35 × 10 –10 mol m –2 s –1 Pa –1 ), respectively, leading to a separation factor of 10. With increasing TMP, the permeance of C 3 H 6 gradually decreased eventually reaching 320 GPU at 32 psi. Similar behavior was also observed by Nair et al. 43 which is associated with the increasing saturation of adsorption sites in ZIF-8. The permeance of C 3 H 8 only exhibited a small drop from 116 GPU at 3.6 psi to 71 GPU at 32 psi, resulting in a separation factor of 4.5 at 32 psi ( Fig. 4b ). Such gas transport trends suggest that defects in this TFC membrane are insignificant as otherwise, we would expect a dramatic increase of permeance with increasing feed pressure due to the presence of pinholes. Despite the relatively low separation factor, the thin nature of the ZIF-8 selective layer contributed to a very high C 3 H 6 permeance (1110 GPU at 3.6 psi and 320 GPU at 32 psi) which is a desirable feature in a real industrial setting. Despite the decrease of C 3 H 6 permeance with increasing TMP, the flux increased linearly within the tested pressure range, leading to a 2.8-fold increase in flux at 32 psi as compared to operation at 3.6 psi ( Fig. 4c ). To more rigorously evaluate the reproducibility of the membrane separation performance, we further tested 5 additional membrane samples. The average permeances of C 3 H 6 and C 3 H 8 and selectivity at ∼3.6 psi were 1044 ± 131 GPU, 104 ± 35 GPU and 11 ± 4 respectively (Fig. S23 † ) indicating good reproducibility. We tested the operation stability of the membrane at 32 psi TMP with an equimolar binary C 3 H 6 /C 3 H 8 feed up to 7 days. No apparent decrease of either the permeance or separation factor was observed suggesting the high robustness of this ZIF-8 membrane for C 3 H 6 /C 3 H 8 separation ( Fig. 4d ). Fig. 4 (a) The cross-section SEM image of a ZIF-8 TFC membrane (inset: corresponding schematic illustration). (b) Binary equimolar C 3 H 6 /C 3 H 8 separation factor (left y axis) and permeance (right y axis) of a ZIF-8 TFC membrane as a function of transmembrane pressure; (c) C 3 H 6 and C 3 H 8 flux as a function of transmembrane pressure; (d) permeance and separation factor of a ZIF-8 TFC membrane operated continuously under an equimolar C 3 H 6 /C 3 H 8 mixture feed at 35 °C, 32 psi transmembrane pressure. Notably, this technique platform is not limited to a specific type of MOF. Given its high chemical and thermal stability, this method should also be applicable to the fabrication of MOF membranes requiring high heat and aggressive organic solvents during synthesis. As a preliminary demonstration, a thin film of a zirconium MOF, UiO-66, 47 was grown onto CNFs-PTFE. The synthesis was carried out at 120 °C in dimethylformamide (DMF) with the addition of acetic acid as a modulator. The SEM image showed a uniform layer of UiO-66 crystallites with 88 ± 2% surface coverage (Fig. S24a † ). Investigation of such a membrane for gas separation is underway."
} | 5,612 |
31180806 | null | s2 | 8,512 | {
"abstract": "Cooperation has fascinated biologists since Darwin. How did cooperative behaviors evolve despite the fitness cost to the cooperator? Bacteria have cooperative behaviors that make excellent models to take on this age-old problem from both proximate (molecular) and ultimate (evolutionary) angles. We delve into "
} | 77 |
32479802 | null | s2 | 8,513 | {
"abstract": "Medium-chain length methyl ketones are potential blending fuels due to their cetane numbers and low melting temperatures. Biomanufacturing offers the potential to produce these molecules from renewable resources such as lignocellulosic biomass. In this work, we designed and tested metabolic pathways in Escherichia coli to specifically produce 2-heptanone, 2-nonanone and 2-undecanone. We achieved substantial production of each ketone by introducing chain-length specific acyl-ACP thioesterases, blocking the β-oxidation cycle at an advantageous reaction, and introducing active β-ketoacyl-CoA thioesterases. Using a bioprospecting approach, we identified fifteen homologs of E. coli β-ketoacyl-CoA thioesterase (FadM) and evaluated the in vivo activity of each against various chain length substrates. The FadM variant from Providencia sneebia produced the most 2-heptanone, 2-nonanone, and 2-undecanone, suggesting it has the highest activity on the corresponding β-ketoacyl-CoA substrates. We tested enzyme variants, including acyl-CoA oxidases, thiolases, and bi-functional 3-hydroxyacyl-CoA dehydratases to maximize conversion of fatty acids to β-keto acyl-CoAs for 2-heptanone, 2-nonanone, and 2-undecanone production. In order to address the issue of product loss during fermentation, we applied a 20% (v/v) dodecane layer in the bioreactor and built an external water cooling condenser connecting to the bioreactor heat-transferring condenser coupling to the condenser. Using these modifications, we were able to generate up to 4.4 g/L total medium-chain length methyl ketones."
} | 396 |
30143627 | PMC6109138 | pmc | 8,515 | {
"abstract": "The mussel cuticle, a thin layer that shields byssal threads from environmental exposure, is a model among high-performance coatings for being both hard and hyper-extensible. However, despite avid interest in translating its features into an engineered material, the mechanisms underlying this performance are manifold and incompletely understood. To deepen our understanding of this biomaterial, we explore here the ultrastructural, scratch-resistant, and mechanical features at the submicrometer scale and relate our observations to individual cuticular components. These investigations show that cuticle nanomechanics are governed by granular microinclusions/nanoinclusions, which, contrary to previous interpretations, are three-fold softer than the surrounding matrix. This adaptation, which is found across several related mussel species, is linked to the level of hydration and presumed to maintain bulk performance during tidal exposures. Given the interest in implementing transfer of biological principles to modern materials, these findings may have noteworthy implications for the design of durable synthetic coatings.",
"introduction": "Introduction Functional coatings, though often inconspicuous and deceptively simple in appearance, are essential in providing the underlying bulk material with durability against physical and chemical stresses. However, the inherent inverse relationship between stiffness and extensibility frequently impedes the design of man-made versions and usually results in the promotion of one of these features at the expense of the other 1 . Biological counterparts on the other hand are known to sidestep this trade-off via remarkable structural and biochemical arrangements 2 . A prime example is the mussel cuticle: this non-mineralized coating 3 , which shields the byssal threads from degradation in both physically and chemically aggressive environments, is capable of extending to strains of up to 120% 4 – 6 despite being as stiff as an epoxy resin 7 . This unusual blend, without a counterpart among synthetic formulations, makes this coating a formidable line of defense against the relentless challenges encountered in littoral habitats 8 and has led to its appeal as a paradigm for designing tough and durable materials 9 , 10 . The functional advantages of the cuticle have been correlated with its microscopic architecture, which has been described as a dual-phase system of both submicrometer inclusions (a.k.a. granules, presumably made of condensed mussel foot protein 1) 11 , 12 and an amorphous matrix. Previously highlighted by Holten-Andersen et al., this arrangement was shown to correlate with the need for energy dissipation 13 , whereas Harrington et al. demonstrated via Raman spectroscopy that dynamic self-healing metal–catechol complexes 14 are concentrated in the granules 15 . As a result, these granules were deduced to be rigid fillers within the apparently soft and extensible matrix and therefore credited for toughening the cuticle in a fashion comparable to that seen in particle-reinforced composites (i.e., by acting to reduce crack propagation and abrasion 13 , 16 ). However, while this view is consistent with modern concepts of multiphase wear-resistant materials, whether it explains in situ performance has not been demonstrated. Moreover, specific questions arise from this scheme, such as why damage should be deflected around highly regenerative domains (for which metal–catechol bonds are especially well-adapted) 17 or as to how structural variations in the granular phase influence functional and mechanical responses. To date, these subjects have remained unexplored, presumably due to the difficulty of discerning the specific phases with the analytical methods currently available, and imply that some cuticle features may still be unknown or elusive. Here we examine the cuticles of four mussel species with different granular morphologies to better understand the material properties of these coatings and explore the correlation between architecture and wear. In situ atomic force microscopy (AFM) is used to identify the nanometer-scale scratch and indentation resistance, while transmission electron microscopy (TEM) images and tomograms provide precise structural information of the cuticles and their precursors. From these analyses, an unexpected discovery is emerging: the granules, not the surrounding matrix, are the softer components of the cuticle. Moreover, they are more hygroscopic than the matrix, suggesting their main function to be akin to that of plasticizers and hydration reservoirs rather than reducing abrasion. While these findings invert our view on cuticle mechanics, they highlight the roles and contributions of the individual phases to overall material performance and offer alternative explanations on how this tough biological coating remains durable in harsh ambient environments.",
"discussion": "Discussion Using different approaches, we have shown at several levels that the nanomechanical relationship between cuticle matrix and granules is opposite to that espoused in earlier models and that the role of the granules is more complex than previously assumed. First, they do not stiffen the cuticle but are instead responsible for keeping it plasticized via hydration. Second, granule hydration and plasticity, in contrast to the surrounding matrix, persist during emersion, preventing the cuticle from rapidly becoming stiff and brittle. Therein, although it stands to reason that these elastic fillers are safety nets implemented to preserve flexibility during periods of low tide, the main advantage of this arrangement would be that they, not the matrix, carry the bulk of deformation energy sustained during tension, as is seen in rubber-toughened epoxies 34 , 35 . Though counterintuitive at first, this concept is plausible when pondering granule composition: as metal–catechol interactions rapidly self-heal when broken 17 , 36 , 37 and increase tensile toughness in polymer networks 38 , it is conceivable that load and damage should be localized to the granules, where their density is highest 15 . The preserved hydrated conditions in the granules would further ensure that regenerative and dynamic processes are continuously active, eventually ensuring that the coating remains functional and durable in and out of water. Although the mussel cuticle resembles cutting-edge synthetic- coatings, its performance, processing, and solvent-free fabrication are still unrivaled. Given these, better comprehending its underlying assembly and functional mechanisms has fundamental scientific and technological value, especially in the development of structures with efficient energy-dissipative, protective, and self-healing properties. Consequently, these implications will hopefully benefit the design of bio-based coatings and support the increasing trends of moving away from petroleum-based, layered materials toward sustainable versions."
} | 1,736 |
37547909 | PMC10400768 | pmc | 8,516 | {
"abstract": "In this article, we describe a proof of concept of the potential use of a biocatalytic process for the functionalization of technical soda lignins from wheat straw through the selective acylation of primary hydroxy groups of lignin oligomers by acetate or hexanoate, thus preserving their free, unreacted phenols. The selectivity and efficiency of the method, although they depend on the structural complexity of the starting material, have been proven on model compounds. Applied to technical lignins, the acylation yield is only moderate, due to structural and chemical features induced by the industrial mode of preparation of the lignins rather than to the lack of efficiency of the method. However, most of the physicochemical properties of the lignins, including their antioxidant potential, are preserved, advocating the potential use of these modified lignins for industrial applications.",
"conclusion": "4 Conclusion This study demonstrates the feasibility of the enzymatic selective acylation of primary hydroxy groups of lignin fractions, using, first, well controlled model compounds (commercial guaiacylglycerol-β-guaiacyl ether and sinapyl alcohol dehydrogenative polymers) in order to design and ensure reproducible efficiency of the analytical protocols to be used for the estimation of the acylation yield, as well as the selectivity of the process strictly inactive toward phenols. The whole process, applied to fractions of commercial wheat straw soda lignins Protobind 1000, led to relatively moderate hexanoylation yields (approximately 100 μmol/g), much lower than those obtained with model compounds. This drastic yield decrease is not only a result of the more complex structure of lignin fractions compared to DHPs, but also the process used for the preparation of the starting lignins (remaining salts and low pH of the resulting lignin fraction solutions, which may favor critical changes in the structure of the lignins). From a methodological point of view, we have also demonstrated that the transmethylation method is a versatile method for the determination of the acylation yield, more so than the direct chromatographic (SEC) and spectroscopic ( 31 P NMR) methods, the former not allowing direct quantification and the latter leading to a large overestimation of the acylation yield due to side-reactions, as demonstrated with DHPs. Moreover, this analytical approach involving transmethylation can be applied to any substrate, regardless of its structural complexity or solubility. Different hexanoylation conditions can be considered (in terms of solvent type, concentration, duration, and CAL-B load); thus, they can be adapted regarding 1) substrate specificities if required and/or 2) cost/efficiency balance target. Different experimental conditions have been tested with modified parameters and how they impact the acylation yield, allowing the design of a large panel of procedures that can be used according to the nature of the substrate (solubility and complexity) and the target efficiency. In future, we will diversify the acyl donor as well as the substrates in order to investigate the influence of biomass botanical origin and the impact of the lignin recovery procedure on the reactivity, assuming that the yield of this acylation reaction may be one of the possible quality criteria for technical lignins for further possible valorization routes. It is important to note that our preliminary results concerning the use of these modified lignins as either antioxidant additives for polymer compounding or as stabilizers for cosmetic emulsions are quite promising and demonstrate the significant advantage of modified lignins compared to starting lignins. These different studies are underway and will be reported in due time.",
"introduction": "1 Introduction Due to the expected gradual depletion of fossil resources, industries have turned to the introduction of polymers, cosmetics, pure natural compounds, or mixtures resulting from the fractionation of the biomass into their formulations. For applications with high added values, the strong consumer demand for environmentally friendly processes and formulations, and a wish for naturalness in commonly used compounds, especially for cosmetics, have strongly stimulated research efforts. Among these compounds and beside secondary metabolites ( Vaishnav and Demain, 2011 ), polysaccharide fractions have received much attention from researchers due to their homogeneity in terms of physicochemical properties and chemical reactivity ( Benna-Zayani et al., 2008 ; Bouyer et al., 2012 ; Robertson et al., 2017 ; Arca et al., 2018 ; Yu et al., 2018 ). Therefore, industries involved in biomass fractionation have mostly favored “polysaccharides first” biorefinery processes. Regrettably, due to the research of optimized polysaccharide or secondary metabolite recovery yields, biomass fractionation processes mostly involve drastic acidic or basic treatments with or without sulfur-containing reagents. Other components of the biomass may be strongly modified structurally, thus highly degrading their chemical structure and their main physicochemical properties of interest. For instance, this is the case with lignins ( Sharma et al., 2020 ; Yao et al., 2022 ), which represent the major biopolymer in vertical terrestrial plants after cellulose ( Ralph et al., 2004 ). These heterogeneous assemblies of polyphenolic oligomers result in plants from the oxidative coupling of p -hydroxy cinnamic alcohols (mainly coniferyl and sinapyl alcohols) and acids ( Figure 1 ) ( Liu et al., 2018 ; Sternberg et al., 2021 ). Thanks to this unique polyphenolic structure, they exhibit protective properties (such as antioxidant, antimicrobial, and anti-UV) and thus play the role of defensive barriers in plants. FIGURE 1 \n p -Hydroxy cinnamic alcohols involved in lignins biosynthesis and putative structure of the resulting biopolymer. The main treatments of lignocellulosic biomass for polysaccharide recovery and lignin residue availability at an industrial scale for further applications are the Kraft ( Chakar and Ragauskas, 2004 ; Gellerstedt, 2015 ), bisulfite ( Duval et al., 2013 ; Miles-Barrett et al., 2017 ), soda ( Takada et al., 2020 ), and organosolv ( de la Torre et al., 2013 ) treatments, which at the end of the process all involve an acidic treatment for the recovery of lignins through precipitation. Thus, large amounts of lignins are produced from paper industries and emerging cellulosic 2G bioethanol industries every year. More than 50 million tons of lignins are produced each year by pulp and paper industries alone ( Bezerra and Ragauskas, 2016 ; Maldhure and Ekhe, 2017 ; Robertson et al., 2017 ), and up to now, lignins were considered more as a residue than as a valuable material, and were mainly burned to produce energy because of their high calorific power ( Zakzeski et al., 2010 ; Chen et al., 2017 ). In order to complete the biorefinery concept by valorizing all the biomass fractions, the resulting lignins must now be considered as potential candidates for industrial chemical applications. With this aim in mind, low-added value applications such as a filler in asphalt ( Wu et al., 2021 ), additive for mineral wool ( Allais et al., 2016 ; Lucia et al., 2020 ), and board binders ( Gravitis et al., 2010 ) have emerged. Nevertheless, higher added value applications such as in carbon fibers or as antioxidant additives in polymers and cosmetic formulations have to be explored for technical lignins, which have already shown antioxidant ( Pouteau et al., 2003 ; Vinardell et al., 2008 ; Ponomarenko et al., 2015 ; Majira et al., 2019 ), antimicrobial ( Alzagameem et al., 2019 ; Cresnar et al., 2022 ), and emulsion stabilizing properties ( Czaikoski et al., 2020 ). For these later applications, incorporation of lignins into most of the usual polymer matrices still faces constraints due to their heterogeneous and complex structure and poor miscibility in apolar matrices, where they show a tendency to form aggregates ( Lora and Glasser, 2002 ; Zhang et al., 2015 ; Romhányi et al., 2018 ; Dias et al., 2019 ). To overcome these problems, it seems that chemical modification of the lignin structure remains the best solution. Degradative depolymerization of lignin has been intensively investigated in past decades ( Weng et al., 2021 ; Roy et al., 2022 ; Zhou et al., 2022 ) but requires, whatever the conditions used, final separative processes to recover pure fractions (or, at least, homogeneous in terms of molar mass). Alternatively, another route to minimize their heterogeneity can be solvent fractionation ( Majira et al., 2019 ; Zhou et al., 2022 ). Some of our previous articles highlighted the potential of phenolic compounds or phenolic fractions derived from lignins constituents as antioxidants ( Reano et al., 2016 ) or as building blocks for polymer chemistry ( Pion et al., 2013 ), and also reported the design of innovative processes for the transformation of technical lignins from various botanical sources and diverse industrial separation processes into valuable fractions ( Thierry et al., 2018 ; Majira et al., 2019 ; Lu et al., 2021 ; Jin et al., 2022 ). In addition, direct modification of lignins to tune their properties may be also a promising pathway, taking into account their valuable chemical functionalities. One example is the acylation of lignin hydroxy groups in order to increase their lipophilicity. Chemical acylation processes in the presence of base catalysts such as 4-dimethyl amino pyridine (DMAP) ( Zhao et al., 2017 ) lead to the formation of byproducts such as carboxylic acids and require organic solvents such as 1,4-dioxane or pyridine. However, most importantly, in such processes, the reactivity of phenolic hydroxyls is higher than that of aliphatic ones, inducing a disappearance of free phenol content and thus a loss of their related properties. The goal of our project was to target antioxidant additive applications mainly for polymer chemistry; we faced the problem of improving the compatibility of lignins with apolar matrices while not affecting their free phenol content. That is why this project aimed to selectively acylate the primary aliphatic hydroxy groups of some lignin fractions using either acetate or hexanoate groups, thus preserving their antioxidant properties afforded by the presence of the free phenolic groups. We therefore turned to the use of Candida antarctica lipase B (CAL-B) as a transesterifying biocatalyst ( Cassani et al., 2007 ; Zhang et al., 2020 ). This enzyme is indeed known to be inactive toward phenols ( Pion et al., 2013 ; Weissbach et al., 2017 ) and to be active in a large range of experimental conditions (solvents and temperatures) ( Cassani et al., 2007 ). CAL-B is commercially available in different forms (expression in Aspergillus niger ( Mustranta et al., 1993 ) or in yeast ( Graber et al., 2003 )), either as free enzyme or immobilized on resins ( Miletic et al., 2010 ) in order to allow its easy separation from the reaction mixture by simple filtration. Other groups succeeded with a similar aim of selective acylation of lignins biocatalyzed by CAL-B, involving either ionic liquid as the reaction media ( Hulin et al., 2015 ) or technical lignins that were post-depolymerized through catalytic hydrogenation ( Martinez-Garcia et al., 2023 ). The present work differs in the sense that it involves technical lignin fractions in the solvent MEK, both being industrially common. Moreover, these published works estimated the resulting acylation yields through the disappearance of the acyl donor in the reaction media (by HPLC) or through the decrease of aliphatic hydroxy groups using 31 P NMR. However, based on our practical experience on lignins analyses and their reactivity, we must consider unpredicted artefacts in such methodologies. Indeed, the lignin spontaneous reactivity that is often noticed in the literature but is not yet fully understood, can generate side-reactions and thus analytical artefacts, leading to misinterpretation. That is why, in this article we furthermore transposed one of the most commonly used method for the quantification of fatty acid composition in complex matrices: transmethylation followed by GC-MS. We follow a robust analytical procedure that is less impacted by the potential biases induced by the unusual chemical reactivity of lignin derivatives upon chemical or thermic treatments. This article describes the feasibility of the combination of a solvent fractionation process and biocatalytic acylation of technical wheat straw alkali lignins (Protobind 1000) to convert them into valuable antioxidant additives for polymers and cosmetic applications as well as a robust analytic methodology to quantify the yield of the acylation step.",
"discussion": "3 Results and discussion 3.1 Feasibility and selectivity of enzymatic acylation 3.1.1 Acetylation of dimer model compound In order to test the feasibility and selectivity of the reaction, the CAL-B activity was first tested on commercial guaiacylglycerol-β-guaiacyl ether dimer, which is commonly used as representative of the β- O -4 linkage between two coniferyl alcohols found in lignin oligomers as the major linkage between two p- hydroxycinnamic subunits. This commercial stereopure dimer exhibits a free phenolic group and both primary and secondary aliphatic hydroxy groups. The acetylation of the dimer ( Figure 2 ) was first investigated in the presence of ethyl acetate and supported CAL-B (10% wt of dimer) in acetonitrile (10 g/L, 50:50 acetonitrile:ethyl acetate) and with continuous removal of the by-produced ethanol by azeotropic distillation. The reaction was easily followed by HPLC and 1 H NMR ( Supplementary Figure S4 ). These first results showed only acetylation of the primary hydroxyl in approximately 90% yield after 8 h of reaction. The reaction could also be monitored by 31 P NMR after phosphorylation ( Supplementary Figure S5 ), where the relative integration of the signal corresponding to the phosphorylated primary OH (triplet at 147.5 ppm) progressively decreased, while the integration of secondary (doublet at 148.2 ppm) and phenolic OH (singlet at 139.5 ppm) remained equal, even if their chemical shift slightly increased (to 148.3 and 139.6 ppm, respectively), thus confirming the selectivity. To further ensure this selectivity, the reaction was conducted over a longer time period (13 days): no peracetylated product was observed by HPLC, 1 H NMR, or 31 P NMR. If CAL-B is well-known to be strictly inactive toward phenols, it may induce acylation of secondary hydroxyls ( Pion et al., 2013 ). In this current case, inactivity toward the secondary hydroxyl might be due either to stereoselectivity toward the involved stereopure commercial dimer, or to the proximity of the hindered phenolic moiety ( Uppenberg et al., 1995 ). Acetonitrile, first selected because it is known for being tolerated by CAL-B ( Dutta Banik et al., 2016 ) has however a rather low ability to solubilize lignins. Thus, it was thereafter replaced by methyl ethyl ketone (MEK), showing similar acetylation kinetics on the commercial dimer ( Figure 3 ). MEK was therefore chosen as the reaction solvent for this process and to efficiently solubilize a large fraction of Protobind 1000 (60% wt), which is our targeted lignin substrate. FIGURE 2 Guaiacylglycerol-β-guaiacyl ether enzymatic acetylation. FIGURE 3 Kinetics of guaiacylglycerol-β-guaiacyl ether (GGE) enzymatic acetylation under reflux in acetonitrile or in MEK (10 g/L), followed by HPLC. 3.1.2 Complexification of the model substrate: DHPs acetylation However, direct monitoring of the reaction on lignins fractions by 1 H NMR or HPLC would be unfeasible due to the complexity of the lignins’ structure and their poor solubility in the appropriate solvents ( Wen et al., 2013 ; Wurzer et al., 2021 ). Therefore, to challenge further analytical methodologies (SEC, 31 P NMR), we turned to the acetylation of a substrate of intermediate complexity: dehydrogenative dehydropolymers (DHPs) ( Lahive et al., 2020 ) of sinapyl alcohol ( Figure 4 ). Such S-type DHPs, exhibiting only two types of inter subunit linkages (β- O -4 and syringaresinol types), are indeed good candidates to adapt our analytical procedures to a more complex substrate representative of lignins. FIGURE 4 Putative structure of S dehydrogenative polymers (S-DHPs) used as model substrate. DHPs were treated in the presence of CAL-B (10% wt compared to DHPs) in ethyl acetate:MEK (50:50 vol, 10 g/L) under reflux for 2 days. Afterwards, the mixture was cooled to room temperature, filtered to recover supported CAL-B, and the solvent evaporated under vacuum to obtain the resulting acetylated DHPs. For the purpose of analytical comparisons, completely acetylated DHPs were prepared through a chemical procedure in pyridine using acetic anhydride. From 1 H- 13 C HMBC NMR (in C 5 D 5 N) of the chemically acetylated DHPs, three acetate groups were observed: one phenolic (2.29 ppm/168.5 ppm) and two aliphatic, corresponding to primary and secondary hydroxyls (1.94 ppm/170.6 ppm and 2.13 ppm/170.0 ppm, respectively), while the enzymatically acetylate showed a single primary aliphatic acetate signal (1.94 ppm/170.6 ppm, Supplementary Figure S6 ). This confirmed the transposability of the selective enzymatic process to more complex substrates, where acetylation is restricted to primary aliphatic hydroxyls. In addition, another characteristic signal was observed at 2.11 ppm/173.3 ppm, corresponding to residual acetic acid (confirmed by comparison with pure acetic acid, which also appears at 134.6 ppm in 31 P NMR, Supplementary Figure S7 ). The molar mass distribution of the samples was also determined through size exclusion chromatography (SEC), assuming that the acetylation could induce an increase in apparent molar mass. Chemically acetylated DHPs showed a slight increase in molar mass distribution when compared to the starting material ( Figure 5 ). Surprisingly, this increase in apparent molar mass was even more significant in the case of enzymatic acetylation, while it should have a lower molar mass gain since its acetylation was partial. Such an observation suggests that side reactions, such as intermolecular cross-coupling reactions, may occur in these reaction conditions in the case of DHPs. FIGURE 5 Size exclusion chromatograms in THF of the starting material and the enzymatic and chemically acetylated DHPs, using UV detection at 280 nm. Signals normalized to the toluene peak eluted at 21.1 min. Therefore, to better understand this unexpected result, two control reactions were run: one in the absence of any acyl donor and another in the absence of the catalyst CAL-B. In the two cases where CAL-B is present (control without acyl donor and enzymatic reaction), a significant increase in molar mass is observed, considering that lipase treatment is able to induce cross-coupling reactions between DHPs residues. In the absence of CAL-B, a similar molar mass increase is observed, but is far less important and probably also due to chemical modifications of the DHP structure upon thermal treatment ( Figure 6 ). The hydroxyl group content of the different samples was then assessed through 31 P NMR after phosphorylation, a method commonly used for the quantification of the different types of lignin hydroxyls ( Table 1 ) ( Meng et al., 2019 ). FIGURE 6 Size exclusion chromatograms in THF for DHPs in different reaction systems (48 h in MEK (10 g/L)), using UV detection at 280 nm. Signals normalized to the toluene peak eluted at 21.1 min. TABLE 1 Characteristics of the substrate DHPs (Entry 1), enzymatically acetylated DHPs (Entry 2), and appropriate controls (Entries 3, 4) kept for 48 h under reflux in MEK (10 g/L); molar mass distributions determined a by SEC. Aliphatic and phenolic hydroxyl contents determined b by 31 P NMR. Entry Enzyme (% wt) Acyl donor M n \n a (g/mol) M w \n a (g/mol) PDI a \n [Aliphatic OH] b (mmol/g) [Phenolic OH] b (mmol/g) 1 — — 1511 2731 1.81 3.32 ± 0.1 1.48 ± 0.05 2 10 Ethyl acetate 1046 3522 3.37 1.98 ± 0.1 1.55 ± 0.07 3 10 — 1101 2733 2.48 2.65 ± 0.05 1.64 ± 0.01 4 0 Ethyl acetate 1109 2676 2.41 2.76 ± 0.02 1.64 ± 0.02 5 10 Ethyl hexanoate 1906 4,658 2.79 2.91 ± 0.04 1.85 ± 0.01 In the case of enzymatic acetylation, the aliphatic hydroxyl content significantly decreased (−40%, Entry 2), suggesting efficient acetylation, as observed by the appearance of an ester spot in 1 H- 13 C HMBC NMR. Nevertheless, even if lower, this phenomenon was also observed in controls (−20% and −17%, Entries 3 and 4), where no acetate was observed through 1 H- 13 C HMBC NMR, thus suggesting that this decrease in aliphatic hydroxyls can also be due to spontaneous side reactions. Moreover, this hypothesis is reinforced by an increase in the free phenol content, probably due to the cleavage of some β- O -4 bonds inside the oligomers. From such global observations it can be assumed once more that DHPs (and probably also lignin fractions) may undergo dramatic structural changes in these reaction conditions (temperature, solvent), including for instance β- O -4 bond cleavage and recondensation reactions. The balance between these two pathways, going from control reactions to the real reaction may be to some extent due to the generation of acetic acid in the medium containing the ethyl acetate and CAL-B, but also to the specific activity of CAL-B on lignin oligomers in the absence of any acyl donor. Although confirming the occurrence of side reactions impacting the hydroxyl content in all cases, 31 P NMR appeared to be nevertheless unsuitable for precise quantification of the acylation yield for DHPs (and probably for lignin fractions), contrary to our expectations. Indeed, the acetylation yield, when estimated on the basis of 31 P NMR data, seemed largely overestimated (approximately 50% mol of the putative primary hydroxy groups). We thus concluded that another analytical method was needed to quantify the acylation yield. We envisaged the use of an indirect method based on either saponification or transesterification methods, allowing the reformation of acids or esters from the acylated lignins, which can be further quantified by GC-MS. However, such a process is incompatible with acetylated compounds since it generates either acetic acid or methyl acetate, which are difficult to quantify by GC-MS. Another longer acyl donor was therefore chosen, ethyl hexanoate, as its excess can still be easily removed by either evaporation or hexane washing, and since its corresponding acid or methyl ester are easily quantifiable by GC-MS. 3.1.3 Hexanoylation of the dimer model compound The hexanoylation process was first tested on the commercial dimer in order to assess the efficiency of this approach with a longer chain and a more apolar acyl donor. Guaiacylglycerol-β-guaiacyl ether dimer was thus reacted with ethyl hexanoate in a ratio of 5:1 mol to ensure an excess of acyl donor at the same conditions (10% wt CAL-B, MEK, 10 g/L, reflux). The kinetics of hexanoylation was monitored by 1 H NMR ( Supplementary Figure S8 ) and HPLC and was found to be very similar as the acetylation ( Figure 7 ), with a hexanoylation yield of 76% after 9 h, thereby proving the versatility of the method in terms of acyl donor. Similar control experiment ran without CAL-B induced no structural changes of guaiacylglycerol-β-guaiacyl ether dimer (results not shown), thus confirming that hexanoylation was indeed only catalyzed by CAL-B. FIGURE 7 Kinetics of guaiacylglycerol-β-guaiacyl ether acylation by ethyl acetate or ethyl hexanoate, in MEK (10 g/L) under reflux, followed by 1 H NMR in CDCl 3 . 3.1.4 Hexanoylation of DHPs and quantification of the acylation yield by a transmethylation/GC-MS procedure DHPs were then reacted with an excess of ethyl hexanoate (5:1 wt) in MEK (10 g/L) under reflux. After the biocatalyzed hexanoylation reaction and the removal of the supported CAL-B by filtration, the reaction medium was concentrated under reduced pressure and first cleaned from residual unreacted acyl donor through precipitation in hexane, and submitted to extended drying under vacuum. 1 H- 13 C HMBC NMR proved the formation of an ester bond ( Supplementary Figure S9 ) as well as the absence of residual acyl donor (no ethyl spots visible), which was confirmed by GC-MS analysis of the hexane supernatant. Saponification of the hexanoylated compounds was first tested but led to very tedious workup procedures and unsatisfying errors in the quantification of the resulting hexanoic acid. We thus turned to a transmethylation process, well-known in the field of triglyceride chemistry. Transmethylation aims to cleave the ester bonds and thus release the hexanoate moieties as methyl esters ( Supplementary Figure S10 ). Thereafter, transmethylation followed by GC quantification established that 184 µmol of methyl hexanoate was released per Gram of reacted DHPs. This value, compared to the aliphatic hydroxyl content of the starting material determined by 31 P NMR (3.32 mmol/g, Table 1 Entry 5), indicates a hexanoylation yield of 5.5% mol of the aliphatic hydroxy groups; nevertheless, such a calculation underestimates the yield as it does not take into account the occurrence of side reactions decreasing the amount of available targeted aliphatic hydroxy groups and does not discriminate primary from secondary hydroxyl content, which may be equivalent according to the putative DHPs structure. Structural modification of DHPs was not investigated further as it was not the aim of this work; however, it appears more reliable to quantify acylation through transmethylation rather than structural analysis ( 31 P NMR or SEC), as we expected initially. For these reasons, the hexanoylation yield will be estimated by transmethylation and expressed in µmol/g. In initial attempts, we compared the quantification through peak area measurement inserted in the equation [molar concentration] = f ([peak area]) determined by injection of CS at different concentrations. Nevertheless, the use of an internal standard of close structure (methyl heptanoate, MQS ) led to higher repeatability and was preferred in the remainder of our study, as described in the material and methods section. This method is indirect and it can be applied to any acylated substrate of complex structure, as far as the residual unreacted acyl donor is quantified or, when possible, carefully eliminated. Therefore, this method was retained for hexanoylation yield estimation in the remainder of the study, aiming to transfer this biocatalyzed selective hexanoylation process to technical lignins and derived fractions. 3.2 Transposition of the enzymatic hexanoylation to lignins 3.2.1 Fractionation of the technical lignins To deal with less heterogeneous samples than the complex technical lignins Protobind 1000, sequential solvent fractionation was applied in order to obtain more defined lignin fractions ( Lu et al., 2021 ; Jin et al., 2022 ). The first step involved ethyl acetate in order to eliminate lower molar mass compounds F1 and to focus on polymeric chains, as they encounter more difficulties in accessing enzyme active sites. The second step involved MEK in order to extract a substrate soluble F2 in the reaction media (MEK), as well as in analytical solvents, by eliminating the insoluble residue F3 . These three fractions showed lower polydispersities than Protobind 1000 and exhibited increasing molar masses as follows: M w ( F1 ) < M w (Protobind 1000) < M w ( F2 ) < M w ( F3 ), while the phenol content decreased in the opposite manner ( Figure 8 ; Table 2 ). FIGURE 8 Size exclusion chromatography in THF of Protobind 1000 technical lignins (pink) and the resulting fractions F1 (blue), F2 (black), and F3 (brown), using UV detection at 280 nm. Signals normalized to the maximum height; toluene used as internal standard, eluted at 20.2 min. TABLE 2 Characteristics of the different fractions (ethyl acetate F1, MEK F2, and insoluble F3 fractions) obtained from Protobind 1000 (PB1000) determined a by gravimetry, b by SEC, and c by 31 P NMR (spectra available in Supplementary Materials). %wt a \n M n \n b (g/mol) M w \n b (g/mol) PDI b \n [Aliphatic OH] c (mmol/g) [Phenolic OH] c (mmol/g) [COOH] c (mmol/g) PB1000 100 573 1744 3.04 1.94 ± 0.02 4.43 ± 0.05 1.17 ± 0.03 \n F1 \n 36 385 813 2.11 1.16 ± 0.04 5.24 ± 0.09 1.27 ± 0.01 \n F2 \n 27 1240 2188 1.76 1.41 ± 0.01 4.13 ± 0.01 1.02 ± 0.02 \n F3 \n 37 1266 2695 2.13 2.50 ± 0.07 3.34 ± 0.12 0.96 ± 0.05 3.2.2 Hexanoylation of lignins fraction F2 The hexanoylation of soda lignins fraction F2 (lignins fraction soluble in MEK but insoluble in ethyl acetate) was thus conducted in MEK with ethyl hexanoate. The first attempt was conducted based on previous experimental parameters: 5 mass equivalent of ethyl hexanoate and 10% wt CAL-B regarding lignins were used in MEK (10 g/L) under reflux for 24 h, which led to a hexanoylation yield of 85 μmol/g. The same experiment conducted in absence of CAL-B showed no hexanoylation; this control informed us that CAL-B activity was responsible for the reaction and confirmed the accuracy of hexanoylation yield quantification through transmethylation followed by GC-MS, inducing no overestimation. In order better understand the reactivity of this new system, different key parameters have been studied to optimize enzymatic hexanoylation: 1) ethyl hexanoate:lignins weight ratio, 2) reaction time, 3) lignins concentration, 4) enzyme load, and 5) the solvent used as the reaction medium. In all following sections, different assays were compared based on their resulting hexanoylation yields determined through transmethylation followed by GC-MS. Some of the following studies were conducted in parallel; therefore, the fixed parameters are not necessarily optimized, aiming to detect tendencies rather than defining an optimized process. 3.2.2.1 Ethyl hexanoate:lignins weight ratio The work on model DHPs was conducted with 5 mass equivalent of ethyl hexanoate in order to ensure a sufficient excess. However, such an excess led to tedious removal of residual unreacted acyl donor at the end of the process. Thus, various ethyl hexanoate:lignins weight ratios (from 0.05 to 5 mass equivalents) were studied at a constant enzyme load (20% wt), lignins concentration in MEK (10 g/L), and reaction time (24 h). Lowering the weight ratio of ethyl hexanoate:lignins from 5:1 to 1:1 gave similar results (102 and 94 μmol/g, respectively), while too low a ratio led to a drop in the hexanoylation yield (25 μmol/g with 0.05:1, Figure 9 ). FIGURE 9 Hexanoylation yield obtained for different ethyl hexanoate:lignins weight ratios at a constant enzyme load (20% wt), lignins concentration in MEK (10 g/L), and reaction time (24 h); yield estimated by transmethylation followed by GC-MS. Thus, a 1:1 weight ratio was preferred as it makes the post-reaction workup easier and the process cheaper and greener by generating less side-products without significantly decreasing the hexanoylation yield. 3.2.2.2 Reaction time Another key parameter can be the reaction time; thus, it was studied from 6 h to 6 days at a constant ethyl hexanoate:lignins weight ratio (1:1), enzyme load (10% wt), and lignins concentration in MEK (10 g/L). 6 h reaction time led to only half of the hexanoylation yield (39 μmol/g) compared to 24 h (81 μmol/g), indicating an incomplete process. Further extension of the reaction time led to a slight increase in the hexanoylation yield ( Figure 10 ). Thus, time can be a parameter of interest but should be balanced according to the energetic cost and lignins side reactions. FIGURE 10 Hexanoylation yield obtained for different reaction times at a constant ethyl hexanoate:lignins weight ratio (1:1), enzyme load (10% wt), and lignins concentration in MEK (10 g/L); yield estimated by transmethylation followed by GC-MS. 3.2.2.3 Lignins concentration in MEK The lignin concentration in MEK was varied from 10 to 50 mg/mL at a constant ethyl hexanoate:lignins weight ratio (5:1), percentage of enzyme (10% wt), and reaction duration (24 h). Similar hexanoylation yields (from 85 to 92 μmol/g) were obtained in the three cases ( Figure 11 ), demonstrating the low impact of concentration on the studied range, thus allowing us to reduce the amount of solvent employed. FIGURE 11 Hexanoylation yield obtained for different reaction concentrations at a constant ethyl hexanoate:lignins weight ratio (5:1), enzyme load (10% wt), solvent nature (MEK), and reaction time (24 h); yield estimated by transmethylation followed by GC-MS. 3.2.2.4 CAL-B load In order to assess the impact of catalyst load on the reactivity, the percentage of supported enzyme was also varied from 10% to 100% wt according to lignins at a constant ethyl hexanoate:lignins weight ratio (1:1), lignins concentration in MEK (25 g/L), and reaction time (24 h). Indeed, the hexanoylation yield increased along with the CAL-B load ( Figure 12 ), even if the phenomenon is far from being linear: increasing the load by 10 times doubled the yield (250 μmol/g). Thus, the CAL-B load is a real lever for increasing the hexanoylation yield; nevertheless, due to its high cost, a balance should be struck between cost and efficiency. FIGURE 12 Hexanoylation yield obtained for different CAL-B loads at a constant ethyl hexanoate:lignins weight ratio (1:1), lignins concentration in MEK (25 g/L), and reaction time (24 h); yield estimated by transmethylation followed by GC-MS. 3.2.2.5 Solvent variation It is known that CAL-B activity and stability are impacted by the polarity and temperature of the medium ( Kitamoto et al., 2015 ; Banik et al., 2016 ). In our case, both are governed by the solvent chosen, which also impacts substrate solubilization. Thus, solvents of different polarities and different boiling points (acetonitrile, acetone, MEK, THF, or a mixture MEK/hexane (50/50), Table 3 ) were tested at a constant ethyl hexanoate:lignins weight ratio (1:1), enzyme load (10% wt), lignins concentration (25 g/L), and reaction time (24 h, Figure 13 ). Acetone, which shows a polarity similar to MEK (Hidebrand parameters of 19.9 and 19.3 MPa 1/2 , respectively) but a lower boiling point (56°C and 80°C, respectively), led to a lower hexanoylation yield (68 μmol/g instead of 102 μmol/g). This lower reactivity might be related to the lower temperature, which decreases the CAL-B activity and/or impeaches ethanol removal. Acetonitrile, known to be compatible with CAL-B ( Arcens et al., 2020 ) and suitable as a dimer acetylation medium, exhibits a boiling point (82°C) similar to MEK, but a higher Hildebrand parameter (24.3 MPa 1/2 ); the significantly lower hexanoylation yield (16 μmol/g) when compared to MEK might be due to the low solubility of lignins F2 in acetonitrile, where only 65% was recovered (generally 90% is recovered). THF shows a similar Hildebrand parameter (18.6 MPa 1/2 ) but a lower boiling point than MEK (66 °C), leading to a similar hexanoylation yield of 97 μmol/g. THF appears therefore to be a choice for substrates of lower solubility. Finally, a mixture of MEK and hexane (50/50) was attempted in order to improve ethanol removal. The resulting hexanoylation yield was only 68 μmol/g, most likely due to the lower solubility of the substrate F2 in this system (only 72% recovered). These overall observations tend to indicate that MEK is a good solvent of choice in our system. When one wants to change for another solvent, attention should be paid to both its boiling point (60°C–100°C) and its polarity (19–20 MPa 1/2 ) in order to preserve a certain reactivity. TABLE 3 Characteristics of the tested solvents and the resulting recovery (after precipitation in hexane) and acylation yields. Solvent Hildebrand parameter (MPa 1/2 ) Boiling point (°C) Recovery yield (%) Acylation yield (µmol/g) Acetone \n 19.9 \n \n 56 \n \n 90 \n \n 68 \n Acetonitrile \n 24.3 \n \n 82 \n \n 65 \n \n 16 \n MEK \n 19.3 \n \n 80 \n \n 89 \n \n 102 \n MEK/Hexane (50/50) \n nd \n 65 \n 72 \n \n 68 \n THF \n 18.6 \n \n 66 \n \n 90 \n \n 97 \n FIGURE 13 Hexanoylation yield obtained for different solvents as reaction media at a constant ethyl hexanoate:lignins weight ratio (1:1), enzyme load (10% wt), lignins concentration (25 g/L), and reaction time (24 h); yield estimated by transmethylation followed by GC-MS."
} | 9,129 |
37390215 | PMC10313163 | pmc | 8,517 | {
"abstract": "Soft robotics offer unusual bioinspired solutions to challenging engineering problems. Colorful display and morphing appendages are vital signaling modalities used by natural creatures to camouflage, attract mates, or deter predators. Engineering these display capabilities using traditional light emitting devices is energy expensive and bulky and requires rigid substrates. Here, we use capillary-controlled robotic flapping fins to create switchable visual contrast and produce state-persistent, multipixel displays that are 1000- and 10-fold more energy efficient than light emitting devices and electronic paper, respectively. We reveal the bimorphic ability of these fins, whereby they switch between straight or bent stable equilibria. By controlling the droplets temperature across the fins, the multifunctional cells simultaneously exhibit infrared signals decoupled from the optical signals for multispectral display. The ultralow power, scalability, and mechanical compliance make them suitable for curvilinear and soft machines.",
"introduction": "INTRODUCTION Natural creatures have fascinating morphing skin appendages with switchable functionalities. Chameleons camouflage by exploiting their pixelated skin appendage (see Fig. 1A ) ( 1 ). Animals use liquid media to control these pixelated appendages. Beautiful skin patterns are created on demand either by stretching ink-filled elastic sacs (chromatophores in cephalopods) ( 2 ) or by secreting/draining liquids from porous photonic structures (tortoise beetle) ( 3 ). Moreover, fluids are used to morph skin appendages, for instance, by bundling thin hairs in the otters fur or the beetles feet for temperature regulation or switchable adhesion, respectively ( 4 – 5 ). The critical advantages to these skin display and texture morphing mechanisms are their multimodality (color, temperature, texture, and adhesion), multifunctionality (camouflage and temperature regulation), and mechanical flexibility, while using only ultralow power compared to energy-intense light-emitting semiconductor devices. In nature, dynamically changing appendages have state persistence, i.e., the ability to use low power to switch the display, and then retain the new state with minimal energy cost. While some of these individual functions have been explored recently using fluid networks ( 6 ) or inflatable skins ( 7 – 8 ), these demonstrations remain limited by preprogrammability, single modality, switching speed, and the high-power requirement for state switching and retainment. Fig. 1. Fapping fins driven by capillarity and hydrodynamics. ( A ) Pixelated skin appendage of veiled chameleon motivates this work. Inset: Close-up view of chameleon skin. Scale bar, 8 mm. ( B ) Bimorophic fin. The middle images show the snapshots for both modes in the side view, where N mode ( q = 1.3 ml/min) starts from center to left and W mode ( q = 9.1 ml/min) starts from center to right (see movie S1). The fin length is 5 mm, and the working fluid is mineral oil. The bottom images are snapshots from the numerical simulation (see movie S2). ( C ) Plot of the drain flow control signal versus time. Positive and negative signs of the y axis correspond to supply and drainage of liquid, respectively. ( D ) Experimental fin flapping regime map with the fixed W = 4 mm. Circle, triangle, and square symbols correspond to l = 5, 6, and 7 mm, respectively. Blue, green, brown, and light pink colors correspond to N = 1.5, 2, 2.5, and 3 mm, respectively. The total width N + W ≍ 5.5 5.5, 6, 6.5, and 7 mm, with negligible fin thickness. The gray color corresponds to the transition between N mode and W mode. ( E ) Mechanism of fin deformation modes. Drainage to the epidermal hole drives two flow paths: along the fin and the across s . The dashed inset corresponds to the profile of the pressure gradient along the fin length. The solid inset illustrates leak velocity driven by the suction pressure within s . ( F ) Universal regime map based on a theoretical model. Top and bottom regimes from the theoretical line correspond to W mode and N mode, respectively. In the theoretical line, we use the prefactor = 0.25 to precisely fit the experimental results. The dashed line corresponds to the numerical model (see section S5). All symbols are described in the caption (D). Here, we studied flexible appendages, which flap to large angles by the hydrodynamic action of liquid droplets, drawing inspiration from the capillary bundling of the hair of otters and beetles. The ultrasoft fins are fixed at the base to wet cells, named flap-phores and spelled here “flap4.” The wet cells are intended to operate in arrays forming pixelated skins, where each cell is individually switched by liquid control via an epidermal pore. The pore is connected to a fluid network, valves, and motorized syringes, but many other sources of liquid pressure can work for these small volumes ( 9 – 10 ). The flapping fin switch between two intriguing stable drained states: straight and bent, which are normal and parallel to the substrate, respectively. The switching is controlled by the rate of drainage from the epidermal hole located next to the fin on one side. At slow drainage, the fins exhibit a capillarity-dominated operation mode where the fin flaps to the narrow side (N mode), and at high drain rate, we observe a suction pressure–dominated mode where the fin flaps to the wide side (W mode). We derive simple scaling law to capture the intriguing physical phenomenon responsible for switching between the N mode and the W mode, which belongs to the broad class of elastocapillarity ( 11 – 12 ), and specifically dynamic elastocapillarity where hydrodynamics plays a dominant role ( 13 – 15 ). In the rest of this report, we show more complex dynamic elastocapillary polymorphism useful for a variety of multimodal signaling devices. We exploit the flap4 cells in constructing soft multipixel display—where each flap4 cell is an individually controlled pixel—and in optical/infrared (IR) multispectral signaling by leveraging the contrast between the fin color and the cells background and by using droplets of different temperatures across the fin to create IR contrast simultaneously. By this demonstration, it is shown that these new capillary flapping phenomena and the multimodal signaling and texture morphing can be added to the repertoire of future reflective display devices, particularly suitable for soft and conformable needs.",
"discussion": "DISCUSSION In summary, we studied the intriguing flapping of ultrasoft fins under the action of capillary hydrodynamic liquid effect. We demonstrated a soft robotic texture morphing cells and imparted contrast switching in addition to the texture change using the contrast of a flapping fin against the background color. These cells are driven by liquid droplets fed to the cells via epidermal pores and connected to a fluidic network underneath the cells. We show that the fin-flapping motion is governed by the balance between capillary forces and dynamic suction forces from the fluid network and show that the negligeable stiffness of the fins enables the flapping and the persistence of the fin state without external power. Various polymorphic cells including two- and four-fin domino-like texture, blooming flowers, and multipixel can display patterns or alphanumerical letters. Droplets of contrasting cold and hot temperature are used for optical/IR multispectral message encoding, a concept that, to our knowledge, only exists in natural creatures. Since the driving force for the shape transition is surface tension, a wide variety of off-the-shelf materials work with this concept reliably, which we view as an advantage over highly specialized stimuli-responsive materials, which require complex synthesis and has a performance that degrade with time. The flap4 cells respond reasonably fast (a fraction of a second to a few seconds), which is considerably faster than diffusion-driven swelling polymers (hundreds of seconds). Since each cell requires only a few milliliters (≍0.2 ml) to operate, this concept can be integrated into large pixels by leveraging recent advances in microfluidic network components such as valves and pumps. This morphing texture can be leveraged in switchable multimodal antenna and transformable electrical circuits. Overall, dynamic skin displays will add critical functionalities to soft machines ( 18 – 19 ), robots ( 20 – 21 ), and morphing fabrics ( 22 ). To accomplish this, we present a performance evaluation of flap4 on curved surfaces (see section S8). Capillary-based morphing texture system can hold the potential to notably benefit the large-scale signage industry by offering substantial improvements in energy efficiency. This system can be used for extensive signage applications, such as those found in stadiums or on building rooftops. The possibility of larger soft display devices is demonstrated in Fig. 3 , where a feasibility test was conducted to showcase a three-by-three multipixel configuration. While the current morphing system necessitates an external light source for displaying visible signs, it can still function as an effective reflective display during daylight hours. A primary advantage of the capillary-based robotic system is its energy efficiency, consuming approximately 10 W for a 100-m 2 screen size when the pixel size is approximately 1 mm. In contrast, light-emitting diode displays consume roughly 1000 W for the same screen size. We expect that miniaturizing the fabrication of each flap4 cell down to 1-mm scale would open more applications. At this scale, the cells will be capable of operating in any orientation with respect to gravity, unlike the current cell size, which is influenced by gravitational effects (see section S9). This will enable uses in billboards, stadium displays, train and bus station signage, and on the facades of structures and buildings and offer significant energy reductions."
} | 2,493 |
34496752 | PMC8425116 | pmc | 8,518 | {
"abstract": "Background One of the ecological impacts of exotic plant invasions may be alteration of the soil microbial community, which may cause changes to the diversity, richness and function of these communities. In order to explore to what extent invasive plants affect the soil microbial community, we performed a meta-analysis based on 46 scientific articles to document the effect of invasive plants on species richness and diversity of bacteria and fungi. We conducted our study across a range of invaded ecosystems including native communities, and evaluated biomass, richness and diversity. We use a random effects model to determine the increase or decrease in the values of the response variables in the presence of invasive plants. Results The results indicated that the response variable that changed with the invasion of plants was the diversity of bacteria. Bacterial diversity in the soil increases with the presence of invasive plants, specifically herbaceous plants producing allelopathic substances growing in forest ecosystems of temperate zones. Conclusions We provide evidence that invasive plants affect the soil biota differentially; however, it is important to consider more variables such as the N and C cycles, since these processes are mediated by soil biota and litter, and chemical compounds released by plants influence them. Changes in bacterial diversity have consequences for the nutrient cycle, enzymatic activity, mineralization rates and soil carbon and nitrogen content. Supplementary Information The online version contains supplementary material available at 10.1186/s12862-021-01899-2.",
"conclusion": "Conclusions Bacterial diversity was the unique microbial variable that was affected by alellopathic substances released by herbaceous invasive plants in temperate forests. A possible explanation for this result is that in the temperate forests, these plants release a smaller variety of secondary metabolites, thus enabling bacteria species to use them as resources. This unexpected result could be considered a significant contribution to invasion ecology. However, further field and greenhouse studies are required to exhaustively evaluate the role of exotic plants on soil bacterial communities. Future studies should also consider a more mechanistic approach, including the nutrient cycles in which soil microorganisms are involved, as well as life traits of the plants, dependence on AMF, land use history and competition between native and exotic plants.",
"discussion": "Discussion The meta-analysis showed that alellophatic substances produced by invasive plants had only significant and positive effects on bacterial diversity. Some of the studies were carried out in greenhouses simulating a comparison of the SMC between invaded and non-invaded areas. One limitation of the recorded studies is the duration of the experiments (< 6 months), thus probably not giving enough time to detect microbial responses to invasive plants [ 44 , 47 ]. For future research, it would be advisable to integrate greenhouse and field experiments, which can complement the information obtained for invasive plants, thus giving us a more realistic picture to understand invasive plant and soil biota interactions. Our study also reveals a geographic research bias; most of the data analyzed comes from studies on herbaceous plants (104 cases) from temperate areas of North America and Europe, while there were fewer studies in tropical areas. This bias seems to be a general pattern in invasion ecology studies [ 35 , 36 ]. We also found that studies of the impact of invasive plants on soil microorganisms have concentrated on six species ( Alliaria petiolata, Berberis thunbergii, Ageratina adenophora, Impatiens glandulifera, Bromus tectorum, Jatropha curcas ), these being 35% of cases (76 studies). Some invasive plant species (24 cases; 9 species) exert significant impact on bacterial diversity. Several studies have suggested that invasive plant species may modify the functioning of ecosystems by altering SMC (e.g., [ 14 , 22 ]. A comprehensive literature review published by Pysek et al. [ 35 ] reveals that invasions by exotic plants tend to increase the richness and abundance of soil biota. Compared to native species, invasive plant species generally produce more leaf litter (49%) that is of better quality (lower C:N ratio) [ 27 ]. The greater quantity and quality of litter increases the C available in the soil, a source of energy for the SMC, which could allow the establishment of a more diverse and abundant SMC [ 52 ]. Our results partially corroborate these results, our categorical model revealed that invasive plant species have a significant effect on the diversity of bacteria in the soil, while they do not generate significant effects on the other components of the SMC [ 17 , 47 ]. A recent meta-analysis comparing the effects of invasive species on the SMC from litter and the rhizosphere reports that litter increases the biomass of soil bacteria due to nutrient intake, while changes generated by the rhizosphere during the invasion decrease the biomass of bacteria [ 52 ]. The authors attribute this result to the fact that litter accumulation can have positive effects on bacterial communities [ 11 , 12 ], while radical exudates (organic acids, allelopathic substances and hormones) could inhibit bacterial biomass. On the other hand, the meta-analysis performed by Meissner et al. [ 30 ] reports null effects of allelopathic substances on the biomass of bacteria in the soil. Our results are somewhat consistent with these findings, we also found that allelopathic substances released by invasive plants have no effect on bacterial biomass [ 22 , 24 ], about 29.4% of invasive plant species in our database were reported to have allelopathic effects (Additional file 1 ), which may partially explain the absence of suppressive effects from the roots of invasive plants on bacterial biomass. Specifically, Meissner et al. [ 30 ] found that neither the litter nor the exudates from the roots of the invasive plants have effects on FLF biomass. This result (and ours) may be attributable to the fact that the effect size values for the different categories were quite variable, indicating that the FLF biomass change is contingent on the kind of invasive species as well as the ecosystem type. Little has been done to investigate how AMF communities can be affected by invasive plants. Our meta-analysis suggests an absence of effects caused by invasive plants on the AMF community; however, certain specific studies indicate significant effects. For example, Vogelsang and Bever [ 49 ] found evidence of a reduction in mycorrhizal fungi density by nonnative plants. More recently, Rezácová et al. [ 38 ] found that invasions by five nonnative plant species altered composition of the AMF community and reduced the diversity of AM fungi in the soil and in the roots of some native plant species. However, neither of the two studies could be included in this meta-analysis because they did not meet the selection criteria established in our study [ 38 , 50 ]. The results obtained in our study can be explained because the invasive plants are associated with a wide range of AMF species widely distributed in regions where they are introduced [ 31 , 33 , 39 ]. This may be favorable to inducing the naturalization and expansion process [ 32 , 37 ] and explain why invasive plants do not alter this community. A result of different evolutionary trajectories of invasive plants is the impressive number of different biochemicals produced by plants [ 6 ], over 100,000 different low-molecular-mass natural products have been identified in plants [ 4 , 13 ]. Unexpectedly, we found that bacterial diversity was positively affected by allelopathic substances produced by invasive herbs in temperate regions; this result occurs because the SMC have an adaptation restricted to a few chemical compounds in these regions. When exudates or secondary metabolites from invasive plants enter the soil, bacteria feed on them and increase their diversity, because in the absence of these, they are not able to use organic matter as a source of energy [ 45 ]. The diversity and concentration of secondary metabolites appears to be greater in the tropics than in temperate ecosystems; in fact, its incidence in tropical flora doubles the flora of temperate zones and declines with elevation [ 29 ]. In contrast, in tropical areas, soil microorganisms have been adapted to a wide variety of substances over time, and are able to tolerate a wide variety of exudates, thus maintaining the diversity and abundance of organisms. Closely related plants and soil microorganisms may differ in their sensitivity to the same biochemical and allelochemical substances when they are from different continents, while distantly related species may have converged to similar sensitivities if they are from the same region. This suggests that plants and soil microorganisms can evolve tolerance to the unique rhizosphere biochemistry of co-occurring species with independent phylogenetic histories [ 6 ]. Physiological traits that contribute to the establishment and expansion of invasive plants can have an impact on ecosystem processes. Allison and Vitousek [ 2 ] evaluated initial leaf litter properties, decomposition rates, and nutrient dynamics in 11 forest plant species of the Hawaiian Islands. They found a 50-fold variation in litter decomposition rates, decomposition in native plants decreased (0.2–2.3 yr −1 ) and that of invasive plants increased (1.4–9.3 yr −1 ) in the forest. In another study conducted in a Long Island forest, New York, USA, Ashton et al. [ 3 ] evaluated the differences in decomposition of the litter of native and exotic plants in mesic hardwood forests. They found that litter decomposition and released nitrogen of alien species were significantly faster than in litter from native species, and the litter from all species types decomposes substantially faster at invaded sites in the forest. The greatest decomposition of the leaves of invasive plants in forest is associated with high specific leaf areas, rapid growth rates, and high leaf nutrient concentrations, which improve leaf litter quality and increase decomposition rates and nutrient cycling [ 2 ]. These results suggest that invasion by exotic plant species in forests alters the decomposition and nutrient cycle of soil, regardless of differences in litter quality specific to native and exotic species [ 3 ]. The addition of new resources that come from invasive plants brings benefits for bacterial diversity. The contribution of these resources could promote short-term changes in the microbial community of the soil [ 23 ] and bacterial reproduction [ 26 ]. The decomposition rate of organic matter in the forest floor is higher than in other ecosystems because there is more moisture and greater presence of disintegrating fauna that will fractionate the material, since vegetation is denser and there is greater microbial potential, which will be the main factor responsible for mineralization [ 21 ]. The characteristics of invasive plants and the taxonomic group they belong to have a significant impact on the diversity of the microbial community [ 35 ]. The plant life form had a positive and significant effect on the diversity of bacteria. A possible explanation for these results is that bacteria recognize the substances produced by invasive herbs as resources, thus enabling an increase of their diversity, further studies are required to test this hypothesis."
} | 2,894 |
32796915 | PMC7429504 | pmc | 8,519 | {
"abstract": "Dark fermentative biohydrogen (H 2 ) production could become a key technology for providing renewable energy. Until now, the H 2 yield is restricted to 4 moles of H 2 per mole of glucose, referred to as the “Thauer limit”. Here we show, that precision design of artificial microbial consortia increased the H 2 yield to 5.6 mol mol −1 glucose, 40% higher than the Thauer limit. In addition, the volumetric H 2 production rates of our defined artificial consortia are superior compared to any mono-, co- or multi-culture system reported to date. We hope this study to be a major leap forward in the engineering of artificial microbial consortia through precision design and provide a breakthrough in energy science, biotechnology and ecology. Constructing artificial consortia with this drawing-board approach could in future increase volumetric production rates and yields of other bioprocesses. Our artificial consortia engineering blueprint might pave the way for the development of a H 2 production bioindustry.",
"introduction": "Introduction Microorganisms thrive in almost all habitats on Earth, where they fulfil important ecosystem functions as complex and highly dynamic microbial communities 1 , 2 . Microbial communities exist in high levels of biodiversity, enabling cooperation and interaction among its members in functional metabolic networks 3 . Compared with mono-cultures, a microbial consortium empowers complex metabolic tasks due to the multitude of possible metabolic reactions and interaction possibilities, which are based on mutualism, commensalism or neutralism 4 , 5 . The streamlined syntrophic interactions or commensal relationships among the microorganisms in microbial consortia were shown to enable an efficient utilization of unrefined substrates, such as cane molasses or beet molasses 6 , 7 , to resist to environmental stressors, e.g., temperature fluctuations or heavy metal exposure 7 – 9 , and to display high productivity or yield 10 , 11 . In nature, a modest undefined consortium may contain thousands of species 12 . However, for efficiently performing bioconversions in natural or artificial ecosystems, the specific metabolic reactions of individual species in the consortium are more relevant than the species richness 13 , 14 . In environmental, biopharmaceutical or energy biotechnology, most of the bioprocesses are developed and optimized through targeted bioprocess development, utilizing metabolically engineered or wild-type organisms, or even undefined microbial consortia of organisms. The emphasis lies in the optimization of productivity and/or yield by using different types of bioreactors and organisms/undefined consortia. However, every organism, even a metabolically engineered organism, possesses specific metabolic bottlenecks, which limit a full substrate to target product conversion. In many cases, the production of the target compound is accompanied by excretion of several metabolic byproducts, which balance cellular homoeostasis, reducing yield and/or productivity. Moreover, bioprocess development relies on established bioreactors and cultivation pipelines. Synthetic or artificial microbial consortia are regarded as part of the solution to debottleneck the inherent physiological limitations of wild-type or metabolically engineered mono-culture and undefined consortia bioprocesses, such as enabling the breakdown of complex carbon sources 15 , efficient substrate utilization 16 , reducing byproduct inhibition through operational stability 17 and high productivities 18 . This can be achieved through selection, design and assembly of microorganisms with specific metabolic (e.g., cellulose utilisers) or ecological (e.g., biofilm forming) functions. In addition, by employing an artificial consortium of selected microorganisms, precision design of a defined microbial co- or multi-culture provides a comprehensive understanding of organismal interactions and allows examining the molecular and eco-physiological basis of community-level functions 19 , 20 . The developments in the field of artificial microbial ecosystem engineering allowed advancing in the aspects of ecology, such as soil bioremediation 21 and biotechnology, e.g., fine chemical 22 , 23 , biopolymer 24 , enzyme 25 , food additive 26 , antimicrobial 27 , biofuel 28 and biohydrogen production 29 – 31 . However, to achieve supreme efficiency of the bioprocess, a precision design strategy to form an artificial consortium of selected microorganisms was not yet considered. Molecular hydrogen (H 2 ) is considered as an alternative source of energy. Biological production of H 2 , referred as biohydrogen production, provides a sustainable and environmentally friendly method for energy generation 32 – 34 . Dark fermentative H 2 production is promising due to high H 2 evolution rates (HERs) compared to photobiological H 2 production processes 32 , 33 . However, the yield of H 2 per substrate consumed (Y (H2/S) ) is limited by metabolic constraints of dark fermentative H 2 -producing microorganisms. According to the theoretical limit, the so-called “Thauer limit”, 4 mol H 2 can be produced per 1 mol of glucose consumed during dark fermentation when acetate is produced as the byproduct 35 . Depending on the microbial group, H 2 formation may occur either via the pyruvate-formate-lyase (PFL) pathway or the pyruvate ferredoxin oxidoreductase (PFOR) pathway 32 . The PFL pathway is operative in Enterobacteriaceae. In this pathway, pyruvate is converted into acetyl-CoA and formate. Formate is either shuttled out of the cell or it can be split into carbon dioxide (CO 2 ) and H 2 by formate hydrogen lyase 32 . The PFOR pathway is operative in Clostridiaceae during H 2 production, which occurs through the action of [NiFe]- and/or [FeFe]-hydrogenases 36 , 37 . Up to now, dark fermentative biohydrogen producing wild-type or metabolically engineered mono-cultures were not successful in improving Y (H2/S) beyond the Thauer limit 29 , 38 , 39 . Therefore, to boost Y (H2/S) , undefined microbial consortia or defined co- and multi-cultures of H 2 -producing microbes were examined in complex or defined medium 40 – 43 . However, control of microbial community composition, media compounds and their concentration through precision design of an artificial microbial consortium were not yet the focus of any study. In our quantitative analysis of pure culture dark fermentative H 2 production, we linked physiological and biotechnological characteristics of H 2 -producing microorganisms through comprehensive meta-data analysis and modelling 32 . Our analysis revealed that Enterobacteriaceae exhibit very high HERs and Clostridiaceae are mesophilic organisms with the highest reported Y (H2/S) on a C-molar level on saccharides. Therefore, we hypothesized that precision design of an artificial microbial consortium composed of Enterobacteriaceae and Clostridiaceae improves Y (H2/S) beyond the Thauer limit. Here we present results from a drawing board-like precision design of artificial microbial consortium of microorganisms with improved HER, and Y (H2/S) beyond the Thauer limit, of two H 2 -producing species, the facultative anaerobic Enterobacter aerogenes and the obligate anaerobic Clostridium acetobutylicum . For the design of this defined artificial consortium, three different major community function-determining parameters were individually and syntrophically investigated: initial substrate concentration of glucose or cellobiose, designing and optimizing a mutual medium, and control of the activity and concentration of initial cell densities. First, initial optimum substrate concentration was investigated for individual strains and a mutual defined medium was designed by applying Design of experiments (DoE). Then, different consortia were created using active inoculum with different initial cell densities of each microorganism. Our interdisciplinary research combines physiology, ecology and biotechnology, provides valuable insights into the ecosystem functionality and enhances H 2 production by constructing a defined artificial consortium.",
"discussion": "Discussion Renewably produced H 2 could be implemented as one of the main energy carriers of the twenty-first century 51 . To gain biological H 2 production at the theoretical Y (H2/S) , different methods (e.g., reactor configurations 52 , metabolic engineering 53 , modelling and optimization 54 , statistical analysis 33 , pre-treatment strategies for spore germination, nutrient formulations, substrate composition and concentration 55 ) were proposed and/or already investigated. Using H 2 -producing defined or undefined consortia was considered as one of the auspicious approaches 38 . However, an undefined consortium fetches many technical problems due to the reaction complexity, process kinetics, difficulties of optimization and various process parameters (e.g., pH and temperature), as well as the ecological and functional aspects of the system 10 . Furthermore, H 2 formation is not the prime aim of microbes, but the microorganism aims on optimizing the energy yield. These two aspects might be in conflict to a certain extent, but a defined consortium allows better control regarding H 2 formation, whereas an undefined mix of microorganisms will tend to optimize energy formation. Therefore, an artificial/defined consortium, with well-studied microorganisms, is essential to further understand the relationship among microorganisms and to allow sophisticated process control, as the physiologies of the members of microbial community can be examined in depth and individually as well as mutually optimized. So far, artificial dark fermentative H 2 -producing consortia were utilized in over 40 studies, which we summarized with respect to dark fermentative H 2 production and their main parameters (Supplementary Data 1 ). The highest reported Y (H2/S) was 4.42 mol (H2) mol −1 (glucose) , which corresponds to 0.74 mol (H2) C-mol −1 ) from a consortium of Caldicellulosiruptor saccharolyticus and Caldicellulosiruptor owensensis 56 , followed by a thermophilic consortium composed of C. saccharolyticus and Caldicellulosiruptor kristjanssonii comprising 3.8 mol (H2) mol −1 (C6 sugar-equivalent) (0.63 mol (H2) C-mol −1 ) 29 . Both studies were conducted on complex medium containing yeast extract. The highest Y (H2/S) reported from a mesophilic consortium of Enterobacter cloacae and Bacillus cereus , was a Y (H2/S) of 3 mol (H2) mol −1 (glucose) , which is the equivalent to 0.5 mol (H2) C-mol −1 57 , followed by a consortium of E. aerogenes and Clostridium butylicum , with a Y (H2/S) of 2.7 mol (H2) mol −1 (glucose) (0.45 mol (H2) C-mol −1 ) 58 . Our study is the first of its kind, which considered and integrated results from several physiological, ecological and biotechnological levels: (1) meta-data analysis and modelling pipeline of dark fermentative H 2 producers 32 ; (2) physiological, ecological and biotechnological aspects of mono- and co-culture design; (3) optimization of H 2 production by subsequently investigating the effect of substrate concentration on growth and gas production; (4) employing DoE method to design a mutual defined medium (E-medium); and, finally, (5) engineering a defined artificial consortium by examining different initial ratios of microorganisms in defined medium. Here, we present an optimum consortium comprising two species with an inoculum ratio of 1 : 10,000 ( E. aerogenes : C. acetobutylicum ) with a Y (H2/S) of 5.58 mol (H2) mol −1 (glucose) (0.93 mol (H2) C-mol −1 ) and 4.38 mol (H2) mol −1 (C6 sugar-equivalent) (0.73 mol (H2) C-mol −1 ) on glucose and cellobiose, respectively. This precisely engineered consortium comprised the highest ever reported Y (H2/S) and clearly surpassed the Thauer limit. Our findings point at a yet unidentified synergistic effect of the two strains that improves H 2 production. The E-medium composition had a major effect on the metabolism of the microorganisms. The obtained metabolic byproducts highlight the active metabolic routes of the microorganisms. On Enterobacter -specific medium, we showed that byproducts of E. aerogenes were mainly acetate and ethanol 59 . In our study, the E. aerogenes mono-culture produced high amounts of 2,3-butanediol, which is an industrial chemical and liquid fuel and is used in food, cosmetics and medicine industries 60 , 61 . It has been reported that 2,3-butanediol is produced by E. aerogenes under molecular oxygen-limiting and anaerobic conditions 62 , 63 , and that the initial acetate source induces butanediol production by catalysing the breakdown of pyruvate to butanediol 64 . E-medium contains acetate and that might be the reason of production of this compound. In addition, a higher level of CO 2 was observed during mono-culture E. aerogenes cultivation, which is again confirming butanediol fermentation. Production of 2,3-butanediol could not be detected during C. acetobutylicum mono-culture cultivation (Fig. 3 and Table 2 ). Compared to the mono-culture experiments, it was observed that during the consortium experiments the release of metabolic end products of the two species changed. Lower amounts of 2,3-butanediol were also detected during the consortium cultivation on glucose compared to mono-culture of E . aerogenes (Fig. 3 and Table 2 ). This is another indication of an operative consortium where both members were metabolically functional. Moreover, during the consortium experiments production of acetic acid was higher and ethanol production was decreased, which most likely provided room for H 2 production. Another aspect of the precision design of the medium was the PB capacity. At pH of 5.5, the consortium was able to produce H 2 due to the activity of C. acetobutylicum . In biohydrogen production, pH values < 4.5 lead to changes in the metabolic pathways towards decreased concentrations of undissociated forms of organic acids, which cause possible inhibition of hydrogenase activity 65 , 66 , affecting ferredoxins’ capacity to donate electrons to reduce protons 66 , 67 and affect microbial growth 66 , 68 . Low pH also induces the sporulation of C. acetobutylicum , which can be observed in the last time point of our FISH images in Supplementary Fig. 4 . The concentration of C. acetobutylicum (coloured pink) at the time point 5 was drastically decreased at the FISH image (Supplementary Fig. 4 ) and the qPCR reads (Fig. 3 ). Another investigated aspect in this study was the initial cell densities of each microorganism. In previous consortium studies, and most of the cases, an equal suspension volume with an unknown amount of living/active cells of each organism has been used for inoculation 29 , 30 . To our knowledge, this is the first study in which consortium was engineered by introducing active microorganisms at initial cell densities of five orders of magnitude difference into the system. The functional co-existence of two bacteria was shown, when they were introduced to the system at all aforementioned inoculum ratios. Expectedly, cell densities of microorganisms and gas production values differed at each inoculum ratio (Supplementary Fig. 3 ). This was an additional indication of the importance for precision design of the consortium including biotic and abiotic factors. Furthermore, H 2 production was initiated earlier in the consortium (during the first 16 h on glucose, 22.5 h on cellobiose) compared to both mono-culture cultivations on each of the substrates. These findings clearly indicate that the engineered consortium with an inoculum ratio of 1:10,000 ( E. aerogenes : C. acetobutylicum ) reached higher HER values on different substrates; thus, H 2 production kinetics are superior over mono-cultures (Fig. 3 ). This study presents an interdisciplinary approach to improve H 2 production beyond the Thauer limit from the molecular to the process level, and enlightens a systematic and engineering understanding and description of the kinetic and mechanistic aspects, which are responsible for design and definition of this efficient artificial microbial consortium. Constructing the consortium with this approach could also improve the productivity/yield of natural or undefined consortia and provide controllable, stable, predictable biotechnological processes over currently existing systems. Precision design of microbial communities might be employed for the targeted enrichment of microorganisms in undefined microbial populations or for the restoration of microbial ecosystems in plant, animal and human health, or in bioremediation. Design of synthetic microbial communities for the targeted conversion of complex biopolymers or surplus electricity to biofuels or intermediate storage molecules such as formic acid will benefit from the specific development of communities of well-characterized pure cultures with known growth, substrate uptake and production kinetics, which are aligned by selecting appropriate concentrations of substrate, pH, reduction potential, salt concentration, inoculum size and co-substrate availability, or the mutual exchange of metabolic byproducts between the syntrophic partners in the synthetic microbial community. The present study is a major leap forward in the design of an artificial microbial consortium through precision engineering. Our improvement route is unprecedented and delivers an active, balanced and highly functional co-existence of two bacteria with improved H 2 production kinetics. The H 2 production characteristics of this defined artificial consortium is superior compared to any mono-, co- or multi-culture system reported to date. The system could be further improved to enhance H 2 production by introducing other microorganisms into the consortium, and the stability of the system can be boosted by H 2 milking technology 69 , 70 , or can be combined with methanogenic archaea to stimulate syntrophic growth. Moreover, precision design of an artificial microbial consortium could even serve as a template for conversion of cellulosic biomass to gaseous and liquid biofuels. Our blueprint for a precision design consortium could hence be further extended for the development of consolidated bioprocesses for targeted conversion of lignocellulosic biomass to liquid biofuels, for the development of start-up communities in anaerobic digestion, for the conversion complex gas mixtures, food waste utilization or (bio)plastic recycling. In conclusion, the precision engineered consortium exhibited highly efficient H 2 production from glucose and cellobiose compared to the mono-cultures of either microorganism under optimal conditions or compared to any consortium reported in literature. Our drawing board-like design of a defined artificial microbial consortium of microorganisms improved HER beyond reported values. The engineered consortium breaking the Thauer limit displayed 6.6- and 2.8-fold higher maximum Y (H2/S) on glucose and 18.3- and 1.7-fold higher maximum Y (H2/S) on cellobiose compared to mono-cultures of E. aerogenes and C. acetobutylicum , respectively. The precision design of artificial microbial consortia, which considers results from a priori physiological and biotechnological knowledge from meta-data analysis will lead to a breakthrough in biotechnology by improving productivity and yield. However, this study indicates that the precision design of artificial microbial consortia might only be efficacious when nutrient demands of the individual members are individually and mutually aligned with the eco-physiological characteristics of the organisms. The eco-physiological requirements of microorganisms in undefined ecosystems have to be considered at a strain level to be able to improve the performance of the individual players in the community and to achieve high production rates and yields."
} | 4,992 |
22830653 | null | s2 | 8,520 | {
"abstract": "Scaffolded DNA origami is a widely used technology for self-assembling precisely structured nanoscale objects that contain a large number of addressable features. Typical scaffolds are long, single strands of DNA (ssDNA) that are folded into distinct shapes through the action of many, short ssDNA staples that are complementary to several different domains of the scaffold. However, sources of long single-stranded DNA are scarce, limiting the size and complexity of structures that can be assembled. Here we demonstrated that dsDNA (double-stranded DNA) scaffolds can be directly used to fabricate integrated DNA origami structures that incorporate both of the constituent ssDNA molecules. Two basic principles were employed in the design of scaffold folding paths: folding path asymmetry and periodic convergence of the two ssDNA scaffold strands. Asymmetry in the folding path minimizes unwanted complementarity between staples, and incorporating an offset between the folding paths of each ssDNA scaffold strand reduces the number of times that complementary portions of the strands are brought into close proximity with one another, both of which decrease the likelihood of dsDNA scaffold recovery. Meanwhile, the folding paths of the two ssDNA scaffold strands were designed to periodically converge to promote the assembly of a single, unified structure rather than two individual ones. Our results reveal that this basic strategy can be used to reliably assemble integrated DNA nanostructures from dsDNA scaffolds."
} | 380 |
29176910 | null | s2 | 8,522 | {
"abstract": "Manganese(IV) oxides, believed to form primarily through microbial activities, are extremely important mineral phases in marine environments where they scavenge a variety of trace elements and thereby control their distributions. The presence of various ions common in seawater are known to influence Mn oxide mineralogy yet little is known about the effect of these ions on the kinetics of bacterial Mn(II) oxidation and Mn oxide formation. We examined factors affecting bacterial Mn(II) oxidation by spores of the marine "
} | 130 |
37124296 | PMC10131207 | pmc | 8,523 | {
"abstract": "The intracellular\napplication of DNA nanodevices is challenged\nby their inadequate cellular entry efficiency, which may be addressed\nby the development of amphiphilic DNA nanostructures. However, the\nimpact of the spatial distribution of hydrophobicity in cell entry\nhas not been fully explored. Here, we program a spectrum of amphiphilic\nDNA nanostructures displaying diverse sub-10 nm patterns of cholesterol,\nwhich result in distinct aggregate states in the aqueous solution\nand thus varied cell entry efficiencies. We find that the hydrophobic\npatterns can lead to discrete aggregate states, from monomers to low-number\noligomers ( n = 1–6). We demonstrate that the\nmonomers or oligomers with moderate hydrophobic density are preferred\nfor cell entry, with up to ∼174-fold improvement relative to\nunmodified ones. Our study provides a new clue for the rational design\nof amphiphilic DNA nanostructures for intracellular applications.",
"conclusion": "Conclusions In summary, we in this study demonstrated the programming of sub-10\nnm hydrophobic patterns on DNA nanorods, which enabled the modulation\nof their aggregation state and consequently their cell binding efficiencies.\nWe conclude that DNA nanostructures carrying modestly hydrophobic\ngroups (rather than highly hydrophobic ones) may show optimal cell\nentry efficiency (∼2 orders of magnitude improvement), which\ncould be due to the restricted aggregation with high accessibility\nof hydrophobic groups to cell membranes. This system also shows\ntwo advantages. First, the six-helix DNA\nnanorod possesses addressability with ∼2 nm resolution (i.e.,\nthe diameter of a DNA double-helix) on its end, allowing high-density\npatterning of hydrophobic moieties at the sub-10 nm scale. Compared\nto wide-space distributions (e.g., placed on different sides of a\npolyhedron 14 ), the focused distribution\nof hydrophobic moieties at one end might contribute to the high cell\nentry efficiency (2 orders of magnitude improvement) in a “point\nattack” manner, as suggested in previous studies. 38 Second, the shape and rigidity of the DNA nanorods\nallow for the formation of flower-like oligomers with discrete numbers\nof branches, enabling quantifiable analysis with AFM. This led to\nconclusions that would be difficult to obtain with conventional DNA\nmicelles composed of indeterminate numbers of single-stranded or double-stranded\nDNAs. 39 In future work, we would endeavor\nto reveal the specific cell entry pathways of these amphiphilic structures\nand the underlying mechanisms. 34 , 40 The possibility of\nlysosomal escaping and the potential of them for cytosol delivery\nof theranostic agents would also be investigated. 41 Overall, our study may provide a new clue for the rational\ndesign of effective nanocarriers based on amphiphilic DNA nanostructures\nfor intracellular applications.",
"introduction": "Introduction The demonstration of DNA nanostructures\nin the field of biology\nand biomedicine has rapidly evolved over the past few decades, showcasing\na wide range of possibilities. 1 Given the\nstructural programmability and addressability with nanometer-scale\nresolution, DNA nanostructures can serve as frameworks to organize\nvarious functional moieties such as small molecules, inorganic clusters,\nand nanoparticles, creating nanoagents, nanodevices, or nanorobots\nwith diagnostic and/or therapeutic functionalities. 2 Particularly, they have been found capable of entering\nliving cells in a size- and shape-dependent manner. 3 , 4 Compared\nto many organic/inorganic nanomaterials, DNA nanostructures are intrinsically\nhydrophilic, anionic, and biodegradable in physiological environments,\nlargely alleviating the concerns about undesired aggregation, cytotoxicity,\nand cumulative toxicity in living organisms. 4 , 5 However,\ncompared to conventional amphiphilic nanocarriers (e.g., liposomes\nand polymeric micelles), plain DNA nanostructures generally show poorer\ncell entry efficiency, which limits their intracellular applications. 6 There have been several approaches to improving\nthe cellular uptake of DNA nanostructures by coating them with materials\nsuch as cationic polymers, 7 peptides, 8 and viral capsid proteins. 9 However, these methods usually compromise the site specificity\nof DNA nanostructures, and may have the risk of causing cytotoxicity\nand undesired immune responses. 10 Recently, a number of rigid amphiphilic DNA nanostructures have\nbeen created via decorating hydrophobic moieties on DNA frameworks.\nThese structures have been employed to interact with lipid membranes\nfor a range of applications, such as the construction of synthetic\ntransmembrane receptors, nanopores, and nanochannels, 11 − 14 the assembly of DNA superstructures on lipid layers, 15 − 18 the shaping 19 − 22 and reshaping 23 , 24 of lipid structures with defined\ngeometries, and the rewiring of intercellular connections. 25 , 26 On the other hand, amphiphilic DNA nanostructures can form micelle-like\naggregates under proper conditions via hydrophobic interactions in\naqueous solutions. 27 − 31 According to several recent studies, 32 − 34 amphiphilic DNA nanostructures\nare more effective at entering cells than plain ones, but the correlation\nbetween their aggregate states and cell entry efficiency is yet to\nbe explored. Here, we report the use of DNA nanorods to encode\nsub-10 nm hydrophobic\npatterns on their terminals, which could result in various aggregate\nstates in the aqueous solution, including monodispersed rods and micelle-like\noligomers with discrete rod numbers ( Scheme 1 ). We found that these different aggregate\nstates in turn led to varied cell entry efficiencies. Scheme 1 Concept\nof Amphiphilic DNA Nanorods (ADNRs) Displaying Programmable\nHydrophobic Patterns for Varying Aggregate States and Cell Entry",
"discussion": "Results and Discussion First, we\ndesigned an amphiphilic DNA nanorod (ADNR) comprising\na hydrophilic DNA nanostructure and a cluster of hydrophobic moieties\non one side of the former. The DNA nanostructure is adapted from a\npreviously described rod-shaped DNA six-helix bundle with ∼22\nnm in length and ∼7.5 nm in diameter, which is assembled from\n20 single-stranded (ss-) DNAs. 33 A given\nnumber of cholesterols are located at the ends of certain DNA helices,\nforming a sub-10 nm-scale hydrophobic pattern displayed on the lateral\nplane of one terminal (or x – y plane, shown in Scheme 1 ) of the nanorod. Compared with previously described polyhedral\namphiphilic DNA nanostructures, 27 , 31 this rod-like structure\nhas a long shape (aspect ratio >1), which may resemble the diblock\namphiphilic molecules known with cell entry capability. 35 In this way, we designed a spectrum of ADNRs\nwith hydrophobic patterns containing 1–6 cholesterols with\ndifferent arrangements, termed ADNR-1, ADNR-2ortho, ADNR-2meta, ADNR-2para,\nADNR-3ortho, ADNR-3meta, ADNR-3para, ADNR-4ortho, ADNR-4para, and\nADNR-6, respectively. Specifically, in ADNR-2para, ADNR-3meta, ADNR-4para,\nand ADNR-6, the cholesterols were distributed in two-, three-, four-\nand six-fold rotational symmetry, respectively (referred to as symmetric\nhydrophobic patterns hereafter); whereas the other structures presented\nasymmetric hydrophobic patterns. We anticipated that these different\nhydrophobic patterns may result in various aggregation states of the\nnanorods in aqueous solutions and consequently vary their cell entry\nefficiency. To prepare the ADNRs with the aforementioned hydrophobic\npatterns,\nwe annealed the component ssDNAs (200 nM each) of the six-helix DNA\nnanorod structure in the synthesis buffer, with different subsets\nof them replaced by cholesterol-labeled ssDNAs (detailed in the SI Methods , with the DNA sequences listed in Table S1 and S2 ). We then characterized the resulting\nstructures with polyacrylamide gel electrophoresis (PAGE) and atomic\nforce microscopy (AFM). In the gel images ( Figure S1 ), ADNR-1,\nADNR-2meta, and ADNR-2para presented a clear band each, with the migration\nrate approximate to that of the unmodified DNA nanorod (close to the\n600-bp marker band), confirming the formation of expected nanorod\nstructures, whereas ADNR-2ortho, ADNR-3ortho, and ADNR-3para presented\nobvious bands with higher migration rates, indicating the existence\nof incomplete byproducts. Meanwhile, when the cholesterol number >2,\nthe monomer yield decreased significantly and many of the products\nwere retarded in the gel loading wells, suggesting the formation of\naggregates under such conditions. The fractions of aggregates estimated\nby measuring gel band intensities were ∼30.1% (ADNR-3ortho),\n28.1% (ADNR-3para) 56.2% (ADNR-3meta), 53.1% (ADNR-4ortho), 60.6%\n(ADNR-4para), and 53.9% (ANDR-6), respectively, suggesting that their\naggregate states varied. Under AFM ( Figure 1 a and Figure S2 ), we observed that ADNR-1,\nADNR-2para, and ADNR-2meta appeared mostly as monodispersed rod-like\nparticles, with a height similar to that of the unmodified DNA nanorods.\nADNR-2ortho formed particle-like structures with a lower height, suggesting\nthat the structures were not intact six-helix nanorods, in agreement\nwith the gel electrophoresis. ADNR-3para and ADNR-3ortho were mostly\nmonomers, with some irregular aggregates ( Figure S2 ). Interestingly, ADNR-3meta, ADNR-4ortho, ADNR-4para, and\nADNR-6 formed many flower-like oligomers with ∼3 (2–3,\nwith 3 the highest frequency), 4 (2–5), 4 (2–5) and\n5 (3–6) branches, respectively (shown in Figure 1 b, c and Figure S2 ), which should be the head-to-head aggregates of the ADNRs driven\nby hydrophobic interaction. Their branches could be unambiguously\nresolved by AFM, allowing us to count the rod number per oligomer. Figure 1 AFM of\nADNRs series structures. (a, b) Representative AFM images\nof ADNRs in the states of (a) monomers and (b) flower-like oligomers,\nwith corresponding line profiles across the structures (marked with\nblank dashes). Scale bar, 20 nm. (c) Frequency distributions of the\nbranch numbers of the flower-like oligomers in b. Based on the above observation, we found that the arrangement of\ncholesterols could affect the formation of the six-helix nanorods\nand their aggregate states. In general, symmetric cholesterol patterns\n(e.g., ADNR-2para, and ADNR-3meta) showed higher yields of monomers\nor flower-like oligomers compared to asymmetric ones with equivalent\ncholesterol numbers (ADNR-2ortho, ADNR-2meta, ADNR-3ortho, and ADNR-3para),\nparticularly when the cholesterol number was 2–3. The reduced\nyields of well-organized structures resulting from asymmetric cholesterol\npatterns may be due to the unevenly high local hydrophobicity, which\ncould interfere with DNA self-assembly in aqueous solutions. On the other hand, we could conclude that the rod number per oligomer\nis generally positively correlated with the cholesterol number in\neach ADNR, which could be attributed to the hydrophobicity increasing\nwith the cholesterol number. When the cholesterol number per ADNR\nwas <3, the concentration of ADNRs in our study was below their\ncritical micelle concentration (CMC), thus mostly presenting as monomers;\nwhereas when the cholesterol number was >3, flower-like oligomers\nformed. Interestingly, oligomers with a rod number >6 were rare\nin\nour observation, suggesting that the scale of amphiphilic assembly\nmight be self-limited due to the strong steric hindrance arising from\nthe rigid DNA frameworks. Collectively, by programming the sub-10\nnm hydrophobic patterns of ADNR, we could obtain different aggregates\nstates with discrete rod numbers (1–6) in the aqueous solution. To investigate the cell entry of the ADNRs with different aggregate\nstates, we incubated Alexa488-labeled ADNRs (100 nM each) with MCF-7\ncells (1640 medium, 0.5% serum, 37 °C, 5% CO 2 ) for\ncytometric analysis. Here, we chose ADNR-1, ADNR-2para, ADNR-3meta,\nADNR-4para, and ADNR-6 for comparison, which could effectively yield\nstructures in the aqueous solution with rode numbers of ∼1,\n1, 3, 4, and 5, respectively. In this experiment, the molar concentration\nof the ADNR monomers was held constant at 100 nM. First, we analyzed\nthe ensemble results measured by flow cytometry ( Figure 2 ). We found that all the ADNRs\nexhibited much higher cell binding than unmodified DNA nanorods. After\n8 h of cell incubation, ADNR-1, ADNR-2para, ADNR-3meta, ADNR-4 para,\nand ADNR-6 resulted in ∼34-, 174-, 122-, 99-, and 20-fold increases\nin cellular fluorescence intensity, respectively, relative to the\nunmodified DNA nanorods ( Figure 2 a, b). In addition, the percentages of fluorescently\npositive cells (fluorescence intensity >10 5 a.u.) at\n8\nh were ∼0.07% (unmodified DNA nanorods), 96.8% (ADNR-1), 99%\n(ADNR-2para), 99.2% (ADNR-3meta), 99.6% (ADNR-4para), and 81.9% (ADNR-6),\nrespectively ( Figure 2 c), suggesting that the vast majority of the cells could bind to\nthese ADNRs. Collectively, these results confirmed that amphiphilic\nDNA structures are overall favored in cell binding and entry, in agreement\nwith previous studies. 32 , 33 , 36 , 37 Notably, ADNR-2para and ADNR-3meta (mostly\nformed monomers and three-branch oligomers) led to higher cell fluorescence\nthan others, suggesting that the cell binding efficiency was not unidirectionally\nincreasing with the rod number. Figure 2 Effect of the cholesterol pattern on cell\nbinding efficiency. (a)\nRepresentative cytometric fluorescence intensity distributions indicating\ncell binding of different ADNRs at 8 h. (b) Mean fluorescence intensities\n(normalized to the highest fluorescence in each replicate) of the\ncytometric results at 1, 2, 4, and 8 h, respectively. Error bars,\nSD of three independent tests. * p < 0.05, ** p < 0.005, *** p < 0.0001, determined\nby one-way ANOVA comparison. (c) Representative scatter plots of cell\nfluorescence distributions after being treated with the structures\nfor 8 h. Gray spots, blank cells. To gain a deeper understanding of the cell entry process of these\nstructures, we employed confocal imaging to inspect individual cells\nthat were treated with the representative ADNRs ( Figure 3 a and Figure S3 ). The results showed that at 1 h, intense fluorescence of\nADNR-2para and ADNR-3meta appeared on the cell membranes and partially\noutlined the cells, whereas the fluorescence signals from other structures\nwere weak. At 4 h, the membrane fluorescence signals of ADNR-2para\nand ADNR-3meta were further enhanced; some signals of ADNR-2para appeared\ninside the cell; while intense signals of ADNR-4para appeared both\non the cell membranes and in the cells. At 8 h, the fluorescence speckles\nwere mostly located inside the cells; meantime, sparse fluorescence\nspeckles of ADNR-6 also appeared inside the cells. During the whole\nobservation, the cellular fluorescence of unmodified nanorods remained\nminimal. These results revealed the processes of the cell entry of\nthese structures. In general, ADNR-2para and ADNR-3meta could more\nrapidly and efficiently enter the cells than other structures; the\nfurther increase of cholesterol number and rod number (ADNR-4para\nand ADNR-6) led to decreased cell entry kinetics. These results were\nconsistent with the flow cytometric measurements. In addition, we\ncan conclude that after 8 h incubation, these structures were mostly\nlocated inside the cells. Hence, the fluorescence intensities obtained\nby flow cytometry at 8 h could quantitatively reflect their cell entry\nefficiency. In addition, we evaluated the cytotoxicity of the ADNRs\nusing CCK8 assay. The results showed that even with much enhanced\ncell entry efficiencies (including ADNR-2para and ADNR-3meta), these\nstructures at the dosage used in this study (100 nM) showed no suppression\nto cell viability after 24 h of cell incubation, suggesting minimal\ncytotoxicity of these structures ( Figure S4 ). Figure 3 Cell entry of ADNRs with different aggregate states. (a) Confocal\nimages of MCF-7 cells incubated with ADNRs for 1, 4, and 8 h, respectively.\nBlue, nucleus stained by Hoechst 33258. Green, ADNRs labeled with\nAlexa 488. Scale bar, 10 μm. (b) Cartoon of the presumed mechanism\nunderlying the difference in cell entry efficiency. These results together suggest that the hydrophobic pattern\ndisplayed\non ADNR plays an important role in its cell entry. Although hydrophobicity\nin general benefits cell entry, their correlation is not unidirectional.\nThe monomers or low-number oligomers with modest hydrophobicity (2\nor 3 cholesterols) show high cell entry efficiencies (∼2 orders\nof magnitude than the unmodified structure). We reason that the modest\nhydrophobicity allows the monomers/oligomers to flip and expose the\nhydrophobic groups to cell lipid membranes easily, leading to effective\ncell binding and entry ( Figure 3 b). On the contrary, an overhigh hydrophobicity (e.g., 6 cholesterols)\nmight lead to more stable micelle-like structures, which are difficult\nto collapse and expose the buried hydrophobic moieties to the cell\nmembrane, leading to decreased cell entry efficiency."
} | 4,232 |
35460693 | PMC9133646 | pmc | 8,524 | {
"abstract": "The mildly thermophilic purple phototrophic bacterium Allochromatium tepidum provides a unique model for investigating various intermediate phenotypes observed between those of thermophilic and mesophilic counterparts. The core light-harvesting (LH1) complex from A. tepidum exhibits an absorption maximum at 890 nm and mildly enhanced thermostability, both of which are Ca 2+ -dependent. However, it is unknown what structural determinants might contribute to these properties. Here, we present a cryo-EM structure of the reaction center–associated LH1 complex at 2.81 Å resolution, in which we identify multiple pigment-binding α- and β-polypeptides within an LH1 ring. Of the 16 α-polypeptides, we show that six (α1) bind Ca 2+ along with β1- or β3-polypeptides to form the Ca 2+ -binding sites. This structure differs from that of fully Ca 2+ -bound LH1 from Thermochromatium tepidum , enabling determination of the minimum structural requirements for Ca 2+ -binding. We also identified three amino acids (Trp44, Asp47, and Ile49) in the C-terminal region of the A. tepidum α1-polypeptide that ligate each Ca ion, forming a Ca 2+ -binding WxxDxI motif that is conserved in all Ca 2+ -bound LH1 α-polypeptides from other species with reported structures. The partial Ca 2+ -bound structure further explains the unusual phenotypic properties observed for this bacterium in terms of its Ca 2+ -requirements for thermostability, spectroscopy, and phototrophic growth, and supports the hypothesis that A. tepidum may represent a “transitional” species between mesophilic and thermophilic purple sulfur bacteria. The characteristic arrangement of multiple αβ-polypeptides also suggests a mechanism of molecular recognition in the expression and/or assembly of the LH1 complex that could be regulated through interactions with reaction center subunits.",
"discussion": "Discussion The multiple α/β-containing LH1 complex of A. tepidum offers a unique model system to investigate the minimum structural requirement for Ca 2+ -binding in bacterial photosynthetic LH1 complexes. This was challenging with the fully Ca 2+ -bound LH1s reported previously from T. tepidum ( 8 ) and Trv. strain 970 ( 14 ) because all LH1 polypeptides in the complexes of these bacteria participated in Ca 2+ -binding and have similar amino acid sequences around the Ca 2+ -binding sites, making it difficult to distinguish the roles of individual residues. However, unlike the LH1 complexes from T. tepidum and Trv. strain 970, the A. tepidum LH1 complex contains both Ca 2+ -bound and Ca 2+ -free αβ-subunits in a single ring, allowing for an unambiguous identification of the key residues responsible for Ca 2+ -binding. Of the total 16 A. tepidum LH1 α-polypeptides, only six copies of the α1-polypeptide bind Ca 2+ to form Ca 2+ -binding sites with β1- or β3-polypeptides. Three amino acids (Trp44, Asp47, and Ile49) in the C-terminal region of A. tepidum α1-polypeptide were identified as the Ca 2+ -ligating residues that form a Ca 2+ -binding WxxDxI motif. This structural motif represents the minimum requirement for an LH1 polypeptide to be able to bind Ca 2+ and is conserved in all Ca 2+ -bound α-polypeptides with reported structures. It is notable that the WxxDxI motif is also present in the LH1 α-polypeptides of Trv. winogradskyi DSM 6702 T and Lamprocystis purpurea ( Amoebobacter purpureus ) DSM 4197 T ( 24 ). Whether the LH1 complexes from these purple sulfur bacteria bind Ca 2+ is unknown, but the presence of this key motif is suggestive. The partially Ca 2+ -bound structure of the A. tepidum LH1 well explains its intermediate properties in terms of thermostability, spectroscopy, and phototrophic growth ( 18 ) compared with the fully Ca 2+ -bound LH1 of the thermophilic T. tepidum and the Ca 2+ -free LH1 of the mesophilic A. vinosum . The native A. tepidum LH1–RC is much more thermostable than that of A. vinosum ( 18 ). The purified A. tepidum LH1–RC has an LH1-Q y transition at 890 nm ( Fig. S1 ) close to that of A. vinosum (884 nm) but far from that of T. tepidum (915 nm). The Ca 2+ -requirements for an LH1-Q y redshift, LH1–RC thermostability, and phototrophic growth of A. tepidum were also less strict than those of T. tepidum ( 18 ). One of the important interactions underling these intermediate properties is the hydrogen bonding between BChl and LH1 polypeptides that is strongly affected by Ca 2+ -binding. The hydrogen bonding strengths between the BChl C3-acetyl group and the Trp residues in the C-termini of LH1 polypeptides measured by our Raman spectroscopy revealed good correlation with the LH1-Q y position ( 18 ). That is, the fully Ca 2+ -bound T. tepidum LH1 forms the strongest hydrogen bonds and exhibits the largest Q y redshift followed by the partially Ca 2+ -bound A. tepidum LH1; hydrogen bonding strength in the latter is much reduced from that of T. tepidum but is larger than that of the Ca 2+ -free A. vinosum LH1. These results can be well explained by our structure. The average hydrogen-bonding lengths between the BChl C3-acetyl oxygen and the Trp nitrogen (ε1) are 2.9 Å and 3.1 Å for the A. tepidum α- and β-polypeptides, respectively. These lengths are longer than the 2.8 Å and 2.9 Å measured for the corresponding T. tepidum α- and β-polypeptides (PDB: 5Y5S ), confirming that the hydrogen bonds formed in A. tepidum LH1 are weaker than those in T. tepidum . The weakened hydrogen bonding incorporated with a partial Ca 2+ -binding in A. tepidum LH1 likely contribute to its relatively lower thermostability than that of T. tepidum LH1. Our structural information combined with biochemical and spectroscopic analyses indicates that these properties are correlated with the Ca 2+ content of the LH1 and is consistent with previous conclusions that Ca ions tightly lock the LH1 rings by forming a Ca 2+ -connected network that strongly contributes to both thermostability and the LH1-Q y redshift ( 8 , 14 ). Our results also support the hypothesis that A. tepidum is a “transitional organism” in the sense of bridging the phenotypic gap between mesophilic and thermophilic purple sulfur bacteria. Besides evidence presented from comparative structural analyses of LH1–RC complexes, this conclusion is supported by comparative physiological and genomic studies of A. tepidum ( 17 ). Effects of charge distribution near the BChl molecules on the LH1-Q y have been well investigated ( 19 , 25 , 26 , 27 ). There is a correlation between the direction of the Q y change and the location and sign of the point charge: positive charges near pyrrole ring I of BChl or negative charges near ring III result in large red-shifts, whereas reversed charges at these locations cause blue-shifts ( 26 , 28 , 29 , 30 ). The Ca ions in A. tepidum and T. tepidum LH1 are located at positions with a distance of ca. 11 Å from the rings I of BChl a ( Figs. 3 B and 5 A ), as measured from the BChl C3-acetyl oxygen. Substitution experiments revealed that these Ca ions indeed alter the electrostatic interaction network and in doing so, influence the LH1 Q y transition ( 30 ). On the other hand, energy transfer from the low-energy LH1 BChl a to the RC has also been investigated mainly using the largely redshifted T. tepidum LH1 and its derivatives ( 31 , 32 , 33 ). Despite an energetically “uphill” process, the energy transfer rate has been shown to be comparable to those for other phototrophic bacteria that show “normal” LH1-Q y absorptions at ∼880 nm, indicating that efficiency of energy transfer from the redshifted LH1 to RC is not significantly compromised by these low-energy pigments. The arrangement of multiple α/β-polypeptides within the A. tepidum LH1 ring is also unique. It differs from that of another multiple α/β-containing LH1 complex from Trv. strain 970 ( 14 ) in (i) higher specificity of pairing of the α/β-subunit and (ii) grouping of the α-polypeptides around the RC. The A. tepidum α1-polypeptides specifically pair with β3-polypeptides to form face-to-face subunits, and similarly the α2-polypeptides specifically pair with β1-polypeptides ( Figs. 2 B and S10 A ). There are no α1/β1- and α2/β3-subunits in the A. tepidum LH1 complex. By contrast, multiple copies of both α1- and α3-polypeptides in Trv. strain 970 form subunits with either β1- or β4-polypeptides ( Fig. S10 B ) ( 14 ). The multiple α1/β3 and α2/β1 pairs in A. tepidum LH1 form two groups, respectively, and are positioned on the opposite sides of the RC core with the α2-polypeptides in close proximity to the RC L- and M-subunits in the transmembrane region. By contrast, all α1-polypeptides in Trv. strain 970 are aligned on one side of the RC and all α3-polypeptides are located on the opposite side with the single copy of α2- and α4-polypeptides in close proximity to the RC in the transmembrane region ( Fig. S10 B ). In the A. tepidum LH1, a split of the clustered Ca 2+ -bound α1/β3(β1) pairs into two groups may also weaken the effects of Ca 2+ -binding on thermostability and LH1-Q y because both properties are highly collective over all αβ pairs in the LH1 ring. The characteristic arrangement of the A. tepidum LH1 αβ-polypeptides is likely regulated in its expression and/or assembly processes through multiple specific interactions within the LH1 complex (among α- and β-polypeptides through N-terminal domains and/or Ca 2+ -binding) and between LH1 and the RC (α-polypeptides with the RC L- and M-subunits in the transmembrane regions and C-terminal domain of the α-polypeptide with the RC C-subunit). In our cryo-EM structure of the A. tepidum LH1 complex, an amino acid insertion of (Thr40) in the C-terminal region of the Ca 2+ -bound α1-polypeptide turned out to have no significant effect on structures near the BChl a -coordinating and Ca 2+ -binding sites ( Fig. 5 ). On the one hand, this implies that the deletion of this residue in T. tepidum α-polypeptides may not be a decisive factor for its LH1-Q y redshift at 915 nm but rather a mechanism to enhance thermostability. On the other hand, insertion of an Ala at the corresponding position in the T. tepidum α-polypeptides resulted in a mutant strain whose LH1 exhibited an absorption maximum at 899 nm ( 23 ). These somewhat conflicting results could be explained by either of two or both possibilities: (i) the genetically modified T. tepidum α-polypeptides may adopt a different conformation in the loop region from that of the native A. tepidum LH1 α1-polypeptide, causing structural changes at the BChl a -coordinating and/or Ca 2+ -binding sites, and (ii) the inserted Ala in the mutant T. tepidum LH1 α-polypeptide may have a different effect from that of Thr in the A. tepidum LH1 α1-polypeptide as seen in the Ca 2+ -binding motif where the sidechain was shown to play an important role for Ca 2+ -binding ( Fig. 2 E ). To clarify this issue, structural details would be required on the native A. vinosum LH1 complex ( Fig. 2 E )."
} | 2,769 |
36001047 | PMC9744196 | pmc | 8,529 | {
"abstract": "Abstract Microbiome studies are often limited by a lack of statistical power due to small sample sizes and a large number of features. This problem is exacerbated in correlative studies of multi‐omic datasets. Statistical power can be increased by finding and summarizing modules of correlated observations, which is one dimensionality reduction method. Additionally, modules provide biological insight as correlated groups of microbes can have relationships among themselves. To address these challenges, we developed SCNIC: Sparse Cooccurrence Network Investigation for compositional data. SCNIC is open‐source software that can generate correlation networks and detect and summarize modules of highly correlated features. Modules can be formed using either the Louvain Modularity Maximization (LMM) algorithm or a Shared Minimum Distance algorithm (SMD) that we newly describe here and relate to LMM using simulated data. We applied SCNIC to two published datasets and we achieved increased statistical power and identified microbes that not only differed across groups, but also correlated strongly with each other, suggesting shared environmental drivers or cooperative relationships among them. SCNIC provides an easy way to generate correlation networks, identify modules of correlated features and summarize them for downstream statistical analysis. Although SCNIC was designed considering properties of microbiome data, such as compositionality and sparsity, it can be applied to a variety of data types including metabolomics data and used to integrate multiple data types. SCNIC allows for the identification of functional microbial relationships at scale while increasing statistical power through feature reduction.",
"discussion": "4 DISCUSSION SCNIC provides a method to measure correlations, find and visualize modules of correlated features, and summarize modules by summing their counts for use in downstream statistical analysis as one method for dimensionality reduction. Using SCNIC with the SMD algorithm for module detection aids in feature reduction in 16S rRNA sequencing data while ensuring a minimum strength of association within modules. As expected, our workflow identified modules in which OTUs tended to be phylogenetically related, especially at relatively high values of R . Using SCNIC, we overall achieved increased statistical power from performing less comparisons, but use of low R ‐value thresholds had the potential to lead to loss of significance by binning loosely correlated features. In these analyses we used SparCC to calculate pairwise correlations in compositional data but Spearman and Pearson are also implemented for cases when the underlying data do not match those well suited for SparCC (e.g., if they are not sparse or with an inverse Simpson index above 13). In these analyses, we also used OTUs as features; however, other microbiome features can be used with SCNIC, such as ASVs, genera, or species defined with a taxonomic classifier, as well as other data types such as metabolome data. SCNIC has also been used in previously published work to perform feature reduction prior to random forest analysis with the microbiome and diverse other data types (Armstrong et al., 2021 ). SCNIC complements existing methods because these either: (1) form correlation networks of microbes for visualization but do not have functionality for selecting and summarizing modules for downstream statistical analysis (Faust & Raes, 2016 ), (2) can select and summarize modules for downstream statistical analysis but are designed for gene expression and not microbiome data (Langfelder & Horvath, 2008 ), only summarize features if they are phylogenetically related (Frøslev et al., 2017 ), or suggest methods for finding modules of correlated microbes but do not provide a convenient implementation (Blondel et al., 2008 ). SCNIC is available both as a stand‐alone application and as a QIIME 2 plugin for easy integration with existing microbiome workflows. SCNIC implements both the LMM algorithm, which had been previously recommended for selecting modules of correlated microbes (Baldassano & Bassett, 2016 ; Jackson et al., 2018 ), and a novel SMD algorithm. The advantage of the SMD algorithm is that all pairs of features in the module have an R ‐value greater than the user‐provided minimum threshold. Using real and simulated data, we showed that SMD produced smaller modules that generally represent sub‐graphs of the larger LMM modules. Since the use of lower R ‐value thresholds similarly produced larger modules including more weakly correlated modules, we speculate that use of LMM might result in a similar trend of identifying more OTUs within significant modules, but with the disadvantage of individually significant OTUs being lost because they are combined with loosely correlated microbes that are not related to the outcome being tested. We illustrate here that varying the R ‐value threshold input by the user has a great impact on the results. However, we have avoided giving specific R ‐value threshold recommendations here, because optimal R ‐values may vary across datasets and data types. Using higher R ‐values thresholds was more likely to identify highly phylogenetically related microbes that likely share overlapping functionality, and in principle could also identify diverse organisms with overlapping niches or highly complementary metabolic functions. Using a lower R ‐value threshold bins a broader community of more loosely correlated features with the risk of bringing together features which should not be grouped and loosing significance of OTUs – as was illustrated in the Great Lakes dataset analysis conducted here. By summarizing correlated features, SCNIC can mitigate overcorrection in multiple test adjustments by reducing the number of taxa and false discovery rate for downstream analysis. However, further work with both real and simulated datasets is required to determine the degree to which network characteristics that are inherent in different microbiome datasets may influence the optimal methods for both selecting and summarizing modules in order to enhance statistical power. The results of our HIV dataset analysis confirm original findings, as well as those of another study (Armstrong et al., 2018 ), but included many new significantly associated taxa. SCNIC also assists in interpretation of microbiome data by identifying correlations among these taxa. Our results recapitulated those of the original publication of these data and previous HIV microbiome studies that all found enrichment of Prevotella with MSM status (Armstrong et al., 2018 ; Dillon et al., 2014 ; Lozupone et al., 2013 ; Noguera‐Julian et al., 2016 ). However, our analyses provide additional insight by identifying correlations between differentiating taxa. For instance, in module‐ 0 , which was more abundant in MSM samples, OTUs assigned taxonomically to the Prevotella genus are correlated with two OTUs identified as Eubacterium biforme (which has recently been renamed Holdemanella biformis ( De Maesschalck et al., 2014 \n ) ). Prevotella copri has previously been associated with increased inflammation (Dillon et al., 2014 ) while in vitro stimulations of human immune cells have found that P. copri did not induce particularly high levels of inflammation but E. biforme did (Lozupone et al., 2013 ). This strong correlation between P. copri and E. biforme in MSM could explain the increased inflammation seen in individuals with higher levels of P. copri , with E. biforme being the true driver. Indeed, MSM status has previously been associated with increased inflammation (Gianella et al., 2012 ; Palmer et al., 2014 ). With the use of SCNIC, this correlation highlighted a route of mechanistic understanding which could be functionally followed up on in further experimental studies. SCNIC detected multiple significant modules, of which none of the OTUs within were significant when analysed separately. Module‐ 20 , which was associated with MSM status, is the fourth most significant feature at R ‐value of 0.2, and is made up of Acidaminococcus , Megasphaera , and Mitsuokella species. These are all from the Veillonellaceae family which is likely the explanation for their correlation. Members of the Veillonellaceae family have been linked with inflammation (Bajaj et al., 2012 ). By increasing statistical power and providing context for the relationships between significant taxa, SCNIC modules open new opportunities for analysis. When a module is associated with a variable of interest, the correlations within the module may imply functional relationships. These can be further investigated with in vitro and in vivo experiments. Studies which aim to test hypotheses generated from correlative analysis will commonly use a single significantly associated microbes. This often does not adequately represent in vivo systems because microbes in isolation often do not affect a disease state or their environment. SCNIC can enhance these confirmatory studies by identifying groups of microbes that may grow better than individual microbes and may better elicit relevant phenotypes than when grown separately."
} | 2,311 |
36001047 | PMC9744196 | pmc | 8,529 | {
"abstract": "Abstract Microbiome studies are often limited by a lack of statistical power due to small sample sizes and a large number of features. This problem is exacerbated in correlative studies of multi‐omic datasets. Statistical power can be increased by finding and summarizing modules of correlated observations, which is one dimensionality reduction method. Additionally, modules provide biological insight as correlated groups of microbes can have relationships among themselves. To address these challenges, we developed SCNIC: Sparse Cooccurrence Network Investigation for compositional data. SCNIC is open‐source software that can generate correlation networks and detect and summarize modules of highly correlated features. Modules can be formed using either the Louvain Modularity Maximization (LMM) algorithm or a Shared Minimum Distance algorithm (SMD) that we newly describe here and relate to LMM using simulated data. We applied SCNIC to two published datasets and we achieved increased statistical power and identified microbes that not only differed across groups, but also correlated strongly with each other, suggesting shared environmental drivers or cooperative relationships among them. SCNIC provides an easy way to generate correlation networks, identify modules of correlated features and summarize them for downstream statistical analysis. Although SCNIC was designed considering properties of microbiome data, such as compositionality and sparsity, it can be applied to a variety of data types including metabolomics data and used to integrate multiple data types. SCNIC allows for the identification of functional microbial relationships at scale while increasing statistical power through feature reduction.",
"discussion": "4 DISCUSSION SCNIC provides a method to measure correlations, find and visualize modules of correlated features, and summarize modules by summing their counts for use in downstream statistical analysis as one method for dimensionality reduction. Using SCNIC with the SMD algorithm for module detection aids in feature reduction in 16S rRNA sequencing data while ensuring a minimum strength of association within modules. As expected, our workflow identified modules in which OTUs tended to be phylogenetically related, especially at relatively high values of R . Using SCNIC, we overall achieved increased statistical power from performing less comparisons, but use of low R ‐value thresholds had the potential to lead to loss of significance by binning loosely correlated features. In these analyses we used SparCC to calculate pairwise correlations in compositional data but Spearman and Pearson are also implemented for cases when the underlying data do not match those well suited for SparCC (e.g., if they are not sparse or with an inverse Simpson index above 13). In these analyses, we also used OTUs as features; however, other microbiome features can be used with SCNIC, such as ASVs, genera, or species defined with a taxonomic classifier, as well as other data types such as metabolome data. SCNIC has also been used in previously published work to perform feature reduction prior to random forest analysis with the microbiome and diverse other data types (Armstrong et al., 2021 ). SCNIC complements existing methods because these either: (1) form correlation networks of microbes for visualization but do not have functionality for selecting and summarizing modules for downstream statistical analysis (Faust & Raes, 2016 ), (2) can select and summarize modules for downstream statistical analysis but are designed for gene expression and not microbiome data (Langfelder & Horvath, 2008 ), only summarize features if they are phylogenetically related (Frøslev et al., 2017 ), or suggest methods for finding modules of correlated microbes but do not provide a convenient implementation (Blondel et al., 2008 ). SCNIC is available both as a stand‐alone application and as a QIIME 2 plugin for easy integration with existing microbiome workflows. SCNIC implements both the LMM algorithm, which had been previously recommended for selecting modules of correlated microbes (Baldassano & Bassett, 2016 ; Jackson et al., 2018 ), and a novel SMD algorithm. The advantage of the SMD algorithm is that all pairs of features in the module have an R ‐value greater than the user‐provided minimum threshold. Using real and simulated data, we showed that SMD produced smaller modules that generally represent sub‐graphs of the larger LMM modules. Since the use of lower R ‐value thresholds similarly produced larger modules including more weakly correlated modules, we speculate that use of LMM might result in a similar trend of identifying more OTUs within significant modules, but with the disadvantage of individually significant OTUs being lost because they are combined with loosely correlated microbes that are not related to the outcome being tested. We illustrate here that varying the R ‐value threshold input by the user has a great impact on the results. However, we have avoided giving specific R ‐value threshold recommendations here, because optimal R ‐values may vary across datasets and data types. Using higher R ‐values thresholds was more likely to identify highly phylogenetically related microbes that likely share overlapping functionality, and in principle could also identify diverse organisms with overlapping niches or highly complementary metabolic functions. Using a lower R ‐value threshold bins a broader community of more loosely correlated features with the risk of bringing together features which should not be grouped and loosing significance of OTUs – as was illustrated in the Great Lakes dataset analysis conducted here. By summarizing correlated features, SCNIC can mitigate overcorrection in multiple test adjustments by reducing the number of taxa and false discovery rate for downstream analysis. However, further work with both real and simulated datasets is required to determine the degree to which network characteristics that are inherent in different microbiome datasets may influence the optimal methods for both selecting and summarizing modules in order to enhance statistical power. The results of our HIV dataset analysis confirm original findings, as well as those of another study (Armstrong et al., 2018 ), but included many new significantly associated taxa. SCNIC also assists in interpretation of microbiome data by identifying correlations among these taxa. Our results recapitulated those of the original publication of these data and previous HIV microbiome studies that all found enrichment of Prevotella with MSM status (Armstrong et al., 2018 ; Dillon et al., 2014 ; Lozupone et al., 2013 ; Noguera‐Julian et al., 2016 ). However, our analyses provide additional insight by identifying correlations between differentiating taxa. For instance, in module‐ 0 , which was more abundant in MSM samples, OTUs assigned taxonomically to the Prevotella genus are correlated with two OTUs identified as Eubacterium biforme (which has recently been renamed Holdemanella biformis ( De Maesschalck et al., 2014 \n ) ). Prevotella copri has previously been associated with increased inflammation (Dillon et al., 2014 ) while in vitro stimulations of human immune cells have found that P. copri did not induce particularly high levels of inflammation but E. biforme did (Lozupone et al., 2013 ). This strong correlation between P. copri and E. biforme in MSM could explain the increased inflammation seen in individuals with higher levels of P. copri , with E. biforme being the true driver. Indeed, MSM status has previously been associated with increased inflammation (Gianella et al., 2012 ; Palmer et al., 2014 ). With the use of SCNIC, this correlation highlighted a route of mechanistic understanding which could be functionally followed up on in further experimental studies. SCNIC detected multiple significant modules, of which none of the OTUs within were significant when analysed separately. Module‐ 20 , which was associated with MSM status, is the fourth most significant feature at R ‐value of 0.2, and is made up of Acidaminococcus , Megasphaera , and Mitsuokella species. These are all from the Veillonellaceae family which is likely the explanation for their correlation. Members of the Veillonellaceae family have been linked with inflammation (Bajaj et al., 2012 ). By increasing statistical power and providing context for the relationships between significant taxa, SCNIC modules open new opportunities for analysis. When a module is associated with a variable of interest, the correlations within the module may imply functional relationships. These can be further investigated with in vitro and in vivo experiments. Studies which aim to test hypotheses generated from correlative analysis will commonly use a single significantly associated microbes. This often does not adequately represent in vivo systems because microbes in isolation often do not affect a disease state or their environment. SCNIC can enhance these confirmatory studies by identifying groups of microbes that may grow better than individual microbes and may better elicit relevant phenotypes than when grown separately."
} | 2,311 |
36001047 | PMC9744196 | pmc | 8,530 | {
"abstract": "Abstract Microbiome studies are often limited by a lack of statistical power due to small sample sizes and a large number of features. This problem is exacerbated in correlative studies of multi‐omic datasets. Statistical power can be increased by finding and summarizing modules of correlated observations, which is one dimensionality reduction method. Additionally, modules provide biological insight as correlated groups of microbes can have relationships among themselves. To address these challenges, we developed SCNIC: Sparse Cooccurrence Network Investigation for compositional data. SCNIC is open‐source software that can generate correlation networks and detect and summarize modules of highly correlated features. Modules can be formed using either the Louvain Modularity Maximization (LMM) algorithm or a Shared Minimum Distance algorithm (SMD) that we newly describe here and relate to LMM using simulated data. We applied SCNIC to two published datasets and we achieved increased statistical power and identified microbes that not only differed across groups, but also correlated strongly with each other, suggesting shared environmental drivers or cooperative relationships among them. SCNIC provides an easy way to generate correlation networks, identify modules of correlated features and summarize them for downstream statistical analysis. Although SCNIC was designed considering properties of microbiome data, such as compositionality and sparsity, it can be applied to a variety of data types including metabolomics data and used to integrate multiple data types. SCNIC allows for the identification of functional microbial relationships at scale while increasing statistical power through feature reduction.",
"discussion": "4 DISCUSSION SCNIC provides a method to measure correlations, find and visualize modules of correlated features, and summarize modules by summing their counts for use in downstream statistical analysis as one method for dimensionality reduction. Using SCNIC with the SMD algorithm for module detection aids in feature reduction in 16S rRNA sequencing data while ensuring a minimum strength of association within modules. As expected, our workflow identified modules in which OTUs tended to be phylogenetically related, especially at relatively high values of R . Using SCNIC, we overall achieved increased statistical power from performing less comparisons, but use of low R ‐value thresholds had the potential to lead to loss of significance by binning loosely correlated features. In these analyses we used SparCC to calculate pairwise correlations in compositional data but Spearman and Pearson are also implemented for cases when the underlying data do not match those well suited for SparCC (e.g., if they are not sparse or with an inverse Simpson index above 13). In these analyses, we also used OTUs as features; however, other microbiome features can be used with SCNIC, such as ASVs, genera, or species defined with a taxonomic classifier, as well as other data types such as metabolome data. SCNIC has also been used in previously published work to perform feature reduction prior to random forest analysis with the microbiome and diverse other data types (Armstrong et al., 2021 ). SCNIC complements existing methods because these either: (1) form correlation networks of microbes for visualization but do not have functionality for selecting and summarizing modules for downstream statistical analysis (Faust & Raes, 2016 ), (2) can select and summarize modules for downstream statistical analysis but are designed for gene expression and not microbiome data (Langfelder & Horvath, 2008 ), only summarize features if they are phylogenetically related (Frøslev et al., 2017 ), or suggest methods for finding modules of correlated microbes but do not provide a convenient implementation (Blondel et al., 2008 ). SCNIC is available both as a stand‐alone application and as a QIIME 2 plugin for easy integration with existing microbiome workflows. SCNIC implements both the LMM algorithm, which had been previously recommended for selecting modules of correlated microbes (Baldassano & Bassett, 2016 ; Jackson et al., 2018 ), and a novel SMD algorithm. The advantage of the SMD algorithm is that all pairs of features in the module have an R ‐value greater than the user‐provided minimum threshold. Using real and simulated data, we showed that SMD produced smaller modules that generally represent sub‐graphs of the larger LMM modules. Since the use of lower R ‐value thresholds similarly produced larger modules including more weakly correlated modules, we speculate that use of LMM might result in a similar trend of identifying more OTUs within significant modules, but with the disadvantage of individually significant OTUs being lost because they are combined with loosely correlated microbes that are not related to the outcome being tested. We illustrate here that varying the R ‐value threshold input by the user has a great impact on the results. However, we have avoided giving specific R ‐value threshold recommendations here, because optimal R ‐values may vary across datasets and data types. Using higher R ‐values thresholds was more likely to identify highly phylogenetically related microbes that likely share overlapping functionality, and in principle could also identify diverse organisms with overlapping niches or highly complementary metabolic functions. Using a lower R ‐value threshold bins a broader community of more loosely correlated features with the risk of bringing together features which should not be grouped and loosing significance of OTUs – as was illustrated in the Great Lakes dataset analysis conducted here. By summarizing correlated features, SCNIC can mitigate overcorrection in multiple test adjustments by reducing the number of taxa and false discovery rate for downstream analysis. However, further work with both real and simulated datasets is required to determine the degree to which network characteristics that are inherent in different microbiome datasets may influence the optimal methods for both selecting and summarizing modules in order to enhance statistical power. The results of our HIV dataset analysis confirm original findings, as well as those of another study (Armstrong et al., 2018 ), but included many new significantly associated taxa. SCNIC also assists in interpretation of microbiome data by identifying correlations among these taxa. Our results recapitulated those of the original publication of these data and previous HIV microbiome studies that all found enrichment of Prevotella with MSM status (Armstrong et al., 2018 ; Dillon et al., 2014 ; Lozupone et al., 2013 ; Noguera‐Julian et al., 2016 ). However, our analyses provide additional insight by identifying correlations between differentiating taxa. For instance, in module‐ 0 , which was more abundant in MSM samples, OTUs assigned taxonomically to the Prevotella genus are correlated with two OTUs identified as Eubacterium biforme (which has recently been renamed Holdemanella biformis ( De Maesschalck et al., 2014 \n ) ). Prevotella copri has previously been associated with increased inflammation (Dillon et al., 2014 ) while in vitro stimulations of human immune cells have found that P. copri did not induce particularly high levels of inflammation but E. biforme did (Lozupone et al., 2013 ). This strong correlation between P. copri and E. biforme in MSM could explain the increased inflammation seen in individuals with higher levels of P. copri , with E. biforme being the true driver. Indeed, MSM status has previously been associated with increased inflammation (Gianella et al., 2012 ; Palmer et al., 2014 ). With the use of SCNIC, this correlation highlighted a route of mechanistic understanding which could be functionally followed up on in further experimental studies. SCNIC detected multiple significant modules, of which none of the OTUs within were significant when analysed separately. Module‐ 20 , which was associated with MSM status, is the fourth most significant feature at R ‐value of 0.2, and is made up of Acidaminococcus , Megasphaera , and Mitsuokella species. These are all from the Veillonellaceae family which is likely the explanation for their correlation. Members of the Veillonellaceae family have been linked with inflammation (Bajaj et al., 2012 ). By increasing statistical power and providing context for the relationships between significant taxa, SCNIC modules open new opportunities for analysis. When a module is associated with a variable of interest, the correlations within the module may imply functional relationships. These can be further investigated with in vitro and in vivo experiments. Studies which aim to test hypotheses generated from correlative analysis will commonly use a single significantly associated microbes. This often does not adequately represent in vivo systems because microbes in isolation often do not affect a disease state or their environment. SCNIC can enhance these confirmatory studies by identifying groups of microbes that may grow better than individual microbes and may better elicit relevant phenotypes than when grown separately."
} | 2,311 |
36001047 | PMC9744196 | pmc | 8,530 | {
"abstract": "Abstract Microbiome studies are often limited by a lack of statistical power due to small sample sizes and a large number of features. This problem is exacerbated in correlative studies of multi‐omic datasets. Statistical power can be increased by finding and summarizing modules of correlated observations, which is one dimensionality reduction method. Additionally, modules provide biological insight as correlated groups of microbes can have relationships among themselves. To address these challenges, we developed SCNIC: Sparse Cooccurrence Network Investigation for compositional data. SCNIC is open‐source software that can generate correlation networks and detect and summarize modules of highly correlated features. Modules can be formed using either the Louvain Modularity Maximization (LMM) algorithm or a Shared Minimum Distance algorithm (SMD) that we newly describe here and relate to LMM using simulated data. We applied SCNIC to two published datasets and we achieved increased statistical power and identified microbes that not only differed across groups, but also correlated strongly with each other, suggesting shared environmental drivers or cooperative relationships among them. SCNIC provides an easy way to generate correlation networks, identify modules of correlated features and summarize them for downstream statistical analysis. Although SCNIC was designed considering properties of microbiome data, such as compositionality and sparsity, it can be applied to a variety of data types including metabolomics data and used to integrate multiple data types. SCNIC allows for the identification of functional microbial relationships at scale while increasing statistical power through feature reduction.",
"discussion": "4 DISCUSSION SCNIC provides a method to measure correlations, find and visualize modules of correlated features, and summarize modules by summing their counts for use in downstream statistical analysis as one method for dimensionality reduction. Using SCNIC with the SMD algorithm for module detection aids in feature reduction in 16S rRNA sequencing data while ensuring a minimum strength of association within modules. As expected, our workflow identified modules in which OTUs tended to be phylogenetically related, especially at relatively high values of R . Using SCNIC, we overall achieved increased statistical power from performing less comparisons, but use of low R ‐value thresholds had the potential to lead to loss of significance by binning loosely correlated features. In these analyses we used SparCC to calculate pairwise correlations in compositional data but Spearman and Pearson are also implemented for cases when the underlying data do not match those well suited for SparCC (e.g., if they are not sparse or with an inverse Simpson index above 13). In these analyses, we also used OTUs as features; however, other microbiome features can be used with SCNIC, such as ASVs, genera, or species defined with a taxonomic classifier, as well as other data types such as metabolome data. SCNIC has also been used in previously published work to perform feature reduction prior to random forest analysis with the microbiome and diverse other data types (Armstrong et al., 2021 ). SCNIC complements existing methods because these either: (1) form correlation networks of microbes for visualization but do not have functionality for selecting and summarizing modules for downstream statistical analysis (Faust & Raes, 2016 ), (2) can select and summarize modules for downstream statistical analysis but are designed for gene expression and not microbiome data (Langfelder & Horvath, 2008 ), only summarize features if they are phylogenetically related (Frøslev et al., 2017 ), or suggest methods for finding modules of correlated microbes but do not provide a convenient implementation (Blondel et al., 2008 ). SCNIC is available both as a stand‐alone application and as a QIIME 2 plugin for easy integration with existing microbiome workflows. SCNIC implements both the LMM algorithm, which had been previously recommended for selecting modules of correlated microbes (Baldassano & Bassett, 2016 ; Jackson et al., 2018 ), and a novel SMD algorithm. The advantage of the SMD algorithm is that all pairs of features in the module have an R ‐value greater than the user‐provided minimum threshold. Using real and simulated data, we showed that SMD produced smaller modules that generally represent sub‐graphs of the larger LMM modules. Since the use of lower R ‐value thresholds similarly produced larger modules including more weakly correlated modules, we speculate that use of LMM might result in a similar trend of identifying more OTUs within significant modules, but with the disadvantage of individually significant OTUs being lost because they are combined with loosely correlated microbes that are not related to the outcome being tested. We illustrate here that varying the R ‐value threshold input by the user has a great impact on the results. However, we have avoided giving specific R ‐value threshold recommendations here, because optimal R ‐values may vary across datasets and data types. Using higher R ‐values thresholds was more likely to identify highly phylogenetically related microbes that likely share overlapping functionality, and in principle could also identify diverse organisms with overlapping niches or highly complementary metabolic functions. Using a lower R ‐value threshold bins a broader community of more loosely correlated features with the risk of bringing together features which should not be grouped and loosing significance of OTUs – as was illustrated in the Great Lakes dataset analysis conducted here. By summarizing correlated features, SCNIC can mitigate overcorrection in multiple test adjustments by reducing the number of taxa and false discovery rate for downstream analysis. However, further work with both real and simulated datasets is required to determine the degree to which network characteristics that are inherent in different microbiome datasets may influence the optimal methods for both selecting and summarizing modules in order to enhance statistical power. The results of our HIV dataset analysis confirm original findings, as well as those of another study (Armstrong et al., 2018 ), but included many new significantly associated taxa. SCNIC also assists in interpretation of microbiome data by identifying correlations among these taxa. Our results recapitulated those of the original publication of these data and previous HIV microbiome studies that all found enrichment of Prevotella with MSM status (Armstrong et al., 2018 ; Dillon et al., 2014 ; Lozupone et al., 2013 ; Noguera‐Julian et al., 2016 ). However, our analyses provide additional insight by identifying correlations between differentiating taxa. For instance, in module‐ 0 , which was more abundant in MSM samples, OTUs assigned taxonomically to the Prevotella genus are correlated with two OTUs identified as Eubacterium biforme (which has recently been renamed Holdemanella biformis ( De Maesschalck et al., 2014 \n ) ). Prevotella copri has previously been associated with increased inflammation (Dillon et al., 2014 ) while in vitro stimulations of human immune cells have found that P. copri did not induce particularly high levels of inflammation but E. biforme did (Lozupone et al., 2013 ). This strong correlation between P. copri and E. biforme in MSM could explain the increased inflammation seen in individuals with higher levels of P. copri , with E. biforme being the true driver. Indeed, MSM status has previously been associated with increased inflammation (Gianella et al., 2012 ; Palmer et al., 2014 ). With the use of SCNIC, this correlation highlighted a route of mechanistic understanding which could be functionally followed up on in further experimental studies. SCNIC detected multiple significant modules, of which none of the OTUs within were significant when analysed separately. Module‐ 20 , which was associated with MSM status, is the fourth most significant feature at R ‐value of 0.2, and is made up of Acidaminococcus , Megasphaera , and Mitsuokella species. These are all from the Veillonellaceae family which is likely the explanation for their correlation. Members of the Veillonellaceae family have been linked with inflammation (Bajaj et al., 2012 ). By increasing statistical power and providing context for the relationships between significant taxa, SCNIC modules open new opportunities for analysis. When a module is associated with a variable of interest, the correlations within the module may imply functional relationships. These can be further investigated with in vitro and in vivo experiments. Studies which aim to test hypotheses generated from correlative analysis will commonly use a single significantly associated microbes. This often does not adequately represent in vivo systems because microbes in isolation often do not affect a disease state or their environment. SCNIC can enhance these confirmatory studies by identifying groups of microbes that may grow better than individual microbes and may better elicit relevant phenotypes than when grown separately."
} | 2,311 |
35947692 | PMC9614513 | pmc | 8,532 | {
"abstract": "Microscopy visualization of the native protein organization and adaptation of the photosynthetic machinery from a fast-growing cyanobacterium reveals its structural heterogeneity and plasticity.",
"introduction": "Introduction In plants, algae, and cyanobacteria, efficient light energy conversion and rapid electron transport rely on the lateral organization and interactions of photosynthetic macromolecular complexes in the thylakoid membrane, including photosystem I (PSI), photosystem II (PSII), cytochrome b 6 f (Cyt b 6 f ), ATP synthase (ATPase), and type-I NADH dehydrogenase-like complex (NDH-1) ( Liu, 2016 ; Mullineaux and Liu, 2020 ). Among these photoautotrophs, cyanobacteria show unique advantages in higher efficiency of energy conversion, faster growth, enhanced biomass production, and genetic tractability ( Knoot et al., 2018 ). Given the global energy crisis and climate change, there is an increasing interest in renewable energy and fossil fuel replacement; cyanobacteria have been rapidly developed as a sustainable chassis for producing valuable chemicals and biofuels ( Jodlbauer et al., 2021 ). \n Synechococcus elongatus UTEX 2973 (Syn2973) has been recently identified as a fast-growing cyanobacterium ( Ungerer et al., 2018a , b ). Although the genomic sequences of Syn2973 and the model cyanobacterium S. elongatus PCC 7942 (Syn7942) are mostly identical except for differences in only 55 genetic loci ( Yu et al., 2015 ), Syn2973 exhibits a three times higher growth rate and more than two-fold higher photosynthetic rate compared with Syn7942 ( Ungerer et al., 2018a , 2018b ). Moreover, Syn2973 is tolerant of high light (HL), whereas Syn7942 could be severely photoinhibited under HL ( Ungerer et al., 2018a ). Studies on the photosynthetic mechanisms of Syn2973 have been performed from the perspective of genetics, physiology, and biochemistry ( Yu et al., 2015 ; Wendt et al., 2016 ; Mueller et al., 2017 ; Ungerer et al., 2018a , 2018b ). However, the lateral organization of membrane complexes in the Syn2973 thylakoid membranes and the structural plasticity of the photosynthetic machinery, which provide the structural basis for efficient energy conversion and photosynthetic adaptation, remain poorly understood. Atomic force microscopy (AFM) has unique advantages in probing the native structures of biological membranes and multi-protein complex assemblies as well as their molecular forces and nanomechanics under physiological conditions ( Liu et al., 2011 ; Liu and Scheuring, 2013 ; Faulkner et al., 2017 ; Miller et al., 2020 ). Previous AFM studies have unravelled the structural landscape of thylakoid membranes from Syn7942, Thermosynechococcus elongatus , Synechococcus sp. PCC 7002, Synechocystis sp. PCC 6803, Prochlorococcus marinus MED4, and cyanobacterial species capable of far-red light-induced photoacclimation ( Casella et al., 2017 ; MacGregor-Chatwin et al., 2017 , 2019 , 2022 ; Ho et al., 2020 ; Zhao et al., 2020 ). Here, we report the native arrangements and interactions of electron transport complexes in the thylakoid membranes from Syn2973 visualized by high-resolution AFM. We also systematically study the organizational variability of photosynthetic supercomplex assemblies from low-light (LL)-adapted, HL-adapted, and iron-deficiency-treated Syn2973 cells. Our results provide insight into the construction principles and inter-complex associations that drive the assembly and dynamics of photosynthetic apparatus in Syn2973. Advanced understanding of efficient photosynthesis and environmental adaptation in cyanobacteria would aid in rational design and rewiring of artificial photosynthetic systems to improve photosynthesis and bioenergy production.",
"discussion": "Discussion The fast-growing cyanobacterium Syn2973 has been considered as a potential candidate for “microbial cell factories” in biotechnological applications. Syn2973 is tolerant of HL and grows three times faster than Syn7942, and its photosynthetic efficiency is more than twofold higher than Syn7942 at optimal growth conditions ( Ungerer et al., 2018a , 2018b ). To compensate for the severe loss of PSI, in HL-adapted Syn7942 the remaining PSI bind with IsiA forming IsiA–PSI supercomplexes to increase the optical absorption cross-section ( Zhao et al., 2020 ). In contrast, IsiA was not expressed in Syn2973 grown under HL ( Figure 1, E–G , Supplemental Figure S1, B–D ), and the HL-adapted thylakoid membranes possess increased PSI content compared with HL-adapted Syn7942, as revealed by AFM imaging ( Supplemental Figure S4M ) and spectroscopy ( Ungerer et al., 2018a ), suggesting a strategy for improving photosynthetic efficiency and growth of cyanobacterial cells. The expression of IsiA in cyanobacteria could be induced by accumulation of reactive oxygen species (ROS) in cells grown under HL ( Havaux et al., 2005 ; Wang et al., 2008 ). The absence of IsiA in HL-adapted Syn2973 may indicate a low level of ROS in HL-adapted Syn2973, resulted from a faster photosynthetic electron transport instead of passing electrons onto oxygen. In addition, the NAD + kinase of Syn2973 exhibits improved kinetics for generating a large pool of NADP + to accept electrons from photosynthetic linear electron flow; Syn2973 has higher ATP-producing activity and an elevated content of NADPH than Syn7942, ultimately facilitating photosynthetic carbon fixation ( Ungerer et al., 2018b ). Our AFM results also revealed that the NDH-1 content increased markedly in HL-adapted Syn2973 thylakoid membranes ( Supplemental Figure S4M ), probably resulting in the increase of NDH-1-dependent cyclic electron flux around PSI ( Hihara et al., 2001 ; Mi et al., 2001 ; Liu et al., 2012 ; Zhang et al., 2020 ). This alteration may lead to the consumption of excess electrons at the acceptor side of PSI, thereby preventing accumulation of ROS ( Martin et al., 2004 ; Latifi et al., 2009 ). In contrast, expression of IsiA proteins in Syn2973 was induced by iron deficiency ( Figure 5, D–F ), as found in other cyanobacterial species ( Burnap et al., 1993 ; Leonhardt and Straus, 1994 ; Singh et al., 2003 ; Chen et al., 2018 ; Zhao et al., 2020 ). It has been suggested that iron-starved cyanobacteria also suffer from oxidative stress ( Latifi et al., 2005 ), and redox imbalance may be the broad spectrum inducer of IsiA expression ( Havaux et al., 2005 ). The electrons produced in the linear electron transport chain may not be utilized effectively due to the decreased PSI content in Fe – Syn2973, and the excess electrons may cause accumulation of ROS, resulting in expression of IsiA. Cyanobacterial thylakoid membranes are structurally heterogeneous and highly dynamic, and are formed in vivo following stepwise biogenesis pathways ( Liu, 2016 ; Mullineaux and Liu, 2020 ; Huokko et al., 2021 ; Zabret et al., 2021 ; Zhang et al., 2021 ; Rahimzadeh-Karvansara et al., 2022 ). AFM imaging has delineated diverse assembly patterns and organizational heterogeneity of photosynthetic complexes in thylakoid membranes from different cyanobacterial species, as reflected mainly by the lateral segregation of PSII and PSI. The PSI trimers showed relatively random orientations in the PSI-enriched thylakoid membranes from Syn2973, Syn7942, Synechocystis sp. PCC 6803, Prochlorococcus marinus MED4, as well as far-red light-acclimated Chroococcidiopsis thermalis PCC 7203 and Chlorogloeopsis fritschii PCC 9212 ( Casella et al., 2017 ; MacGregor-Chatwin et al., 2017 , 2019 , 2022 ; Zhao et al., 2020 ), in contrast to the more regular arrangement of PSI trimers observed in the thylakoid membranes of Thermosynechococcus elongatus , Synechococcus sp. PCC 7002, Prochlorococcus marinus MIT9313 and SS120, as well as far-red light-acclimated Synechococcus 7335 ( MacGregor-Chatwin et al., 2017 , 2019 , 2022 ). The PSII dimers show both random and crystallized arrangements in the Syn2973 thylakoid membranes ( Figures 2 and 3 ; Supplemental Figures S7 and S8 ). Given their distinct absorption and turnover rates, such lateral segregation of PSI and PSII may provide special local membrane environments for energy conversation and electron flow ( Busch et al., 2013 ). The large regions of defined PSII arrays observed in AFM ( Figures 2 and 3 ) are consistent with previous AFM studies on grana thylakoids ( Sznee et al., 2011 ), and on thylakoids from Synechococcus 7335 ( Ho et al., 2020 ). Such arrays were also observed in a previous EM study of crystalline PSII arrays in Synechocystis sp. PCC 6803 ( Folea et al., 2008 ), and they resemble the compartmentalization of PSII in the grana lamellae of plants ( Dekker and Boekema, 2005 ; Kouril et al., 2012 ; Johnson et al., 2014 ; Levitan et al., 2019 ). These regular arrangements of photosystem complexes not only elevate the local PSII content in some thylakoid membrane regions but also provide the framework for docking of multiple phycobilisomes to construct photosynthetic assembly units, as seen in Supplemental Figure S5, B and D and also in Ho et al. (2020) . Such units are composed of phycobilisome arrays, PSII dimer arrays, and surrounding PSI complexes to ensure efficient photosynthesis and state transitions locally. The heterogeneity of cyanobacterial thylakoid membranes was also reflected by the structural variety of PSI–NDH-1 assemblies and mixed membrane-spanning orientations of PSI complexes. In Syn2973 thylakoid membranes, PSI–NDH-1 supercomplexes with various assembly forms were visualized ( Supplemental Figures S4, 10, and 11 ), consistent with the finding on Syn7942 thylakoid membranes ( Zhao et al., 2020 ). The association between PSI and NDH-1 and their flexible, dynamic assembly in native thylakoid membranes could facilitate the NDH-1-dependent cyclic electron transport and physiologically balance the ATP/NADPH ratio required for the Calvin–Benson cycle in response to the changing environments ( Peng et al., 2008 , 2009 ; Kouril et al., 2014 ; Gao et al., 2016 ; Yadav et al., 2017 ). Our study provides evidence for the “upside-down” PSI complexes and variable insertion of photosynthetic complexes into thylakoid membranes. The proper orientation of PSI is essential for electron transfer from plastocyanins in thylakoid lumen to ferredoxins in the cytoplasm ( Supplemental Figure S2 ), and the “upside-down” PSI complexes are unlikely to play the same role as “normal” PSI. Consistently, only a small amount of “upside-down” PSI were observed together with major “normal” PSI complexes in cyanobacterial thylakoid membranes. The mechanisms underlying protein integration into thylakoid membranes and the actual physiological function of “upside-down” PSI complexes merit further investigation. In conclusion, we performed in-depth AFM imaging to unravel the supramolecular organization and variability of photosynthetic complexes in native thylakoid membranes from the fast-growing cyanobacterium ecotype Syn2973, which underpin efficient photosynthesis in different light and iron availability conditions. PSI trimers are predominant in Syn2973 thylakoid membranes, and lateral segregation of PSI and PSII were mainly observed in addition to inter-complex associations of PSI and PSII. HL-adapted thylakoid membranes have a high content of PSI complexes without IsiA assemblies; Fe − thylakoid membranes contain a low abundance of PSI complexes associated with IsiA proteins, forming IsiA–PSI supercomplexes with various structures. The structural plasticity and dynamics of thylakoid membranes were further indicated by the visualization of diverse PSI–NDH-1 assemblies and aberrant membrane orientation of PSI complexes in thylakoids. Advanced understanding of the architecture and modularity of cyanobacterial thylakoid membranes that conduct efficient photosynthesis is essential for unveiling the molecular mechanisms of photosynthetic electron flow and adaptation, and will inform rational design and construction of artificial photosynthetic systems for sustainable biofuel production."
} | 3,019 |
34138740 | PMC8133704 | pmc | 8,534 | {
"abstract": "Direct investigation of gecko setae-substrate contact provides clear evidence for acid-base interactions in gecko adhesion.",
"introduction": "INTRODUCTION The ability of geckos to effortlessly run up and down a vertical wall using their sticky toe pads has garnered considerable attention in the past two decades from biologists interested in the basic understanding of the adhesive system to material scientists interested in developing new synthetic adhesives ( 1 – 17 ). The microstructure of these adhesive toe pads, specifically for one gecko species ( Gekko gecko ), consists of millions of tiny hair-like structures called setae, which further split at the tip into 200-nm-wide and 5-nm-thick nanostructures referred to as spatulae ( 1 , 8 , 13 , 18 , 19 ). These nanostructures come into intimate contact with the substrate upon application of a small perpendicular preload and a few micrometers of parallel drag ( 1 ). This intimate contact between spatulae and the substrate generates adhesive forces that surpass many times the weight of the gecko ( 5 , 13 , 14 , 20 , 21 ). Seminal work by Autumn et al. ( 1 , 2 ) in the early 2000s suggested that van der Waals (vdW) forces (also known as dispersive forces that arise from instantaneous distortions in the electron cloud) govern gecko adhesion. However, several studies have called into question the relative importance of vdW and non-vdW forces in gecko adhesion ( 3 , 4 , 7 , 12 , 17 , 22 ). First, early studies showed that gecko adhesion to hydroxylated surfaces such as glass and alumina was particularly high, which cannot be explained solely by vdW forces because vdW forces are insensitive to surface chemistry. Lower setal adhesion to a silicon wafer coated with a hydrophobic coating compared to a bare silicon wafer further corroborates these results and suggests that forces other than, or in addition to, vdW interactions may contribute to gecko adhesion ( 4 , 15 ). Second, gecko adhesion enhances with increasing humidity at the setal and whole-animal level, implying that capillary forces also play a role in gecko adhesion in humid environments ( 3 , 4 , 7 ). Recent studies show that setal material softening and the resulting increase in viscoelastic dampening with increasing humidity may dominate the adhesion response to humidity, but capillary forces could play a role in certain circumstances ( 10 , 23 , 24 ). Third, the adsorption energy of glycine and cysteine molecules [representative amino acids in corneous beta proteins (CBPs) formerly known as β-keratin, the primary constituent of gecko setae ( 25 – 30 )] calculated using density functional theory (DFT) highlighted that vdW interactions contribute weakly to the overall interaction energy ( 17 ). However, glycine and cysteine molecules may not accurately capture the complex setal surface structure and setal-substrate interactions (e.g., the adhesive interface may contain other chemical constituents). Although the abovementioned studies highlight the possibility that geckos may use non-vdW interactions along with universal vdW interactions, direct molecular-level evidence of a non–vdW-based adhesive mechanism remains elusive. To decipher the type of interactions involved in gecko adhesion, it is important to determine which functionalities are exposed on the setal surface and thereby at the adhesive interface. Previous literature suggests that gecko setae are complex ensembles of cysteine-rich and serine-tyrosine–rich CBPs and a mixture of covalently (bound) and noncovalently (unbound) bonded lipids ( 25 – 32 ). It is clear that unbound lipids make contact with the substrate and are left as a footprint, as revealed by mass spectrometry and surface-sensitive sum frequency generation spectroscopy (SFG) ( 31 ). However, the orientation of these lipids (i.e., headgroups or tails) at the setal surface remains unclear. For instance, the superhydrophobicity of the adhesive gecko toe pad and SFG peaks corresponding to lipid molecules at the contact interface led Hsu et al. to conclude that nonpolar tails of unbound lipids were exposed at the setal surface ( 8 , 15 , 31 , 33 ). However, the water contact angle of gecko setae does not change even after delipidization (chemical treatment to remove unbound lipids), which raises questions about the orientation of the unbound lipids ( 15 ). Orientation of unbound lipids on the setal surface is important because if headgroups are exposed on the setal surface, the setae could interact with substrates using hydrogen bonding, capillary forces, and electrostatic forces, in addition to vdW forces. Thus, identifying whether headgroups or tails are exposed at the setal surface can help resolve the debate on the involvement of non-vdW forces in gecko adhesion. The primary goal of this study is to investigate the adhesive contact interface of gecko setae with the substrate, and to resolve questions about the exclusivity of vdW forces in governing gecko adhesion. Here, we use interface-sensitive SFG, a second-order nonlinear optical technique, to examine the contact interface between gecko setae and a hydroxylated sapphire (one of the crystal forms of alumina) substrate ( 31 , 34 ). We specifically take advantage of the presence of hydroxyl (OH) groups on the surface of sapphire and the innate peak shift of sapphire OH groups in response to the nature and strength of intermolecular interactions between sapphire OHs and the material (gecko setae in this case) in contact ( 35 – 40 ). If solely vdW interactions govern gecko adhesion, we would expect the sapphire peak originally at 3710 cm −1 (in air) to shift by only 20 to 30 cm −1 ( 36 , 38 , 39 ). However, a much higher peak shift would signify the presence of acid-base interactions, a broad term encompassing hydrogen bonding, electron pair donor-acceptor, electrophile-nucleophile, and other polar interactions ( 41 , 42 ). Acid-base interactions have both electrostatic and covalent bonding characteristics, and the relative ratio of these depends on properties of the two interacting materials ( 43 ). By probing the contact with SFG, we plan to determine the nature of interactions (vdW/non-vdW) between gecko setae and hydroxylated surfaces (e.g., glass and alumina). Further, by comparing the shift of sapphire OH groups for gecko setae with model lipid molecules, we will clarify the presence or absence of tails or headgroups at the setal surface and the contact interface. In addition to investigating the adhesive contact interface of gecko setae and the role of vdW forces in gecko adhesion, our secondary goal is to explore the role of unbound lipids in gecko adhesion. Strong adhesion of geckos to hydroxylated surfaces such as glass and alumina could result in considerable damage to the setae; however, no microscopic images have ever shown signs of wear in gecko setae. The only wear is in the form of a lipid footprint left behind on a surface a gecko clings to or walks on ( 31 ). By examining the contact interface before and after detachment using SFG and measuring adhesion forces of pristine and delipidized gecko setae, we explore the role unbound lipids may play in preventing large-scale wear (beyond wear in the form of a lipid footprint) and damage to gecko setae during detachment. This work provides a direct answer to two questions that have not been easy to address in the past two decades: (i) Are there non-vdW interactions at the interface that may contribute to gecko adhesion? and (ii) How does the removal of unbound lipids affects gecko adhesion? The results of this study will expand our understanding of the gecko adhesion mechanism and provide more refined, surface-specific design parameters for designing synthetic gecko-inspired adhesives that adhere more strongly and avoid damage.",
"discussion": "DISCUSSION The results of this study provide direct evidence of strong acid-base interfacial interactions between gecko setae and a hydroxylated sapphire substrate. The SFG spectrum for sapphire OH groups in contact with air shows a peak at ∼3710 cm −1 ( 36 , 38 – 40 ). In contact with pristine gecko setae, the sapphire OH peak shifts by 105 ± 16 cm −1 . The observed peak shift for pristine gecko setae is similar to that of strongly interacting polar fatty acid headgroups, which shifts the sapphire OH peak by as much as ∼120 cm −1 because of strong acid-base interactions ( 40 ). The observation of methylene and methyl C-H signatures in the SFG spectrum of pristine gecko setae, the SFG and mass spectrometry data from footprint residue, and histochemical staining from previous work are all consistent with the hypothesis that unbound lipid molecules are coming in direct contact with the hydroxylated sapphire substrate ( 30 , 31 ). If the hydrophobic tails of these unbound lipid molecules were exposed and making contact with the sapphire OHs, we would expect the sapphire OH peak to shift by only 20 to 30 cm −1 ( 36 , 39 ). However, the observed high-frequency shift (105 ± 16 cm −1 ) indicates that the surface of contacting setal tips (i.e., spatulae) must be covered with unbound lipids exposing polar headgroups, instead of the nonpolar tails. This arrangement of unbound lipids on the setal surface inferred from our SFG results is not contradictory to the previously observed high water contact angles on the toe pad, as recent work by Stark et al. ( 15 ) demonstrated that superhydrophobicity is a consequence of the hierarchical structure of the toe pad rather than the presence or absence of lipids. Moreover, this lipid arrangement is similar to the arrangement of lipids in mammalian stratum corneum ( 45 , 46 ). Exposed headgroups at the setal surface have implications for setal-substrate interactions and thus the adhesive mechanism of geckos. The lipid headgroups, including phosphocholine, phosphorylethanolamine, and carboxyl acid, have the ability to form acid-base interactions with sapphire surface OHs ( 36 , 39 , 40 ) and thus could explain the exceptionally strong adhesion of geckos to clean hydroxylated surfaces (glass and sapphire), contrary to expectations based solely on vdW forces ( 1 , 2 , 7 , 13 , 47 , 48 ). Because strong correlations exist between a material’s acid-base properties and its ability to develop charges ( 49 – 51 ), the observation of acid-base interactions in the SFG spectra could also help explain the charges observed by Izadi et al. ( 12 ) when separating gecko setae from Teflon and polydimethylsiloxane substrates. Upon contact electrification, Lewis bases (or electron donors) tend to become positively charged, while Lewis acids (or electron acceptors) tend to become negatively charged, although the exact charging mechanism remains unclear ( 49 – 51 ). In our study, the sapphire surface acts as a Lewis acid; thus, we would expect it to acquire a negative charge that could result in electrostatic interactions with positively charged setae [proposed by Alibardi et al. ( 28 , 29 ) based on the presence of positively charged CBPs in gecko setae], resulting in appreciably enhanced adhesion, consistent with observed forces. The exposed polar headgroups could also promote the absorption of water molecules on the setal surface, thereby supporting the capillary forces observed in adhesion experiments at setal and whole-animal levels ( 3 , 4 , 7 , 24 ). Geckos likely take advantage of multiple forces (vdW, acid-base, and capillary) to successfully adhere to a multitude of substrates in their natural habitats. Weak vdW forces are universal and present at any setal-substrate interface, while strong non-vdWs are more specific to the nature of substrate. vdW forces may be more relevant for gecko adhesion to hydrophobic leaves and trees. In contrast, nondispersion forces may be more relevant for adhesion to natural rocks and manmade hydroxylated surfaces such as glass. Despite strong adhesion to glass, the interfacial contact of delipidized setae probed by SFG (i.e., sapphire-delipidized setae contact interface) does not detect a signal from functional groups on the surface of delipidized gecko setae and appears similar to the sapphire-air spectrum. This result has two key implications. First, the SFG signal observed from the sapphire-pristine setae interface must be from unbound lipids. Second, the contact area of delipidized setae must be extremely small. To estimate the contact area of delipidized setae pressed and sheared into contact with the sapphire substrate, we model the contact interface with two regions. In the first region, there is direct contact between the setae and the sapphire surface OH groups, which would result in a frequency shift of sapphire free OH peak to ∼3600 cm −1 (assuming similar frequency shifts of sapphire OH in contact with pristine and delipidized setae). In the second region, air is in contact with the sapphire surface OH groups, which would result in the free OH peak at ∼3710 cm −1 . Because the total number of sapphire OHs is fixed, a change in the relative amplitudes of 3600- and 3710-cm −1 peaks can provide information on the percentage of OH groups making contact with the delipidized setae, as shown by Singla et al. ( 38 ) with binary liquid mixtures. Using their method, only 2 to 7% of the total sapphire area is in contact with the delipidized setae (see Supplementary Text for details) ( 38 ). It is remarkable that such a small area of contact can result in a shear adhesion of over 60 to 70 N/cm 2 . Using the measured shear adhesion force for a single seta reported by Autumn et al. (∼200 μN) and estimated setal density of 19,822 ± 490 per mm 2 ( 1 , 52 ), we estimate that ∼15 to 18% of the total sapphire substrate contact area would result in gecko setal shear adhesion of 60 to 70 N/cm 2 , a value similar to the conclusions derived from the SFG experiments (i.e., very small interfacial contact area). Our SFG results are also consistent with nonuniform stress distributions in the gecko toe pad between and within lamellae observed by Eason et al. ( 44 ), highlighting the idea that not all setae are actively used for adhesion. If the true setal area (estimated from the delipidized setal contact spectrum) is so small, it is puzzling why pristine gecko setae display strong methylene and methyl unbound lipid signatures and a frequency shift in the sapphire free OH peak. One reason for this attribute may be related to the process of engaging gecko setae with the sapphire surface (i.e., the load-drag pathway; Fig. 1 ). Specifically, the cohesive failure within the bulk of the unbound lipid layer allows lipids to spread over a much larger area than the actual contact area. The process is analogous to a paint brush (gecko setae) dipped in paint, leaving paint marks (unbound lipids) behind as it is dragged across the surface. This hypothesis is consistent with the observed similarities in the SFG spectra of pristine gecko setae in contact and after separation (i.e., the paint or footprint left behind). Using our previous analysis, we can estimate how much surface area is covered by unbound lipids during contact and after separation of pristine setae ( 38 ). This analysis suggests that ∼98% of the total contact area is covered by unbound lipids during contact and after surfaces are pulled apart. However, this number does not represent the true area of contact for pristine gecko setae, which is difficult to ascertain in the present work because of the process used for engaging setae. In contrast, negligible residue is left behind after delipidized gecko setae are removed from the substrate, despite their strong adhesion relative to pristine gecko setae. We postulate that having this cohesively weak lipid layer on the gecko setal surface results in cohesive failure within this layer, rather than adhesive failure, and, in turn, reduces the probability of damage to gecko setae as a result of high shear forces. The only wear that occurs is in the form of lipid footprints. The continual maintenance of lipid footprints during multiple adhesive events continues to remain unknown and requires further investigation. One possible explanation could be that the small real contact area (2 to 7%) allows the lipid layer to last longer than that expected for a complete contact. The strategy of using lipids as a sacrificial layer allows geckos to use their setae quickly and repeatedly for a period of a few months before they are replaced with a new set of setae during their natural skin-shedding cycle. This strategy is not limited to geckos; several insects use lipid secretions for quick and easy detachment ( 53 – 55 ). Using our spectroscopy results, we can build on the previous gecko setal lipid arrangement model proposed by Jain et al. ( 32 ) to provide further insights into the organization of lipids on the setal surface. To do so, we use the lipid organization in mammalian stratum corneum to predict the lipid organization in gecko (i.e., reptile) skin, which is not as well known ( 45 , 46 ). Our proposed lipid arrangement based on the previous model by Jain et al. , lipid arrangement in stratum corneum, and our SFG results is shown in Fig. 4 ( 32 , 45 , 46 ). In this model, the CBP-based setae are covered with a monolayer of bound lipids similar to the cornified lipid envelope in mammalian stratum corneum, which is typically formed by transesterification reaction of ω-OH fatty acids, ceramides, and glucosyl-ceramides with glutamate residues of corneocyte proteins ( 56 – 58 ). This monomolecular layer of bound lipids exposes polar headgroups on the periphery and coordinates the multilamellar arrangement of unbound lipids, which again exposes polar headgroups. Both the number of unbound lipid layers and the lateral packing shown in Fig. 4 are for the purpose of illustration and require further investigation. However, Fig. 4 conveys the general idea that unbound lipids are present at the setal surface with their headgroups exposed, consistent with our spectroscopy results. When the pristine setae are detached, failure occurs within the cohesively weak unbound lipid layer, rather than the strong lipid headgroup-substrate interface, leaving a footprint on the substrate ( 31 , 55 ). The delipidization process removes this weak unbound lipid layer and likely leaves behind the monomolecular layer of bound lipids. Thus, more force is required to separate the delipidized setae from the substrate than pristine setae, which appear to have a sacrificial lipid layer, consistent with our experimental adhesion results and hypotheses from previous work ( 28 , 31 , 53 – 55 ). Fig. 4 Proposed lipid arrangement at the setal surface. ( Left ) Schematic representation of the protein-lipid arrangement in gecko setae modified from Jain et al. ( 32 ). In this model, the lipids (mixture of bound and unbound lipids; purple) are present as patches in the bulk CBP (formerly known as β-keratin; yellow)–based setae and as a thin coating (unbounding lipids; orange) on the surface of setae. The unbound lipid layer comes in contact with the substrate and is left behind as a footprint ( 31 ). The arrangement of unbound lipid layer on the surface of setae is mediated by a monolayer of bound lipids (green). ( Right ) Sketch of the proposed arrangement of lipids on the pristine and delipidized setal surface derived from our SFG results and lipid arrangement reported in mammalian stratum corneum ( 45 , 46 ). In summary, the results of our study suggest that the adhesive interactions of geckos with hydroxylated surfaces such as glass and sapphire are dominated by acid-base interactions, rather than solely weak vdW forces. The SFG results also suggest that the unbound lipid layers are oriented on the surface of setae with their headgroups exposed, similar to the lipid arrangement in mammalian stratum corneum ( 45 , 46 ). Our results using delipidized gecko setae suggest that the actual contact area of setal hairs is small (∼2 to 7%) and on par with previous estimates based on single setae and whole-animal adhesion measurements ( 1 , 52 ). By comparing observations of pristine and delipidized setae, we highlight the important role unbound lipids play in wear and preventing damage to the setae, and suggest that these lipids help in quick and easy peel during detachment."
} | 5,090 |
35538590 | PMC9088039 | pmc | 8,537 | {
"abstract": "Background The deep sea harbors the majority of the microbial biomass in the ocean and is a key site for organic matter (OM) remineralization and storage in the biosphere. Microbial metabolism in the deep ocean is greatly controlled by the generally depleted but periodically fluctuating supply of OM. Currently, little is known about metabolic potentials of dominant deep-sea microbes to cope with the variable OM inputs, especially for those living in the hadal trenches—the deepest part of the ocean. Results In this study, we report the first extensive examination of the metabolic potentials of hadal sediment Chloroflexi , a dominant phylum in hadal trenches and the global deep ocean. In total, 62 metagenome-assembled-genomes (MAGs) were reconstructed from nine metagenomic datasets derived from sediments of the Mariana Trench. These MAGs represent six novel species, four novel genera, one novel family, and one novel order within the classes Anaerolineae and Dehalococcoidia . Fragment recruitment showed that these MAGs are globally distributed in deep-sea waters and surface sediments, and transcriptomic analysis indicated their in situ activities. Metabolic reconstruction showed that hadal Chloroflexi mainly had a heterotrophic lifestyle, with the potential to degrade a wide range of organic carbon, sulfur, and halogenated compounds. Our results revealed for the first time that hadal Chloroflexi harbor pathways for the complete hydrolytic or oxidative degradation of various recalcitrant OM, including aromatic compounds (e.g., benzoate), polyaromatic hydrocarbons (e.g., fluorene), polychlorobiphenyl (e.g., 4-chlorobiphenyl), and organochlorine compounds (e.g., chloroalkanes, chlorocyclohexane). Moreover, these organisms showed the potential to synthesize energy storage compounds (e.g., trehalose) and had regulatory modules to respond to changes in nutrient conditions. These metabolic traits suggest that Chloroflexi may follow a “feast-or-famine” metabolic strategy, i.e., preferentially consume labile OM and store the energy intracellularly under OM-rich conditions, and utilize the stored energy or degrade recalcitrant OM for survival under OM-limited condition. Conclusion This study expands the current knowledge on metabolic strategies in deep-ocean Chlorolfexi and highlights their significance in deep-sea carbon, sulfur, and halogen cycles. The metabolic plasticity likely provides Chloroflexi with advantages for survival under variable and heterogenic OM inputs in the deep ocean. \n Video Abstract Supplementary Information The online version contains supplementary material available at 10.1186/s40168-022-01263-6.",
"conclusion": "Conclusions This study provides an extensive exploration of the metabolic potential of novel and dominant Chloroflexi lineages retrieved from the hadal sediments of the Mariana Trench. The results demonstrated a high metabolic plasticity of the hadal sediment Chloroflexi , including the complete pathways for hydrolytic or oxidative degradation of recalcitrant OM such as PAHs, PCBs, and organohalides. Our findings expand the current understanding on metabolic capabilities of deep-sea Chloroflexi , and highlight their significance on carbon, sulfur, and halogen cycling in the deep ocean. The metabolic plasticity, the capability to form intracellular storage inclusions, as well as the regulatory modules to respond to nutrient conditions discovered in the MAGs support the notion that the hadal sediment Chloroflexi employ a “feast-or-famine” metabolic strategy. Such a metabolic strategy allows the bacteria to fulfill energy and nutrient requirement via degradation of different substrates according to the nutrient conditions, and regulate the cell activities (e.g., growth, motility) accordingly, providing advantages for their adaptation to the variable OM conditions in the hadal trenches and other deep-sea habitats. This study therefore provides new perspectives on the metabolism and adaptation strategies of Chloroflexi in deep-sea environments.",
"introduction": "Introduction The deep sea harbors around 75% of the prokaryotic biomass and more than half of the prokaryotic production of the global ocean, and is a key site for organic matter (OM) remineralization and storage in the biosphere [ 1 ]. An estimated 1–40% of the photosynthetically fixed carbon in the upper water reach the deep sea [ 2 ], which is generally considered oligotrophic in nature [ 2 , 3 ]. The flux of nutrients varies in intensity and frequency over temporal and spatial scales [ 4 – 6 ], and mass input of particles from surface algal blooms may lead to periodic increases in quantity and bioavailability of OM in the deep ocean [ 3 , 7 ]. Deep-sea microorganisms therefore have to employ special metabolic strategies to cope with the variable OM conditions to ensure their survival and functioning [ 3 , 7 ]. Bacteria of the phylum Chloroflexi are dominant members of microbial communities in the global deep ocean [ 8 , 9 ]. For example, the SAR202 clade of the Chloroflexi on average accounts for > 10%, and in some cases up to 40% of the total prokaryotic community in meso- and bathypelagic water of the Atlantic and Pacific oceans [ 8 , 10 – 12 ]. Chloroflexi have also been shown to account for 25.5–41.3% of total 16S rRNA gene sequences in global marine sediments [ 9 , 13 , 14 ]. Currently, the knowledge on the metabolism of deep-sea Chloroflexi mainly relies on metagenomic or single-cell genomic analysis, due to the lacking of cultivated representatives for dominant deep-sea lineages [ 15 – 17 ]. These studies revealed that Chloroflexi from deep-sea waters harbor genes involved in organosulfur compounds degradation [ 8 , 15 ], sulfite oxidation [ 8 , 15 ], and metabolism of recalcitrant compounds such as cyclic alkanes and aromatic compounds [ 15 – 17 ]. The analysis of deep-sea Chloroflexi from anoxic subseafloor sediments suggests these bacteria have potential for reductive respiration of organohalogen compounds, and for the fermentation of OM combined with CO 2 fixation via the Wood-Ljungdahl pathway [ 9 , 18 ]. These findings suggest that the Chloroflexi play important roles in biogeochemical cycles of the deep ocean. However, existing studies only covered a few seawater or anoxic subseafloor sites [ 8 , 9 , 15 – 18 ]. Given the high heterogeneity of the deep-sea habitats [ 3 ] and great phylogenetic and functional diversity of Chloroflexi bacteria [ 9 , 17 ], the current understanding of the metabolisms of deep-sea Chloroflexi is therefore incomplete, and their genomic basis and metabolic strategies to adapt to fluctuations of OM supply (e.g., OM with different recalcitrancy) in the deep ocean are unclear. The hadal trenches, formed by the subduction of tectonic plates, are the deepest part of the ocean [ 19 ]. Multiple sources of OM inputs combined with frequent OM remobilization due to special topographies, tectonic activities, and intra-trench currents, lead to higher heterogeneity and fluctuation of OM in the hadal trenches than those in other deep-sea habitats [ 19 – 22 ]. However, despite the complex OM supply and extreme environmental conditions, such as high pressure, active microbial carbon turnover in hadal sediments has been frequently reported, making the hadal trenches “hot spots” of OM remineralization in the deep ocean [ 23 – 25 ]. Recently, Chloroflexi have been identified as one of the dominant taxa in seawater and sediment of the hadal trenches [ 15 , 26 , 27 ] and were found to primarily belong to novel lineages [ 27 ]. In addition, hadal Chloroflexi were not only numerically dominant but also highly transcribed in both hadal seawater and sediments (accounting for up to 36.2% of transcribed prokaryotic 16S rRNA sequences), suggesting high in situ activities [ 15 , 27 ]. Co-occurrence network analysis further revealed that Chloroflexi lineages were important in mediating the interactive network within hadal microbial communities [ 26 , 27 ]. To date, only three studies reported the metabolism of hadal Chloroflexi based on 13 MAGs or single amplified genomes (SAGs) recovered from seawaters [ 15 , 17 , 28 ]. These bacteria were shown to encode enzymes to metabolize chitin, dimethyl sulfoxide, aromatic compounds (e.g., phthalate), and osmolytes [ 15 , 28 ], but the detailed degradation pathways were unclear. Moreover, although these studies shed light on the lifestyles of hadal Chloroflexi , but the limited number of investigated genomes restricted the findings primarily to the pelagic SAR202 group II and III [ 15 , 28 ]. The metabolic potentials of other dominant and novel lineages of Chloroflexi living in the hadal zone are thus largely unknown. In this study, we employed a metagenomic approach to fill the knowledge gap on the metabolism of Chloroflexi that are living in surface sediments of the hadal zone. We obtained unique samples from the deepest point of the ocean, the Challenger Deep of the Mariana Trench. Species composition and activity potential of hadal sediment Chloroflexi were revealed using amplicon sequencing of 16S rRNA gene and their transcripts. Representative Chloroflexi MAGs were then retrieved from nine metagenomic datasets, and their phylogeny, distribution, and metabolic potentials were explored. All MAGs were found to belong to novel lineages of Chloroflexi , representing major and widely distributed members of the hadal sediment microorganisms. The recovered Chloroflexi showed the capabilities to degrade a wide range of OM with different levels of recalcitrance. Our analysis also revealed for the first time the presence of hydrolytic and oxidative pathways in deep-sea Chloroflexi for the complete degradation of polyaromatic hydrocarbons (PAHs), polychlorobiphenyl (PCBs), and halogenated organic compounds. Potential metabolic strategies to adapt to fluctuation and heterogeneity of OM in the hadal trenches are proposed based on the metabolic reconstruction.",
"discussion": "Results and discussion Composition and activity of Chloroflexi in sediments of the Challenger Deep In this study, 16S rRNA genes and their transcripts were sequenced for samples from nine different depths of a sediment core retrieved from the Challenger Deep of the Mariana Trench. The results showed that Chloroflexi accounted on average for 20.9% and 19.1% of the total sequences for the bulk (i.e., 16S rRNA gene) and potentially active (i.e., 16S rRNA) bacterial communities, respectively (Fig. 1 A, B, and Additional file 1 : Fig. S1). The distribution of the bulk Chloroflexi population was relatively stable and varied between 18.6 and 24.6% of total rRNA gene sequences in the upper 9 cm (Fig. 1 A, Additional file 1 : Fig. S1). In contrast, the proportion of the transcribed Chloroflexi 16S rRNA sequences varied greatly with depth. Chloroflexi transcripts showed the highest proportion at 3–7 cm below seafloor, accounting for up to 40.6% of the total 16S rRNA sequences (Fig. 1 b, Additional file 1 : Fig. S1). The bulk and potentially active Chloroflexi populations were mainly composed of members from classes Anaerolineae , Dehalococcoidia , Chloroflexia , JG-KF-CM66, and KD4-96, among which Anaerolineae , Dehalococcoidia , and JG-KF-CM66 were the most dominant and highly transcribed ones (Fig. 1 B and C). These results are consistent with previous findings on the microbial composition of hadal trench sediments [ 26 , 27 ], indicating the general significance of Chloroflexi in maintaining the structure and functions of the hadal biosphere. Compared with deeper sediment layers, Chloroflexi showed relatively lower transcription levels at 0–3 cm, and the possible reasons include (1) The physical/chemical conditions (such as high oxygen levels and organic carbon content) of the surface layers might favor the transcriptional activities of other bacterial taxa, leading to lowered proportions of Chloroflexi in the transcript pool [ 23 , 25 ]; (2) A fraction of the microbes at these layers might be sourced from shallower habitats such as sediment of the trench slopes due to lateral transportation [ 19 , 21 ]. These microbes may show low transcriptional activity in hadal trench sediments, due to the extreme environmental conditions, such as high pressure; (3) the potentially higher level of degradation of RNA in the surface layer sediments during sample recovery from the deep trench [ 27 ]. Fig. 1 Composition of the bulk ( A ) and potentially active ( B ) Chloroflexi in the hadal sediments at the order level, revealed by 16S rRNA gene and 16S rRNA, respectively. The relative activities of different orders are shown as the ratio between their frequencies in the 16S rRNA library and the 16S rRNA gene library at each sediment depth ( C ) MAG reconstruction, genome description, and phylogenomic analysis A total of 62 Chloroflexi MAGs with completeness > 50% and contamination < 5% were reconstructed from the nine metagenomes covering different sediment layers (Additional file 2 : Table S1). These MAGs were further dereplicated at 99% average nucleotide identity (ANI) to yield 17 representatives with an average completeness of 68.56% (51.38–92.99%) and contamination ranged from 0.00 to 3.64% (Table 1 and Additional file 2 : Table S1). The genome sizes were estimated to range between 1.85 and 3.90 Mbp, and GC contents were between 58.64 and 69.45% (Table 1 ). Currently, only thirteen genomes of Chloroflexi have previously been reported from the hadal zone (deeper than 6000 m), and 11 of them were from seawater [ 15 , 17 , 28 ]. Only 2 Chloroflexi MAGs (GCA_004356475.1 and GCA_004356815.1) found in the NCBI database were recovered from hadal sediments, but without any description on their metabolism. Table 1 Summary of the 17 representative MAGs retrieved from sediments of the Challenger Deep MAGs Completeness (%) Contamination (%) Contig no. GC% CDS no. Estimated genome size (Mbp) Sequencing depth MT2_13 a 92.99 0.00 275 58.79 2237 2.74 21× MT4_27 89.96 0.31 215 65.69 1934 2.05 41× MT6_15 87.27 3.64 168 58.64 2397 2.85 28× MT4_14 86.30 1.98 408 60.36 2572 3.05 25× MT2_3 85.70 0.00 472 59.61 3150 3.90 24× MT6_13 84.77 0.11 425 60.34 2430 2.99 29× MT1_55 73.57 0.00 386 59.26 1683 2.33 33× MT5_44 67.62 1.19 314 65.58 1512 2.29 14× MT5_40 66.01 0.20 359 58.76 1610 2.61 18× MT1_49 60.51 3.08 400 62.83 2008 3.18 21× MT1_63 56.40 2.97 296 59.82 1319 2.46 15× MT4_29 54.89 2.18 365 62.58 1770 3.09 15× MT8_34 52.62 0.00 467 59.67 1199 2.16 22× MT2_40 52.07 2.20 253 69.45 1108 2.07 25× MT6_44 51.99 0.00 486 59.57 1031 1.85 23× MT9_49 51.53 2.38 257 69.22 1189 2.16 26× MT1_74 51.38 0.61 255 65.29 1016 1.98 18× a MAGs were named using “site + layer + genome number”, for example MT2_13 means the 13th genome from sediment of 2–3 cm below seafloor from the Mariana Trench Phylogenomic analysis showed that the retrieved MAGs belonged to the classes Anaerolineae , Dehalococcoidia , and SAR202 (previously classified as a class) (Fig. 2 , Additional file 2 : Table S2, Additional file 1 : Fig. S2). Taxonomic classification was further conducted with GTDB-Tk toolkits [ 29 ] (Fig. 2 and Additional file 2 : Table S3). The results revealed that these MAGs represent six novel species (MT1_49, MT2_13, MT5_40, MT1_63, MT1_55, MT4_14; relative evolutionary divergence, RED, ranged from 0.89 to 0.99) and four novel genera (MT4_27, MT1_74, MT9_49, MT2_3; RED ranged from 0.76 to 0.89) in the orders SM23-28-2, SAR202 (formerly SAR202 group II), UBA2963 (formerly SAR202 group VII), UBA1151(formerly SAR202 group I), and UBA3495 (formerly SAR202 group III) of the class Dehalococcoidia (Fig. 2 , Additional file 2 : Table S3 and Additional file 1 : Fig. S2). In addition, MT5_44 (RED = 0.52) represents a novel order in the class Dehalococcoidia , and MT6_15 (RED = 0.67) represents a novel family in the order Anaerolineales of the class Anaerolineae (Fig. 2 and Additional file 2 : Table S3). Fig. 2 Maximum likelihood phylogenomic tree of the 17 selected Chloroflexi MAGs. Genome of Mycobacterium tuberculosis was used as the root. Bootstrap values were calculated based on 100 replicates and the values higher than 90% were indicated at the base of the corresponding node. The colored backgrounds show the genomes belonging to the same order. The taxonomy was determined using GTDB-tk and the novelty of the recovered genomes was determined based on GTDB classification. Red square indicates the MAGs with completeness > 80% Six MAGs, i.e., MT6_15, MT4_27, MT2_13, MT2_3, MT6_13, and MT4_14, showed completeness higher than 80% and contamination lower than 3.6% (Table 1 ), and are therefore qualified as type materials according to the criteria defined recently for the taxonomy of uncultivated prokaryotes [ 30 , 31 ]. Following the guidelines developed by Genomics Standards Consortium [ 32 ], Konstantinidis et al. [ 33 ], and Chuvochina et al. [ 34 ], we propose the names Candidatus Hastsouellaceae (fam. nov.) and Ca. Hastsouella hadalis (genus nov. and species nov.) for MT6_15, Ca. Dehalosedimentum haloalkanolyticum (genus nov. and species nov.) for MT 4_27, Ca. Dehaloaerobium marianense (species nov.) for MT2_13, Ca. Dehaloaromaticum marinum (genus nov., species nov.) for MT2_3, Ca. Dehalovulgatia sulfonica (species nov.) for MT6_13, and Ca. Dehalovulgatia fluoreniphila (species nov.) for MT4_14. Explanations on nomenclature can be found in Additional file 3 , and protologues accompanying the proposed names can be found in Table S4. Distribution of the reconstructed MAGs in hadal sediments and other ecosystems Among the 17 representative MAGs, only MT4_27 was found to contain partial 16S rRNA gene. We further searched all of the 62 MAGs recovered, and found 2 additional MAGs that contain partial 16S rRNA gene, i.e., MT9_9 (represented by MT2_13) and MT8_3 (represented by MT2_3) (Additional file 2 : Table S1). These 16S rRNA gene sequences matched with three OTUs from 16S rRNA and 16S rRNA gene libraries constructed in this study (Fig. 3 B). The matched OTUs were mainly distributed in the upper 8 cm of the sediment core and together accounted for 4.2–8.5% and 0.1–4.0% of Chloroflexi sequences in the 16S rRNA gene and 16S rRNA libraries, respectively (Fig. 3 B). Recruitment of shotgun sequencing reads showed that the 17 representative MAGs were present in all depths (0–10 cm below seafloor) of the sediment core, with MT6_44, MT1_74, MT8_34, MT9_49, and MT2_40 being the most abundant (Fig. 3 C). The three MAGs with 16S rRNA gene showed the lowest recruitment values (Fig. 3 C). It is therefore reasonable to postulate that the 14 MAGs without 16S rRNA genes might be more abundant in the bulk and active bacterial communities than the three MAGs with 16S rRNA gene (Fig. 3 B). These results suggest that the recovered MAGs represent major members of Chloroflexi in the hadal sediment of the Mariana Trench. Fig. 3 Distribution of the recovered MAGs in hadal sediments and other natural ecosystems. ( A ) The sampling sites of the datasets included in the analysis; ( B ) the relative abundance of the closest matched OTUs in 16S rRNA and 16S rRNA gene libraries of sediments of the Challenger Deep. The MAGs without any value in 16S rRNA or 16S rRNA gene libraries mean no matched OTUs due to a lack of 16S rRNA gene in the corresponding MAG; ( C ) the reads recruitment of the recovered MAGs in metagenomes from different layers of hadal sediments of the Challenger Deep; and ( D ) reads recruitments in other natural ecosystems. The names of samples from sediments were shown as “sampling site (region)_water depth of the site_depth below seafloor.” The black dots in the heatmap boxes indicate outlier values. Red-colored names at the bottom indicate the MAGs with completeness > 80% The global distribution of the recovered MAGs was evaluated by read recruitments against 58 publicly available metagenomes derived from different natural habitats, including seawater and surface sediments from different depths of the open ocean, sediments of mud volcanos, deep-sea oil spilling sites, deep subseafloor, coastal regions, rivers, and salt-lakes, as well as soils (Fig. 3 A, D and Additional file 2 : Table S5). All of the 17 MAGs showed the highest recruitment values in surface sediments of the Mariana Trench, including the nine samples analyzed in this study (water depth of 10853 m) (Fig. 3 C) and two previous samples with depths of 7942 m and 5838m (Fig. 3 D and Additional file 2 : Table S5), which likely reflect the biogeographic distributions impacted by local environmental selection [ 8 ]. The majority of the MAGs (except the order SM23-28-2) have recruited reads from metagenomes derived from sediments and seawaters of the global deep ocean (Fig. 3 D and Additional file 2 : Table S5), and none of the MAGs was present in seawater or sediments from shallow habitats, including the epi-pelagic zone of the open ocean, coastal regions, river, salt-lake, or soil (Fig. 3 D). The results suggest that majority of the recovered MAGs are widespread in seawater and surface sediment of the deep ocean. However, MAGs from different orders showed apparent preferences on their distributions in different deep-sea habitats. MAGs from the order UBA3495 (formerly SAR202 group III) showed high recruitment values in metagenomes of both deep seawater and sediment (Fig. 3 D). The SAR202 group III has previously been shown to be one of the most dominant Chloroflexi in the water column of the global deep ocean [ 17 , 28 ], and our results highlight the significance of these bacteria in both pelagic and sedimentary habitats of the deep ocean. Expansion of paralogous enzymes, such as flavin-dependent monooxygenases, in SAR202 group III has been suggested to be important for their adaptation to different deep-sea habitats, by diversifying the range of organic molecules that the cells can utilize [ 16 , 17 ]. In contrast to UBA3495, MAGs from the orders Anaerolineales , SAR202 (formerly SAR202 group II), UBA2963 (former SAR202 group VII), UBA1151(formerly SAR202 group I), and the novel order (MT5_44) showed higher recruit values in metagenomes from deep-sea sediments compared to those from seawater, suggesting their preferential distribution in deep-sea sediment habitats. Interestingly, the MAGs of the order SM23-28-2 (particularly MT1_74) only matched with the reads from sediment metagenomes of the Mariana Trench (Fig. 3 D), indicating a potential endemism to the Mariana Trench, which might be a result of long-term adaptation to the special geographic, physical and chemical conditions of the Mariana Trench, such as extreme hydrostatic pressure, tectonic activity, geographic isolation, and nutrient inputs [ 19 , 35 ]. Metabolic potential for degradation of carbohydrates, fatty acids, proteins, and organosulfur compounds Considering the varied completeness (51.38–92.99%) of the retrieved MAGs, the downstream functional analysis was mainly focused on genes and pathways that were successfully identified and annotated from the MAGs. Discussion on absent genes/pathways was avoided to eliminate possible misleading conclusions due to the incompleteness of the genomes. Genome annotation of the recovered MAGs revealed their potentials for organo-heterotrophic metabolisms and utilization of a wide range of OM (Fig. 4 ). Gene sets encoding for complete/near-complete pathways or key enzymes in the central carbohydrate metabolism, including glycolysis, tricarboxylic acid cycle (TCA cycle), pentose phosphate pathway, and β-oxidation of fatty acids, were present in all MAGs with genome completeness > 80% (Fig. 4 and Additional file 2 : Table S6). These pathways allow the degradation/transformation of simple sugars (e.g., glucose), fatty acids, as well as amino acids. On top of this, genes encoding extracellular cellulases (MT4_29, MT2_13, MT1_55), chitinases (occurred in most MAGs), and polygalacturonases (MT2_13) as well as ABC-type transporters for polysaccharides were also present in the MAGs (Fig. 4 and Additional file 2 : Table S7), suggesting the potential to degrade complex polysaccharides, such as cellulose, chitin or pectin. In addition, different types of peptidases as well as ABC-type transporters for amino acids, dipeptides and oligopeptides were found to be present in the Chloroflexi MAGs, indicating their potential to degrade protein detritus (Fig. 4 and Additional file 2 : Table S7). Fig. 4 Overview of the metabolic potentials in the 17 assembled Chloroflexi MAGs. Black arrows show the annotated metabolic pathways and the linkage between different metabolic flows. The pathways with light blue background show the degradation pathways of organosulfur compounds, and those with yellow background show the degradation of different recalcitrant compounds. The red and green values in brackets are numbers of MAGs that encode complete and partial pathways/enzymes, respectively. If only a red value is shown, the associated pathway/enzyme is complete in all of the related MAGs The hadal sediment Chloroflexi MAGs also had the potential capability to degrade various organosulfur compounds (Fig. 4 ). Alkanesulfonate monooxygenase coding genes were present in 13 of the 17 recovered MAGs (Figs. 4 and 5 and Additional file 2 : Table S7). This enzyme cleaves the carbon-sulfur bonds in a wide range of sulfonated alkanes to produce sulfite and aldehyde [ 36 ], with the latter being oxidized to fatty acid by an alkanal monooxygenase, whose coding gene was also present in 13 of the 17 MAGs (Figs. 4 and 5 and Additional file 2 : Table S7). In addition, genes encoding homologs of enzymes involved in dimethylsulfide (DMS) (i.e., DMS monooxygenase and DMS dehydrogenase) and methanesulfonate metabolisms (i.e., methanesulfonate monooxygenase) were found in the MAGs (Figs. 4 and 5 ), and genes encoding the ABC-type sulfonate transporters were also identified (Fig. 4 and Additional file 2 : Table S7). These results suggested the potential of hadal sediment Anaerolineae and Dehalococcoidia to utilize multiple organic sulfur compounds as energy, carbon and sulfur sources, a finding that is similar with previous reports on SAR202 clade (primarily SAR202 group III) from deep seawater [ 8 , 28 ]. Fig. 5 Completeness of the metabolic pathways identified in the Chloroflexi MAGs. “Complete or near complete” indicates pathways that were complete or with one enzyme missing, “partial” indicates pathways with two or more enzymes missing, and “absent” means none of the enzymes in the pathways were identified. The red-colored pathways were further illustrated in Fig. 6 for detailed reaction flows. Red-colored names at the bottom indicate the MAGs with completeness > 80% Pathways for the degradation of phthalate, benzoate, polyaromatic hydrocarbons and polychlorobiphenyl compounds The hadal Chloroflexi MAGs harbored pathways for the degradation of benzoate and phthalate (Figs. 4 and 5 ). Eight MAGs from the orders SAR202, UBA2963 and UBA3495 contained complete or near complete gene clusters encoding six enzymes for the degradation of benzoate to pyruvate or oxaloacetate (Figs. 4 and 5 , Additional file 2 : Table S7 and Additional file 1 : Fig. S3). Eight MAGs from the orders UBA2963 and UBA3495 contained genes encoding complete or near complete pathways for degradation of phthalate to 4-carboxy-2-hydroxymuconate semialdehyde (HCMS) (Figs. 4 and 5 , Additional file 2 : Table S7 and Additional file 1 : Fig. S4), and four MAGs from UBA2963 and UBA1151 contained genes encoding enzymes that can further degrade HCMS to pyruvate (Fig. 5 , Additional file 2 : Table S7 and Additional file 1 : Fig. S4). As benzoate and phthalate are common intermediates in the degradation of many aromatic compounds, we hypothesized that the recovered MAGs might also be able to degrade substrates with more complex structures. Indeed, complete or near complete pathways for the degradation of polyaromatic hydrocarbons (e.g., fluorene) and polychlorobiphenyls (PCBs, e.g., biphenyl and 4-chlorobiphenyl) were found (Figs. 4 and 5 ). Six MAGs from the orders UBA2963 and UBA3495 harbored near complete pathways for the transformation of fluorene to phthalate (Figs. 5 and 6 and Additional file 2 : Table S7). Nine MAGs from the orders UBA2963 and UBA3495 harbored complete or near complete pathways for the transformation of biphenyl to benzoate (Figs. 4 and 5 , Additional file 2 : Table S7 and Additional file 1 : Fig. S5). In addition, the nine MAGs from the orders UBA2963 and UBA3495 also contained complete or near complete pathways for the degradation of 4-chlorobiphenyl to 4-hydroxy-benzoyl-CoA (Figs. 5 and 6 and Additional file 2 : Table S7), which can be further metabolized via benzoate degradation pathway (Fig. 6 ). Fig. 6 The degradation pathways of representative PAHs and POPs identified in hadal sedimentary Chloroflexi MAGs. These pathways were potentially utilized for complete degradation of ( A ) fluorene, ( B ) 4-chlorobiphenyl, ( C ) 1,2-dichloroethane, and ( D ) γ-hexachlorocyclohexane. The illustrated pathways were found to be complete in at least one MAG recovered in this study As labile OM is usually readily utilized by microorganisms in the upper water layers, the remaining OM in the deep ocean often includes a variety of structurally complex compounds, such as aromatic compounds [ 37 ]. Partial pathways of phthalate degradation (phthalate to protocatechuate), and some enzymes involved in the degradation of benzoate (e.g., catechol 2,3-dioxygeenase) have been reported in SAR202 MAGs/SAGs from seawater of hadal trenches and other deep-sea environments [ 15 – 17 , 28 ], and related genes were highly transcribed in situ [ 15 , 28 ]. Our study advances the existing knowledge by identifying the complete pathways for the degradation of phthalate and benzoate to CO 2 by hadal sediment Chloroflexi , and is the first time to show that deep-sea Chloroflexi harbor pathways to completely degrade fluorene, biphenyl and 4- chlorobiphenyl. Pathways for hydrolytic degradation of halogenated organic compounds In this study, we further discovered in hadal sediment Chloroflexi the prevalence of genes encoding haloalkane dehalogenase, haloacetate dehalogenase, and 2-haloacid dehalogenase (Fig. 5 and Additional file 2 : Table S7), which catabolize hydrolytic dehalogenation, replacing the halogen atoms in organohalides with hydroxyl groups [ 38 ]. These enzymes have a broad specificity and participate in the degradation of multiple halogenated OM [ 38 ]. Complete or near-complete pathways for the hydrolytic and oxidative degradation of several chloroalkenes and chlorocyclohexane were further revealed (Figs. 4 and 5 ). Ten MAGs from the orders Anaerolineales , SM23-28-2, UBA2963, UBA1151, and UBA3495 harbored genes encoding the complete or near-complete pathways for the degradation of 1,2-dichloroethane to glycolate (Figs. 5 and 6 and Additional file 2 : Table S7), which can either be further transformed and enter the TCA cycle or be utilized for vitamin B6 biosynthesis. The same pathway also catabolizes the degradation of trans -dichloropropene and cis -dichloropropene to trans -3- and cis -3-chloroacrylic acid, respectively (Figs. 4 and 5 , Additional file 2 : Table S7 and Additional file 1 : Fig. S6). In addition, a pathway for the complete degradation of γ-hexachlorocyclohexane to succinyl-CoA (an intermediate in the TCA cycle) was reconstructed in the Chloroflexi MAGs (Figs. 4 and 5 ). The entire pathway involves 11 enzymes (Fig. 6 and Additional file 2 : Table S7), and complete or near complete sets of genes encoding these enzymes were present in four MAGs from the order UBA3495 (Fig. 5 ). Eight MAGs from orders Anaerolineales , SM23-28-2, SAR202, UBA1151, UBA3495 and the novel order (MT5_44) also encode for the majority of enzymes for γ-hexachlorocyclohexane degradation, but with 2-6 enzymes missing (Fig. 5 and Additional file 2 : Table S7). Currently, deep-sea Chloroflexi have mainly been implied in reductive dehalogenation [ 9 , 39 , 40 ], a strictly anaerobic process which utilizes halogenated organic compounds as electron acceptor to oxidize hydrogen (or formate) [ 41 ]. In contrast, the hydrolytic and oxidative degradation of organohalides are aerobic processes [ 38 ]. The genes coding for haloalkane and haloacetate dehalogenases have been previously reported in Chloroflexi genomes from oxic abyssal sediments [ 42 ], but our study revealed for the first time the complete or near complete pathways for hydrolytic and oxidative degradation of multiple types of organohalides in hadal Chloroflexi (Figs. 4 and 5 ). The MAGs in this study were recovered from surface sediments of the Challenger Deep at depth of 0–10 cm below seafloor, which were well oxygenated as revealed by in situ oxygen measurements conducted at the same site [ 25 ]. Such an environmental condition well supports the potential of the studied Chloroflexi to degrade organohalides via the annotated pathways. In addition, reads recruitment showed that the majority of the MAGs retrieved in this study were also widely distributed in global deep-sea water and surface sediment (Fig. 3 ), which highlights the significance of Chloroflexi in carbon and halogen cycling in oxygenic habitats of the deep ocean. Microbial degradation of persistent organic pollutants in the deepest ocean The metabolic reconstruction of the recovered MAGs in this study revealed the potential of hadal sediment Chloroflexi for the complete degradation of different types of recalcitrant organic compounds, including PAHs (e.g., fluorene), PCBs (e.g., 4-chlorobiphenyl), haloalkanes (e.g., 1,2-dichloroethane and 1,3-dichloropropene), and chlorocyclohexane (e.g., γ-hexachlorocyclohexane) (Fig. 5 and Additional file 2 : Table S7). These findings have important implications for the deep ocean ecosystems, in general, and the hadal trench systems in particular. Many of these compounds are listed as persistent organic pollutants (POPs) by the Stockholm Convention [ 43 ] and their presence and accumulation in deep-sea organisms and environments have been widely reported [ 44 , 45 ]. Recent studies have further revealed that PCBs, microplastics, heavy metals, and halogenated organic pollutants have even accumulated in the deepest trenches of the ocean [ 46 – 49 ], suggesting that the anthropogenic pollutants can be an important part of the OM pool in the hadal trenches. Many types of PAHs and POPs-like compounds can however also be naturally produced via biotic (e.g., biosynthesis via halogenase or haloperoxidase) and abiotic processes (e.g., peroxidative mechanisms, photochemical reactions, volcanic activities) [ 39 , 50 , 51 ], and can be enriched in the deep ocean via the “biological pump” [ 52 ]. The capability to metabolize these recalcitrant OM would likely provide Chloroflexi bacteria with survival advantages in nutrient/energy-limited habitats, which might be one of the reasons for their dominance in the sediments of the hadal trenches as observed in this (Fig. 3 ) and previous studies [ 26 , 27 ]. On the other hand, close interactive networks have been reported to occur between Chloroflexi lineages and between Chloroflexi and other microbial taxa in the hadal trench sediments [ 27 ], which supports the possibility of co-metabolism during the degradation of recalcitrant OM [ 53 ]. For example, the degradation of halogenated organic matter by Chloroflexi (i.e., dehalogenation process) may produce semi-labile intermediates serving as substrates for other taxa in the microbial community [ 41 ]. The capability to degrade recalcitrant OM and the potential co-metabolism relationships with other microbial taxa might be one of the reasons for previous observations that Chloroflexi lineages play keystone roles in interactive networks of the microbial community in hadal trench sediments [ 27 ]. A potentially “feast-or-famine” metabolic strategy in response to fluctuating supply of OM Deep-sea benthic communities experience feast-or-famine conditions due to the periodical and spatial variations in the input of particles in a generally energy-depleted environment [ 7 ]. Deep-sea microbial communities have been shown to respond rapidly to nutrient input, even after long periods of starvation [ 54 ]. However, little is known about the genomic basis and potential metabolic strategies of deep-sea microorganisms for such a lifestyle. Our results showed that the hadal Chloroflexi exhibited capabilities of degrading a wide range of organic carbon, sulfur, and halogen compounds (Figs. 4 and 5 ), including not only labile OM, but also many types of recalcitrant organic compounds (Figs. 5 and 6 ). In addition, the majority of the MAGs harbor genes encoding alpha, alpha-trehalose synthase, trehalose-6-phosphate synthase and trehalose-phosphate-phosphatase (Fig. 5 and Additional file 2 : Table S7), which are key enzymes for the formation of trehalose, a type of intracellular energy storage compound [ 55 ]. Some of the MAGs also harbored genes encoding starch (glycogen) synthase and 1,4-alpha-glucan branching enzyme involved in the biosynthesis of glycogen (Fig. 5 and Additional file 2 : Table S7) [ 56 ], polyphosphate kinase, and phosphate transporter proteins involved in the formation of polyphosphate inclusions (Additional file 2 : Table S7) [ 57 ], or sulfide-quinone reductase, sulfite reductase, and sulfide dehydrogenase involved in the formation of sulfur globules (Fig. 5 and Additional file 2 : Table S7) [ 56 ]. Such features are consistent with a “feast-or-famine” metabolic strategy (Fig. 7 ). During the “feast” condition, such as a pulse input of particulate OM due to diatom bloom in the surface water [ 7 ], the bacteria might preferentially uptake and consume labile OM, and excess energy, carbon, and other elements can be stored as intracellular inclusions (Fig. 7 ). During “famine” condition (i.e., nutrient depleted), the bacteria may enter the “famine” mode to acquire energy from the stored inclusions, and/or from degrading the recalcitrant OM available in the surrounding environments (Fig. 7 ). Fig. 7 The proposed “feast-or-famine” metabolic strategy for hadal sediment Chloroflexi recovered in this study In support of such a “feast-or-famine” lifestyle, the MAGs also encode modules for regulating metabolism in response to changes in nutrient conditions (Fig. 5 ). The majority of the MAGs contained genes encoding pyruvate orthophosphate dikinase, malate dehydrogenase, and malic enzyme (Fig. 5 ), and two of the MAGs also showed potential to encode PEP carboxylase (Fig. 5 ). These enzymes catalyze the inter-conversions between pyruvate, PEP, oxaloacetate, and malate (Additional file 1 : Fig. S7), which interconnect the central carbon metabolic pathways (e.g., the TCA cycle and biosynthesis) and are responsible for regulating carbon fluxes among catabolism, anabolism, and energy supply according to the physiological conditions of the cell (Additional file 1 : Fig. S7) [ 58 ]. In addition, all of the recovered MAGs harbored the lrp gene (COG1522) encoding the leucine-responsive regulatory protein (LRP) (Fig. 5 ), which is one of the “feast-or-famine” regulatory proteins that control the expression of more than 30% of bacterial genes in response to changes in nutrient levels [ 59 , 60 ]. The existence of these regulatory genes suggests the potential of the hadal Chloroflexi to rapidly change the metabolism and physiology under the feast or famine conditions, although the detailed regulation mechanisms can be very complex and are still unknown. Similarly, previous researchers have shown that common marine bacteria (e.g., Vibrio , Alteromonas , Colwellia ) from surface seawaters also followed “feast-or-famine” strategy which allows these bacteria to respond rapidly to the changes in nutrient conditions [ 61 ]. These findings suggest that the “feast-or-famine” lifestyle might be a common strategy for marine microorganisms to adapt to variable nutrient conditions. In addition to the changes in nutrient conditions, other environmental factors such as oxygen levels might also trigger the shift between feast and famine metabolic modes of Chloroflexi in surface sediments. For example, input of fresh particles from surface water algal blooms [ 7 , 62 ] may greatly stimulate the respiration and growth of microorganisms in surface hadal sediment, leading to rapid depletion of dissolved oxygen and the shift from aerobic to anaerobic condition. According to the metabolic potentials annotated (Figs. 4 and 5 ), the hadal Chloroflexi may consume labile organic carbon (such as simple sugars, amino acids) and store the excess energy intracellularly (e.g., in the form of trehalose, glycogen or poly-P) under aerobic condition and degrade the intracellular energy storage compounds for energy generation under subsequent anaerobic condition. In fact, oxygen-triggered “feast-or-famine” metabolic strategy has been well documented in polyphosphate accumulating organisms (PAOs) from wastewater treatment systems [ 63 , 64 ], although the detailed mechanisms might be different from those in hadal Chloroflexi (e.g., PAOs usually synthesize polyhydroxyalkanoate under anaerobic condition but no related genes were found in Chloroflexi MAGs)."
} | 10,412 |
39138250 | PMC11322153 | pmc | 8,539 | {
"abstract": "Previously, we constructed engineered M. circinelloides strains that can not only utilize cellulose, but also increase the yield of γ-linolenic acid (GLA). In the present study, an in-depth analysis of lipid accumulation by engineered M. circinelloides strains using corn straw was to be explored. When a two-stage temperature control strategy was adopted with adding 1.5% cellulase and 15% inoculum, the engineered strains led to increases in the lipid yield (up to 1.56 g per 100 g dry medium) and GLA yield (up to 274 mg per 100 g dry medium) of 1.8- and 2.3-fold, respectively, compared with the control strain. This study proved the engineered M. circinelloides strains, especially for Mc-C2PD6, possess advantages in using corn straw to produce GLA. This work provided a reference for transformation from agricultural cellulosic waste to functional lipid in one step, which might play a positive role in promoting the sustainable development of biological industry.",
"introduction": "Introduction Oleaginous microorganism such as bacteria, mold, yeast, algae, etc. can accumulate and synthesize a large amount of lipids in cells using carbohydrates under certain conditions, and the microbial lipids become an important alternative raw material for biodiesel production 1 – 4 . The fatty acid composition of microbial lipids is similar to that of common vegetable oil, and some also contain rich polyunsaturated fatty acids, such as arachidonic acid (AA), docosahexaenoic acid (DHA), eicosapentaenoic acid (EPA), etc. 5 – 8 . Although the development prospects of microbial lipids are broad, high cost is still the main factor restricting their industrial production. Cellulosic waste is one of the most abundant and cheapest renewable raw materials in nature 2 , 9 , 10 . For example, crop straw is the main agricultural cellulosic waste, which is mostly used as feed, fuel, and organic fertilizer 11 – 13 . Using crop straw to produce microbial lipids can effectively reduce the production cost and realize the sustainable development of agriculture 1 , 14 . The oleaginous fungus Mucor circinelloides , as the first commercially lipid producing strain in the world, can synthesize high level of γ-linolenic acid (GLA) with various important physiological functions 15 – 17 . Previously, we presented a novel approach by co-expression of cellobiohydrolase and delta-6 desaturase in M. circinelloides to facilitate the GLA production from cellulose in one step 15 , 18 . In the present study, an in-depth analysis of lipid production by recombinant M. circinelloides strains Mc-C2TD6 and Mc-C2PD6 using corn straw under solid state fermentation was to be explored. The optimal addition amount of the cellulase and the inoculum of engineered M. circinelloides strains by using corn straw for lipid production were investigated. In addition, we developed a two-stage temperature control strategy for lipid production from corn straw under solid state fermentation. Our study provides a certain foundation for the further application of the engineered M. circinelloides strains in industrial production of microbial lipids from crop straw.",
"discussion": "Discussion Although microbial lipids become an important alternative resource for biodiesel production, and some functional lipids beneficial to people's health are also the main production objects of the food and pharmaceutical industry, high cost is still the main factor restricting their industrial production 1 , 3 , 6 , 13 . The use of cheap alternative raw materials (such as crop straw) is one of the important problems to solve the industrialization and scale of microbial lipids 11 , 13 . The recent researches focus on the strategies of raw material treatment, strain screening and optimization of fermentation to improve the microbial lipid production from cellulosic raw materials 4 , 10 , 13 . However, the adoption of efficient strains that could use cheap cellulosic feedstock is an important way to achieve high microbial lipid production. It is believed that the oleaginous fungus M. circinelloides has the following specific advantages and potential over other oleaginous microorganisms in lipid production. Firstly, M. circinelloides was the first commercially lipid producing strain in the world, and its synthesized lipids are rich in GLA, which has the role of preventing inflammation, softening blood vessels, and lowering blood sugar and lipids 16 , 19 . Secondly, M. circinelloides can absorb and metabolize multiple sugars including glucose, xylose, and cellobiose, so it could make full use of cellulose and hemicellulose in crop straw, which has superiority in reducing production cost 1 , 17 , 18 . Thirdly, M. circinelloides has become a model organism to study the lipid mechanism in fungi with its successfully analyzed genome sequence and perfect gene manipulation system 15 , 20 , 21 . Finally, the ability of GLA synthesis from cellulose in M. circinelloides was enhanced through the genetic engineering, and the simultaneous conversion of cellulose-monosaccharide-lipids was realized in one step, which could not be achieved in other oleaginous microorganisms 15 , 18 . In view of this, the lipid production conditions from corn straw by engineered M. circinelloides strains Mc-C2TD6 and Mc-C2PD6 were investigated. The complex structure of corn straw needs a series of pretreatments to be effectively degraded 22 , 23 . In this study, in order to improve the utilization efficiency of corn straw, cellulase was added in the fermentation process to promote the degradation of corn straw. The addition of cellulase is too little, resulting in poor degradation effect of corn straw. Excessive addition of cellulase can cause excessive cost. The amount of cellulase was added according to its production instructions, and the addition range is 0.5–2.0% (w/w), with 0.0% as a control. The results showed that the optimum concentration of cellulase was 1.5%. However, the lipid yield did not increase significantly when the quantity of cellulase continued to increase. The reason might be that cellulase can assist the engineered M. circinelloides strains to degrade cellulose in corn straw and produce reducing sugars, and meanwhile the stains can further use reducing sugars to synthesize lipids. Thus, with the increase of cellulase addition, the production of synthesized lipids also increased. When the amount of cellulase added reaches a certain value, the final lipid production does not increase or increases slightly due to the limited cellulose that can be degraded in the raw material. In addition, it could also be seen that with the assistance of cellulase, the engineered strains might be more efficient in the decomposition of corn straws, and the resultant lipid yields are much higher than that of the control strain (Mc-2075) and the wild strain (WT). The amount of inoculation is an important factor affecting the growth and lipid production of engineered M. circinelloides strains from corn straw 1 . A larger amount of inoculation can shorten the time when mycelium propagation reaches its peak and make the formation of products arrive in advance 24 . However, too much inoculation will cause insufficient oxygen supply and affect product synthesis, and it is not economical. Generally, the inoculation amount of mold is usually 5–20% (w/w). Therefore, in order to determine the optimal inoculum, we set the inoculum range at dose of 5–20%. The results showed that the optimum inoculation amount was 15%. Generally, the production of lipids from corn straw requires first degradation of corn straw into small molecules of available monosaccharides, and then these monosaccharides are converted into lipids by microbial fermentation 1 , 25 . In the early stage, the engineered M. circinelloides strains cooperated with cellulase to degrade corn straws, and then in the middle and late stages, the strains further used the decomposed small molecules of sugars to grow and synthesize lipids. However, the temperature of enzymatic hydrolysis of cellulose by cellulase was relatively high, while the temperature of growth and lipid accumulation of M. circinelloides was low. Thus, a two-stage temperature control strategy was adopted to investigate the lipid production of engineered M. circinelloides strains from corn straws under the condition of adding 1.5% cellulase and 15% inoculum. The maximum lipid and GLA yields of engineered strains were 1.56 g and 274 mg per 100 g dry medium (for Mc-C2PD6), which were significantly enhanced by 1.8- and 2.3-fold, respectively, compared with the control strain. These results indicated that the engineered M. circinelloides strains could cooperate with cellulase well to synthesize lipids and GLA from corn straw. It should be noted that the pure biological treatment method of microbial and enzyme co-fermentation adopted in this study has incomparable advantages over traditional chemical treatment methods, such as non-toxic by-product inhibition, simple steps, and no environmental pollution, etc. However, it is undeniable that this study also has some limitations, such as long degradation process, incomplete degradation, few available monosaccharide and slow utilization of corn straw. In conclusion, one of the current research hotspots of bioenergy is the production of biodiesel by microbial fermentation using lignocellulosic biomass such as crop straw. In the present study, the effects of cellulase concentration and inoculum amount on lipid production of engineered M. circinelloides strains from corn straw were analyzed. Furthermore, a two-stage temperature control strategy was developed to facilitate the lipid and GLA synthesis from corn straws by engineered M. circinelloides strains. This study laid a foundation for the direct microbial transformation from corn straw to functional GLA in one step, which might be helpful for reducing the production cost."
} | 2,484 |
36837932 | PMC9959831 | pmc | 8,540 | {
"abstract": "Maintaining power consumption has become a critical hurdle in the manufacturing process as CMOS technologies continue to be downscaled. The longevity of portable gadgets is reduced as power usage increases. As a result, less-cost, high-density, less-power, and better-performance memory devices are in great demand in the electronics industry for a wide range of applications, including Internet of Things (IoT) and electronic devices like laptops and smartphones. All of the specifications for designing a non-volatile memory will benefit from the use of memristors. In addition to being non-volatile, memristive devices are also characterized by the high switching frequency, low wattage requirement, and compact size. Traditional transistors can be replaced by silicon substrate-based FinFETs, which are substantially more efficient in terms of area and power, to improve the design. As a result, the design of non-volatile SRAM cell in conjunction with silicon substrate-based FinFET and Metal Insulator Metal (MIM) based Memristor is proposed and compared to traditional SRAMs. The power consumption of the proposed hybrid design has outperformed the standard Silicon substrate FinFET design by 91.8% better. It has also been reported that the delay for the suggested design is actually quite a bit shorter, coming in at approximately 1.989 ps. The proposed architecture has been made significantly more practical for use as a low-power and high-speed memory system because of the incorporation of high-K insulation at the interface of metal regions. In addition, Monte Carlo (MC) simulations have been run for the reported 6T-SRAM designs in order to have a better understanding of the device stability.",
"conclusion": "5. Conclusions This work proposes new circuits and expands on others already in use to largely address the performance issues in FinFET devices. With different configurations for SRAM using different technologies, such as static CMOS (45 nm technology), FinFET (32 nm technology) and Memristor for implementing the 6T-SRAM design in Cadence Virtuoso. On repeated analysis for each of the existing technologies, we conceived of a novel hybrid design of a Silicon (Si) substrate FinFET and Ti/SiO 2 /Si(n++)/TiN Memristor based SRAM. The design is compared with conventional CMOS 6T SRAM, Hybrid CMOS- memristor, and FinFET based 6T SRAM. The reported design is consuming less area as compared to others. Further comparison is done with respect to power and delay. It is much more evident that the proposed hybrid FinFET-Memristor is consuming less power and has minimal delay. Further, from the DC analysis it is still clear that the Static Noise Margin (SNM) is more for hybrid FinFET–Memristor when compared to hybrid CMOS–Memristor design. The Monte-Carlo analysis performed has also substantiated the stability of the proposed device and the results prove that the suggested design variability is relatively higher than that of the hybrid CMOS memristor design but is still superior to static CMOS SRAM.",
"introduction": "1. Introduction SRAMs with low power consumption are becoming increasingly common in VLSI chips. This is particularly true in microprocessors, where on-chip cache capacities are increasing with each iteration to close the growing gap between processor and main memory rates. As integration and running rates have increased, power dissipation has become a critical issue. In recent years, low-wattage IC design has gained attention due to the proliferation of rechargeable devices. The most common method for lowering system power consumption in memory design is to reduce transistor size. Manufacturers of Field Effect Transistor (FET) based scaled CMOS devices are doing everything they can to help achieve this decrease [ 1 , 2 ]. When the approach progresses beyond the 90 nm node and into the sub-nanometer realm, leakage becomes a major worry. When MOS devices hit their scaling limits, the semiconductor industry developed the FinFET, which is now the most popular choice for next generation devices. Moore’s law is being supported by several semiconductor firms investing in FinFET technology. FinFETs feature a completely depleted silicon film (body) that is either undoped or mildly doped. Due to its speedier operation and higher performance, SRAM is employed as a memory element. In SRAM, however, data preservation is critical since data is destroyed when the power is turned off. Due to its instability, systems that use conventional SRAM like primary memory or cache have quite a longer startup time. Nonvolatile memory can assist minimize startup time and energy consumption [ 3 , 4 ]. It is possible that a CMOS Memristor-based SRAM cell would have been an efficient circuit component that would have allowed standard memory cells to sustain data despite the power being disconnected. The proposed research will emphasize the developing novel SRAM architecture that tends to make use of FinFET and Memristor technologies. Further, Monte Carlo simulation is done for the proposed model of the hybrid FinFET-Memristor model. This paper is structured as follows: Section 2 presents the implementation design of SRAM design using CMOS, FinFET and Memristor. In Section 3 , novel Silicon Substrate FinFET–Ti/TiN Memristor hybrid 6T-SRAM design is explored. Section 4 presents the simulation findings while Section 5 concludes the study.",
"discussion": "4. Results & Discussion The basic architecture for the SRAM sense amplifier using CMOS, FinFET, and Memristor are designed and implemented using the Cadence Virtuoso simulation tool. 4.1. Implementation of 6T SRAM Using CMOS Technology The basic 6T SRAM is implemented using Cadence Virtuoso using technology file of gpdk 45 nm, as shown in Figure 5 . In the DC analysis of Figure 8 a, the butterfly curve between both output nodes can be noticed. The inverter features are drawn and mirrored to create the butterfly curve, then calculating the greatest square that can be made between them. The Static Noise Margin (SNM) is determined by the length of the square’s side. The Static Noise Margin determines the SRAM’s stability. The SNM is a metric for determining how stable something is. The greater the SNM, the more stable it is. From the dc analysis, it is well observed that the stability from the 6T SRAM from static CMOS is large as SNM is large. The transient analysis in Figure 8 b shows the operation of write and hold without any voltage drop in the output waveforms. This shows this circuit has excellent stability. From Figure 8 c, the Monte Carlo simulation results for static CMOS SRAM have depicted an average mean value of 15.553 ps and a standard deviation of 4.8163 ps. This curve gives the value of variability of 8.45047 for the random sampling method. 4.2. Implementation of Proposed Ti/SiO 2 /Si(n++)/TiN Memristor The memristor is better in terms of lesser area coverage, and lesser power consumption with high speed. For determining the characteristics of the memristor, a basic circuit is implemented for determining the IV characteristics of the proposed Ti/SiO 2 /Si(n++)/TiN Memristor. The schematic of the same is depicted in Figure 9 a. The IV characteristics clearly shows that the proposed device is properly working as a memristor with an evident hysteresis curve in the DC analysis graph of Figure 9 b. Figure 9 c portrays a schematic view of the hybrid Memristor–MOS based inverter circuit. For the SRAM implementation, it is required to design an inverter circuit using the proposed memristor and N-channel MOS transistor. The input is given into the input of the transistor and output is from the drain end of the transistor and doped side of the memristor. The transient analysis output from the designed circuit is as expected without any drop in voltage, as plotted in Figure 9 d. 4.3. Implementation of Hybrid MOS-Memristor Based 6T-SRAM The schematic in Figure 10 a consumes lesser area as two memristors are used instead of two MOS transistors for SRAM implementation. The stability of the implemented design becomes reduced, as the SNM becomes substantially lesser as compared to the static CMOS transistor based 6T SRAM. This can be seen in the subsequent analysis depicted in Figure 10 b,c. We obtained a Monte Carlo curve for MOS–Memristor SRAM which has an average mean value of 0.401 ps and a standard deviation of 0.0958 ps, as shown in Figure 10 d. This curve gives the value of variability of 4.18580 for the random sampling method. We obtained a reduced amount of variability when compared to the conventional model. 4.4. Implementation of Silicon Substrate FinFET Based 6T-SRAM For smaller channel transistor designs, FinFET is more preferable over the MOS-based design. Hence, using 32 nm technology for FinFET, the design of basic 6T SRAM is implemented in Cadence Virtuoso and its transient response is analyzed. From the above analysis, it is clear that there is a voltage drop issue of FinFET because of leakage current caused due to a lesser threshold of FinFET technology. Figure 11 provides both a schematic representation of the Silicon Substrate FinFET based 6T-SRAM design and an analysis of its transient response. 4.5. Proposed Hybrid FinFET and Ti/SiO 2 /Si(n++)/TiN Memristor 6T-SRAM Design From all the above analysis carried out with the implementation of 6T-SRAM using CMOS transistors, FinFET, and hybrid MOS–Memristor, it is clear that both FinFET and hybrid MOS-Memristor design are quite stable and hence the proposed work concentrates on the 6T-SRAM design using hybrid FinFET and Ti/TiN Memristors, as implemented in Figure 12 a. Memristor and FinFET are good candidates for lesser area and power consumption. In this proposed implementation, CMOS technology is replaced using FinFET technology in the hybrid CMOS–Memristor-based design. The proposed circuit and its simulation analysis are given in Figure 12 a–d. From the DC analysis plot in Figure 12 b, it is well understood that the SNM is quite large for the proposed hybrid design of the FinFET–Memristor which shows that the 6T-SRAM design is much more stable than the existing realization of the static memory using CMOS Transistors and Memristors. Even the voltage drop issue experienced by FinFETs during its transient analysis shown in Figure 11 b is not observed in our proposed hybrid approach. The Monte Carlo (MC) curve for hybrid FinFET–Memristor SRAM is depicted in Figure 12 d, which has an average mean value of 0.988 ps and a standard deviation of 0.153 ps. The MC analysis highlights the value with a variability of 6.4575 through the random sampling method. However, the variability of the proposed hybrid approach is lesser than static CMOS SRAM but greater than MOS–Memristor SRAM. ADE-XL utilizes Monte Carlo to investigate process and mismatch variation. Statistics affect both delay deviations. To evaluate delay variability, 2000 Monte Carlo simulations are undertaken [ 3 ]. During pre-charging to VDD, the Bit Lines are contended high for the read operation. While the pull down fin transistor is high, the bit at the node output of “Inverter 1” implemented using MIM Memristor will become “0”. Again, when the BL releases from VDD, the differential bit line sense amplifier triggers. Read delay is a parameter used to quantify SRAM cell read performance. The read-delay is the time between 50% word line (WL) excitation and 10% bit line pre-charged voltage variance. The read delay readings of the proposed hybrid FinFET and Ti/SiO 2 /Si(n++)/TiN Memristor based 6T SRAM is compared to traditional SRAM designs for various supply voltages in Figure 13 . The comparative results reveal that the proposed 6T SRAM with hybrid design of sense amplifier outstands the conventional designs. Hybrid FinFET-MIM Memristor-based static memory cells can also be configured to write. Two FinFETs NF2 and NF3, depending on the memory cell bit to be written, can accomplish this. Charging BL to VDD will write bit “1” into the cell. By asserting NF2’s gate control input (WL) high, this is achieved. With source input BL high, FinFET (NF2) output (Q) settles to “1” and QB to “0”. The simultaneous assertion of write enable line and high WL triggers this. Deactivating the BLB line to 0 V writes “0” into the intended SRAM cell. This, in turn, activates WL to break the coupling between the two Memristor inverters, causing the node Q to store “0” with the assistance of the second access FinFET (NF3). From Figure 14 , the suggested structure with FinFET and Ti/SiO 2 /Si(n++)/TiN Memristor has a low write delay compared to other SRAM designs reported in the literature. Hold, read and write Static Noise Margin (SNM) measurements for the proposed sense amplifier design of static RAM are displayed in Figure 15 a,b. The findings demonstrate that the suggested device has high noise endurance, as the SNM window frames for read and write operations are sufficiently accurate. Table 1 and Figure 16 show how the suggested designs stack up against the current domino logic topologies in terms of power and efficiency, respectively. Using fewer transistors and activating them only when necessary contributes to the significant drop in power requirements. This exemplifies the swift operation of the proposed hybrid FinFET and Ti/SiO 2 /Si(n++)/TiN Memristor. The reported design has a static and dynamic power dissipation of 6.59 nW and 2.71 µW to deliver an average power dissipation of 2.8 µW."
} | 3,359 |
38274265 | PMC10806880 | pmc | 8,541 | {
"abstract": "Solar-driven biosynthesis\nand bioconversion are essential for achieving\nsustainable resources and renewable energy. These processes harness\nsolar energy to produce biomass, chemicals, and fuels. While they\noffer promising avenues, some challenges and limitations should be\ninvestigated and addressed for their improvement and widespread adoption.\nThese include the low utilization of light energy, the inadequate\nselectivity of products, and the limited utilization of inorganic\ncarbon/nitrogen sources. Organic semiconducting polymers offer a promising\nsolution to these challenges by collaborating with natural microorganisms\nand developing artificial photosynthetic biohybrid systems. In this\nPerspective, we highlight the latest advancements in the use of appropriate\norganic semiconducting polymers to construct artificial photosynthetic\nbiohybrid systems. We focus on how these systems can enhance the natural\nphotosynthetic efficiency of photosynthetic organisms, create artificial\nphotosynthesis capability of nonphotosynthetic organisms, and customize\nthe value-added chemicals of photosynthetic synthesis. By examining\nthe structure–activity relationships and emphasizing the mechanism\nof electron transfer based on organic semiconducting polymers in artificial\nphotosynthetic biohybrid systems, we aim to shed light on the potential\nof this novel strategy for artificial photosynthetic biohybrid systems.\nNotably, these coupling strategies between organic semiconducting\npolymers and organisms during artificial photosynthetic biohybrid\nsystems will pave the way for a more sustainable future with solar\nfuels and chemicals.",
"conclusion": "5 Conclusions\nand Perspectives Here the major advances of organic semiconducting\npolymers for\nthe coupling with the organisms into artificial photosynthetic biohybrid\nsystems began this Perspective and provide an account of the combining\nadvantages between the efficient optoelectrical properties of chemical\nmaterials and the excellent biocatalytic capacity of biological organisms.\nThe biohybrid could enhance the natural photosynthetic efficiency\nof photosynthetic organisms by organic semiconducting polymers. Nonphotosynthetic\norganisms can also be functionalized to have artificial photosynthesis\ncapability in the biohybrid of organic semiconducting polymers. More\nimportantly, customized photosynthetic synthesis of artificial photosynthetic\nbiohybrid systems could be realized for value-added chemicals by\norganic semiconducting polymer-based nanoparticles. Although the significant\napproaches and advances for solar energy conversion have been demonstrated\nto be versatile, sustainable, and distinct in this field of research,\nthe state-of-the-art systems of biosynthesis and bioconversion are\nrequired to further develop and meet the needs of practical applications\nfor foods, pharmaceuticals, and fuels in future. Although inorganic\nsemiconducting materials have heavy-metal cytotoxicity\nand phototoxicity are unavoidable (e.g., Cr, Cd, and so on), high-performance\noptical and electrical properties make them also be used in the design\nof biohybrid systems for solar-driven H 2 evolution, CO 2 reduction, and N 2 fixation. 6 , 32 , 80 − 83 Compared with organic semiconducting\nmaterials, inorganic semiconducting materials also have tunable energy\nlevels and regulated electron transfer in many aspects. And the applications\nof biohybrid systems based on inorganic semiconducting materials could,\nin turn, inspire and support the development of biohybrid systems\nbased on organic semiconducting materials. As one of the value-added\nchemicals and an energy carrier, methane (CH 4 ) could also\nbe converted from the catalyzing CO 2 reduction via the\nwhole-cell microorganisms such as methanogens, which could address\nthe issue of selectivity and oxygen intolerance of chemical catalysts\nand biocatalysts. 84 He et al. constructed\nthe Methanosarcina barkeri ( M. barkeri )-CdS biohybrid by combining the semiconducting nanoparticles with\nnonphototrophic methanogens to achieve CO 2 -to-CH 4 conversion ( Figure 12 a). 85 Owing to the direct band gap (Eg)\nof 2.69 eV for M. barkeri -CdS biohybrid and\nthe LUMO (−0.63 V vs NHE) of CdS nanoparticles, the\nCO 2 reduction to CH 4 was feasible thermodynamically.\nUnder light irradiation, the photoelectrons of solar-driven self-replicating\nbiocatalytic system could be transferred via H 2 ases-mediated\nor cytochromes-mediated pathways possibly from the photoactive CdS\nnanoparticles to the out-membrane proteins of M. barkeri to achieve CO 2 -to-CH 4 conversion with a quantum\nefficiency of 0.34%. Besides the CO 2 fixation to multicarbon\nproducts and the N 2 reduction, microbial H 2 production\nis still significant to pay specific biological catalytic power in\nphotosynthetic biohybrid systems. Compared to the photo(electro)catalytic\nwater splitting in recent years, 86 biological\nhydrogen production has also attracted great attention due to the\nmild reaction conditions, the use of no electron transfer agents,\nand the highly specific biological catalytic power. As the most well-studied\nmicroorganism, E. coli could produce bio-H 2 by endogenous [Ni–Fe]-hydrogenase under dark fermentation\nwithout any genetic engineering. Wong and his co-workers have reported\na whole-cell E. coli –CdS hybrid system for\nenhanced biological hydrogen production by natural sunlight. 32 , 87 More importantly, the E. coli –CdS hybrid\nsystem with excellent light-harvesting ability exhibited increased\nhydrogen production, which is better than other photoheterotrophic\nbacteria. Apart from the semiconducting nanoparticles, the light-absorbing\nSi nanowire arrays could also capture light for efficient reducing\nequivalents (H 2 ) to Sporomusa ovata ( S. ovata ) as nanowire-bacteria hybrid for solar CO 2 reduction into acetic acid under mild conditions ( Figure 12 c). 33 In addition to the previously biological whole-cell biohybrids,\nthe artificial photosynthetic biohybrid system-based enzymes or enzyme\nmimics have also attracted much attention in the field of synthetic\nbiology. Chen and co-workers reported the self-assembled protein hybrid\nnanofibrils by incorporating the [FeFe]–hydrogenase mimics\nand uniform hybrid protein nanofibrils for efficient H 2 generation performance ( Figure 12 d). 20 The lower reduction\npeak signal at −0.8 V (versus Ag/AgCl) of the hybrid protein\nnanofibrils favors the interface electron transfer from the photosensitizer\nof Ru(bpy) 3 Cl 2 to the [FeFe]–hydrogenase\nmimics inside the nanofibrils for proton reduction to hydrogen. In\nview of the above examples from inorganic semiconducting materials,\nit is worth learning that more organic semiconducting polymer-based\nartificial photosynthetic biohybrid systems need to be developed for\nthe CO 2 -to-CH 4 conversion and the H 2 evolution. Also, the organic semiconducting polymers could be designed\nto be the ordered electrode material systems and advanced enzymes/enzyme\nmimics hybrid, and these works provide the possibility for the future\nof based on organic semiconducting polymers. 15 Figure 12 (a) Diagrammatic illustrations of the M. barkeri -CdS biohybrid. Reproduced with permission from ref ( 85 ). Copyright 2019 Elsevier.\n(b) Illustration of whole-cell E. coli –CdS\nhybrid system. Reproduced with permission from ref ( 87 ). Copyright 2017 Wiley-VCH.\n(c) Schematic of the nanowire-bacteria hybrid. Reproduced with permission\nfrom ref ( 33 ). Copyright\n2020 Elsevier. (d) Illustration of the self–assembled protein\nhybrid nanofibrils based on [FeFe]–hydrogenase mimics. Reproduced\nwith permission from ref ( 20 ). Copyright 2023 Chinese Chemical Society. Some of the typical organic semiconducting polymers have\nbeen used\nas excellent candidates as desirable light-harvesters for biohybrids\nby effective light harvesting, promoted photoelectron separation,\nand accelerated electron transfer. More semiconducting polymers, such\nas poly(3-hexylthiophene) (P3HT), polypyrrole (Ppy), and so on, need\nto be further developed and used for artificial photosynthetic biohybrid\nsystems in the future, and more challenges need to be met. For most\nof the phototrophs, for example, cyanobacteria, algae, plants, and\nphotoheterotrophic bacterium, when augmenting photosynthetic conversion\nefficiency using synthetic light-harvesting polymer materials as light\nabsorbers, expanding the solar spectrum utilization, accelerating\nthe electron transfer, and improving photosynthetic efficiency are\nclosely related to the optical absorption and emission wavelengths\nof organic semiconducting materials. For nonphotosynthetic organisms,\nsuch as CO 2 -fixing bacteria, N 2 -fixing bacteria,\nand so on, the appropriate E LUMO and E HOMO favor the energy harvesting, photogenerated\ncharge separation, photoelectrons transport from the LUMO of the polymer\ninto metabolic pathways by mediated or unmediated electron transfer\nmodes between material and microorganism. According to the structure–property–function\nrelationships, the molecular backbone structures of polymers are the\nintrinsic property of the material, and the main optical and electronic\nperformance is the extrinsic property of the material and is determined\nwith the HOMO–LUMO energy levels as the key properties of the\nmaterial. It is therefore proposed that novel organic semiconducting\npolymers can be reasonably designed and synthesized with effectively\ncustomized and subtly tuned absorption and energy levels by modifying\nthe molecular structures of the main and side chains. For example,\nincreasing the effective conjugation length and introducing heteroatoms\nsuch as N, O, and Se will significantly benefit the extent of the\nabsorption region and the red-shifted or blue-shifted absorption peak.\nThe excellent harvesting ability of green light and the optically\nmatched far-red light fluorescence performance of polymer materials\nwith excellent optoelectronic and physicochemical properties would\nenhance the light harvesting of plant photosynthesis to further increase\nagricultural productivity. The selection of the donor and acceptor\nunits with matching energy levels is an effective strategy for the\nmolecular design and synthesis of donor–acceptor (D–A)\nheterojunction type conjugated conducting polymers for the narrow\nbandgap, the strong absorption in the long wavelength region, and\nenhanced charge transport. The introduction of alkyl, aryl, or other\nsubstituents could result in fine-tuning of the optical and electronic\nproperties. Moreover, the desired polymers by in situ catalytic polymerization,\nwhich have a suitable energy level structure and corresponding photoelectrochemical\nproperties, are more relevant for biocatalytic processes. Simultaneously\nthe multifunctional and novel CPNs as promising candidates deserve\nto be further designed and developed with the modification of the\nspecific functional groups or specific functional elements (peptide,\nprotein, enzyme, etc.) for the high fluorescence brightness, efficient\nelectron transfer, and the appended catalytic property. As a crucial\npart of artificial photosynthetic biohybrid systems, all of the above\nfactors make it useful to design reasonably new organic semiconducting\npolymers for constructing novel biohybrids and achieving great feats\nof solar-to-chemical biosynthesis and bioconversion. On the\nother hand, given the synergistic relationship between\nlight, organisms, and semiconducting materials, suitable organisms\nshould be selected to accept the photogenerated electron to suppress\nthe recombination of photoinduced charge carriers in the photoirradiated\nconjugated polymers. And in the fields of synthetic biology, genetic\nengineering technologies, and the De novo protein design will help\nthe construction of efficient metabolic pathways, the optimization\nof custom electron transport chains, and the upregulating of the protein\nexpression levels for the higher biosynthesis efficiency of biohybrid\ncomplexes., 24 88 More tools on the biological aspect should be applied to the in-depth\nmechanism study, such as functional gene analysis, transcriptional\nanalysis, cryo-electron microscopy, and proteomic and metabolomic\nmethodologies, to elucidate the underlying mechanisms of molecular\npathways of electron uptake. 25 , 89 The solar-driven biological\nsymbiosis system of symbiotic algae and microorganisms will go beyond\nnatural synthesis limitations for the revolutionary development of\nagricultural and food industries. More importantly, based on\nthe relationship of structure, interface,\nand energy level, there are many points deserving notice for the assembly\nprocess of the photosynthetic semiconducting biohybrids, such as the\nreasonable design of organic semiconducting polymers, the match mediator\nof biotic/abiotic interface, the thorough understanding of extracellular\nelectron transfer, and so on. Side chain optimization by the addition\nof cationic quaternary ammonium groups and anionic carboxyl groups\nendows the conjugated conducting polymers with sufficient water solubility\nand assembled interactions with the cell membrane of microorganisms\nfor further substance communication and electron transfer. More advanced\nand new characterization techniques (e.g., KPFM to characterize the\nelectronic properties of organic semiconducting polymers and to determine\nthe local electronic structure at the nanoscale, X-ray synchrotron\ntechniques to study the chemical and electronic structure of organic\nsemiconducting polymers, scanning probe microscopy to image the surface\ntopography and to study the electronic behavior of polymer materials,\ninfrared and Raman spectroscopy to identify functional groups and\nto study the vibrational modes of organic semiconducting polymers)\nhave also been developed with particular success to follow the photoelectron\ntransfer and the catalytic reaction process. 38 , 55 , 90 , 91 Moreover,\nmachine learning (ML) and artificial intelligence (AI) have a potential\nrole in predicting polymer properties and designing polymers for desired\nproperties by combining high-throughput synthesis experimentation. 92 − 96 Meanwhile, the rapidly growing field of artificial intelligence\nwill explore interactions between novel nanophotoconductors and microorganisms\nand further benefit the improved performance of the organic semiconducting\npolymers based artificial photosynthetic biohybrid systems to inverse\ndesign the workable polymers for matching with the appropriate microorganisms.\nThe insights would help to fully and conceivably provide feedback\nto guide more appropriate polymer synthesis for better electron flow\npathways between polymers and organisms. Nowadays, the performance,\naffordability, and sustainability of\nsunlight-powered photosynthetic biohybrid systems would support them\nto be further industrialized efficiently and effectively. It will\nbe extremely interesting to ensure a more sustainable bright future\nfor biosynthesis and bioconversion based on biohybrids. The intersection\nof materials science, semiconductor technology, and synthetic biology\nis likely to be a research focus in the coming years. Such disruptive\nimprovements in biological-material hybrid make us confident that\nartificial photosynthetic biohybrid systems will lead the way to attain\nreal-world applications toward a diverse range of solar fuels and\nchemicals in the near future.",
"introduction": "1 Introduction In recent years, the shortage\nof nonrenewable fossil energy sources\nand the exacerbation of environmental problems have become increasingly\nserious, necessitating the development of alternative energy sources. 1 − 4 Solar energy is clean, environmentally friendly, and an abundant\nsource of renewable energy compared to fossil fuels, and it has attracted\na lot of attention from chemists. 5 − 8 Photosynthesis, the paramount chemical reaction\non Earth, encompasses the biosynthesis process in which green plants,\nalgae, and cyanobacteria convert carbon dioxide (CO 2 ) into\ncarbohydrates and liberate oxygen from water under the irradiation\nof sunlight. 4 , 9 − 15 By investigating these natural photosynthetic systems, scientists\nhave constructed a series of solar-driven artificial photosynthetic\nbiohybrids by perfectly combining light-harvesting chemical materials\nwith biologically active components for converting and storing solar\nenergy effectively. 16 − 20 Compared to common organic materials and organic insulating materials,\norganic semiconducting materials are usually chosen for their excellent\nlight-harvesting performance and better electron transport capacity,\nwhich are favored to augment the biosynthesis process, produce high\nvalue-added chemicals, and eventually overcome the limitations of\nnatural photosynthesis. 21 More importantly,\nthese advanced artificial systems can achieve more efficient carbon\nfixation and nitrogen fixation using only sunlight, air, and water\nas starting materials which could outpace natural photosynthesis. 22 , 23 Moreover, some more microorganisms, such as nonphotosynthetic bacteria,\nare also satisfying candidates for converting CO 2 into\nrenewable carbon matter through a diverse array of metabolic pathways\nwhen the extracellular electron has been uptake. 24 − 28 By being integrated with semiconducting materials\ninto a biohybrid system, these microorganisms were endowed with the\nability to harvest and utilize solar light, and utilize photogenerated\nelectrons for high-value chemical production. 6 , 29 − 38 Organic semiconducting polymers are a class of excellent semiconducting\nmaterials with large π-conjugated backbones and a delocalized\nelectronic structure which endow them with promising optical properties,\nelectrical properties, and photoelectric conversion properties. 39 − 45 As early as 2000, the Nobel Prize in Chemistry was awarded to Alan\nJ. Heeger, Alan G. MacDiarmid, and Hideki Shirakawa for their discovery\nand development of conducting polymers. 44 So semiconducting materials have attracted much attention in the\nfield of organic light emitting devices, organic field effect transistors,\nand so on, which is attributed to their unique advantages including\nthe controllable structure and tunable energy levels. 46 − 52 Despite their relatively fixed main-chain structure, the organic\nsemiconducting polymers can be modified, and a range of novel materials\ncan be prepared by tuning the main-chain backbone structures or modifying\nthe functional side-chain groups. 53 Changes\nin structural properties could in turn effectively alter band gap\nenergies and the positions of the lowest unoccupied molecular orbital\n(LUMO) energy levels ( E LUMO ) and the highest\noccupied molecular orbital (HOMO) energy levels ( E HOMO ), further affecting optical absorption and the emission\nand electron transfer process. 21 , 54 Some of the well investigated\nand applied organic semiconducting polymers have been listed in Figure 1 a, namely, poly(fluorene- co -phenylene) (PFP), poly(3-alkoxy-4-methylthiophene) (PMNT),\npoly(p-phenylenevinylene) (PPV), poly(boron-dipyrromethene- co -fluorene) (PBF), poly(fluorene benzothiadiazole) (PFBT),\npoly(fluorene- alt -thienopyrazine) (PFTP), separately. 38 , 55 − 63 These organic semiconducting polymers have been widely used in\nthe fields of chemical, biological, and material sciences in very\nrecent years. However, most of the semiconducting polymers are at\nthe initial stage of constructing artificial photosynthetic biohybrid\nto enhance the biosynthesis and bioconversion effects ( Figure 1 b). 38 , 55 − 66 These recent studies have highlighted the critical role of semiconducting\npolymers in regulating electron transfer and energy conversion in\nconstructed artificial biohybrid systems, leading to the enhancement,\ncreation, and regulation of biological functions. Organic semiconducting\npolymers possess good photostability to maintain their structural\nintegrity and optical/electronic properties under prolonged exposure\nto light. 46 , 52 , 54 In addition,\nseveral strategies have been developed to enhance the photostability\nof organic semiconducting polymers, including the use of stabilizers\nor copolymers that protect against degradation, modification of the\npolymer with improved backbone rigidity or π-conjugated side\nchains that quench excited states and reduce degradation pathways,\nand encapsulation of the active materials. And the toxicity of organic\nsemiconducting polymers is an important consideration when combined\nwith biological systems. 52 , 54 Results from studies\nsuggest that many common organic semiconducting polymers are inherently\nnontoxic and have good biocompatibility, which was assessed mainly\nby the growth of bacteria in the presence of polymer materials. Moreover,\ncharge transfer mechanisms between organic semiconducting polymers\nand organisms play a critical role in biohybrid systems, and different\npathways for energy transfer exist depending on the specific system\nbeing studied, such as redox-active mediators or mediator-free, Förster\nresonance energy transfer (FRET), photon emission (bioluminescence\nor electroluminescence), inner/outer sphere electron transfer, etc.\nAnd the efficiency of energy or charge transfer often depends on the\ndistance between the interacting partners. Proximity generally promotes\nmore efficient transfer. 55 − 60 Thus, the conjugated molecules are also expected to construct artificial\nphotosynthetic biohybrid systems with biological organisms by efficient\nenergy conversion and precise regulation of bioelectronic processes\nfor more desirable high-value chemical products. Figure 1 (a) The common organic\nsemiconducting polymers for artificial photosynthetic\nbiohybrid systems and the corresponding energy levels. (b) Historical\ndevelopment of organic semiconducting polymer-based artificial photosynthetic\nbiohybrid systems, and the corresponding applications of biosynthesis\nand bioconversion. 38 , 55 − 66 Here, we mainly introduce some\nlatest achievements in organic semiconducting\npolymer-based artificial photosynthetic biohybrid systems for augmenting\nbiosynthesis and bioconversion ( Figure 2 ). The structure–activity relationship between\nthe energy levels of the organic semiconducting polymers themselves\nand the biocatalysis and biosynthesis of organisms are highlighted.\nFirst, in the biohybrid containing photosynthetic organisms, the selected\norganic semiconducting polymers with suitable band gap energy and\nphotophysical properties are advantageous for broadening solar light\nabsorption, improving energy conversion, and enhancing the biosynthesis\nproperties of photosynthetic organisms, such as cyanobacteria, algae,\nplants, and photoheterotrophic bacterium. Subsequently, in the biohybrid\ncontains nonphotosynthetic organisms, the structure-dependent energy\nlevel positions of organic semiconducting polymers could determine\nthe possible electron transfer pathways between the chemical materials\nand the nonphotosynthetic organisms, which mean the organic semiconducting\npolymers as electron donors could transfer the photogenerated electron\nto organisms as electron acceptors. The biohybrids consisting of Moorella thermoacetica ( M. thermoacetica ), 63 Sporomusa ovata ( S. ovata ), and others can use the direct or indirect electron\ntransfer process to realize CO 2 fixation, nitrogen (N 2 ) fixation, and hydrogen (H 2 ) production to convert\ninto value-added chemicals and fuels under sunlight irradiation. 60 Finally, organic semiconducting polymer-based\nnanoparticles could be further surface modified to customize the biological\nfunctional applications of artificial photosynthetic biohybrid systems.\nThe advantages in tunability and structural variability of polymer\nmaterials systems are also further illustrated and expanded by learning\nfrom the corresponding inorganic materials systems in recent years.\nLooking ahead, we propose the potential of artificial photosynthetic\nbiohybrid systems enabled by organic semiconducting polymers for enhanced\nenergy conversion efficiency and product selectivity and also suggest\nsome considering methods and techniques for further practical applications\nin the future. Figure 2 Schematic overview of organic semiconducting polymer-based\nartificial\nphotosynthetic biohybrid systems for augmenting biosynthesis and bioconversion."
} | 6,076 |
33398099 | PMC7610452 | pmc | 8,542 | {
"abstract": "Microbial communities often undergo intricate compositional changes yet also maintain stable coexistence of diverse species. The mechanisms underlying long-term coexistence remain unclear as system-wide studies have been largely limited to engineered communities, ex situ adapted cultures, or synthetic assemblies. Here we show how kefir, a natural milk-fermenting community of prokaryotes (predominantly lactic and acetic acid bacteria) and yeasts (family Saccharomycetaceae), realizes stable coexistence through spatiotemporal orchestration of species and metabolite dynamics. During milk fermentation, kefir grains (a polysaccharide matrix synthesized by kefir microbes) grow in mass but remain unchanged in composition. In contrast, the milk is colonized in a sequential manner in which early members open the niche for the followers by making available metabolites like amino acids and lactate. Through metabolomics, transcriptomics and large-scale mapping of inter-species interactions, we show how microbes poorly suited for milk survive in — and even dominate — the community, through metabolic cooperation and uneven partitioning between grain and milk. Overall, our findings reveal how inter-species interactions partitioned in space and time lead to stable coexistence.",
"discussion": "Discussion Several cross-feeding interactions in microbial communities have been previously reported 9 , 11 , 49 , 50 . Here, we systematically unraveled such interactions in a complex community in its natural environment and how these provide a group advantage. Perhaps our most intriguing finding is how L. kefiranofaciens dominates the community despite having no fitness on its own in milk ( Supplementary Fig. 10a ), which is its only known habitat. Its interaction with L. mesenteroides illustrates how L. kefiranofaciens can survive in milk by cooperating with the fellow community members. Furthermore, as L. kefiranofaciens synthesizes the polymeric matrix of the grain, it can maintain its dominance therein, and, since the community is propagated through grain transfer, continues to retain this advantage. The other members of the community must also be carried along to enable utilization of the milk nutrients. The kefir community is thus shaped by a combination of beneficial and competitive interactions that balance between maintaining grain occupancy and nutrient utilization from milk. Highlighting this balance, several species inhibited L. kefiranofaciens on milk plates, while none was inhibited by it ( Fig. 4e , f ). At the community scale, this balance was evident in the shift from competitive to cooperative interactions between solid and liquid ( Extended Data Fig. 8 ). In this scenario, the dominance of L. kefiranofaciens would be expected to decline without the grain; we observed this decline in L. kefiranofaciens abundance within a few transfers when the community was passaged without the grain ( Extended Data Fig. 10 ). We picture the kefir grain as a “basecamp” from which the members colonize milk in an orderly fashion, which is orchestrated by the accompanying metabolite changes ( Fig. 6 ). Basecamp membership is kept constant during grain growth, ensuring renewal of the inoculum for the next fermentation cycle. The spatiotemporal niche separation underlying this basecamp lifestyle may be operational in other microbial systems. For example, species retained in the mucosal layer of the intestine could provide a basis for community maintenance during nutritional changes 51 or after dysbiosis events like antibiotic treatment 2 . Overall, our results demonstrate how community stability emerges through spatiotemporal niche partitioning and provide a roadmap for deciphering complex microbial ecosystems. Fig. 6 Kefir community exhibits a “basecamp lifestyle”. The community in the grain undergoes only minor compositional changes, while the community of microbes that colonizes the milk fraction continuously changes as fermentation proceeds. The grain thus serves as a basecamp that provides inoculum for orderly milk colonization with accompanying metabolite dynamics ( Fig. 1 and 2 ). This basecamp also serves as a reservoir of community members for the next transfer into fresh milk."
} | 1,058 |
39746020 | PMC11733944 | pmc | 8,544 | {
"abstract": "Piezoelectric organic\npolymers are promising alternatives\nto their\ninorganic counterparts due to their mechanical flexibility, making\nthem suitable for flexible and wearable piezoelectric devices. Biological\npolymers such as proteins have been reported to possess piezoelectricity,\nwhile offering additional benefits, such as biocompatibility and biodegradability.\nHowever, questions remain regarding protein piezoelectricity, such\nas the impact of the protein secondary structure. This study examines\nthe piezoelectric properties of lysozyme amyloid fibril films, plasticized\nby polyethylene glycol (PEG). The films demonstrated a measurable\nd 33 coefficient of 1.4 ± 0.1 pCN –1 , for the optimized PEG concentration, confirming piezoelectricity.\nThe PEG was found to hydrogen-bond with the fibrils, likely impacting\nthe piezoelectric response of the film. Polarization imaging revealed\nlong-range alignment of the amyloid fibrils in a circumferential arrangement.\nThese results demonstrate the potential of using amyloid fibrils,\nwhich can be formed from various proteins, to create bulk self-assembled\npiezoelectric materials.",
"conclusion": "Conclusions This\nwork demonstrates the potential of\namyloid fibrils to be assembled\ninto freestanding bulk films that display piezoelectricity and structural\nanisotropy. PEG was added as a plasticizer and was found to form complex\ninteractions with the fibrils, likely in the form of hydrogen bonding.\nWe attribute this as the cause of the initial trend of increasing\nd 33 with increasing PEG concentration. However, a protein:PEG\nratio greater than 1:1.4 was observed to lead to reduced piezoelectric\noutput. The amyloid fibril films were furthermore found to exhibit\ncircumferential alignment as the films demonstrated a Maltese cross\npattern and birefringence attributed to linear retardance.",
"introduction": "Introduction Piezoelectricity, the mechanical coupling\nof charge generation\nand mechanical stress, is often associated with inorganic crystalline\nmaterials that have a unit cell with no center of inversion. Under\nmechanical stress, such noncentrosymmetric piezoelectric unit cells\nelastically deform and lead to the uneven displacement of positive\nand negative atoms, creating a dipole moment. In a bulk piezoelectric\nmaterial that consists of many unit cells in a crystalline arrangement,\nan applied mechanical deformation leads to overall electrical polarization,\nwhereas applying an electric field leads to elastic straining. 1 Piezoelectric materials have many uses in current\ntechnologies, such as in smartphones as fingerprint transducers and\nsensors, and are expected to increase rapidly as the Internet of Things\nmatures. 2 Piezoelectric materials have\nalso been successfully used for a broad range of biomedical applications,\nsuch as health monitoring devices, sources of electrical stimulus\nfor living cell studies, and acting as blood-brain barrier transducers. 3 , 4 However, traditional ceramic piezoelectric materials are stiff,\nbrittle, and often toxic, which limits their use in biomedical applications. 5 , 6 Piezoelectric polymers are considered potential alternatives\nto\ntheir inorganic counterparts in areas that require mechanical flexibility.\nOne example is the synthetic polyvinylidene fluoride (PVDF), which\nis used in wearable electronics as energy harvesters and sensors. 7 , 8 However, piezoelectric polymers generally require preprocessing\nto align the dipoles and polymer chains within the film to achieve\noptimized piezoelectric coefficients. 9 , 10 The piezoelectric\ncoefficients of polymers are generally lower than those of inorganic\npiezoelectric crystals because of the difference in the fundamentals\nof how piezoelectricity arises in the material. 10 For polymers, the piezoelectric mechanism differs from\nthe classical definition and can vary between polymers. Instead of\npolarization from uneven displacement of atoms, such as in ceramics,\npiezoelectric polymers generally already have dipoles within their\nstructure. For PVDF, the dipole moments come from the highly electronegative\nfluorine atoms. Under mechanical strain, the polymer chains and dipole\nmoments reorient themselves which leads to a change in charge accumulation\nat the material surface. 10 Individually,\na polymer chain may possess no symmetry planes. However, in an isotropic\npolymer film, the randomly oriented polymer chains create a symmetry\nplane that cancels the polarization. 10 Another accepted theory on how polymers can exhibit piezoelectricity\nis that the permanent dipoles reorient themselves under applied stress,\nleading to optimal dipole alignment, a theory popular for organic\nbiologically derived polymers. 11 − 13 These piezoelectric biopolymers\nprovide the additional benefit of being easily sourced, biodegradable,\nand nontoxic. 14 For these biopolymers,\nthe dipoles can originate from hydrogen bonding. 15 − 17 In proteins,\nhydrogen bonds and other intermolecular attraction forces stabilize\nthe protein’s secondary, tertiary, and quaternary structures.\nThe complex chemistry and structures of proteins make it difficult\nto determine the specific contributions to piezoelectricity and how\nthese are influenced by protein folding and hierarchical assembly. 12 , 13 Reported piezoelectric coefficients of biological materials vary\na lot. We, and others, have found that they are often inflated, as\nintrinsic piezoelectricity can be convoluted with other extrinsic\nelectromechanical coupling phenomena. 18 , 19 Considering\nthis, it is important to continue to study the mechanisms behind piezoelectricity\nin proteins with robust experimental protocols. Several studies have\nfocused on understanding the details of piezoelectricity in proteins\nto understand the underlying mechanism and whether piezoelectricity\nplays a role in the biological function of the proteins. 20 Collagen was the first biopolymer to be\nstated to be piezoelectric,\nand it has been suggested that the piezoelectricity of collagen is\ninvolved in regulating bone growth. 21 − 23 Several studies on collagen\nsuggest that the ordered hierarchical helical arrangement of individual\ncollagen molecules that combine to create the fibrils is responsible\nfor the observed piezoelectric behavior. 13 Under applied mechanical stress, the permanent dipoles reorient\ndue to stress-induced changes in helical pitch, torsion angles, and\nbond lengths. 24 Another well-studied fibrous\nprotein is silk fibroin. Unlike collagen, this protein can be transformed\nto be rich in β-sheet, and the β-sheet content within\nthe protein has been found to influence its piezoelectric capabilities. 12 , 15 , 25 , 26 Increased β-sheet content requires more hydrogen bonding to\nstabilize the structure, resulting in more dipoles. For example, annealing\nelectrospun silk fibroin mats were found to undergo self-assembly\nto transform their secondary structure of mostly α-helical to\nsemicrystalline β-sheet through ethanol treatment, which was\nreported to cause the increase in d 33 from ∼1 to\n∼3.5 pC·N 1– . 15 With these freestanding silk mats, the authors were able to create\nwearable devices that produced different voltage outputs when placed\non different body parts. Many other protein-rich materials have also\nbeen claimed to be piezoelectric but have often been characterized\nby methods that do not control for extrinsic contributions, such as\ntriboelectricity. 200 , 201 , 202 Although early studies on piezoelectric proteins focused on\nfibrous\nvariants, recent studies found that globular and transmembrane proteins\nare piezoelectric when assembled into crystals. 27 , 28 One such piezoelectric protein crystal is that of the globular,\nα-helix rich protein lysozyme. Lysozyme can also be transformed\ninto fibrils with a completely different secondary structure. Lysozyme\nfibrils are composed of misfolded versions of the original protein\nand occur naturally in vivo or can be induced through\nchemical treatments. 29 , 30 The fibrillar appearance results\nfrom a highly ordered repeating β-sheet secondary structure,\nheld together by hydrogen bonds. 31 There\nare only a few published studies that study the piezoelectric capabilities\nof amyloid fibrils. 32 − 34 The most recent study done on lysozyme amyloid fibrils\nby Dolai et al., reported piezoelectric d 33 coefficients\nmeasured with piezoresponse force microscopy and used the fibrils\nto generate reactive oxygen species for living cell therapy. 32 However, none of these studies looked at creating\na bulk material made from amyloid fibrils. In this work, we\ndemonstrate the ability of lysozyme amyloid protein\nfibrils to possess piezoelectric capabilities when assembled into\na freestanding bulk material. We transformed hen egg white lysozyme,\na cheap and widely available protein, into amyloid fibrils and drop-cast\nthem to form films. The d 33 piezoelectric coefficient of\nthese films was then characterized with a d 33 piezometer,\nwhich is a more straightforward and reliable method compared to other\nstandard techniques used for characterizing piezoelectric proteins.\nTo improve the flexibility of these films, polyethylene glycol (PEG)\nwas added as a plasticizing agent, and its concentration appeared\nto influence the piezoelectric properties of the films. The potential\ninteractions between the plasticizer and fibrils were studied by Fourier\nTransform Infrared Spectroscopy. Since the arrangement of the amyloid\nfibril may impact piezoelectric properties, the arrangement of the\namyloid fibrils within the films was investigated through imaging\nat different length scales, which revealed the film’s birefringent\nproperties and circumferential fiber alignment.",
"discussion": "Results and Discussion Characterization\nof Lysozyme Amyloid Fibrils In literature,\nthe piezoelectric behavior of lysozyme has been investigated by creating\nfilms made up of crystallized protein and glycerol. 27 Within these crystals, lysozyme exists in its native structure,\nwith a large proportion of α-helical secondary structure. 41 To investigate the dependency of the piezoelectric\nproperties of lysozyme on its structure, we increased the amount of\nβ-sheet secondary structure of the lysozyme by transforming\nits quaternary globular structure into amyloid fibrils. The schematic\nin Figure 1 A demonstrates\nthe transformation, which follows a simple incubation process at low\npH and high temperature, i.e., pH 2 and 65 °C, over 7 days. The\nβ-sheet content in the sample was monitored by measuring the\nchanges in fluorescence intensity from a dye, Thioflavin T (ThT),\nwhich specifically binds to the grooves along the long axis of the\nfibrils formed by β-sheet arrangement. 42 Figure 1 B shows the\nThT fluorescence data v.s. time, which illustrates the growth of the\nfibrils in the solution. For the explored incubation conditions, it\nis clear that the exponential growth happens between day two and day\nsix. Incubating the fibrils longer than a week did not result in a\nsignificant increase in fluorescence readings, indicating that a majority\nof the initial globular lysozyme present has transformed into fibrils,\nand the growth is now in the stationary phase. For consistency, our\nlysozyme fibrils were incubated for a week. Figure 1 (A) Simplified schematic\nof transforming the globular helical structure\nof lysozyme into cross-β-sheet amyloid fibrils. (B) Growth curve\nof lysozyme amyloid fibrils with sigmoidal fit. The growth curve was\nobtained by monitoring the increase in β-sheet secondary structure\ncontent via fluorescence of bound ThT. Each day represents an average\nof triplicate measurements for a solution incubated over several days.\n(C) AFM tapping mode height image demonstrating the fibrillar structure\nof lysozyme formed after 7 days of incubation. (D–F) Piezoresponse\nforce microscopy height, amplitude, and phase images of lysozyme fibrils,\nperformed off-resonance and, (G–I) on-resonance frequency. Atomic force microscopy (AFM), in tapping mode,\nwas used to visualize\nthe fibrillar structure of the lysozyme ( Figure 1 C). The average diameter, 3.0 ± 0.6\nnm, of the lysozyme fibrils was calculated through the height images\nand also showed lengths of several microns (>4 μm). The morphology\nof the amyloid fibril, such as length, diameter and number of protofilaments,\nmainly depends on the incubating parameters of the initial protein.\nSince the lysozyme was incubated at low pH conditions for several\ndays, and the solution’s pH was not adjusted postincubation,\nit was expected for them to be several microns long. Adjusting the\npH to neutral conditions has been shown to reduce their contour length. 43 However, it is interesting to note that although\nwe incubated our lysozyme for at least 7 days, their small width is\nindicative that they are composed of only single to few protofilaments.\nIn literature, it has been shown that lysozyme incubated under similar\npH conditions but at a higher temperature for only 30 h led to their\nfibrils having 17 protofilaments with widths measuring at 173 nm,\nbut their heights still at 3.7 ± 0.4 nm. 30 Due to the dimension of the amyloid fibrils, it is appropriate\nto initially evaluate the piezoelectric response at the nanoscale\nthrough Piezoresponse Force Microscopy (PFM). Briefly, PFM performs\nlocalized piezoelectric measurements by applying a sinusoidal voltage\nthrough a conductive cantilever while in contact with the sample surface.\nThree images are obtained during a PFM scan: height, amplitude, and\nphase. The height image shows surface topographical features, and\nthe amplitude and phase images provide piezoelectric information.\nThe amplitude image is the response of a piezoelectric material experiencing\na sinusoidal strain because of the applied voltage. The straining\neffect leads to the cantilever deflecting from its set point. The\nphase image provides information on the polarization direction of\nthe materials relative to the applied field. Typically, biological\nmaterials show weak piezoelectricity, so the measurements can amplified\nby setting the driving frequency of the sinusoidal voltage to the\nresonance frequency of the cantilever-sample system. 44 In our PFM scans, both off and on-resonance imaging\nwere performed\non the amyloid fibrils. Since the scans were performed in contact\nmode, the fibrils look different from our initial AFM height images\nperformed in tapping mode. Off-resonance scans, Figure 1 D–F, do not show any contrast in the\namplitude and phase scans. However, the on-resonance scans, Figure 1 G–I, display\nboth amplitude and phase contrast. To verify these results, we also\nperformed PFM scans on collagen fibrils (Supporting Information, SI, Figure S1A–C ) since collagen is a well-reported\npiezoelectric protein. 23 , 24 The PFM of collagen also showed\ncontrast in amplitude and phase when imaged on-resonance. Such contrast\nis an indicator that the material is piezoelectric. Strictly speaking,\na nonpiezoelectric material does not demonstrate strain in response\nto a potential, and there is no contrast in PFM amplitude or phase.\nAdditionally, concluding that lysozyme and collagen fibrils are piezoelectric\naligns well with the literature, including for many other biologically\nderived materials such as bacteriophages, human teeth, and porcine\narteries. 201 , 45 − 46 However, the use of on-resonance signal amplification makes it challenging\nto attribute the PFM signal to the piezoelectric effect. Performing\nsimple PFM measurements without accounting for extrinsic contributions\nto the PFM signal can lead to wrongly characterizing a material as\npiezoelectric, which has been reported in the literature for porcine\narteries. 202 Nonpiezoelectric effects,\nsuch as electrostatic interactions between the cantilever and sample,\nare often significant when using soft cantilevers (generally required\nfor biological materials) and are not possible to distinguish from\na true piezoelectric response in PFM. Importantly, signal amplification\non resonance, also amplifies such artifacts and requires advanced\ntechniques to decouple the signal. 49 , 50 We attempted\nto use stiffer cantilevers, 40 N/m (compared to 2.7\nN/m as used in Figure 1 ), to reduce the electrostatic contributions as advised in the literature. 51 However, our initial attempts on collagen fibrils\ndid not detect any amplitude or phase contrast, even when performing\nimaging on-resonance ( SI Figure S1D–F ). This may be due to sample damage (as seen in the height image),\nand even if the collagen fibrils were experiencing a strain effect,\na stiff cantilever might not deflect in response to the softer material.\nThe softer material may instead deform around a stiff cantilever.\nFor our amyloid fibril PFM scans, sample damage can already be observed\nwith the comparatively soft 2.7 N/m cantilever, with the outline of\nthe initial 1 × 1 μm scans appearing in the sequential\nlarger scans ( SI Figure S2 ). Overall,\nthese challenges with performing PFM on biological materials\nemphasize that this technique alone cannot conclusively demonstrate\nbiological piezoelectricity. Simple PFM techniques may be prone to\nartifacts that can be easily misinterpreted as intrinsic piezoelectricity,\nwhich is well documented in the literature. 52 Fabrication and Characterization of the Assembled Amyloid Fibril\nFilm A more standardized way to measure piezoelectricity\nis using a d 33 piezometer. 53 This technique, however, is a bulk measurement and requires casting\nthe amyloid fibrils into a film. Furthermore, similar to polymers,\nto prevent having a symmetry plane that can cancel out induced charges,\nthe amyloid fibrils need to arrange themselves with some long-range\norder in the film in order to enable bulk piezoelectric response in\nthe material. A protocol from Knowles et al. was adopted to cast films\nby simply drop-casting a highly concentrated lysozyme fibril solution\non a hydrophobic silicon wafer. 36 The lysozyme\nfibrils made from this protocol were characterized by CD spectroscopy\nand as expected showed the a change in spectra compared to native\nlysozyme ( SI Figure S3 ), confirming conversion\nof majority α-helical to a majority of β-sheet secondary\nstructure. PEG 400 was added to the concentrated fibril solution as\na plasticizer to increase the flexibility of the films. We explored\nthe proper PEG 400 concentration at a range of 0.8% to 5.2% v/v. Our\nattempts to use 0.8% v/v, as referenced in the literature, resulted\nin films that were too fragile to be measured with a piezometer. However,\nat higher concentrations, as shown in Figure 2 A, the amyloid fibril films demonstrated\na measurable d 33 coefficient and a maximum average value\nof 1.4 ± 0.1 pCN –1 was obtained at 3.6% v/v\nPEG. Concentrations higher than 3.6% v/v led to a significant drop\nin the average d 33 coefficient of the films. Furthermore,\nwe also observed that the molecular weight of the plasticizer had\nan impact on the d 33 of the films. At the same concentration\nof 3.6% v/v, a lower coefficient was measured with PEG 1000 compared\nto PEG 400 films. However, using PEG with a lower molecular weight\nof 200 g/mol led to slightly higher d 33 values but with\nlarger variation. Figure 2 Piezoelectric measurements with a d 33 piezometer\non\nvarious lysozyme films. (A) Average measured d 33 and reversed\nd 33 coefficients for freestanding lysozyme amyloid fibril\nfilms with varying PEG concentrations. (B) Average measured d 33 and reversed d 33 coefficients for freestanding\nlysozyme amyloid fibril films with varying PEG molecular weights.\n(C) Images of lysozyme fibril film (Top), Lysozyme in 10 mM HCl Film\nMilli-Q film (bottom left), and Lysozyme in Milli-Q film (bottom right).\n(D) Average measured d 33 and reversed d 33 coefficient\nof various films. For graphs A, B, D, blue bars represent measured\naverage d 33 values and gray bars represent measured d 33 values after flipping samples 180°. The error bars\nshown for graphs A, B and D represent the standard deviation of three\nindependent samples for each condition. To show that the lysozyme in its amyloid fibril\nstate was responsible\nfor the observed response, films with native lysozyme (predominantly\nα-helical secondary structure), dissolved in Milli-Q water and\nlysozyme in 10 mM HCl with the same PEG concentration, were prepared\nas controls. For the HCl conditions, since the solution was not placed\nat high temperatures, the lysozyme is expected to at least partially\ndenature but not fibrillate. While freestanding films could be cast\nfrom both the native and denatured lysozyme (the HCl control), the\nobtained films had different appearances compared to the fibril films,\nas seen in Figure 2 C. The film cast from native lysozyme in water was transparent, and\nthe film cast from lysozyme dissolved in HCl was white and opaque,\nindicating that the lysozyme has denatured and become poorly soluble,\nforming large insoluble aggregates. Additionally, the fibril films,\nregardless of the concentration, dried up to almost symmetric circular\nfilms, while the two other lysozyme-based films dried more asymmetrically. In regards to the piezoelectric behavior, as expected and shown\nin Figure 2 C,D, the\ncontrol films did not have meaningful d 33 coefficients\nmeasured by the piezometer. No signal was measured from the native\nlysozyme film or parafilm. Parafilm, a polyolefin-based polymer, was\nused as the negative control since its symmetric and nonpolar structure\nmakes it unlikely to be piezoelectric. 54 The denatured lysozyme films did show a significant measurable signal.\nHowever, once the films were remeasured after flipping them over 180°\nthe signal did not change polarity, which is indicative of a nonpiezoelectric\nsignal. Such a signal may come from an extrinsic type of electromechanical\ncoupling, such as the electrochemical effect. Unlike piezoelectricity,\nthe electrochemical effect is based on ion migration within a film\nwhen it experiences a mechanical deformation. 55 , 56 For the lysozyme amyloid fibril film with a PEG concentration\nof\n3.6% v/v, the measured average d 33 coefficient of ∼1.4\npCN –1 also demonstrated an average reversed polarity\nof 1.2 ± 0.1 pCN –1 . Additionally, reversed\npolarity was measured for the other fibril-based films with different\nPEG concentrations. The switching of the coefficient polarity is an\nimportant prerequisite for interpreting the signal as piezoelectric,\nsince piezoelectricity is highly directional. Furthermore, we assembled\nan electromechanical device based on a film with this PEG concentration\nand observed peak-to-peak voltages that increase with compression\nload, and reversed polarity when electrodes were switched ( SI Figure S4 ), demonstrating possible future\napplications in energy harvesting. It is interesting to note that\nthe measured d 33 values fall in the same magnitude as the\nreported coefficients for monoclinic lysozyme crystals, 1.07 ±\n0.27 pCN –1 , despite the fact that the protein secondary\nstructure was changed completely. 27 The\nmechanism behind piezoelectricity in native protein crystals is suggested\nto originate from the dipole moment of the structured water surrounding\nthe protein. 28 For our fibril films, the\nmechanism may be different. It is possible that the mechanism governing\nthe fibril films’ piezoelectric behavior may correlate to that\nof other organic piezoelectric polymers, where individual chains contain\na dipole moment from intra or intermolecular bonds. Under mechanical\nstress, these dipole moments may reorient themselves, leading to a\npiezoelectric effect. Amyloid fibrils have been shown to possess an\nelectric dipole moment through numerical calculations and pulsing\nelectric field experiments. 57 The fibrils\nalso possess a highly ordered β-sheet arrangement that is held\ntogether by hydrogen bonding, the same type of bonding that has been\nsuggested to be responsible for piezoelectricity in silk fibroin. 15 In addition to hydrogen bonding within the fibrils,\nintermolecular interactions between fibrils may also influence the\nbulk piezoelectric response. To investigate such potential interactions\nbetween the protein fibrils and the PEG plasticizer, FTIR was used\nto explore the changes in the films. The FTIR spectra of the\ncomposite films of the protein with increasing\namounts of added PEG are shown in Figure 3 B, in stacked formation with PEG levels increasing\nfrom bottom (1.6% v/v) to top (5.2% v/v). The pure spectrum of PEG\nis given in Figure 3 A and pure lysozyme fibril is shown in Figure 3 C for comparison. All spectra of the composites\nin Figure 3 B show spectral\nfeatures from both PEG and protein fibrils. In the fingerprint region\nbetween 900 to 1800 cm –1 , the spectra of the protein-PEG\nfilms show both the amide bands from the protein and the main band\nC–O–C stretching band of PEG at 1096 cm –1 . The amide I band features two overlapping peaks at approximately\n1620 cm –1 , originating from antiparallel β-sheets\nin the fibrils and 1655 cm –1 , which has previously\nbeen assigned to either α-helices or turns/random coils. 58 − 60 While amyloid fibers do not contain α-helices (only β-sheets\nand turns), we cannot rule out the presence of some unfibrillated\nlysozyme. Comparison of the spectrum of the pure PEG in Figure 3 A with the composite spectra\nin Figure 3 B shows\nthat the main peak for PEG, the C–O–C stretching band\naround 1115 cm –1 ( Figure 3 A), has red-shifted to 1096 cm –1 when combined with the lysozyme fibrils ( Figure 3 B) ( SI Table S1 ). It can also be seen that, even at higher PEG concentrations, the\npeak position remains shifted to the same extent. The shift of this\nband to lower wavenumbers indicates that hydrogen bonding has occurred\nbetween the PEG and the protein fibrils. The introduction of hydrogen\nbonding or any intermolecular interactions weakens covalent bonding,\nhence leading to a red-shift. 61 Figure 3 Attenuated\ntotal reflectance-Fourier transform infrared spectroscopy\ncharacterization of (A) PEG, (B) lysozyme amyloid fibril films with\nPEG as a plasticizer (% v/v), and (C) lysozyme amyloid fibril. All\nthe spectra have been ATR and baseline corrected. The spectra for\nthe protein-PEG films were normalized at the same height for the C–H\ndeformation band around 1460 cm –1 . Comparison of the amide II (C–N\nstretch and C–N–H\nin-plane bend) band at 1520 cm –1 in the pure lysozyme\nfibril spectrum in Figure 3 C with the same band in the PEG-lysozyme composites in Figure 3 B shows that an extra\ncomponent has appeared in the amide II band at 1535 cm –1 upon addition of PEG, that closely overlaps with the 1520 cm –1 lysozyme band, giving the appearance of a broader\namide II band. At the same time, in the higher wavenumber regions,\none can see increases in the C–H stretching band intensities\nof PEG at 2890 at 3450 cm –1 with increasing levels\nof PEG relative to the free N–H stretching band of the lysozyme\nfibril amide at 3300 cm –1 . However, the position\nof the 3300 cm –1 lysozyme band does not shift upon\nthe addition of PEG. Similarly, the positions of the two overlapping\namide I components at 1620 and 1655 cm –1 are not\nshifted upon the addition of PEG. They do, however, undergo changes\nin relative intensities with the 1655 cm –1 band\nincreasing relative to the 1620 cm –1 component,\nindicating conformational changes in the protein as PEG levels are\nincreased. The appearance of a new amide II component band at\n1535 cm –1 while a marked red-shift is seen in the\nPEG C–O–C\nstretching mode upon addition to the lysozyme suggests that the added\nPEG molecules undergo hydrogen bonding with the protein fibrils. It\nappears that the slightly negatively charged –O– in\nthe PEG C–O–C functional groups undergo hydrogen bonding\nwith the slightly positively charged –N– in the (C–N)–\namide groups of the lysozyme. This hydrogen bonding occurs already\nupon the addition of 1.6% of PEG. Changes in the degree of intermolecular\ninteractions, such as hydrogen\nbonding have been reported in the literature to have an impact on\nmechanical properties. 15 Additionally,\nany changes in the mechanical properties of the film will directly\naffect their piezoelectric coefficient, d 33 , since piezoelectricity\ninvolves the elastic deformation and straining of the material. Hence,\nwe characterized the mechanical properties of our films with a nanoindenter. Figure 4 shows the average\nYoung’s modulus of our protein films at an indentation depth\npast 4 μm for different levels of added PEG. The overall trend\nis that the Young’s modulus of the protein film decreases as\nthe amount of plasticizer increases, with the only significant change\noccurring between 1.6% v/v and 2.4% v/v PEG. A high variation in the\nYoung’s modulus can also be observed for samples prepared with\n1.6% PEG, which could be attributed to uneven distribution of the\nPEG and localized areas of little or no PEG leading to much higher\nmodulus value. With no or lower concentrations (<1.6% v/v) of PEG,\nit is expected to have a Young’s modulus in the GPa range and\nthe modulus is expected to decrease with the addition of more plasticizer,\nwith plasticizers of higher molecular weights expected to have less\nof an impact. 36 , 62 In our films, the decreased modulus\nmay contribute to the initial increase observed in d 33 .\nHowever, the mechanical properties do not fully explain the changes\nin d33 with PEG content, and other factors (i.e., protein-PEG interactions)are\ndeemed to be at play, leading to the optimal d 33 to be\nat 3.6% v/v. Figure 4 Average reduced Young’s modulus of lysozyme amyloid\nfibril\nfilms with Varying Amounts of PEG as a Plasticizer. The reduced modulus\nwas measured with a Hysitron nanoindenter on three independent samples\nfor each concentration with five measurements on each sample ( n = 15). Another interesting feature\nin the data was a substrate\neffect,\nobserved as a decrease in Young’s modulus with penetration\ndepth. This effect was observed during the first 4 μm of the\npenetration ( SI Figure S4 ). A substrate\neffect is often encountered in thin films in which the measured Young’s\nmodulus ( E eff ) is affected by the Young’s\nmodulus of the measured film on the surface ( E f ) and the substrate ( E s ). 38 I o is a weighing\nfunction that is dependent on the thickness of the thin film and the\ncontact radius of the indenter. For our films, a thin top layer with\nhigher modulus is observed, with a bulk material underneath acting\nas a “substrate”. The presence of a stiffer thin layer\nsuggests a nonuniform distribution of PEG and protein in the films.\nAs PEG concentration increases, the films become more homogeneous,\nreducing the substrate effect. 1 Apart\nfrom PEG, other common plasticizers\nwere trialed (at the optimal PEG concentration of 3.6% v/v) and evaluated\nin terms of their d 33 coefficient ( SI Figure S5 ). A slightly higher but unstable d 33 coefficient was measured when glycerol was used. However, switching\nto a saccharide-based plasticizer, trehalose, yielded no measurable\nd 33 values. A possible explanation could relate to how\nthe plasticizer influences the water content of the prepared films,\nwhich, in turn, may affect the measured piezoelectric coefficient.\nPrevious literature has demonstrated the effect of humidity on the\npiezoelectric response of other fibrillar proteins, such as collagen. 63 We found that further drying of the films containing\nPEG (by exposing the films to higher temperatures (65 °C) or\nto dry N 2 gas) resulted in brittle films with measurable\nmass loss and a zero d 33 response. When these films were\nrehydrated (gaining mass), they generally became less robust and more\nbrittle, with a lower piezoelectric response compared to as-prepared\nfilms. This phenomenon may be due to the brittleness caused by drying,\nwhich prevents the fibrils from elastically deforming. Additionally,\nupon rehydration, some fibrils may have lost their structural integrity,\nleading to reduced d 33 responses. Dried fibrils, upon rehydration,\nhave previously been shown to break into smaller pieces. 64 Overall, the introduction of PEG as a\nplasticizing agent in the\nfilms leads to the complex interaction between the amyloid fibrils\nand PEG in the form of hydrogen bonding and changes in the protein\nstructural conformation. In regard to the d 33 trend observed\nwith PEG concentration, there could be a balancing effect between\nthese two types of interactions. The initial increase in d 33 likely originates from the introduction of hydrogen bonding, providing\nan additional source for piezoelectric charges. Another potential\nimpact of PEG could be its mediation of fibril organization. To investigate\nthe fibril arrangements in the films, we applied various imaging techniques. Characterization of Fiber Alignment within the Assembled Amyloid\nFibril Film The addition of a plasticizer has been reported\nto help in the nematic stacking and alignment of the lysozyme amyloid\nfibrils within the plane of the film. 36 Our measured overall bulk piezoelectric effect of the films indicates\nthe presence of some long-range order in the films. A random fibril\narrangement could lead to the produced electrical charges canceling\neach other and resulting in a zero d 33 coefficient. Initially,\nSEM and AFM were employed to visualize the fibril orientation directly.\nHowever, neither of the two imaging techniques was able to resolve\nfibrils within the films, as shown in Figure 5 A–D. Under SEM and AFM ( Figure 5 B,C), the surfaces of the films\nwere relatively rough with globular features. This is assumed to be\nthe thin layer measured by nanoindentation, likely a mixture of surplus\nPEG and any unfibrillated lysozyme. Additionally, SEM also revealed\nthat the bottom surfaces were relatively flatter than the top ( SI Figure S7 ). For AFM, unsuccessful attempts\nwere made to remove the top surface layer by microtoming ( Figure 5 C). Hence, the surface\nwas rinsed with ethanol to remove surplus PEG ( Figure 5 D). The rinsing method did allow some of\nthe fibrils to be visualized. However, ethanol rinsing was not enough\nto allow for a clear visualization of the fibril orientation. Figure 5 Microscopy\nimages of amyloid fibril films with 3.6% v/v PEG. (A)\nOptical Microscopy in Bright Field Mode and (B) SEM of the film’s\nsurface. AFM height images of films after microtome (C) and ethanol\ntreatment (D). TEM images of fixed and microtome slice (E) with 538\npositive staining and amyloid fibrils with negative staining (F). For TEM, the bulk films were microtomed to 80 nm,\nand sections\nfrom the central portion of the film were imaged, revealing bundles\nof fibrils, as seen in Figure 5 E. The bundles seen are assumed to be the amyloid fibrils\nsince the individual filaments of the fiber bundle appeared to be\nwithin the same scale as individual amyloid fibrils Figure 5 F. The low contrast achieved\nwith the film could be due to the high concentration of the fibrils.\nThe dense films can make it difficult to stain each fibril individually\nto achieve good contrast against the background. Importantly, the\nobserved bundles appear to possess directionality, which is necessary\nfor overall film piezoelectricity. Polarization imaging was\nused to confirm the fibril orientation\nat longer length scales. Under a simple setup with cross-polarized\nlight (CPL), the films demonstrated a Maltese cross pattern ( Figure 6 A,B), often associated\nwith anisotropic birefringence or circumferentially aligned particles. 65 However, using CPL is insufficient to determine\nif the polarization image is from linear retardance due to structural\nanisotropy (linear birefringence) as well as the orientation of the\noptical anisotropy. Mueller matrix imaging, a polarization imaging\ntechnique, can measure and distinguish all the polarization properties\nof a material, namely retardance, diattenuation, and depolarization. 39 , 40 Mueller matrix imaging of the films showed that they predominantly\nexhibited linear retardance and had negligible circular retardance,\ndiattenuation (linear and circular) and depolarization ( SI Figure S8 ). The linear retardance observed\nin the films ( Figure 6 C) indicated an axis of optical anisotropy, given by the optic axis\norientation, that was aligned with the circumferential direction about\nthe center of the circular films ( Figure 6 D,E). Assuming the films exhibit a positive\nlinear birefringence, this would suggest the fibrils are arranged\ncircumferentially. Additionally, polarization imaging on films with\nthe lowest concentration PEG (1.6% v/v) showed no significant difference\nin the retardance intensity or fibrillar arrangement to films with\nhigher PEG, indicating that fibrillar ordering was not dependent on\nPEG concentration. Figure 6 Polarization imaging of lysozyme amyloid fibril films.\nCross-polarized\nlight images for films with (A) 3.6% v/v and (B) 1.6% v/v PEG, demonstrating\na Maltese cross pattern. Mueller matrix imaging maps of linear retardance,\noptic axis orientation, and fiber maps with (C, D, E) 3.6% v/v PEG\nand (F, G, H) 1.6% v/v indicated a circumferential alignment of fibrils.\nThe red lines in the maps in E and H have orientation corresponding\nto the mean optic axis orientation and length scaled by 1 minus the\nvariance in the optic axis orientation for each region (50 pixels\n× 50 pixels). This observed circumferential\narrangement could\nexplain why these\nfibril-PEG films dried up into asymmetric circular films compared\nto the control films ( Figure 2 B). Interestingly, this observed arrangement appears instantaneously\nafter casting the solution when observed under CPL (data not shown).\nThis is in contradiction to our initial assumption that the addition\nof PEG as a plasticizer allowed the fibrils to arrange themselves\nas the film dries. In literature, the circumferential arrangement\nof fibers has been reported in drop-casted films of cellulose nanofibrils. 66 However, in the work by Skogberg et al., the\ndroplet underwent evaporation-induced drying, which was claimed to\nbe the driving force for the alignment. 66 For our films, we assumed that the shear forces drive the alignment\nof the fibrils. The drop-casting method to make our films allows the\nsolution to spread out radially on the hydrophobic substrate, resulting\nin a circular alignment. To further confirm that shear forces can\ncontrol the fibril orientation on the substrate, an experiment was\nconducted where the flow direction of a solution over a silicon wafer\nwas controlled and imaged afterward. Initially the solution was allowed\nto drop diagonally on the wafer, which was then imaged under AFM.\nThe height images obtained for these wafers show the fibrils all aligned\nin the same diagonal direction, Figure 7 A. Subsequently, on the same wafer, another drop of\nthe solution was placed but was made to move cross-diagonally instead.\nThe later height images for the wafers showed fibrils crossing, Figure 7 B. Hence, the obtained\nAFM images demonstrate the ability to align the amyloid fibrils through\nshear forces. Figure 7 Schematic and AFM height images obtained after allowing\nfibril\nsolution to drop (A) diagonally across the hydrophilic silicon substrate\nand (B) cross-diagonally afterward leading to diagonally aligned fibrils. Overall, lysozyme in their amyloid fibril form\ncan self-assemble\nthemselves into freestanding films that demonstrate piezoelectric\nproperties. The addition of PEG as a plasticizer affects not only\nthe mechanical properties of the films but also their bulk piezoelectric\nproperties. FTIR studies on these films highlight that there are complex\ninteractions between PEG and the amyloid fibrils, evidenced by shifts\nin band location and changes in the peak appearance in the FTIR spectra.\nWe propose that these interactions may come from hydrogen bonding,\nwhich contributes to the bulk piezoelectric of our fibril-PEG films,\nin line with the mechanism seen in other piezoelectric biopolymers. 15 , 67 Hence, the ability of these amyloid fibrils to assemble themselves\ninto a bulk material with anisotropy and piezoelectric properties\ndemonstrates their potential to be used in biodegradable piezoelectric-based\ndevices. 6 Additionally, amyloid fibrils\ncan be assembled from a range of other proteins, from functional amyloids\nto globular proteins, which could potentially be sourced from waste\nmaterials. 68 , 69 The current limitation of these\namyloid fibril films is their low piezoelectric coefficients, which\nrestrict their use in broader applications. However, since the fibrils\nare shown to be capable of modifiable bulk alignment, extrusion-based\nmethods may be able to align the fibrils linearly instead of circumferentially\nfor higher coefficients. Furthermore, functionalizing the fibrils,\nfabricating composite films, and selecting the best polymorphs of\nthe fibrils may positively influence the coefficients overall and\nlead to better and more efficient devices."
} | 10,222 |
27602022 | PMC4994424 | pmc | 8,545 | {
"abstract": "The application of next-generation sequencing technology in microbial community analysis increased our knowledge and understanding of the complexity and diversity of a variety of ecosystems. In contrast to Bacteria, the archaeal domain was often not particularly addressed in the analysis of microbial communities. Consequently, established primers specifically amplifying the archaeal 16S ribosomal gene region are scarce compared to the variety of primers targeting bacterial sequences. In this study, we aimed to validate archaeal primers suitable for high throughput next generation sequencing. Three archaeal 16S primer pairs as well as two bacterial and one general microbial 16S primer pairs were comprehensively tested by in-silico evaluation and performing an experimental analysis of a complex microbial community of a biogas reactor. The results obtained clearly demonstrate that comparability of community profiles established using different primer pairs is difficult. 16S rRNA gene data derived from a shotgun metagenome of the same reactor sample added an additional perspective on the community structure. Furthermore, in-silico evaluation of primers, especially those for amplification of archaeal 16S rRNA gene regions, does not necessarily reflect the results obtained in experimental approaches. In the latter, archaeal primer pair ArchV34 showed the highest similarity to the archaeal community structure compared to observed by the metagenomic approach and thus appears to be the appropriate for analyzing archaeal communities in biogas reactors. However, a disadvantage of this primer pair was its low specificity for the archaeal domain in the experimental application leading to high amounts of bacterial sequences within the dataset. Overall our results indicate a rather limited comparability between community structures investigated and determined using different primer pairs as well as between metagenome and 16S rRNA gene amplicon based community structure analysis. This finding, previously shown for Bacteria, was as well observed for the archaeal domain.",
"introduction": "Introduction The investigation of the microbial community composition allows a detailed insight in diversity and potential ecosystem function and fosters understanding of complex microbial processes (Vanwonterghem et al., 2014 ). Recent years have seen a strong increase in sequencing approaches targeting microbial communities via amplicon sequencing or metagenomic and metatranscriptomic approaches (Turnbaugh et al., 2007 ; Hamady et al., 2008 ; Raes and Bork, 2008 ; Caporaso et al., 2012 ; Grosskopf and Soyer, 2014 ; Ininbergs et al., 2015 ). These approaches play an important role in monitoring and comparing large numbers of samples in terms of their microbial composition (Caporaso et al., 2012 ; Kozich et al., 2013 ; Sundberg et al., 2013 ). The by far most often used marker for prokaryotic diversity studies is the 16S rRNA or its corresponding gene. The first to perform extensive research based on the 16S region were Woese and Fox (Woese and Fox, 1977 ; Woese et al., 1990 ). Their profound and passionate work led to the discovery of the third domain of life, the Archaea (Woese et al., 1990 ). Since then, the contribution of the archaeal domain to ecosystem function and diversity was often underestimated in many research fields and studies. While the bacterial fraction of many environments was extensively studied, the Archaea were often not specifically addressed. This underestimation of archaeal contribution to biology can be observed in a variety of studies from Sanger sequencing-based approaches to 454- and MiSeq-based high-throughput sequencing based studies (Frank et al., 2007 ; Herlemann et al., 2011 ; Ding and Schloss, 2014 ; Wang et al., 2015 ). Many reports focusing on Archaea appear to explore extreme environments like hot springs (Beam et al., 2015 ), deep sea volcanos (Reysenbach et al., 2006 ), and black smokers (Takai and Nakamura, 2011 ) to only mention a few, which further promotes the image of Archaea to represent extremophiles. On the contrary, Archaea are ubiquitously found under rather mesophilic conditions like in fresh and marine waters (DeLong, 1992 ; DeLong et al., 1994 ; Karner et al., 2001 ; Stahl and de la Torre, 2012 ), biogas reactors (Sundberg et al., 2013 ), and soil (Leininger et al., 2006 ), the intestinal tract of termites (Paul et al., 2012 ), ruminants (Jeyanathan et al., 2011 ; Kittelmann et al., 2013 ), but also on the human skin (Probst et al., 2013 ; Oh et al., 2014 ), or in the intestine (recently reviewed in Bang and Schmitz, 2015 ), where they complete the microbiome together with their bacterial, eukaryotic and viral partners. Regarding biogeochemical cycles, the archaea harbor the unique trait of the methanogenic pathway (Offre et al., 2013 ). The methane emission by archaeal activity is used in industrial scale as a beneficial source of renewable energy in biogas reactors but is problematic when observed under the perspective of greenhouse gas emission. Two major sources of anthropogenic methane emission is livestock and rice patty fields (Yusuf et al., 2012 ), two habitats known to harbor methanogenic archaeal communities (Janssen and Kirs, 2008 ; Kittelmann et al., 2013 ; Breidenbach and Conrad, 2014 ) and both contributing notable amounts to the overall anthropogenic greenhouse gas emission (Wuebbles, 2002 ; Ripple et al., 2014 ). Additionally, in livestock methane production by the enteric community leads to an energy loss for the host by the emission of the energy rich methane and several studies investigate potential inhibitors of archaeal methane production (Goel and Makkar, 2012 ; Duin et al., 2016 ). The archaeal communities of ruminants has therefore been in the focus of several studies in recent years (Skillman et al., 2004 ; Jeyanathan et al., 2011 ; Kim et al., 2011 ; Singh et al., 2012 ; Tymensen and McAllister, 2012 ; Kittelmann et al., 2013 ; Henderson et al., 2015 ), some of them involving primer evaluation for the archaeal community (Watanabe et al., 2004 ; Gantner et al., 2011 ) or extending the microbiome research by adding results for protozoa and fungi (Kittelmann et al., 2013 ). In an extensive study, Klindworth et al. ( 2013 ) performed a detailed in-silico evaluation of a 16S rRNA primer dataset containing 175 primers and 512 primer pairs, with 72 primers targeting archaeal 16S gene sequences. Primers and primer pairs were tested against the SILVA 16S non-redundant reference database to estimate their accuracy and phylogenetic coverage. Inspired by this study, we tested the experimental applicability of several primer combinations—some recommended in the above mentioned study, others supplemented based on literature review. After initial in-silico validation, the six most promising primer pairs were chosen; three targeting the archaeal, two the bacterial and one overall prokaryotic 16S rRNA gene sequence. These primer pairs showed high in-silico coverage and specificity, and were used to investigate the microbial community of an anaerobic, mesophilic biogas reactor, a habitat known to host a diverse community of Archaea and Bacteria (Eikmeyer et al., 2013 ; Sundberg et al., 2013 ). To eliminate disruptive effects and ensure maximum comparability, we used the same template DNA extracted from one sample of the above mentioned biogas reactor for all approaches. Shotgun metagenomic approaches have been introduced into community analysis (Venter et al., 2004 ) and bear the additional advantage of hinting toward ecosystem potential beside the taxonomic information (Vanwonterghem et al., 2014 ). Renunciation of 16S rRNA gene amplification is another positive effect of shotgun metagenomics, as it rids the data of primer bias (Shakya et al., 2013 ; Logares et al., 2014 ; Tremblay et al., 2015 ). Thus, as an additional and independent approach, we used 16S rRNA gene data obtained in a very comprehensive metagenome sequencing approach of the same biogas fermenter material (Güllert et al., 2016 ) as a reference point for comparison. This study aims to estimate the effect of primer choice on the observed sequence composition of a diverse microbial community. Contrary to other studies focusing on the evaluation of bacterial 16S rRNA primers, we focus here on the evaluation and observation of the archaeal community in more detail. We further critically discuss the reliability of in-silico primer evaluation in terms of unspecific amplification and target specificity in application to environmental samples. Additionally, the 16S rRNA gene amplicon based community profiles were compared to the 16S rRNA gene sequences extracted and assembled from shotgun metagenomic data.",
"discussion": "Discussion The choice of the primer pair for 16S rRNA gene amplification substantially determines quality and perspective on the obtained community data. Numerous recent studies address the topic of comparability between 16S rRNA gene based projects, demonstrating the influences of different effects with the help of mock communities and simulated datasets (Schloss et al., 2011 ; Brooks et al., 2015 ; Tremblay et al., 2015 ). However, complexity of environmental samples cannot be fully mimicked by artificially generated communities and the effects due to the choice of the primer pairs for analyzing complex environmental samples remain in question. Whereas most of the above mentioned studies focus mainly on Bacteria, here we presented comprehensive data generated for the archaeal and bacterial fraction of a complex environment, where we observed similar tendencies of primer effects in both domains. Based on our results five core statements can be formulated: All primer pairs were able to recover and represent a typical complex microbial community of an anaerobic biogas reactor, yet with a different outcome concerning the details of community structure. Sequences of key organisms for major steps in hydrolysis, acidogenesis and acetogenesis mostly belonging to the Clostridia, Bacteroidia , and Actinobacteria were observed in all bacterial datasets. The archaeal datasets provided sequences of species capable of hydrogenotrophic, acetoclastic, and methylotrophic methanogenesis (Wirth et al., 2012 ; Sundberg et al., 2013 ). Especially in the sequences of the archaeal community, a clear abundances ranking of taxonomic orders ( Methanosarcinales > Methanomicrobiales > Methanobacteriales ≥ Methanomassiliicoccales ) was consistently conserved in all tested primer pairs as well as in the sequences obtained from the metagenome. Similarly, ranking order of the two most abundant bacterial classes ( Clostridia > Bacteroidia ) was conserved in the sequence abundance in all bacterial datasets. In the metagenome derived 16S rRNA gene sequences, the classes Clostridia and Bacteroidia were highly abundant (18 and 16%) but in addition sequences of unclassified Cloacimonetes contributed 16% to the dataset (Figure 4 ). This taxon would have been missed using the primer pairs BacV12 or PrkV4 in the analysis of the environment. Organisms of this class were recently found in metagenomic datasets from anaerobic digesters (Solli et al., 2014 ) and are expected to participate in syntrophic degradation of fatty acids and protein intermediates (Pelletier et al., 2008 ; Limam et al., 2014 ). The lower proportion of Clostridia in the metagenomic 16S rRNA gene sequences might indicate an overestimation of this class in the amplicon based approaches. In our case, the most promising combination to analyze the community of the sample would have been the combination of the BacV35 and ArchV46 primer pairs, however the metagenomic sequences still show a different overall bacterial and archaeal community compared to that observed by those primer pairs (Figure 6 ). From a technical point of view, efficiency strongly differed between the evaluated primer pairs, namely due to unspecific amplification by archaeal primer pairs. While 65% of the raw reads from the primer pair ArchV56 could be used for the analysis, the read reduction for the primer pair ArchV34 left only 11% of the raw reads for the final analysis of the archaeal community. Reads filtered from the archaeal datasets mostly belonged to the bacterial domain, with high abundance of the class “ Candidatus ” Cloacimonas . While almost all removed sequences from the primer pair ArchV34 belonged to this class, sequences removed from the datasets of the primers ArchV46 and ArchV56 showed a higher diversity but, as expected, were not comparable in composition to the tested bacterial primer pairs (Figure S1 ). Low specificity of the ArchV34 primer pair toward archaea was not predicted by the Silva TestPrime tool (Klindworth et al., 2013 ), the reason for this being unclear. One of the primers applied in this primer pair, S-D-Arch-0787-a-A-20, showed potential unspecific amplification within the bacterial domain in the in-silico prediction. Inexplicably, the same primer was used as reverse complement in the primer pair ArchV56, which exhibited highest specificity of all archaeal primer pairs tested. This shows that the outcome of sequencing runs is still highly unpredictable and database results cannot be directly transferred to the wet lab application. Alpha diversity (Shannon numbers equivalent 1 D) differed between primer pairs. In general, higher observed alpha diversity for a specific primer pair indicates higher resolution of the present diversity (i.e., better separation of OTUs) in a given sample. The alpha diversity in general as well as the Shannon number equivalents was lower in the archaeal datasets compared to the bacterial. This observation has previously been made for comparable biogas reactors (Francisci et al., 2015 ). The Shannon numbers equivalent was selected as alpha diversity metric as it is more robust in the application in environmental data and can be seen as the number of equally abundant species needed to form the diversity observed within a given dataset (Jost et al., 2010 ). For the primers targeting the archaeal 16S rRNA gene regions, the observed alpha diversity correlates well with the prediction of the database coverage (see Table 3 ). High diversities were observed for the primer pairs ArchV46 ( 1 D = 7.23) and PrkV4 ( 1 D = 8.81) covering the highly variable region V4 (Cai et al., 2013 ). In combination with a high potential coverage of different phyla, ArchV46 thus appears to be quite promising and was recently applied for the analysis of the archaeal domain in a mesophilic anaerobic digesters (Goux et al., 2015 ). The high diversity observed in the PrkV4 dataset correlates well with the high theoretical coverage as predicted by the in-silico evaluation. In our study, the abundances of the domains Archaea and Bacteria were similar in the overall metagenomic (2.4–97.6%) and the amplicon based (3.2–96.8%) 16S rRNA gene sequences generated by primer pair PrkV4, thus showing no strong shift in the proportions of the two domains during amplicon generation. The ArchV56 dataset showed diversity within the same range, a finding which correlated well with the in-silico predicted coverage. The observed positive correlation of in-silico database coverage and observed diversity was also valid for the tested bacterial primer pairs. Beta diversity analysis showed good reproducibility within a primer pair, but poor comparability between primer pairs as shown in the heat map (Figure 6 ). Significant differences between the communities amplified by the tested primer pairs resulted from differential amplification of 16S rRNA gene sequences out of the same starting material as well as the ability of the amplified variable region to discriminate between different taxa (Shakya et al., 2013 ; Tremblay et al., 2015 ). This bias in amplification and classification is also assumed to be the reason for the observed differences compared to the metagenome derived sequences. The comparison to 16S rRNA gene sequences from metagenomic data has previously shown to conform with 16S rRNA amplicon sequences, generated from environmental samples like sheep rumen (Shi et al., 2014 ). For the archaeal dataset generated by the PrkV4, we observed lower clustering in the RDA (Figure 7A ) as well as higher within group variance, which could already be observed in the community composition (Figure 3 ). This resulted from overall lower sequencing depth of the archaeal domain in this dataset. As mentioned before, the overall archaeal abundance within the dataset was 3.2% compared to the also targeted bacterial sequences. As described by Kittelmann et al. ( 2013 ), the application of multiple domain specific primer sets can be beneficial when strong differences in the abundance of the different domains can be observed in a habitat. In the before mentioned study, the rumen communities of several ruminants were investigated by the application for several specific primer sets targeting Bacteria, Archaea, Protozoa and Fungi and pooled in alternating proportions for sequencing to account for the different abundances of the rumen microbiota (Kittelmann et al., 2013 ). In conclusion, to reduce the within group variation of the PrkV4 archaeal dataset an increased number of sequences would be needed for a saturating analysis of the archaeal domain using the general PrkV4 primer pair. As alternative, a separate analysis of the archaeal and bacterial domain might be beneficial for the investigated habitat. To give an advice on which primer pair to use is difficult, if not impossible since the choice depends on the habitat and the research question. However we summarized the results obtained in this study in Table 3 to better compare the dis-advantages of the primer pairs tested in the presented study. Primer pair ArchV34 showed the highest similarity to the archaeal domain of the metagenomic results. Still, it cannot be fully recommended since it was highly unspecific for archaeal sequences. Due to this fact, only 11% of the sequences obtained for this primer pair could be used for analysis. The ArchV46 primer pair showed moderate specificity and high diversity which makes it a reliable candidate for the investigation of new archaeal taxa in diverse environments. It was successfully applied in a recent multi-omics approach investigating the archaeal domain in an anaerobic paddy field. In this complex environment it was able to detect a complex archaeal community, consisting mostly of Methanomicrobia, Methanomassiliicoccales and Methanobacteria , as well as some Thermoprotei (Ogawa et al., 2014 ). The primer pair ArchV56 showed the highest specificity for archaeal 16S rRNA gene sequences. Compared to the metagenomic archaeal community, the observed diversity and similarity was average, which makes this primer pair a decent choice for the detection of archaea, especially in environments with a low abundance of archaea. In combination with a fluorescent probe, this primer pair was originally designed for the quantification of archaea (Yu et al., 2005 ) and was applied in this context in diverse studies (Lee et al., 2008 ; Nettmann et al., 2008 ). Beside good theoretical coverage and high diversity, a clear advantage of the PrkV4 primer pair is the simultaneous amplification of archaea and bacterial 16S rRNA gene sequences which can be helpful in the investigation of synergies between the archaeal and bacterial domain in the environment. This advantage was already confirmed in other environmental studies like in the investigation of the coral-associated microbiota containing substantial amounts of Thaumarchaeota and minor amounts of Euryarchaeota , which were detected in coral mucus for the first time (van Bleijswijk et al., 2015 ). Unfortunately, compared to the bacterial 16S rRNA gene sequences extracted from the metagenome, the sequences generated with the PrkV4 primer pair showed the lowest similarity. One prominent prokaryotic primer pair is applied in the earth microbiome project (Caporaso et al., 2012 ). The PrkV4 primer pair showed higher in-silico coverage compared to the one applied in the earth microbiome project, still this study cannot provide a direct comparison. The primer pair BacV12 has been used in diverse medical and environmental studies so far (Rausch et al., 2011 ; Cozen et al., 2013 ; Langfeldt et al., 2014 ; Mensch et al., 2016 ). Even though the primer pair sowed the highest similarity to the bacterial domain observed in the metagenome, the low theoretical coverage and low observed diversity within the samples may hint toward a possible non-observance of present species. Average results in terms of diversity and similarity to the metagenomic results were observed for the primer pair BacV35. The amplification of the highly variable region 4 of the 16S rRNA gene qualifies this primer pair as a good candidate when the focus of the study lies on the bacterial domain only (Güllert et al., 2016 ). In summary, it appears most beneficial to use the same primer pair when comparing different sites or environments by amplicon sequencing. This assumption has previously been made for bacterial communities (Baker et al., 2003 ; Frank et al., 2008 ; Tremblay et al., 2015 ) and, as shown here, is also valid for archaeal primer pairs. It should be mentioned that additional effects influencing the observed community structure also occur in form of nucleic acid extraction (LaMontagne et al., 2002 ; Brooks et al., 2015 ), the kits applied (Adams et al., 2015 ), PCR artifacts (Schloss et al., 2011 ; Brooks et al., 2015 ), and database bias (Werner et al., 2012 ), as well as bias introduced by selected hypervariable region (Chakravorty et al., 2007 ; Yu et al., 2008 ), the sequencing platform itself (Kim et al., 2011 ; Luo et al., 2012 ; Tremblay et al., 2015 ) or the sequencing center (Hiergeist et al., 2016 ). In addition, comparability between communities analyzed with different primer pairs is bound to taxonomically assigned sequences and therefore limited and biased by the completeness of the database (Werner et al., 2012 ). These aspects further emphasize the need for general standards when planning and conducting environmental microbiological research for the sake of improved comparability, like the human (Turnbaugh et al., 2007 ; Peterson et al., 2009 ) or earth microbiome project guidelines (Gilbert et al., 2010 ) with profound and detailed manual and instruction for the sample preparation, which is of great help for between-study comparability. Finally, a comparison to 16S rRNA gene sequences gained from a corresponding metagenome as presented here appears very helpful and can be recommended as an addition to the mock community testing (Brooks et al., 2015 ; Tremblay et al., 2015 ) for the evaluation of new archaeal or bacterial primer pairs, especially when the community composition of the investigated environment is yet undetermined."
} | 5,776 |
24769753 | PMC4040992 | pmc | 8,546 | {
"abstract": "Corallimorpharia is a small Order of skeleton-less animals that is closely related to the reef-building corals (Scleractinia) and of fundamental interest in the context of understanding the potential impacts of climate change in the future on coral reefs. The relationship between the nominal Orders Corallimorpharia and Scleractinia is controversial—the former is either the closest outgroup to the Scleractinia or alternatively is derived from corals via skeleton loss. This latter scenario, the “naked coral” hypothesis, is strongly supported by analyses based on mitochondrial (mt) protein sequences, whereas the former is equally strongly supported by analyses of mt nucleotide sequences. The “naked coral” hypothesis seeks to link skeleton loss in the putative ancestor of corallimorpharians with a period of elevated oceanic CO 2 during the Cretaceous, leading to the idea that these skeleton-less animals may be harbingers for the fate of coral reefs under global climate change. In an attempt to better understand their evolutionary relationships, we examined mt genome organization in a representative range (12 species, representing 3 of the 4 extant families) of corallimorpharians and compared these patterns with other Hexacorallia. The most surprising finding was that mt genome organization in Corallimorphus profundus , a deep-water species that is the most scleractinian-like of all corallimorpharians on the basis of morphology, was much more similar to the common scleractinian pattern than to those of other corallimorpharians. This finding is consistent with the idea that C. profundus represents a key position in the coral <-> corallimorpharian transition.",
"introduction": "Introduction Understanding the evolutionary history of the Scleractinia and relationships between corals and other members of the anthozoan subclass Hexacorallia should enable a better understanding of how it has been influenced by climate in the past and thus enable better predictions of the likely impacts of climate change ( Romano and Palumbi 1996 ). Of the six Orders of hexacorals, only members of the Scleractinia develop continuous external calcified skeletons ( Daly et al. 2003 ). The Scleractinia suddenly appear in the fossil record in the middle Triassic, about 240 Ma, but the range of morphological variation seen in the Middle Triassic fossils is comparable to that of extant scleractinians ( Romano and Palumbi 1996 ). Molecular phylogenies based on both mitochondrial (mt) and nuclear (nucl) genes imply a deeper divergence (∼300 Ma—in the Late Carboniferous) of extant scleractinians into two major clades, the “Complexa” and the “Robusta” ( Romano and Palumbi 1996 ; Romano and Cairns 2000 ; Chen et al. 2002 ; Le Goff-Vitry et al. 2004 ; Fukami et al. 2008 ; Barbeitos et al. 2010 ; Kitahara, Cairns, and Miller 2010 ; Kitahara, Cairns, Stolarski, et al. 2010 ; Kitahara, Cairns, et al. 2012 ; Kitahara et al. 2012 ; Kayal et al. 2013 ). By adding deep-water species to existing molecular data sets and applying an appropriately calibrated molecular clock, Stolarski et al. (2011) demonstrated that two exclusively deep-sea families, the Gardineriidae and Micrabaciidae, form a “basal” clade that diverged at around 425 Ma, prior to the Complexa/Robusta split, pushing the evolutionary origin of scleractinians deep into the Paleozoic. These results support the scenario that scleractinians are the descendants of soft-bodied (corallimorpharian-like) ancestors that survived the mass extinction at the Permian/Triassic boundary and subsequently gained the ability to deposit calcified skeletons ( Stolarski et al. 2011 ). The “naked coral” hypothesis, first put forward by Stanley and Fautin (2001) to explain the sudden appearance of diverse scleractinian fauna in the middle Triassic, is based on the idea that the skeleton has been an ephemeral trait during coral evolution. Under this hypothesis, the Scleractinia were skeleton-less in the early Triassic, a time when carbonate deposition was suppressed globally ( Stanley 2003 ). Consistent with the idea of skeleton ephemerality, some coral species can undergo reversible skeleton loss under acid conditions ( Fine and Tchernov 2007 ). Strong phylogenetic support for the “naked coral” hypothesis came from analyses based on the alignment of concatenated proteins encoded by 17 complete mt genomes from hexacorallians ( Medina et al. 2006 ); in their analysis, scleractinians were paraphyletic, corallimorpharians being more closely related to the Complexa than are Robusta, the interpretation being that the Corallimorpharia arose by skeleton loss from a scleractinian ancestor at a time (during the mid-Cretaceous) of high oceanic CO 2 levels ( Medina et al. 2006 ). Although the “naked coral” scenario is supported by analyses of protein sequence data, phylogenetics based on mt nucleotide sequences instead strongly support scleractinian monophyly ( Stolarski et al. 2011 ; Kayal et al. 2013 ; Kitahara et al. 2014 ). The fundamental disagreement between phylogenies based on nucleotide ( fig. 1 A ) or amino acid ( fig. 1 B ) sequence data for mt proteins stems from the fact that none of the available models for sequence evolution adequately account for the observed data ( Kitahara et al. 2014 ). One possible explanation for this is the occurrence of a “catastrophic” event—a major and unpredictable change, such as sudden impairment of mt DNA repair processes (which are believed to be an ancestral trait within Anthozoa ( Pont-Kingdon et al. 1998 ; Shearer et al. 2002 ; Brockman and McFadden 2012 ).\n F ig . 1.— Alternative phylogenetic hypotheses for relationships between Scleractinia and Corallimorpharia based on mt genome nucleotide sequences ( A ) or the amino acid sequences of the proteins that they encode ( B ). The trees were modified from Kitahara et al. (2014) . Note that, for both ( A ) and ( B ) scenarios, support for the node separating Corallimorpharia from Scleractinia (the root of the gray part of the tree) was over 97% under both maximum-likelihood analysis and Bayesian inference. Given the intractability of coral/corallimorph relationships using conventional molecular phylogenetics, we explored the informativeness of mt genome architecture in this context. mt gene rearrangements occur relatively infrequently and have proven useful in resolving evolutionary relationships, both shallow and deep, across a broad range of organisms (e.g., Gai et al. 2008 ; Brockman and McFadden 2012 ; Kilpert et al. 2012 ). This study is based on the complete mt genomes of a total of 12 corallimorpharians (8 of which are novel), representing 3 of 4 currently described families ( Daly et al. 2007 ; Fautin et al. 2007 ), and 32 scleractinians, and includes both the early diverging coral Gardineria hawaiiensis ( Stolarski et al. 2011 ), and corallimorpharian, Corallimorphus profundus , which is considered to be the most coral-like of corallimorpharians based on morphological grounds ( Moseley 1877 ; den Hartog 1980 ; Riemann-Zürneck and Iken 2003 ). The results indicate that, by contrast with the Scleractinia, extensive rearrangements of the mt genome have occurred within Corallimorpharia. The most surprising finding, however, was that the mt genome of C. profundus is scleractinian-like, and is organized very differently to those of all other corallimorpharians for which data are available. Both nucleotide and amino acid sequenced-based phylogenetics unequivocally place C. profundus as an early diverging corallimorpharian, indicating that this organism most closely reflects the coral <-> corallimorpharian transition.",
"discussion": "Discussion The most surprising finding of this study was that the mt genome of the deep-sea corallimorpharian, C . profundus , more closely resembles scleractinians in gene organization than it does other corallimorpharians ( fig. 3 A and B ). Although molecular phylogenetic analyses based on nucleotide or amino acid sequence data for mt proteins yield fundamentally different results with respect to the relationship between the “complex” and “robust” scleractinian clades, there is no disagreement concerning the monophyly of the Corallimorpharia nor about the early divergence of Corallimorphus within that clade ( fig. 1 ; Kitahara et al. 2014 ). On morphological grounds, Corallimorphus is also considered the most coral like of corallimorpharians ( Moseley 1877 ; den Hartog 1980 ; Riemann-Zürneck and Iken 2003 ). Several authors ( den Hartog 1980 ; Owens 1984 ; Cairns 1989 , 1990 ; Fautin and Lowenstein 1992 ) have pointed out the level of similarity between Corallimorphus and members of the scleractinian family Micrabaciidae, which are characterized by a reduced skeleton, the fleshy polyp totally investing the rudimentary corallum. Molecular clock estimates imply that the micrabaciids and gardineriids diverged from the scleractinian lineage in the mid-Paleozoic, well prior to the Robusta/Complexa split ( Stolarski et al. 2011 ). The similarity between the earliest diverging members of both the Scleractinia and Corallimorpharia in terms of both morphology and mt genome architecture ( fig. 2 ) implies that Corallimorphus occupies a key position in the corallimorpharian <-> scleractinian transition. Corallimorphus therefore diverged either close to the point of the scleractinian/corallimorpharian divergence (under scleractinian monophyly) or at the point of skeleton loss (under the “naked coral” scenario). If we accept that the organization of the mt genome in Corallimorphus most closely reflects the ancestral pattern ( figs. 1 and 4 ), then extensive reorganizations are required to generate the consensus corallimorpharian architecture (CII in fig. 2 ) and that seen in Corynactis ; in contrast, the rearrangements documented to date within Scleractinia require far fewer steps. In the case of Lophelia , the presence of a 67 bp direct repeat comprising the 3′-end of the ND1 and 5′-end of COB genes ( Emblem et al. 2011 ) implies that the likely mechanism of reorganization was tandem duplication and random loss ( Moritz et al. 1987 ; Zhang 2003 ), which may also account for the COII – COIII inversion seen in Madrepora ( Lin et al. 2012 ). We were unable to identify signatures of duplication-mediated rearrangement in corallimorpharians; however, neither are there obvious examples of inversion of segments of the mt genome in this Order. Rather, extensive segmental reorganization without inversion has occurred within Corallimorpharia, possibly facilitated by the less compact nature of the mt genomes (reviewed in Boore and Brown 1998 ). This contrasts markedly with the situation in octocorals, where many successive inversion events explain the observed diversity of mt gene organization ( Brockman and McFadden 2012 ).\n F ig . 4.— Hypothetical scheme for the evolution of mt genome architecture in the Scleractinia and Corallimorpharia. The scheme is based on the phylogenetic tree shown as figure 5 in Kitahara et al. (2014) , with patterns of gene organization (numbered as in fig. 2 ) indicated in green boxes. Can comparisons of mt genome organization resolve the question of coral monophyly? Although the data presented here are consistent with monophyly of the Scleractinia, they do not exclude the possibility of an origin for corallimorpharians within the coral clade. Phylogenetic analyses based on gene order ( fig. 3 A and B ) were ambiguous. Although both AA- and nt-based molecular phylogenetic analyses unambiguously support monophyly of the Corallimorpharia, the gene order analysis ( fig. 3 A and B ) did not. We interpret the grouping of Lophelia and Corallimorphus in this analysis as an artifact resulting from superficial similarities in gene organization in these two organisms; although gene order is similar, the sequences of those genes are highly divergent. The idea that the grouping of L. pertusa with C. profundus is artifactual is supported by the relatively low DCJ and BPD confidence values (58/49) associated with this node (i.e., well below the 85% confidence interval recommended by Shi et al. 2010 ). When L. pertusa was removed from the analysis, the overall DCJ and BPD statistic performances at the nodes of Corallimorpharia and Scleractinia increased, particularly for the node of C. profundus and Scleractinia/ M. oculata , where support increased from 94/75 to 97/82 ( fig. 3 ). The mt genomes of the Robusta differ from both corallimorpharians and all other corals in several characteristics. First, within the larger Scleractinia/Corallimorpharia clade, the Robusta have the most compact mt genomes (size range 14,853–17,422 bp) as a consequence of having in general shorter intergenic regions and the largest number of overlapping gene pairs (three to six cases of overlaps). In contrast, corallimorpharians have the largest mt genomes (size range 20,092–22,015 bp), longer intergenic regions, and no cases of overlapping genes, with complex corals intermediate in these characteristics (genome sizes 17,887–19,387 bp; 0–2 overlapping gene pairs—most frequently a single case of overlapping genes). Second, the Robusta differ in structural comparisons of the ND5 group I intron ( Emblem et al. 2011 ) as well as in molecular phylogenetics based on this feature. A group I intron interrupts the ND5 gene of all hexacorallians examined to date; these introns typically come and go during evolution but that in hexacorallians contains a variable number of genes and has become an essential feature. The hexacorallian ND5 intron has been “captured” in the sense that it is now dependent on host-derived factors for splicing, as indicated by the substitution of the ωG (the last nucleotide of the intron) by ωA (reviewed in Nielsen and Johansen 2009 ; Emblem et al. 2011 ). Although these characteristics are common across the coral-corallimorpharian clade, the ND5 introns of robust corals have a more compact core and overlapping intron and ND5 -coding sequences ( Emblem et al. 2011 ). In some robust corals, ωA is replaced by ωC, indicating a higher level of dependency on host factors for processing and thus greater integration of intron and host. These qualitative factors, as well as molecular phylogenetics of the ND5 intron sequences, are most parsimoniously accommodated by scleractinian monophyly ( Emblem et al. 2011 ). Third, of the three lineages, the mt genomes of Robusta have the highest (A+T) content and most constrained codon usage, one obvious consequence of which is that phenylalanine is overrepresented in the proteins that they encode, suggesting that mt DNA repair may be reduced in the Robusta ( Kitahara et al. 2014 ). The features outlined above, in which the Robusta differ from complex corals and corallimorphs, are derived characteristics—they serve to resolve the robust corals but do not unambiguously identify the sister group. Scleractinian monophyly explains all of the data most parsimoniously, but the alternative cannot yet be ruled out. The mt genome has been exhaustively mined for answers, but these must likely wait for the availability of appropriate nuclear markers."
} | 3,804 |
28695213 | PMC5498105 | pmc | 8,547 | {
"abstract": "Extracellular polymeric substances play important roles in microbial extracellular electron transfer processes.",
"conclusion": "CONCLUSION Because of the EPS’ strong redox ability and 3D structure, electron transport between microbial cells and extracellular electron acceptors/donors, whether via pili/nanowires, membrane-bound proteins, or electron shuttles, must therefore take the effect of EPS into account. Considering the spatial distances, membrane-bound cytochromes are not able to directly transfer electrons to extracellular electron acceptors when a thick layer of EPS covers the cell surface. It is also important to understand the influence of the redox reactions of EPS in conductive nanowires for EET and if some electrons are transported from the surface of conductive nanowires into the EPS (not extracellular electron acceptors) via redox reactions. Results from the three strains show that EPS attenuates direct EET, although the strains remain electrochemically active ( Fig. 3 , A and B, fig. S5, and table S2). When EPS-depleted cells are exposed to harsh conditions, proteins on the cell surface are easily inactivated, or the cells even killed and EET lost. EPS help cells to attach on solid minerals or other electron acceptors/donors and shorten the gap between microorganisms and solid surfaces. This not only facilitates EET but also saves the energy used to reach the acceptors/donors ( 36 ). We suggest that these strains achieve a successful balance between EET, self-protection, and other natural EPS functions. Indirect EET is another major microbial EET pathway ( 8 ) that may benefit from the presence of EPS, because EPS help microorganisms maintain a high concentration of electron shuttles in the gap between cells and electron acceptors/donors. Unbound flavins as electron mediators have to diffuse within the EPS layer, which shows different fluid mechanics compared with those in bulk solution. Moreover, it is not possible to detect the bound flavin-cytochrome interaction until this EPS layer is removed. In brief, the current work shows that EPS store electrochemically active substances (such as flavins and c-type cytochromes), which act as electron transit media. This enables EPS-enveloped cells to transport extracellularly electrons to acceptors or from donors by EET, where electron hopping is the most likely molecular mechanism for electrochemical ET through EPS.",
"introduction": "INTRODUCTION Electron transfer (ET) is one of the most fundamental life processes, and microorganisms exploit extracellular electron transfer (EET) to exchange information and energy with other microorganisms or with their external environments ( 1 – 6 ). Two mechanistically overarching EET pathways have been proposed ( 2 , 7 – 10 ) ( Fig. 1A ): (i) direct “band-like” electron conductivity, where electrons are transported via extended conductive pili/nanowires or redox proteins located in the microbial outer membrane; and (ii) indirect EET, where ET is mediated by mobile or spatially fixed molecular redox shuttles. However, neither of these pathways has been reported to involve extracellular polymeric substances (EPS). Fig. 1 Representation of microbial EET mechanisms when a microorganism is working as an electron donor. ( A ) A view of previous studies with proposed direct and indirect microbial EET mechanisms. ( B ) Can EET processes be affected when EPS cover the cell surface? ETMs, electron transfer mediators. EPS are fundamental microbial components that determine the physiochemical properties of biofilms ( 11 ). Almost every microbial cell is surrounded by EPS, which assist in biofilm formation and protection from unfavorable environment ( 11 – 13 ). Atomic force microscopy (AFM) shows, for example, that Azospirillum brasilense forms a 2-nm layer of adsorbed EPS on polystyrene substrata after 24 hours of growth ( 14 ), whereas cryo-transmission electron microscopy has shown that Shewanella oneidensis MR-1 (MR-1) in biofilm is enveloped by an EPS layer up to 1 μm after a 110-day cultivation ( 15 ). EPS contribute to biogeochemical cycles of organic matter and nutrient elements. Many components of EPS, such as DNA, humic acids, and some proteins, are redox-active or conductive/semiconductive ( 16 ). Some studies have shown that EPS are able to synthesize metal nanoparticles in the presence of c-type cytochromes ( 17 ) and have redox properties similar to MR-1 ( 18 ). However, it is poorly understood whether EPS are directly involved in EET processes, because previous microbial EET studies were only conducted on biofilms with EPS retained or on cells from the log stage or early steady stage cultures with little EPS. Note that electron transport between the cells and extracellular electron acceptors/donors must pass through the layer of EPS, but it is not known whether these electrons would be captured or whether the ET processes are speeded up by the EPS during the process. Hence, it becomes essential to clarify the EPS function in EET via the direct or indirect pathway ( Fig. 1B ). With a view on possible mapping of EPS-based EET mechanisms ( 16 ), we address here the electrochemical behavior of EPS using three electrochemically active strains: Gram-negative MR-1 ( 8 ), Gram-positive Bacillus sp. WS-XY1 (WS-XY1) ( 19 ), and yeast Pichia stipitis ( Ps ) ( 19 ). Electrochemical analysis of both EPS and whole cells in the presence and absence of EPS was undertaken, followed by morphological and chemical investigations. The results suggest strongly that EPS are crucial in bacterial and yeast EET."
} | 1,389 |
33603148 | PMC8245636 | pmc | 8,548 | {
"abstract": "Pathogens continue to emerge from increased contact with novel host species. Whilst these hosts can represent distinct environments for pathogens, the impacts of host genetic background on how a pathogen evolves post-emergence are unclear. In a novel interaction, we experimentally evolved a pathogen ( Staphylococcus aureus ) in populations of wild nematodes ( Caenorhabditis elegans ) to test whether host genotype and genetic diversity affect pathogen evolution. After ten rounds of selection, we found that pathogen virulence evolved to vary across host genotypes, with differences in host metal ion acquisition detected as a possible driver of increased host exploitation. Diverse host populations selected for the highest levels of pathogen virulence, but infectivity was constrained, unlike in host monocultures. We hypothesise that population heterogeneity might pool together individuals that contribute disproportionately to the spread of infection or to enhanced virulence. The genomes of evolved populations were sequenced, and it was revealed that pathogens selected in distantly-related host genotypes diverged more than those in closely-related host genotypes. S. aureus nevertheless maintained a broad host range. Our study provides unique empirical insight into the evolutionary dynamics that could occur in other novel infections of wildlife and humans.",
"introduction": "Introduction Emerging infectious diseases have led to serious declines in wildlife populations. Heavy population losses have been documented in rabbits following the introduction of Myxoma virus [ 1 ], amphibians from Chytrid fungus [ 2 , 3 ], Tasmanian devils from Devil Facial Tumour Disease [ 4 ] and brown bats from Pseudogymnoascus destructans fungus [ 5 ]. Emergence events can result from reservoir host spill-overs [ 6 ], jumps between host species [ 1 ], the evolution of a new pathogen trait that allows for exploitation of a new host [ 7 ] or by invading a new environment [ 8 , 9 ]. These novel interactions can be initially harmful [ 10 ] or entirely avirulent [ 11 ] to the host, and pathogen virulence and replication rates can evolve [ 12 ]. Pathogen adaptation can play a role in emergence [ 13 , 14 ]. Thus, the need to understand novel pathogen evolution in its new host population is central to predicting and managing the consequences. It has been shown that pathogen evolution can be shaped by many factors in established host–pathogen relationships, such as host genetic diversity [ 15 ], spatial structure [ 16 ] and gene flow [ 17 – 22 ]. Of particular interest is the role of host genotype and genetic diversity. Although genetically homogeneous populations are generally rare in the wild, many conservation efforts focus on declining, isolated and island populations often with low genetic diversity [ 4 , 23 ]. There are also growing numbers of homogeneous populations being managed for agriculture [ 24 , 25 ]. Are these populations hotbeds for increasingly damaging, emerging infections? Host genotype and diversity can affect pathogen evolution in many ways. Individual host genotypes can vary in their immune-based resistance to infection [ 26 , 27 ] but also in other aspects that might impact upon infection success, such as pathogen avoidance [ 28 , 29 ] and starvation responses [ 30 ]. Specialisation, or a narrow variance in infectiousness across different hosts [ 31 ], can evolve when pathogens infect homogeneous host populations of single genotypes [ 21 , 32 ]. An explanation is that pathogen populations acquire mutations that are neutral [ 21 ] or beneficial [ 33 ] in the focal host environment but costly in others. The ability for pathogens to specialise suggests that host genotypes can represent distinct selective environments. At the population level, high host genetic diversity can increase the odds of pathogens encountering resistant genotypes. This outcome has been shown to limit disease spread [ 23 , 34 – 39 ], virulence evolution [ 20 ], evolutionary rates [ 40 ] and also impede parasite adaptation [ 41 ] resulting in host range expansion [ 21 , 22 , 42 ] in established interactions. It is commonly assumed that genetically diverse host populations contain individuals that have protective immunity [ 36 ] (also in some theoretical models predicting pathogen emergence [ 39 ]). However, when pathogens are newly-introduced, most individuals could be susceptible [ 1 , 4 ]. Resistance may not spread until well after emergence when some pathogen evolution will have already taken place [ 12 ]. Yates et al. [ 14 ] found that if pathogen adaptation takes place during emergence, diverse host populations can have a small positive effect given they can contain individuals contributing greatly to transmission (i.e. super-spreaders). It thus remains unclear whether host genetic diversity has an impact on novel pathogen evolution early on in the association when most hosts are susceptible. In this study, we investigated whether host genotype and population-level genetic diversity drive evolutionary changes in the virulence, infectivity and host range of a newly-introduced pathogen. We passaged pathogenic Staphylococcus aureus between populations of a novel host species—the nematode Caenorhabditis elegans —which varied in genotype (24 natural isolates in monoculture) and diversity (polyculture of all 24 isolates). The nematode hosts used were randomly selected across a wild isolate C. elegans phylogeny [ 43 ] to represent a diverse spectrum of genetic backgrounds (Table S1 ). C. elegans are likely exposed to Staphylococcus spp. in natural environments, such as compost [ 44 ] and on button mushrooms [ 45 ] where these nematodes are vectors of bacteria causing blotch disease [ 46 ]. However, S. aureus per se has not been found to date to be a natural, established pathogen of C. elegans . Pathogenic S. aureus strains are known to jump regularly between host species, including a range of domestic animals, wild hosts, such as rodents, non-human primates and bats [ 47 – 49 ]. After a period of evolution, we compared the trajectories of each evolved population across sympatric host genotypes, as well as between host monocultures and polycultures. We measured changes in pathogen-induced host mortality and infection load as these traits relate to S . aureus virulence and infectivity [ 50 , 51 ], respectively. We additionally compared the evolutionary trajectory of host range across treatments by evaluating these traits of evolved populations on a novel host genotype. Our main predictions were that (a) pathogens selected in a host genotype would increase their virulence and infectivity on that genotype, (b) pathogens selected in diverse host polycultures would be more constrained in their evolution of virulence and infectivity and (c) pathogens selected in a host genotype would specialise and show reduced virulence/infectivity on novel hosts. The molecular basis of the adaptive process, and its relationship to host genetic distance, were also explored.\n\nInfection at point of pathogen introduction Previous studies have found that wild C. elegans genotypes differed in susceptibility to infections of Bacillus thruringiensis [ 27 ] and Serratia marcescens [ 20 , 63 ] implying a shared coevolutionary history or significant standing genetic variation for resistance [ 64 ]. Although C. elegans may encounter Staphylococcus spp. on the microbiota of vegetation [ 45 ], human-acquired S. aureus is unlikely to naturally co-occur in the same environment. This likely novelty of the interaction may explain the lack of significant host genetic variation in resistance, as measured by pathogen-induced host mortality (Fig. 1 , Table 1 ) and infection load (Fig. S2 , Table 1 ) across 24 host genotypes at the beginning of the experiment. This pattern held when accounting for the genetic distance between host genotypes ( p > 0.05). It is possible there is host genetic variation for other aspects of fitness during S. aureus infection (e.g. host offspring number, population growth) not measured herein. Infection load was also not associated with pathogen virulence across host genotypes (Kendall’s rank correlation, z = 0.036, tau = 0.002, p = 0.97). Fig. 1 Phylogeny of the 24 C. elegans wild genotypes used in the experiment mapped against the virulent impacts of ancestral and their sympatric, evolved pathogen populations (measured by mean % host mortality caused by infection ± SE). Host mortality for each host isolate was measured across five replicates. The tree is rooted by the most genetically disparate host isolate (QX1121), see Andersen et al. [ 43 ] for full description. Table 1 Extent of variation in pathogen killing ability across host genotypes upon introduction and after experimental evolution. d.f. Test statistic p Host mortality Ancestor 23 X 2 = 28.738 0.19 Evolved 23 X 2 = 39.875 0.016 Infection load Ancestor 23 F = 0.38 0.99 Evolved 23 F = 0.56 0.94 Statistically significant p -values are in bold.",
"discussion": "Discussion The dynamics and outcomes of pathogen evolution are understudied in most emerging disease systems [ 69 ]. Here, we directly tested the selective impact of host genotype and genetic diversity on pathogen virulence and infectivity by tracking their evolution in a novel nematode-bacteria interaction. Throughout the experiment, we found that virulence and infectivity predominately increased, but evolved independently across the range of host genetic backgrounds. More specifically, host killing ability and infection load were not significantly associated at the point of emergence, after evolution, or within host genotypes or diverse host populations. Contrasting evolutionary trajectories between pathogen traits have been found after the initial spread of infection in wild, emerging disease systems [ 12 ] and weak associations between traits are recorded in human infectious diseases [ 70 ]. This larger pattern across the study suggests that different sources or strengths of selection acted on pathogen populations infecting host genotypes and diverse pools of nematodes. We found that higher pathogen virulence evolved, on average, in both single-host genotype monocultures and diverse host polycultures. Our selection regime, whereby pathogens were passaged in a series of naïve hosts still living at the time of pathogen extraction, can account for some of the increase. In many serial passage experiments, the costs of virulence and rapid replication are removed favouring a rise in host harm [ 71 , 72 ]. However, significant variation in virulence arose among host genotypes in monoculture. Host metal ion binding was a common distinction in the gene functions of host genotypes that selected for disparate levels of pathogen virulence. Otherwise known as ‘nutritional immunity’ [ 66 ], hosts can retain metal ions as a strategy for preventing infection. Whilst this might hamper bacterial growth [ 66 , 73 ], fewer metals in the host environment might impose selection on pathogens to increase host metal exploitation and virulence. We did not find higher siderophore production (iron-scavenging molecules) in vitro in the more virulent pathogen populations, but S. aureus can extract host metals, including iron, via other molecular mechanisms [ 74 , 75 ]. Metal ion binding is essential to staphylococci as a component of numerous biochemical processes, including energy production, DNA synthesis, and defence against oxidative stress [ 66 ]. It is also important in the regulation of virulence factors, whereby staphylococcal cells that successfully sequester metal ions from a host exhibit greater pathogenicity [ 66 ]. The evolution of higher virulence within genetically diverse host populations, beyond that observed in host monocultures, is contrary to previous findings that host heterogeneity reduces pathogen virulence [ 20 ]. We hypothesise that pooling 24 host genotypes concentrated hosts able to retain/give up metals—and selected for higher virulence—and/or hosts better able to transmit pathogens. This latter scenario of super-spreading, suggested by Yates et al. [ 14 ] as a potential consequence of host population diversity, could allow for more local transmission within each passage in viscous media. Theory predicts that the virulence of emerging pathogens is predicted to be larger when transmission is high as the pathogen can spread more easily amongst susceptible hosts [ 76 ]. This positive relationship between transmission and virulence has been shown to favour increased virulence in novel infections in nature [ 77 ]. Several studies have shown that host population diversity can limit parasite success [ 34 , 78 , 79 ], even that of emerging pathogens [ 34 , 80 , 81 ], whilst host homogeneity should confer susceptibility. Although this prediction was not realised in the patterns of virulence evolution in this study, we found that pathogen evolution in host monocultures yielded higher infection loads. There was an overall escalation of infectivity within sympatric host genotypes. We conversely found that infection load did not evolve to increase within genetically diverse host populations. Despite the limitation on evolution, the infection loads in these populations were initially higher, on average, than those in host monocultures. This outcome may possibly result from the ecological consequences of population diversity. If resource competition increases with the relatedness in host populations, as predicted by the niche partitioning theory [ 82 ], C. elegans stress pathways could be up-regulated and strengthen their defence mechanisms [ 83 ] in monocultures compared to diverse populations. Alternatively, as stated earlier, we also hypothesise that diverse populations might have experienced more within-culture pathogen transmission. Whether this increased transmission might facilitate increased infectivity and virulence by this pathogen remains to be tested. Aspects of population variation beyond immune-based resistance could be vital to pathogen colonisation. We observed that pathogen populations diverged at the genomic level, and more so following selection in distantly-related host genotypes. This result suggests that host genetic distance corresponds with a difference in selection environments, an assumption sometimes made in predicting the success of host species jumps [ 84 , 85 ]. In the context of host species jumps [ 84 ], host phylogenetic relatedness can determine infection success upon a host switch [ 84 ]—as with primate lentivirus [ 86 ], rabies virus [ 87 ] and Drosophila RNA viruses [ 85 ]—although closely-related species may both be susceptible despite the distance due to the loss or gain of immune functions associated with the pathogen [ 84 , 88 ]. Despite genomic divergence, pathogen populations selected in monocultures of host genotypes did not evolve a contracted host range compared to those in diverse host polycultures. Perhaps the nematode genotypes used were not sufficiently different to favour specialisation, and resistance must spread before specialisation can occur. Or the experimental period was not long enough. However, it has been predicted that a narrower host range should arise in pathogens infecting homogeneous host populations due to faster fixing of beneficial alleles, with slower evolution occurring in diverse host populations [ 40 ]. We found that the host population types drove the same speed of pathogen genomic evolution and SNPs remained at relatively low frequencies in all pathogen populations. Moreover, Gibson et al. [ 21 ] previously found that after 20 nematode host generations and the existence of substantial genetic variation in host resistance, the established pathogen Serratia marcescens did not consistently evolve specialism. The maintenance of a broad host genotype range in our study reflects the generalist strategies of other emerging infections [ 24 , 89 , 90 ]. Generalist pathogens are more likely to emerge than specialist pathogens [ 31 , 91 ]. Through high mutation rates, generalist pathogens can produce diverse genetic variants [ 91 , 92 ] helping them avoid host immune responses and limiting specialised host resistance. Not evolving to specialise on a host genotype can also be explained by the natural history of S. aureus , a prevalent host-shifter [ 93 , 94 ]. This bacterium has a large host range [ 47 , 95 – 98 ] and has emerged in domesticated bovine [ 95 ], poultry [ 99 ], and rabbits [ 97 ], as well as invertebrate species, including C. elegans [ 29 , 100 ]. This study reveals that any changes in pathogen traits and genomes following introduction can be variable across host genotypes and levels of population genetic diversity. As anthropogenic alterations to habitats and geographic ranges increase opportunities for contact between novel pathogens and naive hosts, we should expect more infectious disease emergence [ 37 ]. Increased knowledge on the evolution of novel infectious disease can provide insight on managing the harm caused by pathogens in human medicine, wildlife conservation and agriculture—and on mitigating their spread [ 101 ]. Our results suggest that host differences in metal-sequestration, as well as the ecological implications of host population diversity, warrant further consideration as drivers of infection outcomes over evolutionary time in recent wildlife and human pathogens."
} | 4,373 |
32323059 | PMC7176808 | pmc | 8,551 | {
"abstract": "In this manuscript, recent progress in the area of resistive random access memory (RRAM) technology which is considered one of the most standout emerging memory technologies owing to its high speed, low cost, enhanced storage density, potential applications in various fields, and excellent scalability is comprehensively reviewed. First, a brief overview of the field of emerging memory technologies is provided. The material properties, resistance switching mechanism, and electrical characteristics of RRAM are discussed. Also, various issues such as endurance, retention, uniformity, and the effect of operating temperature and random telegraph noise (RTN) are elaborated. A discussion on multilevel cell (MLC) storage capability of RRAM, which is attractive for achieving increased storage density and low cost is presented. Different operation schemes to achieve reliable MLC operation along with their physical mechanisms have been provided. In addition, an elaborate description of switching methodologies and current voltage relationships for various popular RRAM models is covered in this work. The prospective applications of RRAM to various fields such as security, neuromorphic computing, and non-volatile logic systems are addressed briefly. The present review article concludes with the discussion on the challenges and future prospects of the RRAM.",
"conclusion": "Conclusion This review article provides a brief introduction into the advancement of the memory architecture, the current trends and the limitations while providing a valuable insight into the field of emerging memory technologies. A detailed discussion, highlighting the importance of RRAM, its structure, working mechanism, and classification, has been presented. The key performance parameters and their effect on the RRAM operation has also been detailed within the current manuscript. An elaborate study on the MLC capability of RRAM, along with the methodology have been presented. The manuscript also discusses the important features of the widely accepted RRAM models. The implementation of RRAM for various important applications such as non-volatile logic, neuromorphic computing, security, and non-volatile SRAM have been highlighted. Although, significant success has been achieved in RRAM technology; however, more work is needed as RRAM still suffers from various challenges in terms in terms of high operation current, lower resistance ratios, and reliability issues. More efforts in research should aim to develop methods to achieve faster programming/erasing, lower power consumption, enhancing the storage density by implementing multilevel storage capability and improvement in the fabrication process for enhanced uniformity. In addition, renewed focus should be towards use of RRAM in embedded memory and non-volatile logic applications as breakthroughs in these fields are much more exciting and significant. With continued work and improvements, it is imperative that RRAM devices will be a standout technology for future non-volatile memory applications.",
"introduction": "Introduction Random access memory referred to as RAM can either be volatile or non-volatile. A volatile memory loses its previous stored data on removing the power supply as is the case for dynamic random-access memory (DRAM) and static random-access memory (SRAM). For non-volatile memory, the contents that were stored previously will continue to be retained even after the removal of the supply. Flash memory is a typical example of non-volatile memory. Memory technologies combine the advantages and disadvantages to achieve higher performance, e.g. DRAMs employed in a computer system has high capacity and density, but they are volatile, meaning there is a need to refresh every few milliseconds. Due to this refreshing, the energy consumption of the device increases which is not desirable. SRAM, on the other hand, is fast but it is also volatile just like the DRAM; in addition, SRAM cells are of larger size which hinders its implementation on a large scale. Flash memory, which essentially consists of a metal-oxide-semiconductor field-effect-transistor (MOSFET) in addition to a floating gate in each memory cell, is currently being used extensively particularly for the embedded applications owing to its low cost and high density. Depending upon how memory cells are organized, Flash memory is classified as NOR Flash and NAND Flash [ 1 ]. In NOR Flash, cells are read and programmed individually as they are connected in parallel to bit lines. This resembles the parallel connection of transistors in a CMOS NOR gate architecture. For the case of NAND Flash, the architecture resembles that of a CMOS NAND gate as the cells are connected in series to the bit lines. It must be noted that less space is consumed by the series connection as compared to the parallel one which results in a reduced cost of NAND Flash. However, both types of Flash memories suffer from several disadvantages such as low operation speed (write/erase time: 1 ms/0.1 ms), limited endurance (10 6 write/erase cycles), and high write voltage (> 10 V) [ 2 ]. The memory technologies mentioned above, i.e. DRAM, SRAM, and Flash, are charge storage-based memories. DRAM stores the information in the form of charge at the capacitor, and SRAM is based on the storage of charge at the nodes of the cross-coupled inverters, whereas the Flash memory technology uses the floating gate of the transistor to store the charge. All these existing charge storage-based memory technologies are currently facing challenges to scale down to 10 nm node or beyond. This is attributed to the loss of stored charge at nanoscale, which results in the degradation of the performance, reliability, and noise margin. In addition, requirements of large refresh dynamic power for DRAM and leakage power for both SRAM and DRAM pose serious challenges for the design of future memory hierarchy. Therefore, a new class of memories usually referred to as emerging memory technologies are currently undergoing development and are being actively researched primarily in the industry with the aim to revolutionize the existing memory hierarchy [ 3 ]. These emerging memory technologies aim to integrate the switching speed of SRAM, storage density comparable to that of DRAM, and the non-volatility of Flash memory, thus become very attractive alternatives for future memory hierarchy. To classify a memory device as an ideal one, it should have the following characteristics: low operating voltage (<1 V), long cycling endurance (>10 17 cycles), enhanced data retention time (>10 years), low energy consumption (fJ/bit), and superior scalability (<10 nm) [ 4 ]. However, no single memory to date that satisfies these ideal characteristics. Various emerging memory technologies are actively being investigated to meet a part of these ideal memory characteristics. These memory technologies that depend upon the change of resistance rather than charge to store the information are as follows: (i) phase change memory (PCM), (ii) spin-transfer torque magnetoresistive random access memory (STT-MRAM), and (iii) resistive random access memory (RRAM). In phase change memory, the switching medium consists of a chalcogenide material (commonly Ge 2 -Sb 2 -Te 5 , GST) [ 5 – 7 ]. PCM relies on the difference in resistance between the crystalline phase and amorphous phase for efficient data storage capability. The crystalline phase denotes the low resistance state (LRS) or ON state of the device whereas the amorphous phase denotes the high resistance state (HRS) or OFF state. The SET operation corresponds to LRS generally referred to storing logic value ‘1’, whereas the RESET operation correspond to HRS storing logic value ‘0’ in the device. For SET operation, PCM is heated above its crystallization temperature on the application of voltage pulse, while for RESET operation, a larger electrical current is passed through the cell and then abruptly cut-off so as to melt and then quench the material in order to achieve the amorphous state. In spin-transfer torque magnetoresistive random access memory, the storage capability is due to the magnetic tunneling junction (MJT) [ 8 – 10 ], which consists of two ferromagnetic layers and a tunneling dielectric sandwiched between them. The magnetic direction of the reference layer is fixed, while the application of external electromagnetic field can change the magnetic direction of the free ferromagnetic layer. If the reference layer and the free layer have the same direction of magnetization, the MTJ is referred to be in the LRS. For MTJ, to be in the HRS, the direction of the magnetization of two ferromagnetic layers is anti-parallel. RRAM consists of an insulating layer (I) sandwiched between the two metal (M) electrodes [ 11 , 12 ]. RRAM relies on the formation and the rupture of conductive filaments corresponding to LRS and HRS, respectively, in the insulator between two electrodes [ 13 – 15 ]. A detailed comparison of existing and emerging memory technologies is shown in Table 1 . As is evident from the table, STT-MRAM and PCM have advantages of a smaller area compared to that of SRAM. While STT-MRAM offers fast write/read speed, long endurance, and low programming voltage, on the other hand, PCM has a disadvantage of extensive write latency. RRAM has a lower programming voltage and faster write/read speed compared to Flash and is seen as potential replacement of Flash memory. Among all the emerging memory technology candidates, RRAM has significant advantages such as easy fabrication, simple metal-insulator-metal (MIM) structure, excellent scalability, nanosecond speed, long data retention, and compatibility with the current CMOS technology, thus offering a competitive solution to future digital memory [ 16 ]. The most significant advantages of RRAM are depicted in Fig. 1 .\n Fig. 1 Advantages of RRAM Table 1 Comparison of emerging memory technologies Memory technology SRAM DRAM NAND Flash NOR Flash PCM STT-MRAM RRAM Cell area > 100 F 2 6 F 2 <4 F 2 (3D) 10 F 2 4– 20 F 2 6– 20 F 2 <4 F 2 (3D) Cell element 6T 1T1C 1T 1T 1T(D)1R 1(2)T1R 1T(D)1R Voltage <1 V <1 V <10 V <10 V <3 V <2 V < 3 V Read time ∼1 ns ∼10 ns ∼10 μ s ∼50 ns <10 ns <10 ns < 10 ns Write time ∼1 ns ∼10 ns 100 μ s–1 ms 10 μ s–1 ms ∼50 ns <5 ns < 10 ns Write energy (J/bit) ∼fJ ∼10 fJ ∼10 fJ 100 pJ ∼10 pJ ∼0.1 pJ ∼0.1 pJ Retention N/A ∼64 ms >10 y >10 y >10 y >10 y > 10 y Endurance > 10 16 >10 16 >10 4 >10 5 >10 9 >10 15 ∼10 6 – 10 12 Multibit capacity No No Yes Yes Yes Yes Yes Non-volatility No No Yes Yes Yes Yes Yes Scalability Yes Yes Yes Yes Yes Yes Yes F: Feature size of lithography In this work, recent progress and a detailed overview of RRAM technology are presented. A review of switching materials together with the classification of switching modes and details of the switching mechanism is discussed in the “ Resistive Random Access Memory (RRAM) ” section. The “ Performance Metrics of Resistive Random Access Memory (RRAM) ” section highlights various performance metrics of RRAM. Multilevel cell (MLC) characteristics of RRAM along with various MLC operation schemes and their physical mechanisms are analyzed in the “ Multilevel Resistive Random Access Memory (RRAM) ” section. A detailed discussion on modeling of RRAM device is presented in “ Modeling of RRAM Devices ” section. In “ Applications of RRAM ” section various applications of RRAM are discussed. Finally, challenges and future outlook of RRAM is presented in “ Challenges and Future Outlook ” section. The category wise distribution of papers consulted in the preparation of this review manuscript are presented in Fig. 2 .\n Fig. 2 Category wise distribution of papers consulted for preparation of review on RRAM"
} | 2,940 |
28839216 | PMC5571151 | pmc | 8,552 | {
"abstract": "A synaptic memristor based on IGZO and oxygen-deficient HfO 2 films has been demonstrated. The memristor exhibits a fatigue response to a monotonic stimulus of voltage pulses, which is analogous to the habituation behavior of biological memory. The occurrence of habituation is nearly simultaneous with the transition from short-term memory to long-term memory. The movement and redistribution of oxygen species with the assistance of polarization in HfO 2 layer are responsible for the above results. The observation of habituation behavior proves the potential prospect of memristor on the mimic of biological neuron.",
"introduction": "Introduction The human brain deals with information in parallel, which can easily recognize objects and visual information in complex environment 1 . Therefore, many efforts have been made to realize the neuromorphic computation. The synapse emulation is a key step to build neuromorphic systems that can mimic the human brain 2 . However, the traditional methods, such as the software-based method by conventional von Neumann computers or the hardware-based method by lots of resistor and capacitors in CMOS integrated circuits, occupy large areas and consume much more energy 3 . Nowadays, the realization of a single device with synaptic functions has attracted much attention for the implementation of the neuromorphic system. Among them, the electronic synapses based on memristors have been widely focused 4 – 6 . In 1971, Chua predicted the fourth basic circuit element, namely, memristor 7 . Subsequently, many studies demonstrated that a memristor can be used as an electronic synapse with its conductance representing the synaptic weight 8 – 13 . Although synaptic operation of memristors has been widely demonstrated, the biological properties of habituation/fatigue behavior were not reported for the memristor. An inorganic memristor is so different with a real organic synaptic due to the absence of biological activity, which leads to suspicion for the possibility of achieving a true bio synaptic device. Encouragingly, the paired-pulse facilitation (PPF) behavior has recently been reported 14 , which shows the resemblance between an inorganic memristor with a biological synapse as in response to electrical stimulation. The mechanism responsible for this PPF behavior was demonstrated to be the overlap effect of the two pulses on the memristor 14 . However, when exposed to continuous stimulus, does a memristor show a monotonous increase, that is similar to that of PPF (only a pairs of pulses) measurement? Meanwhile, it has been widely reported that the accumulation of electrical stimulus can make the transition from short-term memory (STM) to long-term memory (LTM) 15 , what is the correlation between this transition with the accumulation of electrical stimulus? In this work, synaptic memristors were fabricated based on the structure of the over-oxidized IGZO and oxygen-depleted HfO 2 (OD-HfO 2 ) films. The habituation in the case of continuous electrical pulse stimulation is observed, which occurs almost simultaneously with the transition from STM to LTM. This interesting behavior of synaptic memristor under continuous stimulation gives the potential application of the simulation of the biological synaptic with the inorganic memristor.",
"discussion": "Results and Discussion The switching layer of the synaptic device consists of two parts: the oxygen-rich IGZO and the oxygen-deficient HfO 2 layer. The two-terminal, bilayer IGZO–HfO 2 memristor is analogy to a biological synapse, and oxygen vacancies are similar to neurotransmitters, as shown in Fig. 1(a) . The device conductivity was treated as synaptic weight in this memristor. Similar to biological synapse, the synaptic weight can be dynamically modified and stored using consecutive spikes 16 . The release and restored back of oxygen vacancies under the stimulation of the pulses play a role similar to that of a neuro-transmitter in the modulation of the strength of the synaptic connection in a biological synapse 17 , 18 . The microstructure and composition of the IGZO films were characterized by SEM and XPS analyses. The films were in high quality and smooth morphology as observed from the AFM and SEM images as shown in Fig. 1 . Figure 1(c) shows the XPS spectra for the IGZO film under various pulses stimulations, which will be discussed further below. Figure 1 Basic characteristics of fabricated simples. ( a ) Schematic illustration of a biological synapse and the IGZO-HfO 2 -based synaptic memristor. ( b ) AFM morphology of the IGZO surface of a pristine sample. ( c ) XPS spectra for the devce under various pulse number stimulations. ( d ) SEM images of IGZO for the pristine sample. ( e ) SEM images enlarged for 50times. ( f ) The lateral profile for the pristine sample. \n As shown in Fig. 2(a) , two successive pulses with fixed intensity and width were applied to the memristor. The postsynaptic current triggered by the second pulse is greater than the first pulse, which is similar to the PPF behavior in the biological synapse. PPF is a plasticity of biological synapse in which spike-induced postsynaptic responses increase when the second spike closely follows the previous spike 19 .When the voltage is turned off, the postsynaptic current does not immediately disappear. In contrast, an attenuation of postsynaptic currents is observed during the off-cycle of the pulse. The attenuation of postsynaptic currents resembles memory loss in biological systems. As is known, PPF is associated with incomplete compensation of oxygen vacancies. After the first pulse, the oxygen vacant filament can be compensated by oxygen, and if the pulse interval is sufficiently small, the oxygen vacancies will not be fully compensated, and therefore the conductive channel will not completely disappear and, after the second pulse, the synaptic current will be greater than the previous one as shown in Fig. 2(a) . This kind of PPF behavior stimulated by a pair of pulses in synaptic memristor provides the possibility to study the post-synaptic currents as a function of the number of pulses. As shown in Fig. 2(b) , The device conductivity continuously increases as consecutive voltages are applied. An interesting phenomenon is that the current increase magnitude is gradually weakened as the number of pulses increases. In other words, although the stimulus still makes the follow-up current larger than the previous one, the increasing magnitude was gradually weakened, which is similar to the habituation of the brain to frequent boring stimuli. This observation of IGZO-HfO 2 device shown in Fig. 2(b) expands the facilitation behavior of memristor, since more than two pulses were applied. Figure 2(c) shows the current as a function of the pulse numbers. When the number of the pulses increases, the increasing magnitude of each current decrease gradually. After a certain number, the current nearly keeps constant. The current decay processes were also compared for the devices under different numbers of stimulation pulses. A phenomenon similar to the brain’s forgetting behavior was discovered as shown in Fig. 2(d) that the forgetting time (the memory decay time from beginning to the saturation) increasingly evolves from several seconds to tens of seconds with increasing number of stimulations, and the corresponding retention ratio increases from around 37% to around 61%, indicating a decreasing forgetting rate. As known, the memory behavior in psychology can be categorized into short-term memory (STM) and long-term memory (LTM) based on the retention time 15 . The transition from STM to LTM can be realized by repeating the pulse stimulus, which is analogous to the rehearsal of the biological brain. In this work, the memristor fatigue achieved through repeated stimulation shows a clear synchronization indication of the STM-to-LTM transition, since the current curves in fatigue region were almost unchanged as shown in Fig. 2(d) . Contrarily, the current has an obvious rapid-decay zone as the pulse stimulation does not make the memristor reach the “fatigue” region, as shown the 20 th and 40 th pulses in Fig. 2(d) . Clear divide for the current curves shown in Fig. 2(d) was observed, which is synchronized with the device habituation/fatigue under sufficient stimulation. STM and LTM are difficult to define from a biological point of view, and the biological definition of STM and LTM are defined usually as follows: The STM (LTM) is a temporary (permanent) potentiation of neural connections, and lasts for a few minutes or less (from hours to years). Besides, STM can be converted to LTM through repeated rehearsals, which involves a physical change in the structure of neurons. Here, one can find the similarity of this definition to our devices, since these temporal characteristics of memory retention were also achieved in the current device. For devices that did not reach the “fatigue region”, the current ratio of ~55% only lasted for less than 20 seconds, whereas for “fatigue” ones, a current ratio of ~55% could reach a time of not less than 70 s (longer time can also be achieved, as shown in Fig. 3(a) ). The 55% ratio was selected because that the current decays into fatigue saturation at approximately this value. The critical threshold is almost synchronized with the occurrence of fatigue behavior (approximately 60–65 pulses, as shown in Figs. 2(c) and Figure 3(a) ). Therefore, once the memristor reaches the “fatigue” zone, the current decay gently and slowly, and the devices were almost unchanged for the current decay characteristics (for example the 80 th , 100 th and 120 th pulses, as shown in Fig. 2(d) ), such invariance could be an indication of reaching the LTM state. Figure 2 ( a ) Synaptic currents of the memristor triggered by a pair of pluses. The pulse intensity, width, and interval are 1 V, 50 ms, and 1 s, respectively. ( b ) The post synaptic current of the memristor in response to the pulse train. As consecutive voltages are applied, on one hand the device conductivity continuously increases, on the other hand, the magnitude of this increase is gradually weakened. ( c ) The current as a function of the pulse numbers. ( d ) Current decays recorded after different numbers of pulse stimuli. In the fatigue region, the current decay rate is almost the same. the current decay curve was significantly different for different pulse counts for the device unreached fatigue due to insufficient pulse training. \n Figure 3 Influence of some parameters on fatigue effect. ( a ) Retention of current decay for the devices with different number of pulse stimulation after 72 h. ( b ) variation of relaxation time constant ( τ ) with the number of stimulation pulses, where the τ is obtained by fitting data from Fig. 2(d) . It is highly resemble with the Fig. 2(c) . ( c ) Current decay for the device stimulated by 120th pulses at different temperatures. Solid lines are fitted curves using Equation 1 . ( d ) Current evolution of pure IGZO device (without OD-HfO 2 layer) for comparison. Current was ceased during the off period of the pulses. \n To quantitatively study the STM behavior, the exponential function including decay term was usually used 20 , 21 , as described below to explain the relaxation process 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}$$R(t)={R}_{s}+({R}_{0}-{R}_{s})\\exp (-t/\\tau )$$\\end{document} R ( t ) = R s + ( R 0 − R s ) exp ( − t / τ ) \n R ( t ), R \n 0 , and R \n s are the retention level of current at time t , t = 0, and saturation state shown in Fig. 2(d) . τ defined as the relaxation time which can evaluate the forgetting rate of memristor. The current retention level ( R \n s ) of LTM should remain stable for a long time, defined as 72 h here as shown in Fig. 3(a) . The relaxation processes after different numbers of stimulation pulses were applied to the device as shown in Fig. 2(b) to Fig. 2(d) . The results indicate that the forgetting rate can decrease and the ratio of current retention can increase through repeated stimulation. In Fig. 3(b) , the derived relaxation time was plotted with pulse numbers. One can find the same saturation characteristic between this time parameter with that of current evolution in Fig. 2(c) , which confirms that the saturation of current retention (fatigue response) could be an indication of LTM. Fig. 3(c) presents the decay curves measured at different temperature. The relaxation process of STM was affected by the temperature. The relaxation time decreases as the temperature increases. According to previous literature 22 , the relaxation time of memory current may simplify to be a linear relationship on diffusion coefficient. On the other hand, considering the dynamic movement of oxygen vacancies (ions), the diffusion coefficient D shows the function of T in the follow: 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}$$D(T)\\propto \\exp \\,(-E/kT)$$\\end{document} D ( T ) ∝ exp ( − E / k T ) where E represents an activation energy and k is the Boltzmann constant. Therefore, the solution of the dynamic diffusion of oxygen vacancy (or charged ions) can be expressed with the equation has the same exponential form as the above decay equation (Equation 1 ) used to fit the relaxation process in Fig. 2(d) , suggesting that the dynamic movement of oxygen vacancies (or charged ions) may be responsible for the transition of STM to LTM. It is noted that the device structure used here, i.e., IGZO-HfO 2 , is substantially different with the reported ones (basically pure single layer) showing PPF behaviors 15 , 23 . It is also interesting that the post-synaptic current persisted upon the cease of the input voltage spikes here, and the current facilitation was more pronounced than previous reported PPF including current keep increasing with a train of pulses (more than two pulses). We thought the highly oxygen deficient HfO 2 layer plays an important role in these functions. Figure 3(d) shows the current evolution of pure IGZO device (the thickness of IGZO is in the same with the device used above) for comparison. Current was ceased during the off period of the pulses, which is consistent with previous reports. HfO 2 with different oxygen (oxygen vacancy) content has ever been shown the surprising properties such as p-type conductivity 24 , luminescence 25 , ferroelectricity 26 , 27 , intrinsic d 0 magnetism 28 , excellent buffer function 29 and so on. Actually, we have also published several works about the ferroelectric behavior of highly oxygen-deficient HfO 2 dielectric films (OD-HfO 2 ) recently 30 , 31 . Highly oxygen-deficient state and/or dopant in HfO 2 films are the primary condition responsible for the formation of the ferroelectrical polarization, since this condition easily leads to lattice distortion to form ferroelectricity 32 . The mechanism of the ferroelectricity of HfO 2 dielectric is still unclear since its first report in 2011, and many works are on it to date 33 , 34 . However, this ferroelectric fact makes some interesting results and shows promising prospect in this work, considering the mature compatibility of HfO 2 to semiconductor industry process. To the best of our knowledge, the reported electronic synapses were usually stimulated with external electrical stimulus; and few artificial synapses based on self-ferroelectric stimulus have been reported. This leads to the interesting phenomena that the post-synaptic current may be persisted upon the cease of the input voltage spikes. The oxygen ion migration can lead to the concentration differences of the oxygen distribution in IGZO layer, and another dynamic process, ferroelectric-induced ion diffusion was also non-negligible. Considering the migration of oxygen vacancy (equivalent to oxygen ions with negative charges, O 2− ), we can understand the memory mechanism as shown in Fig. 4 : i) The conductance of IGZO depends strongly on its oxygen content: the higher oxygen content, the lower the conductivity 15 . When a positive bias voltage is applied to the top electrode (oxygen-rich IGZO side), the electric field induced motion of oxygen ions compresses the highly resistive oxygen-rich IGZO layer, thus increasing the device conductivity, and the polarization also began to form in the OD-HfO 2 layer. The movement of oxygen vacancy induced by electric field (also with the assistance of ferroelectric polarization in OD-HfO 2 layer) can change the relative thicknesses of oxygen-deficient and oxygen-rich layers, thus modulating the device conductance. The above process was illustrated in Fig. 4(a) and (b) . ii) When the bias is removed, the oxygen vacancy (ions) was expected to be back restored. However, the spontaneous polarization of OD-HfO 2 layer restricts this process, therefore, only a partial retreat of the conduction front could be present, which reducing the device conductivity. Such a dynamic process corresponds to the current decay during cease of pulse. The above process was illustrated in Fig. 4(c) . iii) When a further pulse is applied after the former one, polarization still exists in the case of short pulse interval, thus the superposition of the two identical dynamic processes pushed the conduction front more forward, thus resulting in the increase of the current. This process was illustrated in Fig. 4(d) . High-rate stimulation, where the idle time between pulses is very short, can suppress the back-diffusion of oxygen vacancy. This may be a possible reason why frequent stimulation can prompt the transition of the STM to LTM, because in this process, the HfO 2 polarization was also stressed to be a more solidified state. The back restore dynamic makes a fraction of the oxygen ions recombine with oxygen vacancies (V o \n 2+ ). This process makes the compensation of local structural change which responsible for the STM. However, in our device, the STM may easily transformed to LTM state due to the existence of spontaneous polarization of OD-HfO 2 layer, and with its further stabilization for the pulse stimuli, which may lead to the observation of novel nonvolatile “training-memory” behavior in our device. Combination of two film is important that the highly oxygen deficient HfO 2 thin film provides the environment of the ferroelectric polarization response to external voltage, while the sufficient oxygen content in over-oxidized IGZO gives the sufficient negatively charged ions. The conductive path based on oxygen vacancy induces an increase of the post-synaptic current directly 35 . Figure 4 Diagram for the dynamic process for the devices. IGZO layer ( a ) in original state ( b ) during pulse, ( c ) during off period of pulse and ( d ) during further pulse. And HfO 2 layer ( e ) in original state ( f ) during pulse, ( g ) during off period of pulse and ( h ) during further pulse. \n In addition, our recent work 36 highly resonates with that in ref. 37 , i.e., the existence of locally accumulated clusters in HfO 2 film (based on oxygen vacancy, analogy to Ag clusters in Prof. Yang’s work) was confirmed using a series of XPS measurements. Therefore, another dynamic process may not be excluded which leads to the PPF-kind of current increasing in HfO 2 layer together with the mechanism shown in Fig. 4(e-f) . The clusters shrink back and stretch out in high resistance state (HRS) and low resistance state (LRS), respectively. Via connecting/rupture between the neighbor cluster sites, the switching of the resistance was formed. Ag nanoparticle and oxygen vacancy may have similar behavior to some extent in mimic synaptic influx and extrusion of Ca 2+ . When a voltage pulse is applied, the local temperature increases due to Joule heating and the potential is tilted by electric forces acting on oxygen vacancy (or charged oxygen ions) clusters with induced charge, both of which cause larger clusters to break up. As the clusters become more uniformly distributed in the active layer, the resistance drops, the current and temperature increase, and a positive feedback results in the formation of a conductive channel. As soon as the power is turned off, the temperature drops, and the oxygen vacancy start to recover back with the loss of the applied voltage. Eventually, most of the oxygen vacancies have recovered back into clusters, and the high-resistance state is re-established along with the original conductive path distribution almost restored, leading to volatility. The model predicts interesting conductance evolution similar to synaptic behavior when a train of pulses is applied. First, when the initial voltage pulse is applied, electric field-assisted diffusion pumps some of the oxygen vacancies out of the cluster and they start to bridge each other. If the time between pulses is shorter than the restore time of stretched oxygen vacancies, more oxygen vacancies are pushed to form conductive channel which leads to the gradual increase of device conductance, similar to the paired-pulse facilitation (PPF) phenomenon in bio-synapses. As for the reason that device conductance finally reaches a saturated conducting (fatigue region), the reason is, as the electric field pumps more and more oxygen vacancies to form connected conductive path, the number of vacancies in original clusters decreases (where the distribution peak decreases as more and more pulses arrive). As a result, the number of oxygen vacancies available decreases and the increasing of device conductance starts to decay. This dynamic process is directly related with the external field, i. e., the “E” illustrated in Fig. 4 , rather than internal ferroelectric characteristic of HfO 2 layer. And it is not responsible for the current presentation during the gap of the pulses. Therefore, it is not conflict with the diagram explanation. This process could exist at the same time, leading to the PPF-kind of current increasing in HfO 2 layer together. Generally, the intensity of a pulse is characterized by three parameters: amplitude, interval, and width. In order to investigated the impact of pulse intensity on the increasing magnitude of current, these three parameters were compared as shown in Fig. 5 . The ordinate Y (increasing ratio, IR) is defined as follow, 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}$${\\rm{IR}}=({{\\rm{I}}}_{{\\rm{n}}}-{{\\rm{I}}}_{n-1})/{{\\rm{I}}}_{{\\rm{1}}},$$\\end{document} IR = ( I n − I n - 1 ) / I 1 , where I n is the current for the n th cycles. To investigate the impact of each parameter on current increasing ratio, we change one of the parameters of the pulse and fix the remaining two parameters. As shown in Fig. 5 , The IR is increased by increasing the intensity of the pulse, i.e. by increasing the pulse width, decreasing the spacing, or increasing the width. However, as the pulse intensity increases, the IR trend is almost constant, i.e., keep decreasing and tend to be saturated. A stronger pulse allows the memristor resistance to be effectively reduced, resulting in a higher current. However, the “fatigue” effect of memristors based on the dynamic mechanism of oxygen vacancies is always present within limited pulse interval, which resembles the memory fatigue of the human brain in repetitive single stimuli. The time interval was varied between the two spikes, and the facilitation is decreased with the interval increase, as shown in Fig. 5 . the device conductance increases (PPF) from its initial conductance and interval leads to a reduced rate of increment when the time interval between pulses is long. No further curves shown is because the conductance of the device starts to show no facilitation with longer interval. It is reasonable that sequential pulses with a long enough interval, may form a conducting path first, but before the next pulse arrives the path breaks and the oxygen vacancies (or ions) are re-accumulated back to original position (IGZO) or clusters (HfO 2 ). They were previously reported for the cluster distribution (HfO 2 ) 36 and the line forward/back distribution (IGZO) 15 . Figure 5 The increasing ratio (IR) versus the cycle numbers for the memristor. ( a ) interval increasing, ( b ) width increasing and ( c )magnitude increasing. \n In addition, Oxygen ions migration was not only influenced by the electrical filed (drift), but also influenced by the concentration distribution (diffusion). With the charged oxygen ions movement, a barrier in IGZO near the interface region may be decreased/increased with the application/premovement of pulse voltage. The interfacial barrier evolution under the inertial polarization filed and the external voltage field, respectively. Both energy hight and physical width of interfacial barrier were present under these internal and external filed 31 , 38 . Figure 1(c) shows the results of the XPS analysis on O 1 s spectrum in a-IGZO film. the 80 th XPS spectra was different with the fresh and 40 th ones, though the latter spectra were basically consistent. This may a strong evidence for the STM to LTM transition. This indicated the fatigue state (80 th ) truly formed with altered the oxygen bonding and composition in the a-IGZO. Three distinct components of O 1 s peak were fitted by Gaussian Lorentzian deconvolution, which centered at 530.6, 531.5 and 532.6 eV, respectively 38 – 40 . The binding energy of the spectrum at 531.5 was associated with the oxygen deficient state within the a-IGZO film. The rest two binding energy (530.6 and 532.6 eV) were usually attributed to the presence of stoichiometric and loosely bound oxygen on the surface or interstitial pores of the a-IGZO film. One can find that the subpeak denoted the oxygen deficient state (531.5 eV) was very limited even in the fresh sample, which confirms that the IGZO layer was over oxidized as expectation. Oxygen ions were pushed towards the tops surface under pulse stimulations, then, at fatigue state (80 th ), the surface chemical bonding may structurally change leading to the LTM formation, as shown in Fig. 1(c) . It could be observed that the oxygen deficient state was vanished for the subpeak of 531.5 eV absent. In conclusion, Habituation/fatigue response to pulse stimulations, resembling memory functions of biological systems, have been demonstrated in the IGZO/HfO 2 memristor. The electrical conductivity increased with the stimulation of continuous pulse, but the degree of increase gradually decreased. Temperature dependence was observed for the relaxation processes of current retention.Dynamic of oxygen vacancy (ions) under external pulse and internal polarization fied is a dominant mechanism of memory evolution in STM. The STM could be enhanced and transferred to LTM by repetitive stimulation training, which is thought to be related to the local structure transition from unstable to solidified with the assistance of pulse train. Though further investigation is keep going to make the synaptic simulation more accurate and comprehensive, the observation of habitual behavior demonstrates the potential of memristors for biological neuron mimics."
} | 6,959 |
36004569 | PMC9825943 | pmc | 8,554 | {
"abstract": "Abstract The efficient valorization of lignin is crucial if we are to replace current petroleum‐based feedstock and establish more sustainable and competitive lignocellulosic biorefineries. Pulp and paper mills and second‐generation biorefineries produce large quantities of low‐value technical lignin as a by‐product, which is often combusted on‐site for energy recovery. This Review focuses on the conversion of technical lignins by oxidative depolymerization employing heterogeneous catalysts. It scrutinizes the current literature describing the use of various heterogeneous catalysts in the oxidative depolymerization of lignin and includes a comparison of the methods, catalyst loadings, reaction media, and types of catalyst applied, as well as the reaction products and yields. Furthermore, current techniques for the determination of product yields and product recovery are discussed. Finally, challenges and suggestions for future approaches are outlined.",
"conclusion": "4 Conclusions and Perspectives The need for renewable precursors in the production of chemicals, fuels, and materials, together with the large and underutilized lignin side‐stream from the pulp and paper industry, motivate efforts to valorize technical lignin. A win–win situation could be reached by addressing the bottlenecks in pulp production in combination with the extraction of lignin from black liquor. Economic advantages could be achieved, while simultaneously diversifying the product portfolio from lignin, for example, by recovering aromatic monomers produced via depolymerization. Although lignin depolymerization is not a new area of research, several challenges remain to be overcome before a sustainable and economically viable process can be achieved. Catalytic oxidative depolymerization is a promising route for lignin conversion to value‐added products due to its mild reaction conditions and the possibility of using inexpensive and green oxidants such as air or oxygen. Heterogeneous catalysts have the advantages of easier separation and regeneration of the catalyst than homogeneous systems. Although numerous studies were dedicated to this approach, the majority of these studies employed lignin model compounds as substrates, and the findings often cannot be extrapolated to real lignin or native lignin. The starting material strongly affects the results of oxidative lignin depolymerization, and can lead to operational problems, such as catalyst deactivation. Attention should be directed to technical lignin in order to achieve efficient catalytic oxidative depolymerization that can be applied industrially for the utilization of the aromatic feedstock provided by lignin side‐streams. One of the significant challenges highlighted in this Review is the stability and recyclability of the heterogeneous catalysts used in oxidative lignin depolymerization. A high number of catalyst systems exhibited metal leaching or deactivation resulting from the deposition of organic material or other impurities. Additionally, the detailed mechanisms behind catalytic deactivation and regeneration have remained unclear. Therefore, the emphasis in future research should be on developing stable and robust catalytic systems for oxidative lignin conversion. The complexity of lignin and lignin‐derived product mixtures still limits the insight that can be obtained from individual analytical techniques. The development of analytical methods will be needed to obtain comprehensive information on suitable pathways and products. Furthermore, little attention has been paid to the separation of depolymerized lignin products, \n [107] \n whereas developing an efficient separation process should go hand in hand with depolymerization in order to improve the upstream and downstream processes for oxidative lignin valorization. The recovery of aromatic products resulting from oxidative lignin depolymerization will involve a complex sequence of separation and purification processes, since the reaction mixture usually contains other products with similar physical and chemical properties. Most processes for the separation and purification of the aromatic monomers of interest, such as vanillin and syringaldehyde, are very labor‐ and energy‐demanding, and involve environmentally harmful solvents, and high material losses. Developing an efficient and environmentally friendly separation process is thus one of the most important tasks towards the industrial application of lignin‐derived aromatics.",
"introduction": "1 Introduction Lignin is the most abundant naturally occurring aromatic polymer on Earth and has excellent potential as a renewable feedstock for the production of chemicals, fuels, and functional materials. Lignin is also a major side‐product from the pulp and paper industry and cellulosic ethanol production, presenting a bulk raw material commonly referred to as technical lignin. \n [1] \n Despite its immense potential, lignin is still vastly underutilized and is primarily burned for energy recovery. The annual global chemical (Kraft) pulp production has been estimated to be 130 million tons, and about 50 million tons of lignin are obtained as a side‐product.[ \n 2 \n , \n 3 \n ] The potential of lignin as a sustainable feedstock for various applications is derived from its high abundance in nature, making it the most abundant aromatic compound in nature. The valorization of lignin has the potential to improve both the economic viability and the environmental performance of forest‐based industries. \n [4] \n \n The development of successful strategies for lignin valorization necessitates efficient technologies for lignin depolymerization. Lignin depolymerization strategies reported in the literature involve acid‐catalyzed, \n [5] \n base‐catalyzed, \n [6] \n biochemical, \n [7] \n oxidative, \n [8] \n reductive, \n [9] \n and thermal \n [10] \n methods. Each method is associated with its own benefits and drawbacks, and complex product mixtures are usually obtained.[ \n 3 \n , \n 4 \n ] The variations in operating conditions between the different depolymerization methods can also impact the compositions and yields of the resulting products significantly. Moreover, each kind of technical lignin is unique with regard to its purity, dispersity, molecular weight, and chemical structure.[ \n 11 \n , \n 12 \n ] Thus, selecting the appropriate depolymerization method for every lignin type is key to ensure product efficacy. Among different depolymerization methods, oxidative depolymerization is generally attractive due to the relatively mild operating conditions and their ability to produce targeted products with multiple functionalities. Oxidative lignin depolymerization can be used to produce valuable chemicals such as aromatic aldehydes and acids, as well as aliphatic carboxylic acids.[ \n 13 \n , \n 14 \n , \n 15 \n , \n 16 \n ] The oxidative approach has also been recommended as one of the most promising methods for the transformation of lignin into high‐value chemicals that are suitable for production on commercial scale. \n [17] \n Due to its considerable potential, the oxidative depolymerization of lignin is the main subject of the present Review. Most oxidative lignin depolymerization studies have been carried out using oxygen as the oxidizing agent in alkaline media, thus enabling the selective production of aromatic aldehydes (i. e., vanillin and syringaldehyde).[ \n 13 \n , \n 18 \n ] A number of Reviews have already been published on the oxidative valorization of lignin, and the reader is referred to these Reviews for general references and detailed information on this topic.[ \n 4 \n , \n 8 \n , \n 9 \n , \n 14 \n , \n 15 \n , \n 19 \n , \n 20 \n , \n 21 \n , \n 22 \n , \n 23 \n , \n 24 \n ] Here, focus is on providing a critical Review of the challenges and opportunities in understanding the conversion of technical lignin, via heterogeneous catalysis under oxidative conditions, on the molecular level. The different heterogeneous catalyst systems are summarized and discussed in relation to the type of lignin, the operating conditions used, and the resulting products. Aspects associated with this kind of conversion, including methods for product analysis, separation, and purification, are also surveyed. Thus, the present Review is complementary to existing literature, providing important insights into parameters and methodologies especially relevant for the oxidative valorization of technical lignin."
} | 2,104 |
29401493 | PMC5798774 | pmc | 8,556 | {
"abstract": "Impacts of global climate change on coral reefs are being amplified by pulse heat stress events, including El Niño, the warm phase of the El Niño Southern Oscillation (ENSO). Despite reports of extensive coral bleaching and up to 97% coral mortality induced by El Niño events, a quantitative synthesis of the nature, intensity, and drivers of El Niño and La Niña impacts on corals is lacking. Herein, we first present a global meta-analysis of studies quantifying the effects of El Niño/La Niña-warming on corals, surveying studies from both the primary literature and International Coral Reef Symposium (ICRS) Proceedings. Overall, the strongest signal for El Niño/La Niña-associated coral bleaching was long-term mean temperature; bleaching decreased with decreasing long-term mean temperature (n = 20 studies). Additionally, coral cover losses during El Niño/La Niña were shaped by localized maximum heat stress and long-term mean temperature (n = 28 studies). Second, we present a method for quantifying coral heat stress which, for any coral reef location in the world, allows extraction of remotely-sensed degree heating weeks (DHW) for any date (since 1982), quantification of the maximum DHW, and the time lag since the maximum DHW. Using this method, we show that the 2015/16 El Niño event instigated unprecedented global coral heat stress across the world's oceans. With El Niño events expected to increase in frequency and severity this century, it is imperative that we gain a clear understanding of how these thermal stress anomalies impact different coral species and coral reef regions. We therefore finish with recommendations for future coral bleaching studies that will foster improved syntheses, as well as predictive and adaptive capacity to extreme warming events.",
"conclusion": "Conclusion Understanding how El Niño/La Niña events impact coral reefs is crucial for developing strategies for coral reef conservation, which is important not only for biodiversity conservation, but also because tens of millions of people in over 100 countries rely on coral reefs for subsistence and to maintain their livelihoods [ 86 ]. The additional benefits that coral reefs provide are extensive, including protection against wave action, provision of fish habitat [ 87 ], recreation and tourism, and aesthetic and cultural benefits [ 88 – 90 ]. The resilience of these benefits is incrementally being eroded, as local stressors decrease baseline resilience [ 91 , 92 ] and climate change disables coral bleaching protection by shifting ocean warming trajectories on reefs from “protective” (trajectories that include a moderate amount of warming followed by a period of recovery before more intense heating instigates a bleaching event, essentially priming the corals to better respond to the heat), to “lethal” (trajectories that either spike rapidly and/or remain above bleaching thresholds) [ 93 ]. Our meta-analysis confirms that El Niño and La Niña-associated heat stress, as measured by maximum DHW, is a likely contributor to patterns of coral cover loss across the world's oceans. We also found that this trend is mediated by a temperature gradient in ENSO-associated coral cover loss, suggesting that higher latitude reefs may experience a smaller amount of El Niño and La Niña-associated decline compared to equatorial reefs. We show that there is a dearth of published studies reporting taxa-specific responses and changes in broad-scale ecosystem metrics to El Niño and La Niña, and we recommend that future studies should incorporate a broader range of resilience metrics in order to cope with measurement uncertainty and ecological surprise [ 94 ]. Future syntheses of recent and emergent bleaching events will allow us to discover where reefs are doing better than expected, and to more accurately focus global research, management, and conservation efforts.",
"introduction": "Introduction Climate change poses an imminent threat to the persistence of the world's coral reefs. Anthropogenic ocean warming is fundamentally altering marine ecosystems [ 1 ], exacerbating chronic local stressors such as overfishing, eutrophication, and coastal pollution, and threatening the resilience of marine ecosystems [ 2 , 3 ]. With increasing anthropogenic stressors, many coral reef ecosystems have transitioned from “safe operating spaces”–ecosystem states that are resilient to periodic stress events—towards “zone(s) of uncertainty” in which natural variability limits our prediction of ecosystem response, and “zone(s) of high risk” in which the ecosystem and its associated functions are already degraded [ 4 ]. Global surface warming manifests not only as gradual increases in overall temperature, which are predicted to exceed 2°C by 2100 [ 5 ], but also as intense pulse heat stress events such as the warm phase of the El Niño Southern Oscillation (ENSO). ENSO is a quasi-periodic fluctuation in oceanographic and atmospheric conditions, which transitions among El Niño, neutral, and La Niña conditions. El Niño is associated with increases in sea surface temperature (SST) that are primarily centered in the Central and Eastern Tropical Pacific Ocean [ 6 ]. Short-term positive warm temperature events caused by both El Niño and La Niña have instigated coral bleaching, so including both types of events allows us to investigate ENSO-related warming both in the Eastern/Central Pacific and in the Western Pacific, respectively. Major El Niño events have triggered three global coral bleaching events over the past four decades, the most intense of which unfolded over the course of 2015 and 2016 [ 7 ]. Extreme El Niño events are predicted to double in the future due to greenhouse warming [ 8 ], and the frequency of severe coral bleaching events is expected to increase even under moderate warming scenarios [ 9 ]. This projected increase in pulse warming events further threatens reefs which are already facing a multitude of stressors. Short-term thermal stress events, such as El Niño, primarily impact corals by inducing coral bleaching. Under normal conditions, the coral animal lives in symbiosis with endosymbiotic Symbiodinium (previously, zooxanthellae), single-celled algae that reside in the coral's tissue and provide metabolic products necessary for coral survival [ 10 , 11 ]. Coral bleaching results from a loss of Symbiodinium , which can occur under stressful conditions as the symbiosis breaks down leading to the coral losing pigmentation as Symbiodinium lose photosynthetic functionality and are ejected from the coral tissue [ 12 ]. If the symbiosis is not reestablished before the coral is depleted of metabolic products and energy reserves, subsequent mortality can occur [ 13 ]. Coral bleaching can occur in response to a variety of environmental stressors, the most documented of which is increased water temperatures [ 14 ]. Substantial natural gradients of environmental stressors, including sea surface temperature regimes, exist at varying spatial and temporal scales [ 15 , 16 ], and corals begin to bleach when ocean temperatures exceed local thermal thresholds. Coral bleaching severity and the extent of subsequent coral mortality is highly variable both within and across regions. Moreover, coral bleaching can affect any reef, occurring not only in areas with coastal human populations but also on remote reefs [ 17 ] and in protected areas [ 18 ]. While bleaching variability may be partially attributed to coral species differences, many other factors appear to influence the magnitude of changes in the coral reef community after pulse warming events. Coral bleaching and mortality caused by extreme El Niño/La Niña events can induce catastrophic changes in foundational coral reef ecosystem structure [ 19 , 20 ]. The impact of extreme El Niño events on coral communities worldwide has been observed and quantified since the early 1980s [ 21 – 23 ]. Notable regional examples include estimates of up to 95% coral mortality in some locations in the Eastern Pacific [ 24 ] and nearly 100% local coral mortality at some sites in Indonesia [ 25 ] during the 1982–83 El Niño event, and up to 90% coral mortality on individual shallow Indian Ocean reefs during the 1997–98 El Niño event [ 26 ]. Many early studies and bleaching reports were based upon underwater visual observations of bleaching severity and coral mortality. When conducted without replication, visual estimates can preclude quantitative meta-analysis because they may be prone to intra-observer variability and may not include statistical measurements such as error (e.g. standard deviation, standard error, or confidence intervals) necessary for many quantitative analyses. Depending on local conditions and the frequency of thermal stress, there is evidence that coral reefs can recover, even after an extreme El Niño/La Niña event [ 23 , 27 – 29 ]. Despite the potential for recovery, the effects of a single El Niño/La Niña event can be permanent. For example, one reef-building hydrocoral became locally extinct, while another was driven to probable extinction as a direct result of the 1982/1983 El Niño event [ 30 ], and Panamanian reef structure damaged by the 1982/1983 El Niño had not returned to pre-El Niño levels nearly twenty years later [ 31 ]. Even when the damage is not permanent, reefs can take more than a decade to recover from the impacts of an intense El Niño [ 32 ]. Despite these widely recognized and extensively cited effects of ENSO-related pulse warming events, a quantitative global analysis of the effects of El Niño/La Niña on coral communities has not yet been conducted. Our research builds upon previous reviews which investigated the impact of single El Niño events on coral reefs [ 33 – 35 ], differential bleaching of corals between two El Niño events in one region [ 36 ], and the long-term recovery of coral reefs after El Niño [ 37 ]. Here, we address three questions of relevance to understanding climate change impacts on coral reefs: 1) How much coral bleaching and mortality have been observed and documented during previous El Niño/La Niña events?, 2) Does the coral stress metric Degree Heating Weeks (DHW) accurately predict El Niño/La Niña-driven changes in coral bleaching and cover on a global scale, and what other factors are influential?, and 3) How does the 2015–2016 El Niño event compare to prior events in terms of severity (maximum DHW) and geographic extent? We conducted a global meta-analysis to quantify the effects of El Niño/La Niña events on coral communities. Although it is widely accepted that the increase of SST that results from El Niño/La Niña events induces coral bleaching, the consistency of the effects of El Niño/La Niña fluctuations on coral bleaching and mortality has yet to be quantified through multiple ENSO oscillations at a global scale. To evaluate the relationships between DHW and coral bleaching and mortality, we build upon previous research ( Box 1 ) using the Reynolds OI Level 4 AVHRR 0.25° sea surface temperature product to compute a comprehensive coral heat stress data set. This method allows for calculation of remotely-sensed coral heat stress for any date since 1982 using a consistent data set and climatology, allowing comparison of coral heat stress indices and response throughout the majority of published El Niño/La Niña-related coral bleaching events. Box 1. Degree Heating Week (DHW) products and coral bleaching prediction Understanding how El Niño- and La Niña-related heat stress affects coral reefs at a global scale requires quantification of the magnitude of thermal stress for each reef location. Quantification of thermal stress on coral reefs using satellite observations began with the definition of ocean \"hot spots\" [ 38 ] and in situ validation of satellite temperature observations at the reef scale [ 39 ]. In tandem with the development of \"hot spot\" analyses, researchers developed the concept of Degree Heating Weeks (DHW) as a cumulative metric of thermal stress on coral reefs [ 19 , 40 ]. The US National Oceanographic and Atmospheric Administration (NOAA) Coral Reef Watch Program (CRW) further developed these methods, using high-resolution (9km satellite pixel) HotSpot anomaly mapping [ 41 ] which improved correlation of satellite temperature observations with in situ measurements of coral reef temperature fluctuations [ 42 ]. NOAA CRW DHW products now include both an updated 50-km product [ 43 ], and a newly-released 5-km product [ 44 ]. The NOAA CRW products have been successful in detecting many coral bleaching events around the globe (e.g. [ 45 ]; [ 46 ]), and can now be used to forecast upcoming thermal stress on coral reefs [ 47 ]. Despite these major developments in understanding and predicting thermal stress and coral reef bleaching, a study that compared two DHW metrics to ReefBase bleaching reports found poor congruence between DHW magnitude and bleaching events both expected and observed, which was attributed to localized temperature variability and related coral adaptation [ 48 ]. The extent of coral bleaching in a given location has been suggested to vary in relation to historical variability in maximum SST and local climatological maximums, although lack of adequate bleaching reports has hindered our understanding of these relationships and of global coral bleaching trends [ 49 ]. Most recently, nowcasting and predictive tools have allowed managers to prepare for and respond to coral bleaching events, and cutting-edge climatological heat stress models and analyses are continually being developed at the US NOAA CRW program [ 50 ] and the Australian Bureau of Meteorology [ 51 ]. As local and regional coral bleaching reports are increasingly quantified, analyzed, and published, we have the ability to better understand how events such as El Niño and La Niña affect coral reefs at a global scale.",
"discussion": "Results and discussion El Niño heat stress Comparisons of individual El Niño/La Niña thermal stress anomalies, using the consistent 33-year climatology in our new data product, reveal considerable differences in the overall intensity of heat stress during past El Niño events and in the locations of the maximum heat stress amongst these events ( Fig 2 and S2 Fig ). The El Niño events of 1982/1983 and 1997/1998, which were the largest prior to the 2015/2016 event [ 6 , 67 – 70 ], show a typical El Niño pattern, with maximum heating occurring along the equator and the western coast of South America ( Fig 2 and S2 Fig ). In contrast, the Central Pacific El Niño or “Modoki”, with maximum heating along the equator near the dateline, can be seen clearly in the 2002/2003 and 2009/2010 El Niño events ( Fig 2 and S2 Fig ). Additionally, global maps of the maximum cumulative El Niño for each reef location illustrate stark differences in severity and areal impact of El Niño-associated heat stress ( Fig 2 ). 10.1371/journal.pone.0190957.g002 Fig 2 Maximum heat stress (DHW) for each reef location in the world (calculated at a 0.25° spatial resolution from AVHRR satellite data) during each of the eight El Niños that occurred in the past 35 years. Studies of El Niño/La Niña impacts on coral reefs Our literature search identified a total of 773 unique publications, of which 36 fit all of our search criteria. From the original articles returned from the searches, only 144 articles addressed coral bleaching or cover losses in relation to ENSO-associated warming ( Fig 1 ). Many of the initially-excluded studies were returned from our search but were irrelevant to the current meta-analysis, such as analyses of coral skeletal composition for reconstruction of historic climate. There is some overlap between coral bleaching papers (n = 50), and coral cover papers (n = 108) when split from relevant papers (n = 144), because a few studies (n = 7) provided data on both coral bleaching and coral cover. After all exclusion steps were complete, there were 5 coral bleaching studies which included \"before bleaching\" data, 15 coral bleaching studies which included data only during or after the bleaching event, and 28 coral cover studies which included all necessary attributes for analysis ( Fig 1 ). In total, 6 studies reported on coral bleaching associated with ENSO, and 30 studies reported on coral cover losses associated with El Niño/La Niña ( Fig 1 ). Our meta-analysis included a total of 453 data points globally ( Fig 3 ), which we subset into coral bleaching (n = 251) and coral cover (n = 202). The majority of data points were taken from graphs (n = 158); however, data points were also obtained from tables or directly from the text (n = 72). 10.1371/journal.pone.0190957.g003 Fig 3 Study locations included in this global meta-analysis. Studies reporting changes in coral bleaching due to El Niño/La Niña warming are marked in white, and studies reporting changes in coral cover due to El Niño/La Niña warming are marked in red. The background color scale represents the number of data points that were extracted from each location. Data from non-El Niño/La Niña bleaching events, and from papers excluded from this meta-analysis are not included on this map. The included studies employed a variety of survey types using both quadrat and transect methods, but in general these can be summarized into two categories: visual estimates and photo surveys. Included studies ranged in depth from 0.5m to 40m, although depth was often reported as a range nearly as large, hindering our ability to conduct any depth-specific analyses. We also were unable to examine taxonomic-specific effects of El Niño/La Niña-related bleaching analysis because even in the most commonly reported families (e.g. Poritidae , Pocillloporidae , Acroporidae ) there were few (<20, generally <5–10) data points for each family. Additionally, most data points for each family were extracted from a maximum of 1–3 studies, which were often spatially clustered (e.g. Pocilloporidae on the Pacific coast of Central/South America). Similarly, we did not conduct analyses by biogeographic region or local sub-regions, as this is a broad-scale coral bleaching study focused on El Niño/La Niña-related warming events, and because we analyze underlying environmental factors (e.g. mean long-term temperature and variance) rather than coarse large-scale biogeographic regions. While we concede that factors such as depth and taxonomy are or may be important drivers of coral bleaching and cover changes during warming events, these specific analyses were not possible using currently available peer-reviewed research. This meta-analysis focuses on peer-reviewed research on coral bleaching events which occur during El Niño and La Niña events. Although a great many other regional bleaching events have occurred outside of El Niño/La Niña events, the goal of this manuscript is to specifically investigate the influence of thermal stress on coral communities during El Niño and La Niña events. This may also, to some extent, limit the observed underlying environmental variability (e.g. by default our method excludes studies investigating bleaching due to cold water events, or point-source pollution). Additionally, if the \"Web of Science\" search is conducted similarly as above, but excluding the terms (El Niño OR El Nino OR ENSO), a total of 8,686 potential papers are returned for inclusion compared to the 773 reviewed in the current study. While this would be an admirable meta-analysis, it is outside of the scope of the current study. El Niño/La Niña effects on coral bleaching and coral cover As expected, El Niño- and La Niña-related warming causes an increase in coral bleaching ( Fig 4a ), with maximum coral bleaching reaching 100% in some locations. The best model for the partially-simulated before bleaching data set both using glmulti and reverse stepwise ANOVAs included long-term mean temperature, SSTmean (pseudo-R 2 = 0.058, Table 2 ). Decreasing SSTmean lowered the effect of El Niño/La Niña heat stress on coral bleaching, with equatorial latitudes experiencing the most coral bleaching, and coral bleaching decreasing further away from the equator although this effect was small. We suggest that the most likely driver of this phenomenon is adaptation of corals to intra-annual temperature variability on higher latitude reefs [ 49 ]. Despite the fact that SSTmean accounts for a relatively small amount of variability (~6%, pseudo-R 2 = 0.058), it is still notable that this moderator appears to influence coral bleaching, given the variety of other factors that also contribute to patterns in coral bleaching. When variability of effect sizes between different sites/time points within a paper were taken into account, TimeLag was removed as a significant moderator. This is most likely due to the fact that we expect the effect size (i.e. how much bleaching occurred) to vary within a study in direct relation to time since maximum heating occurred. Consequently, taking this into account in our model construction essentially masks the true effect of TimeLag on coral bleaching by accounting for this in within-study variation. 10.1371/journal.pone.0190957.t002 Table 2 Top model results for coral bleaching (including measured and simulated before-bleaching values) and coral cover loss (up to one year after maximum heat stress). Model Moderators (Top Model) QM df QE df Bleaching SSTmean, TimeLag:MaxDHW 7.5*** 2 1652*** 140 Cover MaxDHW, SSTmean 21*** 2 796*** 153 QM is the test of moderators (i.e. whether the moderators explain a significant amount of variance in the model), and QE is the test for residual heterogeneity (unexplained variance) after accounting for all included variables (p-value < 0.001 noted with ***). MaxDHW is maximum DHW experienced by reef during the present El Niño event, SSTmean is the long-term mean temperature, and TimeLag is the time since maximum DHW occurred. A colon represents an interaction between two moderators. See S3(b) and S3(c) Fig for the top ten model results. 10.1371/journal.pone.0190957.g004 Fig 4 Effect size and moderators of top coral bleaching and coral cover models (p-value < 0.001 noted with ***, p-value <0.05 noted with *). a) Overall effect size (standardized mean difference ± 95% confidence intervals) for coral bleaching (black; including measured and simulated before-bleaching values) and coral cover loss (red; up to one year after maximum heat stress). El Niño/La Niña warming significantly increases coral bleaching and significant decreases coral cover. Significant moderators in b) the coral bleaching model and c) the coral cover model. MaxDHW is maximum DHW experienced by reef during the present El Niño event, SSTmean is the long-term mean temperature, and TimeLag is the time since maximum DHW occurred. A colon represents an interaction between two moderators. Overall, percent coral cover significantly decreased after an El Niño/La Niña event ( Fig 4a ). The maximum reduction of local reef cover among studies was 100%, which was due to taxa-specific losses estimated within studies (absolute loss across the entire coral community = 80.5%). The model moderators with the best consensus for explaining coral cover loss were MaxDHW and SSTmean (pseudo-R 2 = 0.047, Fig 4c , Table 2 ). As expected, coral cover loss increased as cumulative heat stress (MaxDHW) increased ( Fig 4b ). Long-term mean temperature is also a significant moderator, with increasing SSTmean related to decreases in coral cover. Only the top two consensus moderators were included in the main presentation of results ( Table 2 , Fig 4 ). Although three of four coral cover models agreed on these two moderators, one model that included the full two years of data (after the El Niño/La Niña peak, constructed with glmulti) also included a small, but statistically significant interaction between MaxDHW and TimeLag ( S3a Fig ). The model including full two years of data (constructed with glmulti) did not include MaxDHW, although the other models (full two years of data constructed with glmulti, one year of data only and full two years of data constructed with reverse stepwise ANOVAs) did include MaxDHW as a significant moderator. Since SSTmean is a significant moderator in both bleaching and cover models, if we accept long-term mean temperature as a proxy for latitude, then low-latitude coral communities experience both higher bleaching and coral cover loss during El Niño/La Niña events ( Fig 4 ) than corals at higher latitudes. However, the thermal gradient in coral cover loss we demonstrate in this meta-analysis (and the proposed corresponding decrease in coral loss with increasing latitude) is in contrast to a previously observed latitudinal gradient coral bleaching (increasing bleaching with increasing latitude) [ 71 ]. This lends support to the hypothesis of increased recovery at higher latitudes, where corals bleach more frequently, but are either acclimated or adapted to recover from bleaching stress. Residual heterogeneity Within all models, there was a significant degree of heterogeneity remaining after accounting for all significant moderators ( Table 2 ). This suggests additional factors not included within these models have a significant influence upon coral bleaching and cover loss during ENSO. This may also be due to non-linearities in moderator-response interactions, which we were unable to test with the available data. With more data, or with a more constrained question (e.g. local- or regional-scale analyses), exploring non-linear models may be instructive for understanding residual heterogeneity not resolved with linear modelling. Factors not included within our models that could account for this heterogeneity include coral taxa, local adaptation, and coral depth, as well as a suite of abiotic factors including wind, upwelling and currents, and localized (i.e. meter to kilometer scale) thermal anomalies. The intensity and distribution of coral bleaching can be affected by the taxonomic composition of individual reefs, based on species-specific bleaching susceptibilities [ 72 ]. Reef location, both depth and distance to shore, are also important factors determining the vulnerability of corals to thermal bleaching and subsequent mortality [ 73 ]. Local acclimation and adaptation were not considered in our study, but almost certainly influence coral bleaching patterns as well [ 74 ]. For example, local oceanographic conditions affect susceptibility to bleaching, as high water flow [ 75 ] and upwelling [ 76 ] potentially limiting acclimation and decreasing coral tolerance to warming. Many additional factors contribute to heterogeneity in coral bleaching patterns, ranging from cloud cover [ 77 ] and water quality [ 78 , 79 ], to basic coral biology and micro-complexity [ 80 ]. These factors are fundamentally important to understanding patterns in coral bleaching, survival, and resilience, and continued investigation and synthesis are encouraged as more data become available. The novelty of the 2015–2016 El Niño event Prior to 2015, the most extreme El Niño events observed to impact coral reefs globally occurred in 1982–83 and 1997–98. The 2015–2016 El Niño surpassed these events both in terms of ocean warming intensity and extent [ 81 ], causing unprecedented ecological consequences worldwide. While areas affected by typical Eastern Pacific El Niño events (i.e. the coast of Central America) were still not affected worse than the massive El Niño in 1997–1998, the 2015–2016 El Niño now dominates tropical waters as the highest cumulative stress on corals globally on record ( Fig 5 ). In fact, for several reefs in the Central Pacific region, the 2015–2016 El Niño exceeded the threshold \"Not experienced by reefs as yet\" (24 Degree Heating Weeks) described by Hoegh-Guldberg merely 6 years ago [ 82 ]. As the return time between bleaching events decreases, we expect that the balance of long-term influence of coral resistance versus recovery from bleaching events will shift, making coral resistance incrementally more important than recovery [ 83 ]. 10.1371/journal.pone.0190957.g005 Fig 5 El Niño events with the greatest heat stress. Both figures show which El Niño event caused the greatest maximum DHW for each area. Note that this figure does not demonstrate bleaching response, only maximum cumulative heat stress per El Niño event. The events are color-coded by year. The 1997/1998 El Niño event (green) was the most severe event in the Eastern Pacific around the South American coast. a) All El Niño events from 1982–2010, showing how much heterogeneity there is in the geographic distribution of the most extreme heat stress. b) All El Niño events since 1982, including the 2015–2016 El Niño event, demonstrating the coral heat stress homogenization that occurred during this most recent El Niño/La Niña warming event. Recommendations Based on the outcomes of this systematic review of the published ENSO-warming related coral literature, we make several recommendations for future coral reef bleaching studies. First, researchers should include at least the following El Niño/La Niña warming parameters: Magnitude of warming: current local Degree Heating Week at the time of field sampling; Timing and trajectory of warming (e.g. include figure with temperature or DHW trajectory leading up to sampling time point); A history of bleaching events for the study location(s). We suggest including a supplementary figure of heat stress over time since 1982 (e.g. by extracting data from our new DHW data product). This would allow for examination of historic heat stress, as well as the potential for local acclimation and adaptation due to previous heat stress conditions. Additionally, it is imperative that researchers present all study parameters including: Exact sampling dates for each location sampled; Exact GPS coordinates for each location sampled, including multiple GPS coordinates in cases where there is more than one study site; Sample size, at the smallest measured scale (e.g. samples per site); Sampling error (e.g. standard deviation or standard error); Any available before-impact data: coral cover, coral bleaching, including corresponding sampling errors, dates, and methods; Exact coral survey depth(s). Large depth ranges (e.g. 5- 20m depth), and even moderate ranges at critical depths (e.g. 2-10m depth) obfuscate patterns in coral bleaching by greatly increasing unexplained variability in light exposure. If a study examines coral bleaching at different depths, this should be reported, with data and results specified by depth. As well as study-specific taxonomic information including: Taxonomic composition of the surveyed corals, as well as species-specific coral responses to heat stress; Overall coral community response, in cases where the primary purpose of the study is to investigate single (or a few) coral species. Finally, we encourage researchers to make their data and results fully open upon publication of their study. This includes, but is not limited to: photo/video survey methods which can produce archives of coral reef status. Reproducibility, and consequently future synthesis work, would be enhanced by making images available online (through sites like CoralNet [ 84 ]), and by making analyses transparent by providing data, code, and results (via sites like GitHub). We found that only a subset of papers considered for this meta-analysis included both the temperature stress (i.e. either temperature anomaly at the time of sampling or cumulative DHW) and a specific time period of when the sampling took place. It will be much easier to identify patterns in bleaching and mortality if we can rectify the data we already have to a quantitative time frame of thermal stress. Additionally, we found that 20 out of 44 coral bleaching papers and 20 out of 61 coral cover papers did not include sample size, sampling error, or both. Finally, we note that long-term monitoring data sets are important, as they provide a baseline against which to compare changes to coral reef ecosystem structure and health, and we strongly support the development of such data sets. Building a mechanistic understanding of how local variability in baseline coral cover and bleaching changes during El Niño/La Niña warming will allow us to identify the processes that give rise to bright spots [ 85 ] that foster coral reef resilience and recovery over the long term. Conclusion Understanding how El Niño/La Niña events impact coral reefs is crucial for developing strategies for coral reef conservation, which is important not only for biodiversity conservation, but also because tens of millions of people in over 100 countries rely on coral reefs for subsistence and to maintain their livelihoods [ 86 ]. The additional benefits that coral reefs provide are extensive, including protection against wave action, provision of fish habitat [ 87 ], recreation and tourism, and aesthetic and cultural benefits [ 88 – 90 ]. The resilience of these benefits is incrementally being eroded, as local stressors decrease baseline resilience [ 91 , 92 ] and climate change disables coral bleaching protection by shifting ocean warming trajectories on reefs from “protective” (trajectories that include a moderate amount of warming followed by a period of recovery before more intense heating instigates a bleaching event, essentially priming the corals to better respond to the heat), to “lethal” (trajectories that either spike rapidly and/or remain above bleaching thresholds) [ 93 ]. Our meta-analysis confirms that El Niño and La Niña-associated heat stress, as measured by maximum DHW, is a likely contributor to patterns of coral cover loss across the world's oceans. We also found that this trend is mediated by a temperature gradient in ENSO-associated coral cover loss, suggesting that higher latitude reefs may experience a smaller amount of El Niño and La Niña-associated decline compared to equatorial reefs. We show that there is a dearth of published studies reporting taxa-specific responses and changes in broad-scale ecosystem metrics to El Niño and La Niña, and we recommend that future studies should incorporate a broader range of resilience metrics in order to cope with measurement uncertainty and ecological surprise [ 94 ]. Future syntheses of recent and emergent bleaching events will allow us to discover where reefs are doing better than expected, and to more accurately focus global research, management, and conservation efforts."
} | 8,632 |
27486438 | PMC4949241 | pmc | 8,557 | {
"abstract": "Endosymbioses between animals and chemoautotrophic bacteria are ubiquitous at hydrothermal vents. These environments are distinguished by high physico-chemical variability, yet we know little about how these symbioses respond to environmental fluctuations. We therefore examined how the γ-proteobacterial symbionts of the vent snail Ifremeria nautilei respond to changes in sulfur geochemistry. Via shipboard high-pressure incubations, we subjected snails to 105 μM hydrogen sulfide (LS), 350 μM hydrogen sulfide (HS), 300 μM thiosulfate (TS) and seawater without any added inorganic electron donor (ND). While transcript levels of sulfur oxidation genes were largely consistent across treatments, HS and TS treatments stimulated genes for denitrification, nitrogen assimilation, and CO 2 fixation, coincident with previously reported enhanced rates of inorganic carbon incorporation and sulfur oxidation in these treatments. Transcripts for genes mediating oxidative damage were enriched in the ND and LS treatments, potentially due to a reduction in O 2 scavenging when electron donors were scarce. Oxidative TCA cycle gene transcripts were also more abundant in ND and LS treatments, suggesting that I. nautilei symbionts may be mixotrophic when inorganic electron donors are limiting. These data reveal the extent to which I. nautilei symbionts respond to changes in sulfur concentration and species, and, interpreted alongside coupled biochemical metabolic rates, identify gene targets whose expression patterns may be predictive of holobiont physiology in environmental samples.",
"conclusion": "Concluding remarks This study of deep-sea vent snails is unique in that it examined the transcriptional responses of bacterial symbionts under well-controlled environmental conditions. These results show that while I. nautilei symbionts can use either sulfide or thiosulfate for energy during chemoautotrophic growth (as observed in Beinart et al., 2015 ), the transition between these substrates appears to have surprisingly little effect on symbiont transcription patterns. Periplasmic nitrate reduction appears critical at higher sulfide or thiosulfate conditions, likely reflecting a dependence on nitrate respiration at elevated sulfur species concentrations, as well as enhanced nitrogen assimilation to meet potentially elevated growth rates. Notably, when sulfide is low or absent, there are marked and unexpected changes in symbiont transcription. First, I. nautilei symbionts express genes that are consistent with mixotrophy, suggesting the possibility of transitioning to heterotrophic growth when energy for CO 2 fixation is limiting. The origin and composition of the organic carbon for symbiont heterotrophy is unknown in I. nautilei , but could involve recycling of organics leaked from host cells or the catabolism of internally stored carbon (Kleiner et al., 2012b ). Second, these data suggest that at lower sulfide concentrations I. nautilei symbionts may experience oxidative stress. We posit that the symbionts may be experiencing a reduction in reactive oxygen species scavenging when electron donors are scarce, or alternatively may be exposed to elevated reactive oxygen species concentrations caused by stressed hosts. While the causal factor remains unclear, it is important to note that these conditions are comparable to those found in situ , especially around lower flow diffuse vents. Third, a striking commonality among many of the differentially expressed genes identified in this study, including form II RubisCO, oxidative TCA cycle enzymes, and genes involved in iron uptake, is that redox potential is implicated in their regulation based on studies in other bacteria; thus environmental redox conditions may control many processes in I. nautilei symbionts. Our results also highlight the variability of I. nautilei symbiont community composition across the Lau Basin vent system and provide the first molecular evidence for the presence and activity of methanotrophic symbionts in Ifremeria . The persistence of methanotrophy transcripts in our incubations, despite the lack of a clear methane source, may reflect an adaptation to the dynamic vent environment, with the Ifremeria -associated bacterial community being poised to rapidly consume methane when this energy source becomes available. Together these data reveal that vent geochemical dynamics affect a broad range of metabolic systems and subsystems, such as carbon, nitrogen and sulfur metabolism. Such variation will in turn govern the extent to which the holobiont's (host + symbiont) activity influences environmental geochemistry. Finally, the apparent response to low or no sulfide was quite striking, and future studies should further examine the extent to which oxidative stress influences symbiont function in situ . It should be noted that the relative abundances of transcripts, for example those encoding RubisCO, AprAB, or stress proteins, may be reasonably accurate indicators of substrate conditions and symbiont metabolic state. Other loci, however, appear less transcriptionally responsive to environmental change and their expression patterns should be interpreted with caution. In the future, similar controlled experiments that couple measurements of chemical flux (e.g., carbon fixation, sulfur and methane oxidation) to biomolecule abundances are necessary for determining which molecular patterns accurately reflect symbiont physiology, and therefore the extent to which meta-omic datasets can be used for predictive models of symbiont contributions to holobiont fitness.",
"introduction": "Introduction Hydrothermal vents are dynamic ecosystems where the vigorous emission of hot, chemically reduced fluid from the seafloor into the surrounding seawater results in a temporally and spatially variable physico-chemical environment. The dominant animals in these habitats live in partnership with bacterial symbionts that oxidize chemical reductants from the venting fluid, including hydrogen sulfide, methane, and hydrogen (Tivey, 2007 ), as well as thiosulfate, an intermediate oxidation product of sulfide that accumulates at some vents (Gru et al., 1998 ; Waite et al., 2008 ; Gartman et al., 2011 ). The symbionts oxidize these reduced compounds to fuel carbon fixation (chemoautotrophy), which provides primary nutrition for themselves and their animal hosts (Cavanaugh et al., 2006 ; Dubilier et al., 2008 ). To acquire the chemical substrates needed for symbiont metabolism, symbiotic vent animals live where they can access both emitted fluid and oxygenated seawater (Stewart et al., 2005 ). Vent symbioses thrive in mixing zones where fluid turbulence causes conditions to change rapidly over small spatial scales such that adjacent animals may experience vastly different chemical and physical environments (Johnson et al., 1986 ). Sulfur chemistry in these zones can be particularly dynamic, with concentrations of hydrogen sulfide and derivatives of sulfide oxidation (e.g., thiosulfate) varying temporally and in proximity to vent sites (Gru et al., 1998 ; Mullaugh et al., 2008 ; Waite et al., 2008 ; Gartman et al., 2011 ). Furthermore, venting can halt, either transiently or permanently, due to changes in volcanic activity or the path of fluid flow (Butterfield et al., 1997 ). The dynamic nature of hydrothermal vents thus exposes organisms to fluctuating concentrations of diverse reductants and oxidants (Shank et al., 1998 ), and may drive variation in the metabolism and growth dynamics of symbiotic partners (Duperron et al., 2007 ; Dubilier et al., 2008 ; Robidart et al., 2011 ). Descriptive and experimental studies over the past 30 years have enabled a broad understanding of the chemoautotrophic metabolism of vent symbionts. Physiological experiments with intact animal hosts (e.g., Childress et al., 1986 , 1991 ; Girguis et al., 2000 ; Girguis and Childress, 2006 ; Nyholm et al., 2008 ; Petersen et al., 2011 ; Goffredi et al., 2014 ; Beinart et al., 2015 ) and excised symbionts (Belkin et al., 1986 ; Fisher et al., 1987 ; Wilmot and Vetter, 1990 ; Childress et al., 1991 ; Nelson et al., 1995 ) have established that vent symbionts can fix inorganic carbon via the oxidation of hydrogen sulfide, thiosulfate, hydrogen and/or methane. Separately, analyses of symbiont gene content (Kuwahara et al., 2007 ; Newton et al., 2007 ; Robidart et al., 2008 ; Nakagawa et al., 2014 ), gene and protein expression (Markert et al., 2007 , 2011 ; Nyholm et al., 2008 ; Robidart et al., 2011 ; Gardebrecht et al., 2012 ; Wendeberg et al., 2012 ; Sanders et al., 2013 ), and enzyme activity (Felbeck, 1981 ; Stein et al., 1988 ; Robinson et al., 1998 ) from vent symbioses have clarified the pathways that symbionts employ for these metabolisms. While it has been speculated that shifts in chemical availability are coupled to rapid physiological responses by the host, the symbiont, or both (Girguis and Childress, 2006 ), inferences made from freshly collected animals reveal at best a general picture of the conditions experienced in situ , making it challenging to identify the relationships between environmental conditions, symbiont/host physiology, and shifts in the metabolic coupling of the symbiont and host. A greater understanding of how the environment shapes the physiology of vent symbionts requires controlled experiments that couple measurements of metabolic rates with analyses of gene expression. Accordingly, we conducted high-pressure respirometric experiments on the symbiotic vent snail Ifremeria nautilei to characterize changes in symbiont gene expression (metatranscriptomes) under varying regimes of reduced sulfur species and availability. I. nautilei lives in mixing zones around hydrothermal vents in the southwestern Pacific, deriving its nutrition from intracellular gill symbionts that can oxidize both sulfide and thiosulfate to fuel autotrophy (thioautotrophy; Beinart et al., 2015 ). Previous characterization of the I. nautilei gill symbiont community revealed associations with a lineage of γ-proteobacterial sulfur oxidizers from the Order Chromatiales (Urakawa et al., 2005 ; Suzuki et al., 2006 ). Methanotrophic bacteria may also be present (Gal'chenko et al., 1992 ; Borowski et al., 2002 ), although their representation across host populations is inconsistent (Windoffer and Giere, 1997 ; Suzuki et al., 2006 ) and molecular characterization of I. nautilei methanotrophs is lacking (Petersen and Dubilier, 2009 ). We herein present 16S rRNA gene and metatranscriptome data elucidating symbiont community composition and metabolism over a range of environmentally relevant conditions: (i) no electron donor (ND), (ii) 105 μM hydrogen sulfide (LS), (iii) 350 μM hydrogen sulfide (HS), and (iv) 300 μM thiosulfate (TS). These experiments clarify pathways of sulfur, nitrogen, and carbon metabolism in the Ifremeria -associated bacterial community and provide the first molecular evidence for the activity of methanotrophic bacteria in the gill-associated community. The results are derived from experiments designed to study the effects of geochemical dynamics on sulfur oxidation and carbon fixation rates (Beinart et al., 2015 ). An analysis of symbiont transcripts in light of the observed metabolic rates reveals how I. nautilei symbionts sustain chemoautotrophic activity at low sulfide concentrations, but exhibit a marked change in their physiology as sulfur availability increases.",
"discussion": "Results and discussion Metabolic activity Ifremeria nautilei individuals were collected in four separate dives from four locations within the ABE and Tu'i Malila vent fields of the Eastern Lau Spreading Center (Table 1 and Figure S1 ). All individuals were subjected to an 8-h acclimation period in pressurized aquaria before the treatment incubations. Snails used for the HS, TS, and LS treatments were acclimated without an electron donor, whereas those used in the ND treatment were acclimated with 300 μM sulfide. After acclimation, treatment conditions were established in the aquaria and remained constant for 27–40 h. The same snails that were used for rate measurements were used for subsequent symbiont community composition and metatranscriptome analysis. Sulfur oxidation and carbon fixation rates for all individuals were reported in Beinart et al. ( 2015 ) and are summarized here for convenience in interpreting the gene expression patterns. Sulfur metabolism rates were measured by assessing the net flux of sulfide, thiosulfate, and polysulfide through high-pressure aquaria under each treatment. I. nautilei demonstrated net uptake (oxidation) of sulfide or thiosulfate when these compounds were supplied across all tested conditions (Table 2 ). Mass-specific sulfur oxidation rates were highest in I. nautilei exposed to 300 μM of thiosulfate (TS), with the next highest rates in the HS followed by the LS treatments (Table S2 ). The measurement of net sulfur flux through the aquaria also revealed the excretion of polysulfides by I. nautilei when exposed to the HS treatment (Beinart et al., 2015 ). Oxygen (O 2 ) was detectable in HPRS outflow of all treatments during the experiments, indicating that complete anaerobiosis was not established in any treatment. Symbiont carbon fixation was observed in treatments exposed to both sulfide and thiosulfate (Table 2 ). In contrast to sulfur metabolism rates, which were based on the collective uptake by all experimental individuals (i.e., sulfide disappearance from aquaria) and therefore reflected an average metabolic rate across individuals per treatment, carbon fixation rates were measured by the incorporation of stable isotope labeled ( 13 C) inorganic carbon into individual snails. Compared to symbiont-free foot tissue (average δ 13 C −28.0 ‰± 0.95 s.d; Beinart et al., 2015 ), elevated carbon isotopic compositions were detected in the gills of I. nautilei individuals exposed to HS, LS, and TS (Table 2 ), whereas carbon incorporation was not observed in any individuals in the no electron donor (ND) treatment. Although mass-specific incorporation rates were relatively consistent among individuals in the TS treatment, rates differed widely among individuals within the HS and LS treatments. Notably, an individual in the HS treatment demonstrated the highest rate across all treatments, while other HS individuals showed low or undetectable carbon incorporation (Table 2 ). This pattern, seen also in the LS treatment, reflects the position of the individuals in the aquaria, with the most productive individuals closest to the incoming water, suggesting that metabolism by downstream individuals was limited by the activity of those upstream (Beinart et al., 2015 ). Symbiont taxonomic composition—16S rRNA gene amplicons At the conclusion of each incubation, gills of three host individuals were immediately sampled to assess symbiont phylogenetic affiliation and gene transcription profiles. Illumina sequencing of 16S rRNA gene (DNA) amplicons revealed that all 12 experimental I. nautilei hosted highly related communities of symbionts (Figure 1 ). A single OTU (97% similarity cluster) within the order Chromatiales of the γ-proteobacteria was dominant across all individuals, accounting for 90.1% of all sequences (Figure 1 ). A representative sequence from this OTU clustered phylogenetically with the known sulfur-oxidizing I. nautilei symbiont, differing from published sequences by 1-2 nucleotides (<0.8%; Figure 2 ). Figure 1 Distribution of amplified 16S rRNA genes (A) and 16S rRNA transcripts (B) in Ifremeria nautilei samples from experimental shipboard incubations . Figure 2 Maximum-likelihood phylogenetic tree of the top 5 most abundant 16S rRNA gene OTUs . The 5 OTUs shown here accounted for 98% of all 781 662 reads analyzed. The tree is based on 252 unambiguously aligned nucleotide positions. Two ε-proteobacterial Alviniconcha hessleri symbionts are used as the outgroup. The numbers to the left of each node are bootstrap values obtained from 1000 replicates; only values greater than 50% are indicated. Amplicon sequencing also detected putative methanotrophic bacteria in association with I. nautilei used in the treatments, although at much lower abundance than the dominant Chromatiales symbionts. Sequences related to the Methylococcaceae, a γ-proteobacterial family of methanotrophic bacteria, represented 7.7% of all amplicons (all individuals combined) and three distinct OTUs (Figure 2 ). These OTUs were abundant in the TS, LS, and ND incubations (up to 16.8% of amplicons from a single host), but represented <0.02% of sequences from individuals in the HS incubation (Figure 1 , and discussed in more detail below). The Methylococcaceae OTUs did not cluster with the methanotrophic symbionts of other animals, but rather formed a separate cluster with a cloned 16S rRNA sequence from a shallow hydrothermal vent system (Hirayama et al., 2007 ; Figure 2 ). These findings represent the first published DNA sequences identifying a γ-proteobacterial methanotrophic symbiont in I. nautilei . Previous electron microscopy studies have observed multiple bacterial morphotypes in the gill cells of some I. nautilei individuals, including a low abundance morphotype with stacked cytoplasmic membranes characteristic of type I methanotrophs (Gal'chenko et al., 1992 ; Borowski et al., 2002 ). However, prior molecular analyses detected only the Chromatiales symbiont (Windoffer and Giere, 1997 ; Borowski et al., 2002 ; Urakawa et al., 2005 ; Suzuki et al., 2006 ). The fifth most abundant OTU recovered from the amplicon library was affiliated with the genus Endozoicomonas of the order Oceanospirillales , and accounted for 0.5% of all sequences (Figure 2 ). Endozoicomonas bacteria are aerobic heterotrophs that appear commonly in association with marine invertebrates (Dubilier et al., 2008 ). There is evidence that Oceanospirillales are intranuclear parasites that cause host cell death in symbiotic bathymodiolin mussels (Zielinski et al., 2009 ). Intracellular Oceanospirillales have also been observed in the hydrothermal vent gastropod Alviniconcha that co-occurs with I. nautilei in the Lau Basin, although the nature of the association is not clear (Beinart et al., 2014 ). To investigate whether differences in symbiont diversity resulted from incubation conditions or from in situ variability among the sampled populations, we extended the 16S rRNA gene analysis to include 10 snails collected from the same vent fields as the experimental animals. This analysis revealed a similar dominance of the single Chromatiales OTU, but also considerable variation in the relative abundance of Methylococcaceae among host individuals (Figure 3 ). Notably, Methylococcaceae OTUs composed < 0.05% of the symbiont population in snails from the ABE vent field collected during the same dive as snails HS1-3 used in the HS treatment, while the symbiont population within snails from the Tu'i Malila vent field contained 5–17% Methylococcaceae . These results suggest that the low proportion of Methylococcaceae in the HS 16S rRNA gene datasets may reflect environmental variation among symbiont populations rather than an effect of incubation conditions (Figure 1 ). Our results, from both experimental and environmental specimens, highlight the potential for local variation in the proportional abundances of Chromatiales and Methylococcaceae symbionts in I. nautilei . Additional studies are necessary to confirm whether such variation is linked with environmental sulfide and methane availability, as shown for other dual chemosynthetic symbioses (Duperron et al., 2011 ). Figure 3 Bar graph showing the distribution of 16S rRNA OTUs in Ifremeria nautilei from ABE and Tui Malila vent fields . Metatranscriptome characteristics and taxonomic composition Each metatranscriptome was derived from a single snail and consisted of 1 to 5 million high quality paired sequence reads representing both symbiont and host RNA (Table 3 ). Metatranscriptomes were generated from the same 12 individuals (3 from each treatment) that were used in the metabolic experiments and phylogenetic analyses. Of the reads generated, 69–77% represented rRNA, while 5–9% represented mRNA transcripts with matches to protein-coding genes in the NCBI nr database. The remaining 16–24% of the reads not identified as either rRNA or mRNA may represent unidentified protein-coding genes or enzymatic and regulatory RNA. Approximately 51–73% of mRNA transcripts identified in MEGAN were eukaryotic in origin, with only 8–12% of these transcripts successfully assigned a functional annotation. The single most abundant transcript identified in all 12 metatranscriptomes (2–7% of all identified protein-coding transcripts) matched the cytochrome oxidase subunit I gene of I. nautilei with > 99% identity, confirming the species identity of the host individuals. Other abundant eukaryotic transcripts identified in all host individuals regardless of treatment included cytochrome oxidase subunit II, tRNAs, NADH dehydrogenase complex I, fumarate reductase, cytochrome bc1 complex, and ATP synthase subunits, suggesting that all snails were metabolically active at the time of sampling and aerobic conditions were maintained throughout all experiments. Table 3 Metatranscriptome sequence characteristics . Snail Total reads rRNA reads Non-rRNA reads BLASTX Hits MEGAN assigned proteins Eukaryotic assigned proteins Bacterial assigned proteins Other assigned proteins a HS-1 2,106,871 1,500,965 605,906 (28%) 175,997 114,267 74,424 (65%) 36,481 (32%) 3,362 (3%) HS-2 3,097,455 2,399,502 697,953 (23%) 215,535 150,694 87,022 (58%) 56,710 (38%) 6,962 (4%) HS-3 2,753,601 2,074,095 679,506 (25%) 226,213 156,536 79,203 (51%) 75,514 (48%) 1,819 (1%) TS-1 1,138,408 870,307 268,101 (24%) 63,447 42,524 29,811 (70%) 11,654 (27%) 1,059 (3%) TS-2 3,140,132 2,412,371 727,761 (23%) 192,802 138,917 85,890 (62%) 48,067 (35%) 4,960 (3%) TS-3 2,903,835 2,219,159 684,676 (24%) 188,561 138,659 83,649 (60%) 50,116 (36%) 4,894 (4%) LS-1 1,280,435 974,592 305,843 (24%) 69,872 42,796 29,272 (68%) 13,051 (30%) 473 (2%) LS-2 3,511,480 2,583,832 927,648 (26%) 265,621 184,359 109,752 (60%) 67,796 (37%) 6,811 (3%) LS-3 2,958,987 2,295,224 663,763 (29%) 189,656 138,931 92,533 (67%) 41,765 (30%) 4,633 (3%) ND-1 2,115,702 1,462,295 653,407 (31%) 153,995 96,528 70,884 (73%) 23,290 (24%) 2,354 (3%) ND-2 4,982,423 3,854,530 1,127,887 (23%) 318,447 229,893 128,470 (56%) 88,102 (38%) 13,321 (6%) ND-3 3,775,724 2,717,615 1,058,109 (28%) 298,774 203,180 137,423 (68%) 59,246 (29%) 6,511 (3%) a Includes proteins predicted to be of viral, archaeal, or unknown origin . Bacterial 16S rRNA transcripts composed 19–53% of the total small subunit (SSU) rRNA reads (16S + host 18S) in each metatranscriptome. Interestingly, 16S rRNA represented a higher proportion of total SSU transcripts in metatranscriptomes from the HS treatments (average 45 ± 7%) compared to those from the other treatments (average 26 ± 6%; Table 3 ). This difference may reflect variation in either symbiont density or activity. This is consistent with the aforementioned metabolic data, and suggests that at elevated sulfide concentrations the bacterial symbiont population is more active. The vast majority of 16S rRNA transcripts (91.6–99.1% in all metatranscriptomes) matched sequences related to the Chromatiales (Figure 1 ). Consistent with the 16S gene amplicon results, a proportion of 16S rRNA transcripts (1.6–5.1%) from the TS, LS, and ND metatranscriptomes were classified as Methylococcaceae , whereas Methylococcaceae represented < 0.02% of total 16S transcripts in the three HS transcriptomes (Figure 1 ). In all specimens, the proportion of Methylococcaceae sequences in the 16S rRNA datasets derived from metatranscriptomes was approximately 1/4 the proportion observed in the DNA-amplified 16S gene datasets. This consistently lower proportion in the transcript data could indicate a low cellular RNA to DNA ratio, suggesting that Methylococcaceae bacteria were less metabolically active than the sulfur-oxidizing population during the incubations. This is consistent with the assumption that all of the incubations lacked appreciable methane or other C1 compounds that methylotrophs typically metabolize (neither methane nor any other C1 compounds were added to these incubations). Alternatively, the discrepancy between the DNA and RNA data could result from PCR bias and uneven amplification of different templates. Finally, consistent with the amplicon data, a minor fraction (0.2–0.5%) of 16S rRNA transcripts in each metatranscriptome matched sequences of the Oceanospirillales order (Figure 1 ). Differential gene expression To detect differential expression, protein-coding transcripts were assigned to functional SEED subsystems (Level 1 and 2) and proteins (Level 4) in MEGAN5 (Huson et al., 2011 ). The baySeq R package was then used to determine which model (treatment grouping) best explained subsystem and protein expression patterns (Table 4 ). Differential expression patterns were best explained by dividing the metatranscriptomes into two groups based on the concentration of electron donor in the incubation experiments. Grouping transcriptomes according to high electron donor (HD: HS and TS treatments) vs. low electron donor (LD: LS and ND treatments) conditions explained an estimated 39% of Level 1 subsystem differential expression, compared to 24% when transcriptomes were grouped according to the four individual treatments (Table 4 ). This trend was also observed when transcripts were assigned to Level 3 SEED subsystems and functional proteins (Table 4 and Tables S1 , S2 ). Other models, including one comparing the thiosulfate treatment to the sulfide-containing treatments, explained less than 20% of variation at all subsystem levels. The availability of electron donor, and not donor compound type, therefore appears to be the strongest driver of symbiont differential gene expression in these experiments. Table 4 Estimated proportion of differential expression based on posterior likelihoods of differential expression among treatments . Estimated proportion of differential expression Model SEED level 1 (28 subsystems) SEED level 3 (344 subsystems) SEED level 4 (1086 genes) (HS,HS,HS,TS,TS,TS) (LS,LS,LS,ND,ND,ND) 0.39 (8) * 0.29 (23) 0.49 (62) (HS,HS,HS) (TS,TS,TS) (LS,LS,LS) (ND,ND,ND) 0.24 (6) 0.05 (10) 0.04 (2) (HS,HS,HS,TS,TS,TS,LS,LS,LS) (ND,ND,ND) 0.18 (1) 0.07 (0) 0.16 (0) (HS,HS,HS,LS,LS,LS) (TS,TS,TS) 0.10 (0) 0.09 (0) 0.21 (1) HS, High Sulfide; TS, Thiosulfate; LS, Low Sulfide; ND, No e- donor . * Number of subsystems differentially expressed (FDR = 0.05) shown in parentheses . While our data indicate electron donor availability as the major driver of symbiont transcriptional variation, it is possible that factors independent of treatment conditions also affected the observed transcription patterns. First, some genes may be constitutively transcribed at uniform levels among treatments. Second, both the host and symbionts may produce storage molecules that could—over the duration of the incubations—dampen the transcriptional and physiological response to our treatments. For example, some chemoautotrophic symbionts produce elemental sulfur that can be stored and may be used as energy to drive carbon fixation when exogenous reductants are absent or low (Vetter, 1985 ; Wilmot and Vetter, 1990 ; Windoffer and Giere, 1997 ; Vetter and Fry, 1998 ; Pflugfelder et al., 2005 ). The presence of such molecules could enable symbionts to maintain thioautotrophic growth and associated gene expression patterns without exogenous reduced sulfur. It is possible that the snails used here may have contained varying amounts of storage compounds due to differences in sulfide availability at the collection sites. Finally, the half-life of mRNA varies widely, generally from minutes to hours (Rauhut and Klug, 1999 ; Deana and Belasco, 2005 ), raising the possibility of a decoupling between transcript profiles and the inferred physiological outcomes linked to transcribed genes. However, the long duration during which snails were maintained in HPRS (8 h acclimation + 27–40 h incubation) makes it unlikely that our results reflect transcriptional variation due to differences in in situ conditions at the collection sites. Eight of the 28 Level 1 SEED subsystems differed significantly in expression (FDR < 0.05, Table 4 and Table S1 ). Sample clustering based on transcript representation in these 8 categories grouped high electron donor (HD) samples separate from low electron donor (LD) samples (Figure 4 ). In agreement with the higher carbon fixation rates measured directly in HD samples, transcripts associated with nitrogen metabolism, respiration, and nucleosides/nucleotides were more abundant in metatranscriptomes from HD treatments. Closer inspection of differentially expressed genes in these categories indicated that the availability of reduced sulfur compounds increased the representation of transcripts involved in nitrogen assimilation, ATP synthesis, cytochrome oxidase biogenesis, and dissimilatory nitrate reduction (Table S2 ). In contrast, LD incubations were enriched in transcripts involved in stress response, protein metabolism, phages/prophages/transposable elements/plasmids, clustering-based subsystems, and iron acquisition/metabolism (Figure 4 ). Figure 4 Heat map displaying hierarchical distance clustering of the eight SEED Subsystems differential expressed under HD vs. LD conditions based on normalized transcript abundance . Each row corresponds to a SEED subsystem, and each column to an individual experimental Ifremeria nautilei . Relative expression levels are indicated by colors shown in the scale at the top left. The dendrogram at the top shows the clustering of the individual snail samples. The dendrogram at left demonstrates the clustering of the SEED Subsystems. It should be noted that many proteins are listed in multiple SEED subsystems. For example, the chaperone protein DnaK is found in both protein metabolism and stress response subsystems. Other proteins are listed in seemingly unrelated subsystems, as is the case for the Fe-S cluster scaffold protein SufB, which is included in the phages/prophages/transposable elements/plasmids subsystem. Thus, drawing definitive conclusions from Level 1 SEED subsystem expression is difficult. Below, we use the results of the differential expression analysis of SEED functional proteins combined with supplemental manual queries of BLASTX results to describe in detail the expression patterns for key processes of symbiont physiology, especially those identified as differentially expressed in baySeq. Based on the 16S rRNA gene and transcript patterns described above, the vast majority of bacterial transcripts described are likely derived from the sulfur-oxidizing symbiont population. However, in the absence of genome data, we are not able to definitively localize all transcripts to a particular symbiont phylotype. S metabolism Sulfur oxidation genes typical of chemoautotrophic γ-proteobacteria were transcribed in all individuals (Figure 5 ). The canonical sulfur oxidation pathway in chemoautotrophic γ-proteobacteria is believed to start with the incomplete oxidation of sulfide or thiosulfate to elemental sulfur in the periplasm by the SoxABXYZ proteins (Ghosh and Dam, 2009 ). Through manual searches of BLASTX results (bit score > 50), we identified transcripts matching the s oxABHWXYZ genes (Table S4 ). Typical of γ-proteobacterial sulfur oxidizers, transcripts corresponding to the SoxCD proteins, which are necessary for complete oxidation of sulfide to sulfate by the Sox system, were not found, which may indicate that I. nautilei symbionts form elemental sulfur in the periplasm (Ghosh and Dam, 2009 ). SoxH, SoxK, and the thioredoxin SoxW were also detected in all metatranscriptomes and are commonly found in γ-proteobacterial sulfur oxidizers, but their roles in sulfide oxidation are unclear. In addition, manual searches of BLASTX results detected sulfide:quinone (oxido)reductase (Sqr) and sulfide dehydrogenase (Fcc) transcripts, both of which oxidize sulfide to polysulfides in the periplasm, although only Sqr has been shown to be necessary for sulfide oxidation (Dahl and Friedrich, 2008 ). The abundance of sqr, fcc , and s oxABHWXYZ transcripts was relatively uniform in all metatranscriptomes, demonstrating no differential expression based on metabolic activity or the concentration or type of electron donor (Tables S3 , S4 ). Figure 5 Expression patterns for key proteins involved in proposed sulfur oxidation, nitrogen reduction and assimilation, and carbon metabolism pathways in chemosynthetic Ifremeria nautilei symbionts . Proteins and complexes are colored based on transcript abundance in snail metatranscriptomes: green, constitutive expression (transcripts detected in all metatranscriptomes); yellow, HD expression > LD expression (FDR < 0.05); red, LD expression > HD expression (FDR < 0.05); white, sporadic or low expression (transcripts detected in fewer than 9 of the 12 metatranscriptomes); only proteins for which transcripts were detected are shown. Sulfur oxidation: Sox, Sox multi enzyme complex; AprAB, adenylylsulfate reductase; DsrL, Sulfur oxidation-associated protein DsrL; DsrEFH, putative sulfurtransferase complex; DsrC, putative bacterial heterodisulfide; DsrAB, reverse-type dissimilatory sulfite reductase; DsrJKMOP, sulfite reduction-associated complex DsrMKJOP; QmoABC/HdrABC, putative quinone-interacting membrane-bound oxidoreductase; RnfG, Rnf electron transport complex; Sat, sulfur adenylyltransferase; Sqr, sulfide quinone (oxido)reductase. N metabolism: Nap, periplasmic nitrate reductase; NirS, membrane-bound respiratory nitrite reductase; NirBD, NADH-dependent siroheme nitrite reductase; NorCB, nitric oxide reductase; NosZ, nitrous oxide reductase; GlutSyn, glutamine synthetase + glutamate synthase. TCA cycle: Pyr dehy, pyruvate dehydrogenase; Cit syn, citrate synthase; Acon hydr, Aconitate hydratase; Iso dehy, Isocitrate dehydrogenase; 2oxo dehy, 2-oxoglutarate dehydrogenase; Suc ligase, Succinyl-CoA ligase; Suc dehy, Succinate dehydrogenase; Fum hydr, Fumarate hydratase class I; Mal dehy, Malate dehydrogenase. Carbon fixation: RubisCO, Ribulose bisphosphate carboxylase; PGK, Phosphoglycerate kinase; GAPDH, NAD-dependent glyceraldehyde-3-phosphate dehydrogenase; TPI, Triosephosphate isomerase; FBA, Fructose bisphosphate aldolase; PfkA, Reversible pyrophosphate-dependent phosphofructokinase; HppA, membrane-bound proton-translocating pyrophosphatase; TKL, Transketolase; RPE, Ribulose-phosphate-3-epimerase; RPI, Ribose-5-phosphate isomerase A; PRK, Phosphoribulokinase; RuBP, Ribose-1,5-bisphosphate; 3PG, 3-phosphoglycerate; G3P, Glyceraldehyde-3-phosphate; DHAP, Dihydroxyacetone phosphate; FBP, Fructose-1,6-bisphosphate; F6P, Fructose-6-phosphate; X5P, Xylose-5-phosphate; Ru5P, Ribulose-5-phosphate; R5P, Ribose-5-phosphate; S7P, Sedoheptulose-7-phosphate; PPi, inorganic pyrophosphate; Pi, inorganic phosphate. Following the production of elemental sulfur, sulfur-oxidizing γ-proteobacteria use the reverse dissimilatory sulfite reduction pathway (rDSR) to produce sulfite, followed by sulfite oxidation to sulfate by APS reductase (AprAB) and sulfate adenylytransferase (Sat; Dahl et al., 2005 ). Both MEGAN SEED protein classification and manual searches of BLASTX results showed that DSR genes dsrABCEFHJKLMOPNR were similarly represented in all metatranscriptomes (Tables S3 , S4 and Figure 5 ). Transcription of aprAB , encoding the subunits of APS reductase, was correlated with sulfur metabolism rates ( R 2 = 0.99), and aprAB was the only sulfur metabolism gene identified by the baySeq model as significantly up-regulated in HD incubations (FDR = 0.006; Table S2 ). Multiple studies of free-living and endosymbiotic dissimilatory sulfur-metabolizing bacteria have found that AprAB expression is high in relation to other proteins (>6% of total proteome; Markert et al., 2007 , 2011 ) and correlated with the energy state of the cell (Zhang et al., 2006 ; Pereira et al., 2008 ). Transcription of the genes encoding the quinone-interacting membrane-bound oxidoreductase QmoABC complex, which is proposed to transfer electrons from AprAB to the periplasmic quinone pool, was sporadic among treatments (Figure 5 ). However, both SEED identifications and manual searches showed that all metatranscriptomes contained abundant sequences encoding heterodisulfide reductase HdrABC, which may be homologous to the QmoABC complex (Ramos et al., 2012 ). Indeed, genomic analysis of Thiolapillus brandeum strain Hiromi 1, a free-living strain closely related to I. nautilei sulfur-oxidizing symbionts, identified a chimeric cluster of qmoAB-hdrBC genes (Nunoura et al., 2014 ) and metagenomic/proteomic analysis of Riftia pachyptila γ-proteobacterial symbionts found abundant proteins annotated as heterodisulfide reductases with high similarity to Chlorobium tepidum QmoAB proteins (Markert et al., 2011 ). Thus, differentiation of these homologous protein complexes is problematic. In addition, the nature of electron transfer from AprAB to the periplasm is unclear and may involve heterodisulfide reductase and flavin-based electron bifurcation in order to conserve energy (Ramos et al., 2012 ). Evidence for expression of multiple proteins of the Rnf electron transport complex, which could couple with HdrABC to generate an ion motive force and thus ATP (Biegel et al., 2011 ; Ramos et al., 2012 ; Buckel and Thauer, 2013 ), was also observed in all symbiont metatranscriptomes (Figure 5 and Tables S3 , S4 ), with transcripts for RnfG significantly more abundant in HD vs. LD incubations (FDR = 0.04; Table S2 ). In summary, we observed expression of the full suite of expected sulfur oxidation genes in I. nautilei symbiont metatranscriptomes; however, expression of most sulfur oxidation genes, excluding AprAB, was not impacted by electron donor concentration or type. This result could be explained by several factors, including (i) stability and persistence of sulfur-oxidation transcripts produced prior to our experiments (i.e., during in situ activity), (ii) oxidation of elemental sulfur molecules stored by the symbionts to maintain thioautotrophic growth in the absence of sufficient exogenous reduced sulfur, or (iii) constitutive expression of most symbiont sulfur oxidation genes. Of these, the latter option may be most likely given that the predicted half-life of most mRNA molecules is considerably shorter than the duration of our experiments (Rauhut and Klug, 1999 ), and that there is no evidence that I. nautilei produce sulfur storage granules (Windoffer and Giere, 1997 ; Borowski et al., 2002 ). Rather, it has been suggested that indispensable metabolic pathways in bacteria are often constitutively expressed and are not repressible (Salmon et al., 2003 , 2005 ). Thus, if I. nautilei symbionts are obligate sulfur oxidizers and cannot oxidize other electron donors, it is not surprising that sulfur oxidation genes are transcribed constitutively. Here, APS reductase (AprAB) and RnfG were the only sulfur oxidation genes whose transcription increased in response to reduced sulfur, suggesting that these proteins are tightly regulated by environmental conditions and are important for electron transport and energy conservation when either sulfide or thiosulfate is abundant. Indeed, other studies have also found that the transcription of AprAB is relatively responsive to environmental conditions compared to that of other sulfur oxidation genes (Zhang et al., 2006 ; Markert et al., 2007 ; Wendeberg et al., 2012 ). AprAB may therefore be a particularly good indicator of sulfur oxidation activity. Carbon metabolism Previous work has shown that most γ-proteobacterial chemoautotrophic symbionts fix carbon via the Calvin-Benson-Bassham (CBB) cycle (Woyke et al., 2006 ; Duperron et al., 2007 ; Markert et al., 2007 ; Kleiner et al., 2012a ; Sanders et al., 2013 ). Both SEED classification in MEGAN and manual searches of BLASTX data detected transcripts corresponding to all CBB cycle enzymes, although expression levels depended on the enzyme and incubation conditions (Figure 5 and Tables S3 , S4 ). Furthermore, transcripts encoding the three key enzymes of the rTCA cycle (ATP citrate lyase, 2-oxoglutarate oxidoreductase, and fumarate reductase) were not detected, indicating that I. nautilei symbionts use only the CBB cycle for carbon fixation. Based on SEED classifications, form II Ribulose-1,5-bisphosphate carboxylase/oxygenase (RubisCO) was the most highly transcribed CBB cycle gene and the only CBB gene significantly more abundant in HD incubations (Table S2 and Figure 5 ). RubisCO transcript abundances and average 13 C incorporation rates were both substantially higher in HD vs. LD incubations, although individual values were only moderately correlated ( R 2 = 0.24), suggesting that post-transcriptional regulation, redox poise, or other factors may control activity of the RubisCO enzyme and carbon incorporation rates (Beinart et al., 2015 ; Table 2 ). One individual in particular, snail HS3, exhibited a relative high abundance of RubisCO transcripts, but did not incorporate appreciable levels of 13 C. Form II RubisCO (CbbM), the only form detected in our data, is sensitive to oxygen and is hypothesized to function best in environments with low O 2 and high CO 2 concentrations (Tabita et al., 2008 ). However, it is thought that cbbM expression is unlikely to be directly regulated by CO 2 concentration but may instead be regulated by redox conditions (Badger and Bek, 2008 ; Tabita et al., 2008 ; Alfreider et al., 2012 ). Our data are consistent with the hypothesis that cbbM transcription is induced when oxygen is low, and suppressed when oxygen concentrations are high, such as likely to be the case in the LD treatments if reduced access to electron donor decreases the symbionts' requirement for oxygen. Thus, transcription levels may be tightly linked to redox conditions experienced by the symbionts. Further investigation on the regulation of RubisCO in symbionts is needed to understand how this critical enzyme is regulated and optimized for conditions within the host. Transcripts matching phosphoribulokinase, fructose-bisphosphate aldolase, transketolase, and ribulose-5-phosphate isomerase were detected in all 12 metatranscriptomes, whereas transcripts for other CBB cycle enzymes were detected at lower levels and more sporadically among datasets (Table S3 ). As has been reported for symbionts of deep-sea tubeworms ( Riftia pachyptila ), clams ( Calyptogena magnifica ), and shallow water gutless marine worms, we did not detect transcription of the genes encoding sedoheptulose-1,7-bisphosphatase and fructose-1,6-bisphosphatase, which are necessary to regenerate ribulose-1,5-bisphosphate (Markert et al., 2007 ; Newton et al., 2007 ; Kleiner et al., 2012b ). However, manual searches of BLASTX results detected transcription of genes for reversible pyrophosphate-dependent phosphofructokinase (PfkA) and a membrane-bound proton-translocating pyrophosphatase (HppA), which together have been proposed to replace sedoheptulose-1,7-bisphosphatase and fructose-1,6-bisphosphatase in some chemolithotrophic symbionts (Table S4 ). Using reversible PfkA in combination with HppA in the CBB cycle reduces the net ATP consumption of carbon fixation as the membrane-bound proton-translocating pyrophosphatase establishes a proton gradient across the cytoplasmic membrane, allowing for ATP generation by ATP synthase (Reshetnikov et al., 2008 ; Kleiner et al., 2012a , b ). In support of this hypothesis, PfkA enzyme activity was detected in strain Hiromi 1, a strain closely related to I. nautilei sulfur-oxidizing symbionts, pfkA is conserved in all complete genomes of sulfur-oxidizing γ-proteobacteria available in public databases, and pfkA:hppA are co-localized in the genomes of many free-living and symbiotic chemoautotrophic bacteria (Kleiner et al., 2012a , b ; Nunoura et al., 2014 ). Evidence for expression of bacterial oxidative TCA enzymes was observed in all metatranscriptomes, and transcripts corresponding to citrate synthase, 2-oxoglutarate dehydrogenase, and malate dehydrogenase were significantly more abundant in LD vs. HD treatments based on the baySeq differential expression analysis (Table S2 ). In addition, transcripts encoding two enzymes involved in the biosynthesis of lipoyl cofactors (octanoate-[acyl-carrier-protein]-protein-N-octanoyltransferase and lipoate synthase) essential for the activation of oxidative TCA cycle enzymes and a long chain fatty acid coenzyme A ligase involved in activation of fatty acid breakdown were also significantly enriched in LD metatranscriptomes (Table S2 ). Activation of fatty acids and expression of oxidative TCA cycle enzymes when electron donors are limiting could provide a number of advantages for symbiont survival. Incorporation of fatty acids and other organic carbon through the TCA cycle is less energy intensive than fixing inorganic carbon and also produces NADPH, which presumably would be in short supply when electron donors are not available (Nunoura et al., 2014 ). However, net heterotrophic growth requires uptake of organic compounds, and we did not find evidence for the expression of genes involved in the transport of organic compounds into the cell (Kleiner et al., 2012b ). Alternatively, increased expression of TCA cycle enzymes could be driven by the oxidizing conditions in the LD incubations, as has been observed in multiple studies in Escherichia coli (Salmon et al., 2003 , 2005 ; Shalel-Levanon et al., 2005 ; Toya et al., 2012 ). It is not clear from our data whether I. nautilei symbionts are performing net heterotrophy or are using internal carbon molecules such as fatty acids and glycogen as a source of carbon and reducing equivalents when electron donors are limited. Free-living strains related to I. nautilei symbionts have been observed to incorporate complex organic carbon sources such as fumarate, formate, citrate, pyruvate, and peptone (Sievert and Vetriani, 2012 ; Nunoura et al., 2014 ). Our data raise the possibility that I. nautilei symbionts are mixotrophic, fixing inorganic carbon when reduced sulfur is available, but switching to chemoheterotrophy when energy supplies are limiting. Nitrogen metabolism Inorganic nitrogen compounds are generally abundant at hydrothermal vents, while organic nitrogen compounds are scarce (Johnson et al., 1986 ). Vent fluids may contain ammonium, which can be assimilated by many microorganisms. However, nitrate is typically more abundant and can be used as both an oxidant for energy generation (respiration) and a primary nitrogen source for assimilation. Both processes begin with nitrate reduction to nitrite, but proceed through different enzymatic pathways (Potter et al., 1999 ; Klotz and Stein, 2008 ). The next step in nitrogen assimilation is the cytoplasmic reduction of nitrite to ammonia by the complex NirBD and assimilation by glutamine synthetase/glutamate synthase or asparagine synthase. In the dissimilatory pathway, nitrite may be further reduced to nitrogen gas (denitrification) in three steps: nitrite reduction by NirS or NirK, nitric oxide reduction by the NorCB complex, and finally nitrous oxide reduction by NosZ. Alternatively, nitrite can be reduced to ammonium by Nrf to generate energy in a process of dissimilatory nitrate reduction to ammonia (Potter et al., 1999 ; Klotz and Stein, 2008 ). Metatranscriptomes from HD incubations were significantly enriched in transcripts encoding proteins of both assimilatory and dissimilatory (respiratory) nitrate reduction and ammonium assimilation compared to LD treatments (Figure 5 and Table S2 ). Interestingly, no transcripts corresponding to the membrane-bound dissimilatory Nar or cytoplasmic assimilatory Nas nitrate reductases were identified. We only found transcripts encoding the periplasmic nitrate reductase NapCFGH proteins, which may be used in either assimilation or respiration, and in some cases may serve as an electron shunt when electron flow through the aerobic respiratory chain is restricted due to low oxygen concentrations (Potter et al., 1999 ; Klotz and Stein, 2008 ). The relative abundance of Nap transcripts was on average over 20X higher in HD compared to LD conditions (Table S3 ). Transcripts encoding a complete suite of respiratory denitrification [ nirS, norCB , and nosZ ] and cytoplasmic nitrogen assimilation [nirBD, glutamine synthetase-glutamate synthase] enzymes were also significantly enriched in HD metatranscriptomes (Table S2 ). These results suggest that when sufficient electron donor is available I. nautilei symbionts reduce nitrate to nitrite in the periplasm and subsequently use it for respiration (denitrification) in the periplasm or transport the nitrite to the cytoplasm for assimilation (Figure 5 ). Metatranscriptomic analysis of sulfur-oxidizing symbionts of the vent snail Alviniconcha found evidence for the same nitrate assimilation and denitrification pathways reported here (Sanders et al., 2013 ). Furthermore, a recent study showed that a close free-living relative of the I. nautilei symbiont coupled complete denitrification to sulfur oxidation and reached higher cell densities with nitrate as electron acceptor compared to oxygen (Nunoura et al., 2014 ). Using the periplasmic Nap complex for nitrate reduction provides considerable adaptability in nitrogen metabolism and respiration in response to dynamic nitrogen concentrations and redox conditions. After reduction of nitrate to nitrite in the periplasm, nitrite can be transported into the cytoplasm only when it is needed for assimilation, keeping cytoplasmic nitrite concentrations low and thus avoiding nitrite toxicity (Rowe et al., 1979 ). Additionally, unlike the respiratory Nar proteins, the Nap complex is not inhibited by oxygen (Potter et al., 1999 ). Consequently, the Nap complex can be expressed under aerobic conditions, potentially allowing symbionts to use nitrate as an electron acceptor and avoid competition with the host for oxygen (Sanders et al., 2013 ). Our results confirm that sulfur-oxidizing symbionts tightly regulate the expression of nitrogen assimilation and denitrification genes in response to reduced sulfur and oxygen availability. Stress response Symbiont populations in LD incubations express genes that are consistent with increased oxidative stress (Chen et al., 2009 ). Of 19 protein-coding genes identified as significantly more abundant in LD incubations using the baySeq model, 10 were related to the bacterial stress response (Table S2 ). In bacteria, a highly conserved “heat-shock” stress response occurs in response to heat, starvation, radiation, and oxidative agents (Neidhardt et al., 1984 ). The heat-shock response involves production of a set of heat-shock proteins, many of which are molecular chaperones involved in transit across membranes, targeted proteolysis, and polypeptide folding (Genevaux et al., 2007 ). Stress-induced transcription of heat-shock genes is typically controlled by binding of the sigma factor RpoH to RNA polymerase (Bukau, 1993 ). Based on SEED protein classification in MEGAN, metatranscriptomes from LD incubations contained 9X more RpoH-related transcripts than those from HD incubations, indicating that the heat-shock response was induced in symbionts without sufficient electron donor in the presence of oxygen (Table S2 ). In accordance with an increase in RpoH, transcripts encoding proteins in the DnaK chaperone system and the Clp/Hsp100 family of ATP-dependent protein remodeling machines also increased in LD incubations, notably representing the most abundant transcripts in ND datasets. The DnaK chaperone system is composed of DnaK, DnaJ, and GrpE, which work together with Clp/Hsp100 family proteins to catalyze protein disaggregation and refolding or protein degradation during physiologic stress (Dougan et al., 2002 ). Here, DnaK, DnaJ, and GrpE transcripts were significantly enriched in LD incubations, although DnaJ and GrpE transcripts were 10 to 100X less abundant than DnaK (Tables S2 , S3 ). Transcripts encoding ClpA and ClpX, as well as FtsH subunits and peptidases associated with ATP-dependent protein degradation (Dougan et al., 2002 ), were also significantly enriched under LD conditions (Table S2 ). Other subunits associated with the Clp/Hsp100 family, such as ClpB and ClpS, were highly expressed in LD incubations, but were not significantly differentially expressed under the criteria of the baySeq model (Table S3 ). Transcription of genes involved in iron regulation and 4Fe-4S cluster biosynthesis were also significantly more abundant under LD conditions, potentially due to the importance of iron homeostasis for cell survival during oxidative stress (Table S2 ). Ferrous iron reacts with the hydrogen peroxide present during oxidative stress to form damaging peroxide radicals. The rate of oxidative damage is therefore elevated when cytoplasmic iron concentrations are high (Imlay, 2013 ). Transcripts related to the MerR family of metalloregulators, which detect and respond to reactive metals and changes in redox conditions, were significantly enriched in LD metatranscriptomes. These included transcripts encoding the oxidative stress metalloregulator SoxR, which is known to induce expression of multiple proteins that mitigate cellular damage, including superoxide dismutase and the regulatory Fur protein (Brown et al., 2003 ; Imlay, 2013 ). Both the Fur protein, which suppresses iron import and utilization, and superoxide dismutase, were significantly more abundant in LD incubations (Table S2 ), suggesting that I. nautilei symbionts without sufficient electron donating compounds were attempting to limit iron uptake and neutralize superoxide to avoid oxidative damage. Proteins containing 4Fe-4S centers and mononuclear iron also are very susceptible to damage from reactive oxygen species (Imlay, 2013 ). Transcripts from two Fe-S cluster assembly pathways, the Isc (iron sulfur cluster) system and the Suf (sulfur formation) system, were detected in all snail metatranscriptomes (Table S3 ). The Suf system has been found to operate specifically to protect and assemble Fe-S clusters during oxidative stress (Ayala-Castro et al., 2008 ). In accordance with this suggested function, transcripts encoding a key scaffold protein (SufB) in the Suf system were on average 20X more abundant in LD vs. HD incubations (Table S2 ). LD transcriptomes were also significantly enriched in transcripts for the DNA repair enzyme RecA (Imlay, 2013 ) and the stress-induced morphogene BolA (Santos et al., 1999 ). Together, these results indicate significant oxidative stress in symbiont populations without sufficient electron donor, suggesting that I. nautilei cannot shield its symbionts from fluctuations in redox conditions that may negatively impact symbiont survival. Methane (C1) metabolism Despite a presumed lack of exogenous methane in these experiments, we detected transcripts indicating that methanotroph-related I. nautilei symbionts were active during the incubations. Phylogenetic analysis of 16S rRNA transcripts and amplified rRNA genes placed the methanotroph-related I. nautilei symbionts in a monophyletic group within the γ-proteobacterial family Methylococcaceae (Figure 2 ) . Methylococcaceae are type I methanotrophs typically characterized by the use of the ribulose monophosphate (RuMP) pathway for incorporation of C1 carbon into biomass and a membrane-associated particulate methane monooxygenase (pMMO) that catalyzes the oxidation of methane to methanol (Chistoserdova, 2011 ). Manual searches of BLASTX files revealed transcripts matching the pMMO genes pmoCAB , accounting for 0.1-1.5% of non-rRNA reads in the TS, LS, and ND datasets (Table S4 ). Recovered pMMO sequences shared >85% amino acid similarity to pMMO proteins from Methylomicrobium album, Methylomicrobium buryatense , and Methylomarinum vadi , supporting the phylogenetic placement of these symbionts within the Methylococcaceae , but not specifically related to methanotrophic symbionts of deep-sea bathymodiolin mussels. No transcripts indicative of sMMO were detected (Hanson and Hanson, 1996 ; Chistoserdova, 2011 ). PmoC-related reads represented the most abundant methanotroph-related functional gene transcript observed in the TS, LS, and ND metatranscriptomes, and were 5–10X more abundant than PmoA or PmoB transcripts. Two previous transcriptomic studies of alpha-proteobacterial methanotroph pure cultures observed that the pmoCAB operon is the most highly expressed operon in the genome when sufficient copper is present, with pmoC gene expression 6–7X higher than pmoA/B (Matsen et al., 2013 ; Vorobev et al., 2014 ). Evidence for the further oxidation and assimilation of methane was also detected, although transcripts from these pathways were one to two orders of magnitude less abundant than pMMO-related transcripts (Table S4 ). Transcripts encoding methanol dehydrogenase proteins (MDH), which catalyze the oxidation of methanol to formaldehyde (Chistoserdova, 2011 ), were present in three of the nine methanotroph-containing metatranscriptomes. In addition, six metatranscriptomes contained transcripts matching methanotroph-related genes for the biosynthesis of the methanol dehydrogenase cofactor pyrroloquinoline quinone. Expression of unique genes for formaldehyde oxidation and assimilation through the RuMP pathway (3-hexulose-6-phosphate synthase and 3-hexulose-6-phosphate isomerase) were also detected in six of the nine methanotroph-containing metatranscriptomes (Table S4 ). Collectively, these results confirm the transcriptional activity of a methanotrophy pathway in I. nautilei symbionts. The presence of methanotrophy-related transcripts despite a lack of C1 compounds in these experiments is surprising as no methane was introduced into these experiments, and the chance of methane being present in any of the injected mixed gases is very low. Our data may therefore be explained by a constitutive transcription of methanotrophy genes, or a relatively high stability of methanotrophy transcripts. Indeed, both constitutive expression and high transcript stability have been reported for the pmoCAB operon (Lieberman and Rosenzweig, 2004 ; Deana and Belasco, 2005 ; Chen et al., 2008 ; Wendeberg et al., 2012 ; Matsen et al., 2013 ). Alternatively, putatively methanotrophic I. nautilei symbionts could be growing on substrates other than methane, such as methylated sulfur compounds or even multi-carbon substrates. However, this explanation seems unlikely, as all Methylococcaceae characterized to date are obligate methanotrophs (Chistoserdova, 2011 ). The functional contributions of I. nautilei methanotrophs to holobiont fitness remain to be determined."
} | 14,891 |
34745072 | PMC8569243 | pmc | 8,558 | {
"abstract": "Biofilm formation within the process of bioleaching of copper sulfides is a relevant aspect of iron- and sulfur-oxidizing acidophilic microorganisms as it represents their lifestyle in the actual heap/dump mining industry. Here, we used biofilm flow cell chambers to establish laminar regimes and compare them with turbulent conditions to evaluate biofilm formation and mineralogic dynamics through QEMSCAN and SEM-EDS during bioleaching of primary copper sulfide minerals at 30°C. We found that laminar regimes triggered the buildup of biofilm using Leptospirillum spp. and Acidithiobacillus thiooxidans (inoculation ratio 3:1) at a cell concentration of 10 6 cells/g mineral on bornite (Cu 5 FeS 4 ) but not for chalcopyrite (CuFeS 2 ). Conversely, biofilm did not occur on any of the tested minerals under turbulent conditions. Inoculating the bacterial community with ferric iron (Fe 3+ ) under shaking conditions resulted in rapid copper recovery from bornite, leaching 40% of the Cu content after 10 days of cultivation. The addition of ferrous iron (Fe 2+ ) instead promoted Cu recovery of 30% at day 48, clearly delaying the leaching process. More efficiently, the biofilm-forming laminar regime almost doubled the leached copper amount (54%) after 32 days. In-depth inspection of the microbiologic dynamics showed that bacteria developing biofilm on the surface of bornite corresponded mainly to At. Thiooxidans , while Leptospirillum spp. were detected in planktonic form, highlighting the role of biofilm buildup as a means for the bioleaching of primary sulfides. We finally propose a mechanism for bornite bioleaching during biofilm formation where sulfur regeneration to sulfuric acid by the sulfur-oxidizing microorganisms is crucial to prevent iron precipitation for efficient copper recovery.",
"conclusion": "Conclusion The main copper minerals worldwide correspond to primary sulfides chalcopyrite and bornite, so it is crucial to understand the microbial variations affecting their efficient copper extraction. In this sense, this work shows that the standard approaches of bioleaching tests under stirring conditions prevent biofilm formation on bornite and chalcopyrite at 30°C. In addition, we determine that biofilm development is not feasible on chalcopyrite even under laminar flow mesophilic conditions, posing a great challenge for copper hydrometallurgy. Successfully, the biofilm-forming laminar regime improves almost twice the leached copper content of bornite compared with stirring conditions in the presence of ferrous iron (Fe 2+ ), reducing the time from 48 to 32 days of cultivation. For the first time, we demonstrated that the main microorganisms comprising the biofilm development on bornite are sulfur-oxidizing bacteria, while the ferrous-oxidizing bacteria are in planktonic form. This lifestyle of acidophiles explains the accelerated pace of copper recovery during biofilm formation, where the action of sulfur-oxidizing bacteria maintains the low pH via sulfuric acid formation, thus, preventing iron precipitation. The next step toward sustainable copper hydrometallurgy is to develop a process that promotes biofilm formation on the surface of chalcopyrite under mesophilic or moderate thermophilic conditions to extract copper from this bulk material efficiently, pointing to the creation of novel acidophilic microbial consortia.",
"introduction": "Introduction One of the most important challenges that face the copper industry is the development of sustainable technologies to leach complex ores composed of mixed copper mineral species in low grades, with a prevalence of primary copper sulfides. The major component of these materials is usually chalcopyrite (CuFeS 2 ), followed by bornite (Cu 5 FeS 4 ) ( Panda et al., 2015 ). To this end, bioleaching has emerged as an efficient technology to leach secondary sulfides under mesophilic conditions (ambient temperature) ( Bustos et al., 1993 ) and primary sulfides mostly under thermophilic conditions above 60°C ( Duarte et al., 1993 ) or at moderate thermophilic (45–50°C) in stirred tank bioreactors ( Cancho et al., 2007 ; Hedrich et al., 2018 ). Recently, acidophilic bacterial consortia recovered 43% of copper from chalcopyrite using shake flasks at room temperature (25–30°C) ( Ma et al., 2018 ). In another study utilizing a bioleaching column coupled with a fuel cell, the system yielded 244 mg L –1 of copper from 150 g of chalcopyrite at 30°C in 320 days ( Huang et al., 2019 ). In the case of bornite, studies have shown bioleaching activity under mesophilic conditions compared with abiotic controls ( Jun et al., 2008 ; Bevilaqua et al., 2010 ). Besides, researchers have proposed a galvanic effect with chalcopyrite accelerating the oxidative dissolution of bornite in the presence of microorganisms ( Wang et al., 2016 ). Inoculation of the moderate thermophilic Leptospirillum ferriphilum and Acidithiobacillus caldus in shaking flasks at 45°C enabled the conversion of bornite into several intermediates including covellite (CuS) and isocubanite (CuFe 2 S 3 ) through the bioleaching process, with a final copper extraction of 70% ( Hong et al., 2019 ). Finally, several works have established the high bioleaching efficiency of thermophiles over chalcopyrite ( Petersen and Dixon, 2002 ), but none of them uncovered the role of biofilm formation in this process. Recently, a study reported biofilm development over chalcopyrite under moderate thermophilic conditions, where L. ferriphilum was inoculated at a high concentration (above 10 8 cells/g mineral) in the presence of under-stoichiometric iron concentrations ( Bellenberg et al., 2018 ). A few studies have specifically addressed mesophilic biofilm formation on chalcopyrite, with researchers postulating that high precipitation of jarosite and elemental sulfur restrict biofilm development ( Lei et al., 2009 ), while others identified microcolonies of At. ferrooxidans and L. ferriphilum forming a monolayered biofilm under short contacting periods ( Africa et al., 2010 ). Here, we demonstrated that mesophilic inoculation of bacteria with iron- and sulfur-oxidizing capacity in a flow cell chamber enables the detailed study of biofilm development on bornite resembling heap/dump bioleaching conditions when assuming a homogeneous packed bed behavior, where liquid percolation is driven by gravity through the ore particle surface. The biofilm-promoting laminar regime improved the copper leaching rate compared with stirred conditions in the presence of ferrous iron (Fe 2+ ) and highlighted the importance of biofilm formation and its microbial composition to explain the mode of action and lifestyle of acidophiles carrying iron and sulfur oxidization during bornite bioleaching for copper extraction.",
"discussion": "Results and Discussion Mesophilic bioleaching has been used at an industrial scale for secondary copper sulfide ores, but not for the primary ones like chalcopyrite, given the low copper recovery and slow kinetics ( Córdoba et al., 2008 ). For low-grade primary copper sulfide ores, a few industrial bioleaching applications are in the advanced stages to make the copper extraction economically feasible, to some extent, because of our inability to capture the lifestyle and growth development of the leaching microorganisms during copper recovery from actual primary ores. Thus, we investigated the capabilities of Leptospirillum spp. and At. thiooxidans to form biofilm on primary copper sulfides utilizing a standardized and reproducible flow cell approach that resembles heap/dump bioleaching conditions. It is important to highlight that most studies today employed primary copper sulfide high-purity minerals with concentrates added at low pulp densities and high cell loads in shaking flasks ( Corkhill et al., 2008 ; Wang et al., 2016 ; Bellenberg et al., 2018 ). These growth conditions are optimal for homogeneous mixing and planktonic microbial growth, but certainly not for biofilm development ( Tolker-Nielsen and Sternberg, 2011 ). First, we set the screening assay under turbulent flow regimes to mimic shaking flask liquid homogeneous conditions ( Figure 1A ). On the other hand, the method for studying bacterial biofilm development uses cell flow chambers under laminar regimes ( Tolker-Nielsen and Sternberg, 2011 ; Crusz et al., 2012 ). Figure 1B depicts the components that comprise the cell chambers to study metal-leaching acidophiles. We found that the bubble trap is essential for attaining a homogeneous flow and proper propagation of acidophilic microorganisms ( Crusz et al., 2012 ). We next placed on top of the cell chamber, in various sets of experiments, high-purity natural primary copper sulfide minerals to evaluate biofilm formation and copper extraction and finally compare it with turbulent conditions. Importantly, the minerals were not cleansed by acid solutions (HCl) nor autoclaved since these procedures alter the surface of the material, provoking initial leaching of copper and eliminating native microorganisms inhabiting these primary copper sulfide materials ( Cerda et al., 2018 ). FIGURE 1 Diagrams showing the flow chamber for biofilm formation and the screening assay under shaking conditions. (A) Screening procedure scheme to bioleach copper from primary copper sulfides in mesophilic conditions (30°C) under shaking using six-well plates with inoculation of bacterial acidophiles. (B) Flow cell model and assembly diagram of bioleaching copper from primary copper sulfides under laminar regimes and buildup biofilm composed of acidophilic bacterial consortium growing at 30°C. Chalcopyrite Bioleaching Mineralogical analyses were done on the head mineral to confirm chalcopyrite purity ( Figure 2A , time 0) and at the end of the bioleaching assay to record possible transformations ( Figure 2A , 30 and 60 days). Chalcopyrite inspection before bioleaching by QEMSCAN-BMA indicated a purity of 70% (previously estimated over 90% by optical microscopy), with a significant 16% of unclassified minerals ( Figure 2A ), with a clean surface as observed by SEM ( Figure 2C ). Under shaking conditions, precipitates completely covered the chalcopyrite surface after 30 days of bioleaching, partially classified as potassium and/or sodium jarosite based on the EDS elemental identification ( Figure 2D ). This was not the case for chalcopyrite bioleached under laminar regimes as the micrographs show low precipitates, and EDS analysis remained constant ( Figure 2E ). FIGURE 2 QEMSCAN-BMA analysis of high-purity natural chalcopyrite before bioleaching (t0) and after flow cell bioleaching assay (30 days) for (A) turbulent and (B) laminar conditions. (C) SEM image and EDS spectrum of high purity chalcopyrite before bioleaching. SEM images and EDS spectra of high-purity chalcopyrite after 30 days of (D) turbulent (screening method) and (E) laminar flow mesophilic bioleaching (flow cell method). Low CR: no mineral detection. Using the laminar flow chamber, the final mineralogical analysis showed significant variations with an increment in chalcopyrite proportion to 78% and reduced unclassified material to 11%, probably indicating the removal of initially present impurities from the mineral surface. On the contrary, in the turbulent flow regime, the mineralogical analysis was almost unable to identify chalcopyrite at the end of the assay (only 2% identification at day 30), while the unclassified raised to 94% attributed mainly to not identified mixed phases ( Figure 2A , 30 and 60 days), indicating major mineral surface coverage by formed precipitates including potassium and/or sodium jarosite as indicated by EDS ( Figure 2D ). The screening bioleaching assay and the laminar flow condition showed a low copper recovery of 6% in 30 days incubated at 30°C and maintaining this copper level until day 60 ( Figures 2A,B ). Biofilm formation did not occur under turbulent or laminar flow regimes ( Figures 2D,E ). These results are in agreement with those previously published by Lei et al. (2009) . Some studies describe biofilm buildup on the altered surface of the chalcopyrite in a short period where chalcopyrite was either previously electro-oxidized resulting in S 0 , CuS, and S 2– that promote bacterial attachment ( García-Meza et al., 2013 ), or crushed down to a size of less than 80 μm and finally polished ( Yang et al., 2015 ). Other works using batch stirred bioreactors set at 42°C and inoculated with L. ferriphilum , Sulfobacillus spp., and A. caldus showed 94% extraction of the available copper in the concentrate, where chalcopyrite remained intact during the leaching process (6% of the remaining Cu) ( Spolaore et al., 2011 ; Hedrich et al., 2016 ). This confirms that even at moderate thermophilic conditions and employing a different acidophilic consortium from this study, it is challenging to leach the copper content from this primary sulfide mineral. Bornite Bioleaching Considering that chalcopyrite is commonly associated with bornite in nature and that a galvanic couple between these two minerals has been described ( Wang et al., 2016 ), the next step was to analyze the mesophilic bioleaching process correlated with highly pure bornite. We first inoculated the same proportion of the acidophilic bacterial consortium used for chalcopyrite along with the addition of ferric iron (0.7 g Fe 3+ /L), which enabled the rapid recovery of copper from bornite (40%) after 10 days of cultivation under stirring conditions ( Figure 3A ). Previous studies showed accelerated bornite leaching in the presence of external ferric iron addition ( Hidalgo et al., 2019 ), and on the contrary, the screening bioleaching assays carried out with the addition of ferrous iron (0.7 g Fe 2+ /L) exhibited a copper extraction close to 30% after 48 days at 30°C ( Figure 3B ). In parallel, the laminar regime leached 54% of the Cu content at day 32, with no further Cu recovery until day 48 ( Figure 3C ), illustrating the importance of the process set up and bacterial attachment to accelerate copper bioleaching from bornite at 30°C. FIGURE 3 Mineralogical dynamics, SEM images, and EDS spectra of high-purity natural bornite bioleaching under mesophilic (30°C) inoculation. (A) Turbulent flow bioleaching with addition of ferric iron (screening method), (B) turbulent flow bioleaching with addition of ferrous iron (screening method), (C) laminar flow bioleaching (flow cell method). SEM images and EDS spectra for (D) bornite before bioleaching, (E) bornite after 10 days (40% Cu recovery) under turbulent flow mesophilic bioleaching (screening method), (F) bornite after 48 days (30% Cu recovery) under turbulent flow mesophilic bioleaching with addition of ferrous iron (screening method), (G) bornite after 46 days (54% Cu recovery) under laminar flow mesophilic bioleaching (flow cell method). Red, orange, and green arrows indicate exopolysaccharide (EPS) network, altered mineral, biofilm formation, respectively. Low CR: no mineral detection. QEMSCAN-BMA recorded 32% bornite purity before bioleaching (estimated at 75% by optical microscopy) with an unclassified mineral fraction close to a third (time 0 in Figures 3A–C ). The mineralogical dynamics of mesophilic bioleaching indicated in both regimens (turbulent and laminar flow) the intermediate formation of idaite (Cu 3 FeS 4 ) in the first stage of bioleaching. Chalcopyrite detection was attributed to ferric ion precipitation masking further bornite/idaite detection ( Figures 3A–C ), though former studies consider it as part of the bornite (bio)leaching mechanism ( Pal-ing, 2018 ). Turbulent and laminar flow regimes were initially inoculated including external ferrous iron (Fe 2+ ) addition; in the turbulent regime, only a slight iron-oxidizing activity was observed until day 20 ( Figure 4A ) with a concomitant microbial growth, stopped by a significant pH increase (pH 4.25, data not shown) and consequent iron precipitation ( Figure 4A ). Interestingly, biofilm formation was observed only under laminar flow conditions, with a significant alteration in the bornite surface and an exopolysaccharide (EPS) network ( Figure 3G , indicated with red arrows), compared with turbulent flow conditions where no biofilm formation was encountered with the initial addition of ferric nor ferrous iron ( Figure 3F ). Biofilm development mediated initially by EPS production during bioleaching depends on several factors including mineral type ( Gehrke et al., 1998 ), process temperature ( Zeng et al., 2010 ; Barahona et al., 2014 ), and the presence of organic compounds such as D-galactose ( Saavedra et al., 2020 ) and sodium glucoronate ( Bellenberg et al., 2015 ). For instance, pyrite is an excellent sulfide mineral for biofilm formation under acidic conditions as a strong electrochemical interaction occurs between the synthesized EPS-Fe 3+ and the negatively charged mineral surface ( Gehrke et al., 1998 ; Bellenberg et al., 2015 ). It differs highly compared with chalcopyrite, where EPS synthesis occurs only when inoculated at a high cell density ( Lei et al., 2009 ; Yang et al., 2015 ; Bellenberg et al., 2018 ), with some studies using more than 10 9 cells/g mineral as initial biomass ( Lei et al., 2009 ). Unfortunately, this biomass loading is impractical at the industrial mining scale for cooper bioleaching of primary sulfur-bearing ores. It is clear that further investigation aiming at promoting biofilm development on chalcopyrite at mesophilic conditions is necessary under different flow regimes, varying the type and abundance of acidophilic consortia, resembling industrial operations, and taking care of the mineral characteristics, initial cell density, and avoiding major mineral pretreatments to make the process economically feasible. FIGURE 4 Ferric ion (Fe 3+ ) formation and oxidation–reduction potential (Eh) during (A) turbulent flow bioleaching with addition of ferrous iron (700 mg/L) (screening method) and (B) laminar flow mesophilic bioleaching with addition of ferrous iron (700 mg/L) (flow cell method). Microbial Analysis During Bornite Bioleaching Considering the extensive copper recovery from bornite under the laminar flow regime, which was not seen here for the turbulent condition in the presence of ferrous iron (Fe 2+ ), we next performed a dynamic microbial analysis during biofilm formation. Initial inoculation of the acidophilic consortium in a proportion 3:1 ( Leptospirillum spp.: At. thiooxidans ) was followed under the laminar flow condition, revealing the predominance of the iron oxidizers in solution until the end of the assay and the appearance of unidentified (possibly mineral native bacteria) in planktonic form ( Figure 5A ). After day 20, the iron present in the solution was detected in the ferric form ( Figure 4B ), correlating with the increment in the total iron-oxidizing bacteria in the recirculating solution ( Figure 4B ). However, FISH analysis indicated the prevalence of Acidithiobacillus spp. in the biofilm ( Figure 5B ), opening the discussion concerning the interaction and relevance of both planktonic and sessile microbial populations. On the one hand, the planktonic iron oxidizers recycle the ferric ion that actively (bio)leach the mineral surface, a phenomenon initially favored under turbulent conditions ( Hong et al., 2019 ) based on the screening copper recovery kinetics observed ( Figure 3 ). However, the lack of biofilm formation with the presence of microbes with sulfur-oxidizing activity did not generate the required acid production to balance its consumption during the microbial iron-oxidation process ( Figure 4A ), avoiding further copper extraction from bornite ( Figure 3B ). Clearly, this is not representative of the industrial heap/dump copper mining operations, where the kinetics of microbial iron and sulfur oxidation occur at a different pace. In this sense, the laminar flow biofilm analysis shows that the attached microbial population is mainly composed of sulfur-oxidizing bacterial species ( Figure 5B ), most likely due to mineral surface elemental sulfur accumulation ( Lei et al., 2009 ; Bobadilla-Fazzini, 2017 ). This phenomenon can be considered useful in terms of initial mineral surface exposure to the ferric leaching agent, though detrimental due to extensive biofilm growth and mineral surface hindering at a later process stage, perhaps requiring specific conditions to tailor microbial composition for efficient copper extraction. FIGURE 5 Microbiological determinations during high-purity natural bornite bioleaching under mesophilic (30°C) inoculation with external ferrous iron (Fe 2+ ) addition in laminar flow bioleaching (flow cell method). (A) Microbial dynamics in the recirculating solution via qPCR and cell counts. (B) Representative biofilm composition by FISH (30 days) where the image corresponds to overlap of FISH-Cy3 (red, Acidithiobacillus spp.) and FISH-FAM (green, Leptospirillum spp.). EDS analyses confirmed the expected bornite atomic composition before bioleaching ( Figure 3D ), and the altered bornite surface with the impoverishment of copper during bioleaching ( Figure 3F ), validating the identification of idaite by QEMSCAN. Moreover, EDS analysis showed that during the turbulent flow screening method, the precipitates formed over the surface of bornite appears to be sodium jarosite [Na Fe 3 (OH) 6 (SO 4 ) 2 ] due to notable increments in oxygen and sodium ( Figure 3E ), as recently reported for ferric sulfate pretreatment followed by moderate thermophilic bioleaching ( Liu et al., 2020 ). On the other hand, the surface EDS analysis of laminar flow cell bornite showed no sodium at all but high carbon proportions indicating the organic nature of the biofilm formed ( Figure 3G ). Mechanism for Bornite Bioleaching During Biofilm Formation Combining the biofilm development, ferric profile, and the mineralogical QEMSCAN and EDS analyses, it is possible to propose a mechanism for bornite bioleaching under laminar flow conditions ( Figure 6 ). Previous studies have shown two stages for bornite ferric leaching ( Pesic and Olson, 1984 ), with an initial leaching rate significantly dependent on ferric ion concentration and with no elemental sulfur formation leading to idaite: \n (1) \n Cu FeS 5 + 4 4 Fe ⟶ 3 + Cu FeS 3 + 4 2 Cu + 2 + 4 Fe 2 + \n FIGURE 6 Proposed bioleaching mechanism for primary copper sulfides bornite and chalcopyrite under mesophilic inoculation of bacteria with iron- and sulfur-oxidizing activity forming biofilm. Indeed, the ferrous iron oxidation is mediated by Leptospirillum spp. yielding the ferric form: \n (2) \n 2 F e + 2 + 1 / 2 O + 2 2 H ⟶ + 2 F e + 3 + H O 2 \n This first stage is observed in our analysis from the beginning of the process until days 34 and 13 for turbulent and laminar flow with ferrous iron addition, respectively, with a corresponding copper extraction close to 30% ( Figure 3 ). In the second stage, a slow leaching rate of idaite arises independent of ferric ion concentration ( Figure 3C ) where the sulfur produced is proportional to copper dissolved and tending to envelop the residue mineral particles, therefore, being the limiting step: \n (3) \n Cu FeS 3 + 4 8 F e ⟶ 3 + S + 0 3 C u + 2 + 9 F e 2 + \n Here, the iron- and sulfur-oxidizing microorganisms exert a catalyzing role, with the pH maintenance as a critical aspect in order to prevent ferric ion precipitation ( Figure 4B ), a role exerted by the sulfur-oxidizing biofilm controlling the leaching kinetics ( Figure 5B ): \n (4) \n S + 0 3 O + 2 2 H O 2 ⟶ 2 S O 4 + 2 - 4 H + \n These results confirm that mesophilic biofilm occurs extensively over bornite under laminar flow regimes, modifying its surface in a very different way than ferric leaching does under turbulent flow conditions ( Hong et al., 2019 ; Liu et al., 2020 ). In addition, as the buildup of biofilm comprises several stages, initiating with the bacterial adherence to the surface, then the cells aggregate entering a phase of irreversible attachment, maturation, and finally dispersion ( Garnett and Matthews, 2012 ). This process could explain the significantly higher copper recovery levels compared with shaking conditions ( Figures 3B,C ). More precipitates are observed under turbulent conditions masking the proper quantification of the resulting iron sulfide compounds ( Figure 3 ). Overall, the flow cell chamber is a more appropriated approach to studying biofilm formation and mineralogic dynamics that resemble the heap bioleaching of this abundant primary copper sulfide."
} | 6,159 |
35377816 | PMC9169740 | pmc | 8,559 | {
"abstract": "Significance The development of novel degradable biocomposites can contribute to answering the increasing global demand for sustainable materials. We present a method to obtain self-bonded biocomposite materials from cultured plant cells. Subjecting cells to a cold-compression molding process creates hierarchical biocomposites that have stiffness and strength comparable to commodity plastics, while being 100% biodegradable in soil. Introducing fillers expands the attainable functionalities, demonstrating the versatility of the proposed platform. The use of fast-growing plant cells offers the benefits of short harvest time, zero biomass waste during processing, in situ manufacturing, and no arable land requirement. The approach allows the possibility of further tuning the final material properties by genetically engineering the processed cells.",
"discussion": "Results and Discussion We harvest plant cells from a suspension culture and compress them in a permeable mold, to achieve a densified dehydrated structure ( Fig. 1 A and Materials and Methods ). During compression, water diffuses through the plant cell wall, and the cell volume is gradually reduced. When the cells reach a dry state, corresponding to an approximate 98% weight loss, the resulting bulk materials ( Fig. 1 B ) consist of a hierarchical lamellar stack of compacted cell walls. Cross-section scanning electron microscopy (SEM) images of the resulting material ( Fig. 1 C and D ) illustrate the obtained anisotropic microstructure. Fig. 1. ( A ) Schematic of the fabrication method. Plant cells are cultured, harvested, and subjected to a controlled compression and dehydration, resulting in a lamellar densified architecture when dried. ( B ) Photograph of the biocomposite. ( C and D ) SEM top and cross-sectional views of the anisotropic microstructure. We characterize the cell morphology with light and laser scanning confocal microscopy ( Fig. 2 B – D ), which shows that the plant cells are elongated, with a mean length of 170 ± 60 µm and a mean width of 45 ± 10 µm, and are surrounded by a thin primary cell wall. By staining the cells we confirm that the cell walls contain cellulose, pectin, and phenolic compounds ( Fig. 2 B – D ) as expected from this cell type ( 26 , 27 ) ( Materials and Methods ). Compositional analysis of the dry biocomposite material confirms that it is composed of 15% cellulose, 20% hemicelluloses, 6.8% pectins, and 6.3% phenolic compounds. Literature suggests that in tobacco plant cells the remaining components are lipids, nucleic acids, proteins, and inorganics (ash content), which together account for about 45% of the dry mass ( 28 ). Residual water within our samples is gravimetrically determined to be 7 ± 3 wt %. Thus, the process results in a biocomposite material, composed of a heterogeneous mixture of the natural cell wall biopolymers. Fig. 2. ( A ) Photograph of the cell culture. Microscopy images of the cells stained for ( B ) pectins, ( C ) cellulose, and ( D ) phenolics. (Scale bars in B – D , 20 µm.) ( E ) SEM image of a cross-section, demonstrating the lamellar microstructure. ( F ) TEM and ( G and H ) HRTEM images of cross-sections (arrows pointing to dangling fibrils tethering adjacent cell walls, marked as CW1 and CW2). ( I ) FTIR spectra of 1) hydrated cells, 2) dried cells, and 3) dry biocomposite. Color coding for biopolymer peak assignment: gray, all carbohydrates; blue, cellulose; green, pectin; brown, phenolics; and yellow, proteins. ( J ) MD simulation of a fraction of two neighboring cell walls ( Inset shows pectin and cellulose chains diffused in the adjacent cell wall space). Carbon atoms are colored gray in cellulose, orange in phenolics, yellow in hemicellulose, and blue in pectin chains for clarity. Oxygen atoms are colored red, and hydrogen atoms are white. ( K ) MD results from tensile loading of the compacted cell walls system: 1) total energy and 2) hydrogen bonding energy. Inset shows chains being unentangled and pulled out from the space between the two cell walls when subjected to tensile forces. ( L ) SEM of a fractured surface of a tensile-tested biocomposite. ( M ) XRD pattern of the biocomposite with marked contributions from cellulose polymorphs I α , I β , II, and III. SEM and transmission electron microscopy (TEM) observations of the biocomposite materials reveal their hierarchical, anisotropic, and lamellar microstructure composed of compacted plant cells ( Fig. 2 E – H ). TEM demonstrates that the nanofibrillar structure of the primary cell walls is preserved during cell compression and dehydration ( Fig. 2 F and G ). Accepted models suggest that the primary cell wall is a multilayered structure consisting of cellulose nanofibrils, arranged in various orientations within each plane (from entirely isotropic to helically aligned, depending on cell type and developmental stage), bound in a matrix of hemicelluloses, pectins, and proteins ( 29 ). Even in the case of randomly distributed cellulose nanofibrils in the plane of the wall, the structure is considered highly anisotropic across thickness ( 30 ). TEM images of our biocomposites show an average dehydrated cell wall thickness of 185 ± 57 nm and cellulose crystalline nanofibril bundles with diameters 1 to 30 nm being conformed across the consecutive parallel planes ( Fig. 2 G and H ). We observe a hierarchical structure: at the supracellular level (micrometer scale), a lamellar microstructure consisting of compacted cells ( Fig. 2 E ), and at the subcellular level (nanoscale), an anisotropic, multilayered structure, derived from the natural organization of the cell wall components ( Fig. 2 G and H ). High-resolution TEM (HRTEM) images show that the outer nanofibrils of the cell walls tether to the adjacent walls ( Fig. 2 H , dangling fibrils pointed by arrows). Fourier-transform infrared (FTIR) spectroscopy of hydrated and oven-dried cells and of the compacted biocomposite ( Fig. 2 I ) reveals the predominant vibrations of carbohydrates (cellulose, hemicelluloses, and pectin), proteins, and phenolic compounds ( 26 , 27 , 31 ) in all samples (see SI Appendix , Table S1 , for detailed bond assignment). Comparing the spectra reveals that after processing, the biocomposite maintains all the native cell carbohydrate components and retains the same degree of pectin esterification [indicated by the ratio of the 1,735/1,414 peak intensities ( 26 )] but has a slightly lower amount of protein compared to the living cells [lower intensity of the 1,650 cm −1 peak ( 26 )]. Moreover, the red-shifted hydrogen bonding band at 3,000 to 3,700 cm −1 , in the dried biocomposite, compared to the hydrated cells, reflects the strong intermolecular hydrogen bonding between the cell wall biopolymers ( SI Appendix ) ( 32 , 33 ). Based on HRTEM ( Fig. 2 H ) and FTIR ( Fig. 2 I ) analyses, we postulate that the cell wall adhesion in our biocomposites is provided by two mechanisms: 1) fibrillar interlocking and 2) intermolecular interactions of the polymer chains in the adjacent parts of the cell walls. Further, considering the role of pectins in the structural adhesion of cells in plants ( 34 ), we hypothesize that esterified pectins may facilitate this intercellular bonding. We conduct reactive molecular dynamic (MD) simulations (ReaxFF) to verify our hypothesis and study the molecular interactions between adjacent cell walls. Following the fabrication process, we simulate the compression of the outer parts of two neighboring cell walls ( Fig. 2 J and K and SI Appendix , Figs. S1–S3 ). Cell walls are modeled as mixtures of cellulose, hemicellulose, pectin, and phenolics at the ratios identified from chemical analysis. The modeled cell wall polymers encompass all functional groups (-OH, -CO 2 CH 3 , -CH 2 OH, and -COOH) available for intermolecular interactions within the cell walls. The compacted and equilibrated system has a volume of 3.6 × 2.7 × 2.5 nm 3 and shows that all polymer chains at the outer layers of neighboring cell walls interact and diffuse in each other’s structure upon compressing. Thus, the compression process leads to molecular interlocking between adjacent cell walls. Then, we subject the system of compacted cell walls to tensile testing, which shows that the total energy has a substantial hydrogen bonding energy component, in agreement with literature on self-adhesive cellulose materials ( 35 ) ( Fig. 2 K and SI Appendix , Fig. S3 ). The fibrillar interlocking leads to chain unfolding (unentanglement) and cascading hydrogen bond breaking and reformation events upon tensile loading ( Fig. 2 K , Inset ) ( 35 , 36 ). These results match SEM observations of the tensile tested fracture surfaces ( Fig. 2 L ), which show fibrils bridging neighboring parts of the matrix. X-ray diffraction (XRD) patterns of the biocomposite reveal multiple polymorphs of semicrystalline cellulose (I, II, and III, marked in Fig. 2 G ) ( 23 ). Native cellulose from plant species crystallizes in the type I polymorphs ( I α , I β ). Regeneration and mercerization, ball-milling in presence of water and other methods ( 36 ) lead to the more thermodynamically stable cellulose polymorph II, while ammonia treatment followed by thermal and pressure treatments is known to convert either cellulose I or II to III ( 37 – 39 ). In the densified biocomposites, cellulose microfibrils partially undergo phase transformations to form crystal structures II and III, likely in response to the pressure applied during dehydration and the changing chemical environment during cell dissociation. We postulate that upon compaction and cell death, protein and nucleic acid losses, as suggested by FTIR, lead to the diffusion of amine-rich compounds from the protoplasm to the extracellular space through the cell walls, thus facilitating the phase transformations, together with the extended period of compression. We perform tensile and three-point bending tests to characterize the mechanical performance of the dehydrated biocomposites. Since our process results in bulk biopolymer structures, which can serve as a standalone structural material as well as a polymer matrix to be reinforced with fillers, we compare their mechanical properties to 1) bulk synthetic polymers of similar density, which also serve as both a host matrix and unfilled bulk material, and 2) natural and engineered wood, which share similar composition, have a hierarchical structure, and are also used as structural bulk materials. We choose different softwoods (pine), hardwoods (poplar, oak, and walnut), commercial plywood and MDF, and synthetic plastics of similar density (polystyrene [PS], polypropylene [PP], and low-density polyethylene [LDPE]) ( Fig. 3 A – C and SI Appendix , Fig. S4 ). Stress–strain plots obtained from the biocomposites ( SI Appendix , Fig. S5 ) show an initial linear elastic response upon loading, both under tension and bending, followed by a brittle failure at small strains (1 ± 0.3%). The Young’s modulus, calculated from the initial linear elastic part of the tension experiments, is 2.5 ± 0.4 GPa, and the ultimate strength is 21.2 ± 3 MPa. The flexural modulus is 4.2 ± 0.4 GPa, and the modulus of rupture is 49.3 ± 3.2 MPa. Testing the flexural properties of the biocomposite on the two perpendicular planes (see schematic in SI Appendix , Fig. S6 ) reveals that stiffness varies by a factor of ca. 1.75 in the two directions, while strength remains unaffected by orientation. The measured difference in stiffness is due to the anisotropic microstructure of the biocomposite, as discussed above. Tension tests show that our biocomposites are stiffer than the other tested materials ( Fig. 3 A ). However, natural woods have higher strength ( Fig. 3 B ), which can be explained by their different cellular architectures, cell wall compositions, and component arrangements within the secondary cell walls. The cells used in our biocomposites originate from the herbaceous plant Nicotiana tabacum , and they naturally develop a thin, unlignified primary cell wall (we confirm a low phenolics amount of 6.2 wt %). These cells do not form secondary cell walls and cannot self-organize in a hierarchical microstructure in our in vitro cultures. Regardless, the macroscopic mechanical performance of the biocomposites is comparable to that of commercial engineered woods and commodity plastics. They surpass literature-reported values for bulk, three-dimensional composite biomaterials manufactured through bottom-up methods by eukaryotic organisms that include plant cells, mycelium, or yeast matrixes. All these biocomposites have not been treated to remove native components, and the microorganism cells serve in fact as the building blocks of those materials ( 12 , 13 , 18 , 20 , 40 , 43 , 44 ) ( Fig. 3 D ). We note, however, that in particular, mycelium-based composites have lower density than our biocomposites ( 18 ). Recently, microbial cultured cells (from Escherichia coli ) were used to create cell-based rigid films through a simple casting and ambient drying method ( 45 ). The localized mechanical properties of the microbial films were measured through nanoindentation, revealing a reduced modulus of 5 to 42 GPa and strength values of 60 to 800 MPa (calculated from the measured hardness values). While those properties cannot be directly compared to the plant cell bulk properties we report in our work as they address different length scales, they demonstrate another promising direction for cell-based materials. Fig. 3. ( A ) Young’s modulus and ( B ) tensile strength of the biocomposite and reference materials. ( C ) Material density. ( D ) Comparison of mechanical properties of this work (blue circles are results from bending tests, and blue squares are results from tensile tests) and literature-reported biocomposites in which microorganism cells serve as a matrix, and their composition and nanostructure have not been altered: yeast-based composites represented with rhombuses and mycelium-based materials with circles ( 12 , 13 , 18 , 40 – 42 ). ( E ) Biodegradation of the biocomposite and natural pine. Sample notation: BC, pure (without fillers) biocomposite; 1, pine; 2, poplar; 3, oak; 4, walnut; 5, plywood; 6, MDF; 7, PS; 8, PP; 9, LDPE. (Error bars indicate SE for n > 5 measurements.) I α , I β , II, and III. CNF, cellulose nanofibrils; CNT, carbon nanotubes. A key factor in the design of sustainable materials is their end-of-life fate. The realization of biological matrix materials, such as those described here, offers an environmentally friendly alternative to nondegradable materials, which typically survive in landfills. To assess the biodegradability of our plant-based biocomposites, we perform agricultural soil incubation tests ( Materials and Methods ), comparing their mass loss with that of natural wood ( 46 ). Results show an initial mass gain corresponding to water uptake from the soil, in both natural wood and biocomposites ( Fig. 3 E ). The detectable mass loss due to biodegradation of the biocomposites begins 3 wk after incubation, while for natural wood it begins about 7 wk later. This can be associated to the presence of lignin in natural wood, which is known to provide resistance to pathogen attacks on cell walls ( 47 ) and the water sensitivity of our material. We observe an almost complete biodegradation of the biocomposite 14 wk after initial incubation. A drawback of rapid biodegradation is water sensitivity. We perform water uptake and thickness swelling tests ( SI Appendix ) and find that our biocomposites respond similarly to mycelium/wood biocomposites ( 18 ). After more than 24 h of incubation our materials completely disperse in water. However, when stored in ambient conditions, they do not exhibit any swelling or fouling. In applications, the water uptake could be mitigated with surface treatments or water-resistant coatings. The use of cell cultures for materials fabrication allows on-demand property tuning by interfacing with additive particles ( 48 , 49 ). We demonstrate the ability to tune and introduce properties in the composites, by incorporating filler additives. The addition of different amounts of natural nanoclay (NC) platelets, for example, changes the biocomposites’ compressive modulus and strength ( Fig. 4 A ). Upon the introduction of 0.15 wt % NC the Young’s modulus and strength increase by 36 and 87%, respectively. At 0.5 wt % NC the improvements are less significant, 10 and 5%, respectively. At higher NC concentrations both properties are reduced below the values of the unfilled material, which is often observed in polymer nanocomposites because of fillers’ aggregation acting as a stress concentrator ( 52 ). Different filler particles expand further the biocomposites’ property space ( Fig. 4 B ). We plot the Young’s modulus as a function of density of different plant cell-based biocomposites: pure cell matrix (BC), biocomposites containing various amounts of carbon fibers (CF), halloysite and montmorillonite NC, and graphene (G). Their properties lie at the intersection of natural cellular materials, including wood-based materials, and commercial plastics ( Fig. 4 B ), presenting Young’s moduli spanning over one order of magnitude. We note that our biocomposites are outperformed by pure cellulose materials and densified wood products. This is because our approach preserves the entire natural cell wall composition through mild processing. Filler additives also endow functionalities, such as electrical conductivity or magnetic properties. The electrical conductivity of plant cell/CF composites, for example, can be tuned varying the CF content ( Fig. 4 C ). Similarly, the addition of 13.5 wt % iron oxide nanoparticles (IN) in the plant cell matrix conveys ferromagnetic properties, which allow the biocomposite to support more than five times its weight when attracted by a magnet ( Fig. 4 D ). We foresee possible use of our densified plant cell biocomposites in panels for packaging or non–load-bearing automotive applications or space manufacturing. Fig. 4. ( A ) Compressive modulus and strength of biocomposites with NC platelets. ( B ) Young’s modulus versus density for various materials and our biocomposites. Blue groups correspond to bending experiments, and red groups correspond to compression. The cellulose* area corresponds to pure cellulose fibers, papers, and nanocellulose-based products, including bacterial cellulose ( 37 , 50 , 51 ). Densified wood data are from ref. 20 . ( C ) IV curves for biocomposites with 1 and 20 wt % CF. ( D ) Biocomposite with IN exhibiting magnetic properties. We introduce a method to create natural biocomposite materials based on cultured plant cells. The method capitalizes on the plant cell’s ability to synthesize intricate multilayered structures of cellulose, hemicellulose, phenolics, and pectin in their cell walls, which we preserve as nanostructured building blocks. In the future, the use of different cell cultures and/or genetically modified species ( 53 ), as well as the modulation of processing conditions to modify the cell arrangement at the microscale, may allow the fabrication of materials with significantly altered properties. Similar fabrication approaches can be envisioned for many other biological systems (e.g., algae and fungi) that can provide complex nanostructured elements as building blocks for advanced composite biomaterials."
} | 4,881 |
38375242 | PMC10875449 | pmc | 8,560 | {
"abstract": "In this work, we present a novel stretchable bimodal sensor that can simultaneously detect temperature and humidity changes based on poly-hydroxyethyl acrylate (PHEA) elastomer infused with 1-ethyl-3-methylimidazolium tetrafluoroborate (EMIM-BF4) ionic liquid. The sensor exhibits high transparency, stability, and biocompatibility, as well as excellent mechanical and sensing properties. The sensor can achieve a maximum strain of 761%, a sensitivity of 4.5%/°C at room temperature, a detection range from −35 to 120 °C, and a response time of 10 ms. The sensor is able to provide acute response to movement of human hand at close range and can detect temperature changes as small as 0.004 °C in the range of 20–30 °C. The sensor also responds to humidity change, showing a high sensitivity to humidity change of 4.4%/RH% under the temperature of 30 °C. The sensor can be used for various applications in wearable electronics, human-machine interfaces, and soft robotics.",
"conclusion": "4 Conclusion In this work, we have presented a novel stretchable temperature and humidity sensor based on a composite made with elastomer and ionic liquid. The sensor shows high transparency, flexibility, and ionic conductivity, as well as excellent stability and self-adhesiveness. The sensor can monitor both temperature and humidity changes with excellent sensitivity higher than most existing works at around room temperature, and thus is applicable in room temperature sensing and breath sensing. We believe that this polymer has great potential in both academic and applicational uses.",
"introduction": "1 Introduction Stretchable sensors that can detect multiple stimuli such as temperature and humidity are highly desirable for applications in wearable electronics, human-machine interfaces, and soft robotics. Temperature sensing is especially important for monitoring human health and physiological activities, as well as environmental changes [ [1] , [2] , [3] , [4] ]. However, most of the existing stretchable sensors are based on electronic conductors and semiconductors, which have limitations in transparency, stretchability, stability, and biocompatibility [ [5] , [6] , [7] , [8] ]. Ionic conductors, on the other hand, offer advantages such as high transparency, low impedance, and good biocompatibility [ [9] , [10] , [11] ]. However, they often face challenges due to the high ion concentration and the presence of water in the matrix for ion dissolution, which can result in poor mechanical properties and water evaporation [ [12] , [13] , [14] ]. Therefore, there is a pressing need to develop novel ionic conductors that can combine high stretchability, transparency, stability, and sensitivity for bimodal sensing. Various attempts have been made to create sensors based on hydrogels and conducting elements, utilizing different measurement techniques [ 6 , 15 , 16 ]. However, many of these sensors rely on complex structures or incorporate conducting particles, such as multiwall carbon nanofibers, rendering the material opaque and compromising its mechanical properties [ 17 ]. Additionally, some previous works involved the use of metallic salts, introducing the challenge of corrosion over extended periods, which can damage the polymer structure [ 18 , 19 ]. In 2018, S. Ding and colleagues [ 20 ] introduced a non-corroding, single-phased ion gel material that has high sensitivity to temperature. Nonetheless, their utilization of a hydrophobic matrix limited the material's capability to simultaneously detect moisture. There also have been other trials to make stretchable temperature and humidity sensors by application of stretchable designs [ 21 , 22 ]，but by incorporating such designs, the manufacturing process becomes more complicated. Consequently, the development of a bimodal sensor capable of sensing both temperature and humidity remain an unsolved challenge. The elastomer PHEA (Poly Hydroxyl Acrylate) catches our attention due to its superior hydrophilicity and biocompatibility. The hydrophilicity of PHEA allows it to absorb water molecule from ambient air and to develop a humidity sensing mechanism based on it. And the high biocompatibility of PHEA allows long period of contact of PHEA matrix to human skin, which allows the fabrication of breath sensors from it. In addition, the high biodegradability of PHEA will cause little environmental impact upon commercialization. 1-Ethyl-3-Methylimidazolium Tetrafluoroborate (EMI-BF4), as a novel kind of ionic liquid, has been used in lithium batteries to improve their ionic conductivity and ion transportation behaviors at room/low temperatures [ 23 ]. However, the utilization of EMI-BF4 as an ionic conductor when coupled with hydrogel remains a relatively unexplored area of research. In this work, we report an ultrasensitive stretchable bimodal sensor that can sense temperature and humidity simultaneously based on novel elastomer and ionic liquid samples. The elastomer can provide mechanical strength and elasticity to the sensor. During mechanical property tests, PHEA (Poly Hydroxyl Acrylate) sample doped with 10 wt% ionic liquid is capable of a maximum strain of 761%. The ionic liquid infusion provides exquisite sensing ability to both temperature and humidity change, with a high sensitivity of 4.5%/°C at room temperature range. wide detection ranges from −35 to 120 °C, fast response time of 10 ms. These characteristics render it well-suited for a variety of applications in soft electronics.",
"discussion": "3 Results and discussion 3.1 Mechanical properties of the polymer The initially fabricated PHEA (Poly Hydroxyl Acrylate) samples demonstrated commendable mechanical properties, with remarkable stretchability exceeding 790%. This underscores PHEA's suitability as a high-quality elastomer. In our work, we incorporated various ionic liquids, including EMI-BF4 (1-Ethyl-3-Methylimidazolium Tetrafluoroborate), to impart conductivity. Notably, the addition of a relatively modest amount of ionic liquid within the range of 10–20 wt% had a minimal impact on mechanical properties. As depicted in Fig. 1 b, the elastomer enriched with 10 wt% EMI-BF4 exhibited an outstanding maximum stretchability of 761%. 3.2 General properties of the material We doped the PHEA samples with various ionic liquids, each at 10 wt%, and compared their respective room temperature resistivity values. We used an undoped PHEA sample as a reference, which exhibited insulating properties with a resistivity of 1.5 × 10 7 Ω/m. After doping with different ionic liquids, the resistivity of the samples decreased, with some falling into the resistivity range of semiconductors. As depicted in Fig. 2 a, the resistivity of the samples doped with 10 wt% EMI-BF4 was compared with those doped with EMI-TFSI (1-Ethyl-3-methylimidazolium bis (trifluoromethyl sulfonyl) imide) and EMIES (1-Ethyl-3-methylimidazolium ethyl sulfate), which are commonly used ionic liquids. The results revealed that the addition of EMI-BF4 led to the lowest resistivity among the three samples at room temperature (25 °C) under an ambient relative humidity of RH = 80%. Notably, increasing the EMI-BF4 content to 20–30 wt% had minimal impact on mechanical properties, ensuring high elasticity was maintained. Fig. 2 (a)The resistivity of HEA ionogels with different ionic liquid. (b)Strain-resistivity curve of the material. Fig. 2 As reported in previous studies [ 24 , 25 ], conducting elastomers can serve as effective strain sensors. We selected the PHEA samples doped with the 10 wt% EMI-BF4 to do the test. Subsequently, we measured the gauge factor of this sample, as depicted in Fig. 2 b, the figure illustrates that the resistivity of the sample varies in response to elongation (strain) up to 200% with a gauge factor of 5.7. 3.3 Electric properties of the material Our initial research objective was to measure resistivity concerning temperature for temperature sensing applications. As shown in Fig. 3 a, our findings revealed that the selected sample (PHEA doped with 10 wt% EMI-BF4) maintained conductivity at low temperatures down to −35 °C and remained stable even at elevated temperatures up to 120 °C. Notably, the sample retained relatively low resistivity even at a low temperature of −20 °C. This temperature response closely aligns with similar behavior observed in other previously reported conducting hydrogels [ 26 , 27 ]. Furthermore, when compared to existing materials of the same class [ 6 , 18 ], our sample exhibited a broader sensing range. Fig. 3 (a)The upper and lower limit temperature sensing range of EIL. (b)The response to temperature of sample with different EMI-BF4 concentration. Fig. 3 In addition to its wide temperature stability range, the sample also exhibited high sensitivity to temperature changes. Fig. 3 b illustrates the varying responses of differently EMI-BF4 concentration samples (ranging from 5% to 40%) to a temperature change from 25 to 120 °C. To maintain consistent water content during these tests, a hydrophobic tape (Polytetrafluoroethylene) was used to seal the samples after pretreatment. It is evident that the sample with 10 wt% EMI-BF4 displayed the highest sensitivity to temperature changes, while the 40 wt% sample exhibited the least sensitivity. To quantitatively measure the sensitivity, we defined the thermal response as the relative resistance variation (ΔR/R0) concerning the initial resistance level (R0). Sensitivity was calculated from the slope of the fitted response-versus-temperature curve using Equation 1. Eq.1 S ( % ) = ( Δ R ) / ( R 0 * Δ T ) * 100 % The sample with 10 wt% EMI-BF4 showcased the highest sensitivity at 4.5%/°C at room temperature (25 °C), surpassing most previously reported materials [ [28] , [29] , [30] , [31] , [32] , [33] , [34] , [35] , [36] , [37] , [46] ]. A detailed comparison is provided in Table S1 . However, it is important to note that the sample with 5 wt% EMI-BF4 exhibited lower sensitivity (2.5%) compared to the 10 wt% sample (4.5%). This observation can be attributed to the nature of PHEA as a long-chain elastomer, which undergoes coiling and relaxation as temperature changes. When the dopant concentration is reduced to very low levels, the polymer network formed by these coiled chains becomes denser, potentially inducing cross-links [ 31 ]. This significantly reduces the ability of conducting ions to move within the polymer, thus reducing sensitivity. Conversely, with very high dopant concentrations, the restriction on ion movement inside the polymer network is minimal at room temperature, resulting in little change as temperature varies. After studying the samples’ response to the temperature change, we then conducted research on its repeatability of results. First the PHEA sample doped with 10 wt% EMI-BF4 is deprived of water content and sealed with PTFE tape as mentioned before and situated in an air tight box. Then the temperature inside the box is heated from 70 to 100 °C. As shown in Fig. 4 , the sample is able to reproduce the same resistance value at a given temperature after a few cycles. This points out that the sample has reproductible resistivity values when subjected to a slow cyclic temperature change. Fig. 4 Cyclic temperature and resistance behavior of EIL. Fig. 4 Our next study was to test the response of the sample to changes in humidity at a specific temperature. As illustrated in Fig. 5 a, the PHEA sample doped with 10 wt% EMI-BF4 attains various equilibrium weights when exposed to different relative humidity levels, ranging from 20% to 80%, at room temperature. Subsequently, we measured the resistivity of the sample at room temperature as the ambient humidity increased from 30% to 80%, with results depicted in Fig. 5 b. As shown in the figure, there is a substantial change in resistivity as the environmental humidity increases from 30% to 80%. Fig. 5 (a)The equilibrium weight with the increase of humidity at room temperature. (b)The resistivity versus humidity curve of the sensor at room temperature. Fig. 5 The sample is then subjected to humidity change under different temperatures from 30 to 80 °C as shown in Fig. 6 . The data collected is arrayed into a temperature humidity matrix comprised of resistance curves under different temperature. The matrix is later used in the fabrication of the bimodal sensor. From the TH-matrix data, it's found that the sample has a highest sensitivity of 4.4%/RH% under the temperature of 30 °C, which is 3 times more than previously reported resistive moisture sensors, which have their sensitivities in the range of 1–2%/RH% at maximum [ [38] , [39] , [40] , [41] ]. Detailed comparison is shown in Table S2 . Fig. 6 The response to humidity of the sensor at different temperature. Fig. 6 3.4 Potential applications of the material Based on the properties of the PHEA sample doped with 10 wt% EMI-BF4, we have developed a bimodal sensor capable of simultaneously sensing temperature and humidity. By detecting the movement of a human hand and monitoring the behavior of a lit candle, our sensor has demonstrated high sensitivity to both temperature and humidity. After this, we have extended the applications to motion detection and respiratory sensing. From the tests, it is shown that the sample has the potential to be used in various areas. 3.5 Bimodal sensor As stated in the “Introduction” section, the bimodal sensor is made of one sealed and one bare PHEA strip doped with 10 wt% EMI-BF4. The strips are situated on a transparent plastic substrate. The circuit is then connected to the substrate. In a typical measuring cycle, the resistance value of the sealed strip is measured first. Then the resistance value is substituted into the data curve shown in Fig. 3 a to acquire the temperature value. After acquiring the temperature value, the resistance value of the bare strip is measured. Then the resistance value of the bare strip is substituted into the data curves shown in Fig. 6 together with the temperature value acquired above to calculate the humidity. Then the combined temperature and humidity data is sent to the screen by a Wi-Fi module. A photo and schematic of the bimodal sensor is shown in Fig. 7 . Fig. 7 Schematic and photo of the bimodal sensor. Fig. 7 Our sensor boasts an exceptional temperature resolution, capable of detecting temperature changes as small as 0.004 °C, at the temperature range from 20 to 30 °C. This stands in stark contrast to the majority of commercial products ( Table S3 ) and academically reported stretchable sensors, which typically exhibit resolutions within the range of 0.1–0.5 °C [ 42 , 43 ]. We first demonstrated this high resolution by employing the temperature sensor to detect a human hand. We placed the hand at a distance of 15 cm from the sensor and detecting the temperature is 27.648 °C (see movie S1 ). As the hand gradually approached to a distance of 5 cm, the temperature changed to 27.664 °C. From the test data depicted in Fig. 8 , it is evident that the bimodal sensor consistently generated temperature change signals throughout this process and could clearly discern the movement of the hand, thus affirming the high temperature resolution. A comparison of cost of our sensor with commercial sensors that have a similar resolution is shown in Table S4 . It is shown that, the elastomer sensor costs much less to produce than existing high-resolution sensors. Fig. 8 The response of the sensor to hand movement. Fig. 8 In addition to its remarkable temperature resolution, the bimodal sensor also exhibits a high humidity resolution. We conducted a candle test to compare our bimodal sensor with the commercial temperature and humidity dual sensor, DHT22 (see movie S2 ). As shown in Fig. 9 , we firstly placed a lit candle 3 cm in front of the EIL sensor, with the L value set to zero. Then, we gradually moved the candle to L = 15 cm and recorded the changes in sensor output during this process. It is noticeable that the DHT22 failed to detect humidity changes during this phase, maintaining a stable humidity output of 81.6%. However, our bimodal sensor demonstrated humidity variations ranging from 81.592% to 81.605%, showing our robust humidity sensing capabilities. Simultaneously, in comparison to the DHT22, our bimodal sensor was able to precisely detect detailed temperature changes during this test. Fig. 9 The response of the sensor to hand movement. Fig. 9 3.6 Breath detection As previously introduced in the section on the bimodal sensor, it exhibited remarkable sensitivity, capable of detecting changes in ambient temperature as small as 0.004 °C. This exceptional sensitivity allowed us to integrate the sensor with an N95 mask, creating a breath sensor, as depicted in Fig. 10 . In this figure, the sensor is positioned on the front arc of an N95 mask and connected to a computer via a small Wi-Fi module as shown in Fig. 10 a. The data in Fig. 10 b demonstrates the consistency in temperature and humidity variations when the subjects are breathing. Fig. 10 (a)Schematic and photo of the breath sensor. (b) Breath testing results. Fig. 10 Upon further comparison with other existing works [ 44 , 45 ], it becomes evident that our bimodal sensor performed comparably to current commercial breath sensors. Furthermore, the high sensitivity enabled the sensor strip to be placed on breath masks rather than on the upper lip. This adaptation made it possible to utilize breath sensors on patients requiring oxygen support, while avoiding potential skin irritation often associated with conventional wearable sensors that need skin contact [ 47 ]."
} | 4,408 |
20442869 | PMC2861701 | pmc | 8,561 | {
"abstract": "The human brain efficiently solves certain operations such as object recognition and categorization through a massively parallel network of dedicated processors. However, human cognition also relies on the ability to perform an arbitrarily large set of tasks by flexibly recombining different processors into a novel chain. This flexibility comes at the cost of a severe slowing down and a seriality of operations (100–500 ms per step). A limit on parallel processing is demonstrated in experimental setups such as the psychological refractory period (PRP) and the attentional blink (AB) in which the processing of an element either significantly delays (PRP) or impedes conscious access (AB) of a second, rapidly presented element. Here we present a spiking-neuron implementation of a cognitive architecture where a large number of local parallel processors assemble together to produce goal-driven behavior. The precise mapping of incoming sensory stimuli onto motor representations relies on a “router” network capable of flexibly interconnecting processors and rapidly changing its configuration from one task to another. Simulations show that, when presented with dual-task stimuli, the network exhibits parallel processing at peripheral sensory levels, a memory buffer capable of keeping the result of sensory processing on hold, and a slow serial performance at the router stage, resulting in a performance bottleneck. The network captures the detailed dynamics of human behavior during dual-task-performance, including both mean RTs and RT distributions, and establishes concrete predictions on neuronal dynamics during dual-task experiments in humans and non-human primates.",
"introduction": "Introduction A ubiquitous aspect of brain function is its modular organization, with a large number of processors (neurons, columns, or entire areas) operating simultaneously and in parallel. Human cognition relies, to a large extent, on the ability to perform an arbitrarily large set of tasks by flexibly recombining different processors into a novel chain (e.g. respond with the right hand to the red square) [1] – [3] . Yet this flexibility does not happen without a cost. Chaining individual computations is done at a very slow pace (100–500 ms per step) and with a considerable temporary tying-up of the brain's resources, generating what is known as “dual-task interference” – the inability to perform several tasks at once [4] – [8] . Several cognitive theories support this view, arguing that while most mental operations are modular and parallel, certain specific processes which establish flexible links amongst existing processors impose a serial processing bottleneck [3] , [9] – [15] . The psychological refractory period (PRP) provides a classic and clear demonstration in experimental psychology of the coexistence of parallel processing and serial processing bottlenecks within a cognitive task. When performing two tasks in rapid succession on two successively presented targets T1 and T2, delays are observed in some but not all of the T2 processing stages. Analysis of these delays suggests that a “central decision stage” suffers from seriality while perceptual and response operations occur in parallel [4] , [6] , [7] , [16] , [17] . Despite the fact that the PRP has been one the most widely studied paradigms to investigate dual-task interference, no network implementation had been proposed which provides a plausible implementation of its underlying mechanisms. Boxological and schematical models of the PRP [4] , [18] , [19] have successfully determined a theoretical framework which provides a synthesis of two basic aspects of cognitive architecture: 1) its chronometric organization, 2) its components that can act in parallel and those that impose seriality. According to these models, each task involves three successive stages of processing: a perceptual, a central, and a motor component. The perceptual stage of sensory processing - which is performed in a modular (parallel) fashion - does not provide a major contribution to temporal variability. A subsequent stage of serial processing involves a stochastic integration process, traditionally used to model decision making in single tasks [20] – [23] and is a main source for the variability in response time. In contrast, the last motor processing stage has only a small contribution to response variability and can be performed in parallel without interfering with other processing stages from concurrent tasks. Despite their simplicity, these models have been very successful in explaining a broad range of behavioral data, including the complex response time distributions of dual-task experiments, which can be precisely predicted only after untangling the serial and parallel stages of each task [18] . Until now, the modeling of dual tasks is only specified at a level of mathematical description and functional cognitive architecture [4] , [18] , [24] , [25] . At the neurophysiological level, understanding what kind of collective neural organization leads from massively parallel single-unit processing to a serial unfolding of two successive decisions has not been established. This situation is, to a large degree, due to the fact that there have been detailed monkey electrophysiology of single-task decision making [26] , [27] , but no comparable investigation of dual-tasks. Here we present an effort to bridge this gap between an abstract mathematical description and the underlying complex neurophysiology. We present a detailed model, based on realistic properties of spiking neurons which is capable of flexibly linking processors to form novel tasks. As a consequence of this flexibility, the network exhibits a functional serial bottleneck at the level of the “router” circuit needed to link processors. The model presents detailed predictions for future electrophysiological studies of dual-tasks and serial computations in the human and non-human primate brain.",
"discussion": "Discussion Summary of Results The present model constitutes, to our knowledge, the first spiking-neuron model of a global architecture capable of simulating the entire sensory-motor chain of processing in a dual-task setting. We could explain the detailed dynamics of behavior (including both mean RTs and RT distributions) during dual-task-performance, by simulating a large-scale network of realistic neurons, comprising about 20.000 spiking neurons and 46.000.000 synaptic connections. For consistency with the majority of previous PRP experiments, we simulated an experimental design in which stimuli involve distinct sensory modalities and the responses distinct effectors. Under these circumstances, interference occurs exclusively at the routing stage, commonly referred to in psychology as the response selection stage [4] . The central aspect of our model is a detailed neuronal implementation of this flexible “routing” and how it manages to change from one task to another in hundreds of milliseconds, using an area that maps stimuli onto responses which we have termed the router network. The model capitalizes on a number of existing elements: (1) perceptual attractor networks capable of encoding stimuli and maintaining them in an exponentially decaying buffer [62] , [71] , (2) an accumulation-to-threshold mechanism, comprising both recurrent neuronal assemblies [36] and a thresholding device inspired by the architecture of basal ganglia [81] ; (3) a control network comprising rule-coding units capable of modulating other areas in a top-down manner [32] , [45] , [55] , [82] – [85] ; (4) the concept of a routing circuit implemented by neurons with broad connectivity, capable of transiently interconnecting other brain processors in a flexible manner [33] , [47] , [86] – [89] . The novel aspect of the present simulations is to integrate these theoretical constructs into a global functional architecture. We observed that the interplay between these control and routing mechanisms resulted in a central limitation during dual-task processing, which manifested itself either as a delay in the second task (PRP), or a complete interruption of the processing of a second target (Attentional Blink). Based solely on the known dynamics of neurotransmitter receptors, the model reproduces, in a quantitative manner, a large number of behavioral observations of dual-task interference (see [17] , [18] , [35] ): A sequential delay in RT2. This delay decreases with a slope of −1 as SOA increases reflecting a sequential bottleneck. The absence of any effect of the second task on response times to the first task (mean and distribution). Strong correlations between RT1 and RT2 which progressively diminish as SOA increases. Distinct interference patterns associated with different task manipulations: changes which affect the sensory delay processing of Task 2 are absorbed during the slack time separating task 1 and task 2, while changes which affect the accumulation time (i.e. central processing in the router) propagate additively. Switch from the PRP (delayed response to T2) to the blink (an absence of the response to T2) by adding a mask after the T2 stimulus. An increase in blink probability when T1 visibility is reduced. These results are in full accordance with the central interference model [17] , [35] , [90] , by which certain processes are carried out in parallel and routing and accumulation are intrinsically serial. Our model provides a detailed neuronal implementation of this classical psychological model and makes many new predictions for the neurophysiological correlates of the PRP. Comparison with Previous Neuroimaging Studies Several brain-imaging experiments implicated a number of cortical systems in the PRP phenomenon. The cerebral basis of processing bottlenecks has been investigated with Event Related Potential studies (ERPs), which have shown that the PRP results in reduced and/or delayed components [91] – [97] . Using time-resolved fMRI [98] – [100] , Dux and collaborators showed a slight delay in the peak fMRI activity in prefrontal cortex during a PRP paradigm [101] , implying that the PFC was one of the fundamental nodes responsible for the central bottleneck of information processing. Recently, using both time-resolved fMRI and high density ERP recordings we could fully parse the execution of two concurrent tasks in a discrete sequence of processing stages. The ERP analysis demonstrated that a late P3-like complex is in fact delayed by an amount comparable to the PRP effect on RTs, and time-resolved fMRI confirmed that the PRP delayed parietal and prefrontal activation by several hundreds of milliseconds [77] . The notion that the global P3 indexes a late capacity-limited central stage fits with results from the AB. As we could show in the simulations the main difference between the PRP and the AB can be accounted for solely by the masks used to produce the AB, which interfere with the local memory of T2. The result is that T2 processing is not merely delayed (PRP), but erased and it therefore escapes from consciousness. During AB, the initial ERP components up to about 270 ms are essentially intact, but the P3 component is essentially abolished [73] , [76] , [102] , [103] . The P3 component can only be detected in seen trials, in an all-or-none fashion [73] , [104] . We observed this precise dependence for the activity of routing neurons and the onset of task-setting neurons, suggesting that the P3 is likely to constitute a large-scale electrophysiological marker of the router system. Also, as indicated by our simulations, increased latencies in T1 processing resulted in higher probability of the second target being blinked [73] , [105] , [106] . Direct comparison of AB and PRP paradigms suggests that both affect the same P3 component [95] . The spatial resolution of EEG is very imprecise and thus a better characterization of the locus of central processing bottlenecks in the brain comes from fMRI studies, which have pinpointed a broad parietofrontal network that exhibits various manifestations of central capacity limits [67] , [107] , including the AB [67] , [105] , [108] and the PRP [77] , [101] , [109] , [110] . This network is ubiquitously activated by a large variety of goal directed tasks [107] suggesting that it plays an important role in flexible routing information between remote neuronal representations. Our network postulates a hierarchical organization of this system: neurons controlling the whole-task structure (order network) gate neurons controlling the individual tasks (task-setting network), which, in turn, gate the routing from the sensory representations to the motor intention stage. Such a hierarchical organization has been demonstrated in humans in the prefrontal cortex as the Broca region and its homologue in the right hemisphere implement executive processes that control start and end states as well as the nesting of task segments that combine in hierarchically organized action plans [52] , [111] – [114] . A hierarchical organization involved in planning of complex sequential tasks has also been found in non-human primates [113] , [115] . Emergence of Seriality in Cortical Networks which Perform Flexible-Task Settings and Scaling of the Model Understanding the emergence of serial behavior in the human brain is an important and central theoretical question in cognitive psychology as modularity and parallel processing are hallmarks of brain computations. Different authors have proposed cognitive architectures that can explain how components of the mind work to produce coherent cognition [14] , [24] , [86] , [116] – [118] . Concrete implementations of these ideas have shown that these coherent states which transiently bind together existing modular processors naturally result in serial behavior [14] , [43] . Here we have tentatively proposed that seriality in dual (or multiple) task performance results from the necessity to establish a task set through the activation of a “router” network. This router network is shared by all sensory-motor mappings and its activity can, potentially, code for a virtually infinite number of possible tasks. A task-setting program acts as a gate, permitting routing neurons to propagate information if they receive the appropriate sensory input. This system acts as a control mechanism that avoids erroneous, conflicting or unwanted stimulus-response associations. We showed that a concrete implementation of such a control system results in serial behavior of the routing process when probed in dual-task situations. In our network, seriality and its behavioral manifestations, the PRP and the Attentional Blink, emerged from competition between task-setting neurons which, through a lateral inhibition process, prevented the simultaneous activation of two task settings. This form of control is necessary to ensure correct task performance in conflicting mappings - as classically demonstrated in the Stroop paradigm in which the same stimulus may lead to distinct responses according to task requirements [119] . While this mechanism is strictly required only in conflicting response mapping situations, which is not the case in our present simulations, it is possible that it has emerged as a ubiquitous mechanism in control networks to assure correct task performance. Seriality in non-conflicting tasks would therefore emerge as a consequence of the need for a flexible mechanism linking stimuli with multiple responses according to context [28] , [29] . Another possible origin of seriality relates to the coding properties of the router (for a simple illustration see Figure S6 ). Here we have explored a comparatively simplified situation of a small number of tasks, stimuli and responses in which all possible routings were coded by distinct neural populations. This mechanism would result in a combinatorial explosion in a more realistic setup, arguing that the code of router neurons should be distributed, i.e. each routing scheme should be encoded in a large population of neurons. This is consistent with many findings in prefrontal cortex neurons which have found that a large fraction of neurons respond to virtually all tasks [83] . In this scheme, the precise pattern of active and inactive neurons determines the code and thus superposing two routing configurations (of two distinct tasks) should result in a mixture leading to erroneous mapping properties. Avoidance of incorrect mappings in a combinatorial router can be implemented by the same mechanism shown here, leading to serial routing in the composition of flexible task settings ( Figure S6 ). Comparison with Alternative Implementations and Existing Models Previous modeling efforts have established cognitive architectures which can account for human complex problem solving [14] , [24] , [116] . The adaptive control of thought–rational (ACT-R), for example, proposes a theory of distinct modules that interact with each other to produce coherent cognition [14] . While ACT-R is based on a sequential scheme, the temporal constant of the sequential step in ACT-R and in the PRP are not comparable: in ACT-R, productions (if-then structures representing procedural knowledge) fire approximately every 50 ms, about five times faster than the PRP delay. The 50 ms delay of individual productions is consistent with other experimental approaches which have suggested a discrete organization of cognition at a frequency close to 13 Hz [120] . These observations of ∼50 ms productions and the comparably slower ∼300 ms PRP delay can be reconciled by modeling the entire routing program as a sequence of productions, as in the ACT-R implementation of the PRP of Byrne and Anderson [25] . Sensory modules in the ACT-R involve a two-layer structure, a visual module (mapped to occipital/temporal regions) and a visual buffer (mapped to parietal regions). The visual buffer incorporates a selection mechanism that determines the contents of the visual system which will be available to other processors. Our model provides a concrete neuronal implementation of these mechanisms. In our model, the sensory hierarchy acts as a module which can select and maintain information locally (unless a subsequent element such as the mask overrides the buffer). This information can be broadcasted to the rest of the network. Similarly, in ACT-R the selection of actions is achieved by a loop that mimics the Basal-Ganglia- cortical connections. By building up on previous architecture for thresholding and gating sensory information through striatal-cortical interactions [44] our model provides a neuronal implementation of these mechanisms. The router circuit in our model builds on previous computational models which have studied the role of contextual signals on transient sensory-motor mappings [30] , [33] , [121] , [122] . Salinas (2004) showed that a linear read-out of sensory input could result in arbitrary sensory-response mappings if sensory responses are modulated by (a non-linear) contextual influence. A concrete implementation of flexible mapping by rule-setting contextual signals was developed by Deco and Rolls [47] , [123] . In the present model, the router binds sensory and motor representations. Similar conceptions of flexible routing circuits have been applied to other instances of information binding such as, linking the attributes of an object in pattern recognition [89] or linking discrete objects to temporal contexts through distributed representations as recently proposed by Wyble and Bowman [124] . Olshausen and colleagues implemented a routing scheme in a set of control neurons which rapidly modify the strength of intra-cortical connections to implement the attentional gating of information flow from early visual representations to a higher level object-centered reference frame [89] , [125] . The SAIM model of selective attention [88] , [126] has shown how this ‘dynamic routing’ model can be extended to account for a wide range of results of visual experiments with competing stimuli in space, i.e. neglect [127] or in time, i.e. inhibition of return [88] in both normal and impaired subjects. The SAIM model [88] shares many features with our network: it implements a routing neuron which is modulated by a control (task-setting) network and thus acts as a coincidence-detector of a task-setting program and current sensory state. Recently, Heinke and collaborators showed how the SAIM model can be implemented with spiking units [126] . Our network provides an implementation of simple boxological models of dual-task execution in the PRP [17] , [34] , [35] . While very simple, these models have established a vast range of predictions in behavioral experiments regarding the precise functional dependence of RTs with SOA and how these functions should change with different manipulations. By incorporating ideas of models of decision making, we previously generated a schematic model that accounts for the entire distribution of RTs and how it changes in the interference regime [18] . Here we have shown that these ideas can be implemented robustly in realistic network architecture. A critical aspect of our network is that while the router is occupied by T1, the T2 stimulus was maintained in the recurrent activity of high-level sensory units, thus forming a memory which remains local because it cannot activate the router. This coexistence of parallel mechanisms – a cascade of sensory processes which encode the stimulus - and of serial bottlenecks – queuing by the routing process - constitutes a hallmark of PRP observations. Our network implemented this local memory as a local attractor showing progressive integration and exhibiting a metastable form of memory that could be maintained for a few hundred milliseconds. According to this proposed mechanism, the memory trace remains stored in a local network and is relatively fragile as it can readily be overridden by a mask. The critical observation is that the mask can only override processing of T2 if it the router is occupied by T1. To our knowledge, our model is the first one to propose a concrete neural implementation of the mechanisms leading to the PRP. In contrast, several computational models have been recently proposed for the attentional blink [43] , [78] , [128] – [130] . Two current explanations include the simultaneous type serial token (ST2) model [78] which proposes that access of sensory representations to working memory is gated by an episodic-driven attentional signal and the boost and bounce model [130] which suggests that a target initiates an attentional boost which is interrupted when the trailing task-irrelevant stimulus is accidentally boosted. Our model shares with the ST2 model the idea of gating of a router-system and with the boost and bounce model that task-setting activation is not a phasic event, but rather, can stay active until it is inhibited by a termination signal. We emphasize that our model does not intend to give a detailed account of all the findings from attentional blink experiments, but instead to show how the same mechanisms that lead to delayed responses in the PRP can lead to missed targets in the AB. Recent reviews of the extensive AB literature argue for a multifactor origin in this processing deficit [131] , and thus it might be impossible to pinpoint a single mechanism behind the full diversity of experimental findings (although see [132] , [133] ). Nevertheless, our results show that limited capacity operations – as the one implemented by our router/task-setting network – may play a central role in the attentional blink [72] , [134] . One aspect of the attentional blink phenomena which our model fails to replicate is the relative increase in performance observed at very short SOA (∼100 ms), an effect known as lag-1 sparing [5] . This effect is not observed when T1 and T2 involve different modalities [135] (as in our simulations of the AB) or spatial locations [136] . Recent experiments show that the sparing can even be spread to several targets presented rapidly without intervening distractors [137] , [138] , suggesting that the unit of selection of a serial attentional process is not the individual target but an extended event which may include several rapidly presented targets [132] , [139] , [140] . This grouping does not happen without a cost, since order swapping and performance tradeoffs between different targets do occur [78] , [141] . In our model, the task-setting configuration is sustained until information is routed to the motor system, and thus it might be possible to extend the present model such that more than one target in a RVSP benefits from the same task-setting configuration. Processing a temporally extended event encompassing several targets would require broadening – in feature space - the action of the task-setting network as well as making the router/task-setting complex capable of flexibly routing information not only to motor areas but also to mnemonic [142] or sensory areas in order to achieve recursive computations. In fact, we see the extension of the present model along the lines just discussed: the different types of neurons used in our implementation (briefly reviewed in the next section) have been found in the awake behaving monkey and may serve as a basis from which to construct complex cognitive programs, as those implemented in systems like ACT-R [3] or SOAR [143] - but with a stronger grounding on neurophysiological findings [144] . In this implementation, we see router neurons as capable of accumulating evidence not only towards a motor response, but implementing a full production system [145] , [146] where stochastic rules are selected according to the information contained in different mnemonic systems which are in turn updated by external stimuli and by the action of the productions themselves. These ideas will form the basis for a future extension of the present model to flexible series of chained tasks. Comparison to Previous Electrophysiological Studies and Novel Predictions Most, if not all, types of neurons used in our implementation have been observed in studies that measured single-neuron activity in awake behaving monkeys during single-task performance. Here we will briefly mention the main types of neurons in the various areas of our model and compare them to neurophysiological data, a comparison that will have to remain somewhat superficial as we cannot attempt to discuss the precise relationships between the variety of tasks employed in the neurophysiological studies and the PRP task implemented here. Firstly, the properties of the sensory areas of our model are consistent with what is known about representations in areas of sensory cortex. Neuronal activity in low level sensory cortex is largely (but not entirely) determined by the incoming sensory information [147] , while neurons in higher areas carry information about the behavioral relevance of stimuli, as well as traces of stimuli to be remembered [148] . Secondly, neurons in areas of parietal and frontal cortex have response properties consistent with the routing process proposed by our model. Many of these cells are tuned to categories of stimuli that are associated with a particular behavioral response [149] – [151] and integrate evidence in favor of one of a number of possible actions until a threshold is reached, just as is required by the model's router [152] – [154] . Thirdly, some neurons in the frontal cortex only respond if a particular stimulus maps onto a particular motor response, but not when the same stimulus or response is part of a different stimulus-response mapping [60] , and yet other prefrontal neurons code abstract rules [84] . Clearly, the response properties of these neurons are in accordance with the model's task-switching network. Finally, neurons in the motor response selection stage of our model have either a gradually increasing activity before the response or they respond with a sharp burst at the time of the response. Neurons with gradually increasing activity before the motor response and cells with a motor burst are indeed observed in areas of the motor cortex [155] , [156] as well as in the basal ganglia [157] . These results, taken together, indicate that the types of units required by our implementation are broadly consistent with the types of neurons that are observed in neurophysiological experiments. Our network can also explain timing and latencies of the sequence of events identified in single-task physiological experiments in monkeys [158] – [160] and humans [161] . Accumulation of information about the upcoming response influences the firing rate of routing neurons at a latency of about 200 ms, a latency that may be relatively fixed for a given task [162] . This latency cannot be explained solely by synaptic delays, since measurements of conduction velocity of cortical feedforward and feedback connections showed that they can be rapid, even faster than intrinsic connections within a cortical area [163] , [164] . A previous neurophysiological study showed that the onset of response modulation in the visual cortex depends of the sequencing of subtasks, with later modulation for subtasks that occur later in a sequence [165] . Our model grasps this observation: the latency of the response of routing neurons depends on the order in which the two subtasks are executed ( Figure 3B–C ). The present results suggest that the latency of feedback modulation may reflect the time required by the network to settle into a brain-scale state of coherent activity [18] , [87] , which in our model is reflected by a coherent pattern of activity across sensory, router, and task-setting networks coding different aspects of the same subtask. Our observations also raise a note of caution on the interpretation of processing latencies from physiological data. A concrete example is conveyed in our model by the measurement of activity in the routing neurons. Spiking activity shows a clear sequential scheme: routing neurons of T2 start integrating only once routing of T1 has completed ( Figure 3B ). Thus, the latency at which spiking activity exceeds a certain threshold constitutes a physiological marker of the PRP effect. The picture is quite distinct if one would measure synaptic router activity ( Figure S3 ). During the time in which T1 is being routed and T2 is being buffered, T2 sensory neurons spike and project silently (i.e. without evoking spiking responses) to router neurons. Hence synaptic activity in T2 router neurons increases during T2 compared to baseline. A consequence of this observation, which may be of relevance beyond the specifics of this study, is that timing analysis based on synaptic or spiking activity yield qualitatively different observations. Various studies have simultaneously measured different markers of neurophysiological activity such as multi-unit activity (MUA), laminar current-source density (CSD) and local field potentials (LFP) [166] and fMRI [167] or EEG [168] . Multimodal interactions have been shown to display such a mixed effect in response latencies. Primary auditory cortex shows a clear CSD response to somatosensory stimulation, without observable changes in the spiking response as measured by MUA [169] . Computational models may be a useful link to bridge information gathered at different scales. Our data showed that fluctuation in response time could be accounted by the dynamics of noise fluctuations in relation to the timing of stimulus routing ( Figure 6 ). When noise is oscillatory, this is determined by a precise phase relation. Our model does not explain how this relation can be entrained. Neurophysiological data of multi-sensory integration suggests that somatosensory stimuli can reset the phase of ongoing oscillations in primary auditory cortex such that auditory stimuli are boosted if presented during the high excitability phase [169] , [170] . Also, it has been shown that neuronal oscillations can entrain to environmental rhythms improving discriminative performance and decreasing response times [65] , [66] . As mentioned, these aspects lie outside the scope of the present model. The correlates of the bottleneck have yet to be studied at the single cell level and our simulations therefore generated a number of new predictions that could be tested in future neurophysiological experiments. First the model establishes the existence of routing and task-setting neurons with well distinct dynamics and connectivity with different neuronal populations. At the anatomical level, routing neurons should receive inputs from all sensory modalities and from task setting neurons. At the functional level, they should be characterized by their firing in response to specific conjunctions of stimuli and responses, a preference which may change dynamically according to task context, on a time scale of about 100 ms or more (for supporting evidence, see [60] , [113] ). Task-setting neurons should engage in a competition such that two task-setting programs or routing schemes cannot coexist in time. This should avoid unwanted mappings but also causes an inertia which results in relatively slow switching (>100 ms) from one task-setting to another leading to seriality in the routing process. In a PRP experiment, neurons coding for the memory T2 stimulus should show a characteristic temporal profile, comprising (1) a phasic sensory response, time-locked to actual stimulus presentation, (2) a sustained response exhibiting a slow exponential decay, and (3) a late amplification at the time when task 1 routing is completed and the router neurons of task 2 become active. On the contrary, the onset of router and task-setting neurons of Task 2 should be delayed at short SOA, with a delay that should decrease with SOA because task 2 router neurons are released from the inhibition of task 1 as soon as it is completed. In trial-by-trial comparisons, at short SOA values, the onset of router and task-setting neurons of T2 should be locked to the response time of the first task. While sharing the onset, the model predicts distinguishable time-courses of activations for router and task-setting neurons. Task-setting neurons should show sustained high-levels of activation throughout the duration of the task while router-neurons activity should ramp to a critical threshold. In an AB experiment task-setting neurons of T2 should be active both in seen and unseen trials. Only in unseen trials should the memory of T2 fade below a threshold (either due to fluctuations in transient response or in the durations of the memory due to the extension of T1) impeding routing and broadcasting to the rest of the network. These predictions will become testable once an awake animal model of dual-task performance is defined."
} | 8,762 |
30424359 | PMC6187360 | pmc | 8,563 | {
"abstract": "Microfluidic automation technology is at a stage where the complexity and cost of external hardware control often impose severe limitations on the size and functionality of microfluidic systems. Developments in autonomous microfluidics are intended to eliminate off-chip controls to enable scalable systems. Timing is a fundamental component of the digital logic required to manipulate fluidic flow. The authors present a self-driven pneumatic ring oscillator manufactured by assembling an elastomeric sheet of polydimethylsiloxane (PDMS) between two laser-engraved polymethylmethacrylate (PMMA) layers via surface activation through treatment with 3-aminopropyltriethoxysilane (APTES). The frequency of the fabricated oscillators is in the range of 3–7.5 Hz with a maximum of 14 min constant frequency syringe-powered operation. The control of a fluidic channel with the oscillator stages is demonstrated. The fabrication process represents an improvement in manufacturability compared to previous molding or etching approaches, and the resulting devices are inexpensive and portable, making the technology potentially applicable for wider use.",
"conclusion": "4. Conclusions The authors fabricated a plastic pneumatic oscillator circuit that showed both long-term stability of frequency and robustness of operation, although the resolution of the laser engraver used did not allow for high-density fabrication. The out-of-phase three-valve fluidic device established oscillator suitability for self-driven control logic for microfluidic devices. A syringe was sufficient to drive the oscillator for a considerable period of time, demonstrating its applicability in portable lab-on-a-chip diagnostics. The fabrication process, based on laser engraving and APTES treatment, was reliable for the fast, efficient prototyping of these circuits during the design and development phase, as it took approximately 10–20 min to complete the laser engraving of PMMA and then four hours for the assembly. Higher-resolution laser engravers should allow for fabricating higher-density, smaller valve devices responding at higher frequencies. No additional tool adjustments or calibration is required as in techniques such as CNC milling or hot embossing. Provided that the resolution can be improved, the uniformity of features and the low cost of the process make these circuits potential candidates for simplifying the control circuitry of lab-on-a-chip devices and for driving all-plastic autonomous robots [ 25 ].",
"introduction": "1. Introduction Microfluidic automation can feature large arrays of valves and pumps performing multiple analyses for diagnostic assays and biological applications [ 1 ]. The development of complete systems that integrate sample preparation, fluidic manipulation, and detection mechanisms [ 2 ] is associated with increasingly complex interconnections and microvalve automation. Microfluidic very large-scale integration (mVLSI) [ 3 ] has enabled parallel fluidic manipulations using thousands of microvalves, thereby alleviating some of the burden. However, the associated control hardware requires many external connections, thereby making the management of its operation prone to errors (leaks, etc.) and cumbersome to operators unfamiliar with the technology. Designs that minimize the number of control signals [ 4 ] have provided an impetus to parallelization, but they have not eliminated entirely the challenge of external connection complexity. In order to drastically reduce off-chip connections, researchers have explored the integration of microfluidic logic control elements on chip of varying complexity [ 5 , 6 , 7 , 8 ] with the aim of facilitating the development of simple-to-use, portable devices. These circuits have been fabricated with time and labor-intensive techniques such as soft lithography [ 9 ] and glass etching, limiting the manufacturability and mass production capability. The expansion of microfluidic circuits into the realm of point-of-care testing (POCT) hinges on the simplification of the manufacturing process and the automation of the steps involved [ 10 ]. To this end, computer numerical control (CNC) milling [ 11 ] and laser engraving [ 12 ] have been explored as techniques for microfluidics fabrication. Micro-milled three-dimensional microfluidic flow cells have demonstrated the potential of milling for medical diagnostics [ 13 ]. Laser-engraving has emerged specifically as a very promising technology due to factors such as fastest turnaround time, comparatively low operational costs and ease of design processing. A wide range of polymers can be processed by a laser cutter, including thermoplastics such as polymethylmethacrylate (PMMA). Thermoplastics are extensively used for microfluidics due to their high manufacturability and properties amenable to microfluidic analysis [ 14 , 15 ]. Valves and pumps have been demonstrated by sandwiching polydimethylsiloxane (PDMS) between channel structures created on PMMA [ 16 ]. Implementing self-driven microfluidic devices requires an on-chip timing reference or clock generation circuit [ 17 ]. Pneumatic oscillators fabricated with glass and PDMS have been characterized by Duncan et al. [ 7 ]. Scalability of microfluidic logic circuits using prototyping techniques such as CNC milling has also been shown [ 18 ]. In this paper, the authors demonstrate an all-plastic (PMMA/PDMS) ring oscillator chip with features manufactured using laser engraving and driven with a syringe which acts as a simple and portable pneumatic power source. The device is assembled by permanently bonding the three layers using chemical surface activation in contrast to mechanical approaches to keeping the layers together, such as binder clips [ 19 ] which cannot form a direct bond between the PDMS and PMMA [ 20 ].",
"discussion": "3. Results and Discussion 3.1. Frequency Measurement The variation in frequency of oscillations with the magnitude of vacuum applied (via Elveflow pressure controller, Elveflow, Paris, France) across devices fabricated and assembled in different batches is plotted in Figure 3 . Frequency increases were directly proportional to vacuum over a vacuum supply range of 3–6 psi (3.08 Hz at 6 psi being twice the 1.48 Hz oscillations at 3 psi). The curve started to level off at VAC = 8 psi attaining an average 4.15 Hz, suggestive of the saturation observed in an equivalent MOSFET’s voltage transfer characteristics. A low pressure threshold of 2.5 psi was observed below which oscillations cease, corresponding to the threshold voltage required to turn on a NMOS. The frequency–vacuum supply relationship observed correlates to the alpha power law commonly used to estimate delay in CMOS circuits with short channel transistors [ 22 ], implying an analogous “short-seat” effect in the valve. Near the vacuum threshold of oscillation (2.5 psi), a higher variation of frequency across devices was observed, indicating a sensitivity to process variations, particularly in the adhesion of the PDMS to the PMMA over the valve seat area after bonding. An examination of valve operation through the captured videos at frequencies below 2 Hz confirmed this hypothesis. As frequency increased, the lower variation across devices could be due to the adhesion being less of a dominating factor. As the sensitivity did not appear to be a limitation for higher frequency generation, the devices could be suitable for medical applications that require rapid analysis. A frequency of 4.15 Hz (at VAC = 8 psi), for example, would be useful for a controlled pump sequence. After observing the oscillation frequency stability over a period of a few hours, the authors made a preliminary assessment about the usefulness of the oscillator as a practical timing reference. As Figure 4 illustrates, the frequency increased by 25% after more than two hours of operation at a constant vacuum supply and stayed at the new value for another two hours. This drift implied a lowering of the threshold pressure required to open the valve, attributable most probably to the decrease in adhesion of PDMS to the PMMA over the long duration [ 7 ]. The continuous frequency increase after a period of constant oscillations suggested an apparent threshold of surface adhesion deterioration for the membrane. The deterioration continued linearly and stabilized after an hour, beyond which oscillations became constant again. Beyond the five hours of oscillations plotted in Figure 4 , the frequency remained steady after further observation suggesting the possibility of long-term stable operation. After the cessation and subsequent initiation of the oscillations, the starting frequency was unaltered (pre-drift value of 3 Hz as shown in Figure 4 ), implying that the original PDMS adhesion was restored. This behavior reflected an initial settling time for the oscillator and supported the concept of a pre-operational running period to ensure stable oscillations for on-chip applications. The frequency vs. vacuum pressure applied relationship for an oscillator with a smaller pull-up resistance is depicted in Figure 5 . Halving the resistance did not produce a proportionate decrease in oscillation period. At VAC = 4 psi and 6 psi, the corresponding average frequencies were close to that of the oscillator with pull-up resistance R (see Figure 1 C). A reason for the frequency remaining largely unaffected could be the contribution of the parasitic resistance of the interconnect lines (connecting the valve stages) and valve resistance dominating the overall resistance in comparison to the pull-up resistance. Reducing the valve and interconnect line lengths enabled the tuning of the frequency with the pull-up resistor length and simultaneously achieved area optimization. 3.2. Portability Experiment The portability of the device was demonstrated by powering the oscillator with a syringe. The vacuum level applied was measured by connecting a tube at the mouth of the syringe to an analog pressure gauge. Applying a steady vacuum by clamping a pulled 60 mL syringe generated oscillations that could be sustained for as long as 14 min at an average frequency of 7.5 Hz, as shown in Figure 6 which depicts the three characteristic performances of seven oscillators tested. Although the average frequency of five oscillators was between 3 and 4.2 Hz, two pulsated at approximately double the value (7.5 Hz), manifesting a property of the laser-engraving process: the effect of channel roughness on their average depth. Features made using the same raster mode laser settings on different substrates had a variability in roughness depending on the scanning pattern of the beam for the particular run. The resultant variation in average channel depth factors into fluidic resistance by an inverse cube. A depth of ~85 µm would reduce the channel resistance by almost a third and the oscillation time period by approximately 60%. After the initial period of steady oscillations, the eventual frequency decline observed in Figure 6 could be caused by the syringe losing vacuum. During valve closing, air was pulled into the valve chamber from the atmosphere. When the valve opened, air in the chamber flowed out through the pull-up resistor into the supply line which was attached to the syringe. The time interval for frequency drop-off to zero from the initial value was 13–14 min across devices, indicating a uniform rate of air flow into the syringe. This unvarying air flow was due to the dependency of the flow on the pull-up resistor rather than the frequency of valve operation. 3.3. Flow Control in Fluidic Channel A circuit wherein the individual ring oscillator inverter outputs are driving three valves in series in a fluidic channel is shown in Figure 7 . The figure also depicts the relative phases of the three inverter outputs and their dependence on the time delay of each inverter stage and the interconnecting channel delay: If τ l = RC , τ x l = R x C , and τ y l = R y C ( R x C x , R y C y , RC x , and RC y are negligible), the equivalent time delay at the input of each stage is T D + τ l , T D + τ l + τ xl , and T D + τ l + τ y l , where T D is the time delay of each inverter. The corresponding relative phase shift at each stage is as follows: 0°, τ x l T D + τ l + τ x l × 360 ° , τ y l T D + τ l + τ y l × 360 ° . By modifying the channel dimensions and consequently the resistance between the output of one stage and the input of another (specifically R x and R y ), the phase difference can be suitably varied. The circuit serves to demonstrate the possibility of a timed sequence of opening and closing of valves, useful in an assay, for example, where specific reagents need to be mixed in at different times. Red-colored water assisted in the visualization of the phase shifts of the pulsating fluidic valves (see Figure 8 , Multimedia view). The three phases corresponding to the state of the valve in sequential order are the following: partially open, open, and closed. The time period of fluidic valve operation with a vacuum supply of 8 psi for the oscillator circuit was 0.2 s with each phase corresponding to ~0.067 s. As seen in Figure 8 , the open-to-close transition was sudden, whereas the close-to-open transition was phased-out with an intermediate, partially-open state. This behavior is attributed to the elasticity of the PDMS and the strength of its adhesion to the PMMA valve seat surface. The closed valve remained in that state due to the PDMS valve seat stiction until a threshold vacuum to open was reached. Thicker PDMS membranes were less elastic and hence separated from the rest (closed) position at a slower speed, but on the other hand, they snapped back faster to the closed state than thinner ones. With one full cycle of the valve operation equivalent to 360°, the phase difference between partially open and open/closed states was 90°. Actuation patterns depended on the phase difference between the stages and could be set by changing the channel length between stages. Adding inverter stages to the oscillator and using the output from the appropriate nodes to drive the fluidic valves is another method to achieve the desired actuation outcome. 3.4. Benchmarking of Oscillators Comparing the oscillators used in the study to previously demonstrated pneumatic and hydraulic circuit implementations provided a benchmark for assessment. The semi-autonomous fluid handling by means of on-chip digital logic was initially demonstrated by Hui et al. [ 23 ]. Pneumatic oscillators fabricated with glass/PDMS and controlling the liquid pumps formed a section of the control circuitry having five external inputs. Kim et al. [ 24 ] developed micro-hydraulic oscillators using a gravity water-head as the power source that had the advantage of low-pressure actuation and enabled parallelization without pressure drops. However, a device relying on gravity is dependent on orientation and movement, affecting its suitability for portable applications. Additionally, the lower pulse frequencies achieved by such circuits (<3 Hz) versus the higher frequencies (up to 50 Hz) attainable via pneumatic oscillators are characteristic of high-resistance liquid-medium devices (resistance being approximately two orders of magnitude greater in the case of water compared to air). Rhee and Burns [ 9 ] demonstrated digital circuits constructed with three-layer PDMS valves. Their pneumatic clock generator was implemented as a single inverter fed back through a resistance (narrow, long channel) and capacitance (chambers in the flow and control layers separated by a membrane) in series. Clock frequencies between 2 and 4 Hz for varying RC values and high stability over a ~2 min time period of observation were reported. One reason for the relatively short period may have been the use of a one-inverter configuration which is inherently less stable than the three-inverter structure. Loss of power in the device due to air-permeability of the PDMS flow and control layers likely contributed to the rapid decline in oscillations as well. Pneumatic oscillators fabricated by sandwiching a PDMS membrane between etched glass flow and control chambers have been shown to work at frequencies up to approximately 5 Hz depending on the pull-up resistance, demonstrating a logarithmic drop in frequency with vacuum supply [ 7 ]. Robustness against vacuum fluctuations at small strengths has been claimed based on this behavior. The oscillators have a frequency drift of <1%/h. The authors achieved the same frequency range with a comparatively simplified fabrication process. Their oscillators showed a linear drop in the frequency of oscillations with vacuum supply with the rate of decrease becoming smaller below 3 psi vacuum, which correlated to robustness in the case of the glass-etched oscillators. Their oscillators displayed zero frequency drift for operation periods less than 2 h, which made them deployable (as an example) for analysis involving certain yeast cell cycles or measurements in cell density changes."
} | 4,266 |
38794516 | PMC11124873 | pmc | 8,566 | {
"abstract": "The synthesis of conventional plastics has increased tremendously in the last decades due to rapid industrialization, population growth, and advancement in the use of modern technologies. However, overuse of these fossil fuel-based plastics has resulted in serious environmental and health hazards by causing pollution, global warming, etc. Therefore, the use of microalgae as a feedstock is a promising, green, and sustainable approach for the production of biobased plastics. Various biopolymers, such as polyhydroxybutyrate, polyurethane, polylactic acid, cellulose-based polymers, starch-based polymers, and protein-based polymers, can be produced from different strains of microalgae under varying culture conditions. Different techniques, including genetic engineering, metabolic engineering, the use of photobioreactors, response surface methodology, and artificial intelligence, are used to alter and improve microalgae stocks for the commercial synthesis of bioplastics at lower costs. In comparison to conventional plastics, these biobased plastics are biodegradable, biocompatible, recyclable, non-toxic, eco-friendly, and sustainable, with robust mechanical and thermoplastic properties. In addition, the bioplastics are suitable for a plethora of applications in the agriculture, construction, healthcare, electrical and electronics, and packaging industries. Thus, this review focuses on techniques for the production of biopolymers and bioplastics from microalgae. In addition, it discusses innovative and efficient strategies for large-scale bioplastic production while also providing insights into the life cycle assessment, end-of-life, and applications of bioplastics. Furthermore, some challenges affecting industrial scale bioplastics production and recommendations for future research are provided.",
"conclusion": "9. Concluding Remarks and Future Perspectives The use of microalgae as a bio-factory for the synthesis of bioplastics has attracted significant attention due to the ability of these photoautotrophic organisms to grow rapidly with less nutrients. Microalgae produce a variety of biopolymers, including PHA, PLA, PU, cellulose-based polymers, starch-based polymers, and protein-based polymers, when cultivated under different conditions. These polymers have great potential and interesting properties, including biodegradability, biocompatibility, and non-toxicity. Techniques such as genetic engineering, metabolic engineering, the use of photobioreactors, artificial intelligence, and machine learning are currently being employed for large-scale and inexpensive production of bioplastics from microalgae for applications in the agriculture, healthcare, packaging, electrical and electronic, and construction industries. The recommendations for future directions include the following: (i) Further studies on microalgae biorefinery involving the use of genetic engineering and metabolic engineering as vital tools for enhanced biomass production and purity should be carried out to achieve high quality novel bioplastics at lower costs. (ii) The development of energy efficient and cost-effective photobioreactors will provide controlled culture conditions for enhanced microalgae biomass yields for bioplastics synthesis. (iii) Further research on bioprospecting for novel hyperactive microalgal strains and the application of consortium of microalgae is crucial for industrial scale production of bioplastics with less additives, thus promoting circular economy for a sustainable future. (iv) A proper understanding of the mechanisms of bioplastics accumulation in microalgae is imperative to pave the way for more research opportunities. (v) The use of different compatible natural reinforcing agents should be the focal point of future research for the synthesis of bioplastics with greater tensile strength and robust thermal stability.",
"introduction": "1. Introduction Plastics are produced from oil, natural gas, coal, or petrochemicals. These carbon-based polymers have transformed our lives in diverse ways by opening avenues for vital developments in many industries. In recent years, plastic production has increased tremendously owing to rapid population growth and advancements in the use of technologies [ 1 ]. Worldwide plastic production is expected to reach 445 million tons, with an additional increase to 589 million tons, by 2050 [ 1 ]. These synthetic materials are stable, transparent, lightweight, versatile, durable, affordable, and resistant to corrosion, with high strength [ 2 , 3 ]. However, despite their immense benefits, overuse of these fossil-based polymers results in serious impacts on the environment, causing pollution, global warming, and fossil fuel depletion, due to their hydrophobicity and huge resistance to biodegradation [ 4 , 5 ]. In addition, synthetic plastics are recalcitrant in nature and release toxic chemicals to the environment, especially when indiscriminately disposed of, thereby polluting water bodies, and adversely affecting ecosystems [ 6 , 7 , 8 ]. To overcome the abovementioned challenges, there is a need to produce plastics from natural renewable biomass sources. Bioplastics are degradable or non-degradable biobased polymers [ 9 , 10 , 11 ]. They are produced from natural polymers of plant, animal, or microbial origin. Microorganisms serve as an excellent source for bioplastics production due to their ease of cultivation, rapid growth rate, high productivity, ease of genetic manipulation, etc. [ 12 ]. The use of microalgae as a feedstock for bioplastic production is highly preferred, owing to the ability of these photoautotrophic organisms to grow at a faster rate with high biomass. Unlike plant-based bioplastics, the use of microalgae does not lead to food competition for human consumption [ 13 ]. In addition, microalgae have fewer nutritional demands and thrive well in non-arable environments (e.g., wastewater) [ 14 ]. Microalgae consume inorganic compounds for growth and production of certain metabolites (e.g., proteins, carbohydrates, and lipids). These metabolites are utilized for various applications, including the synthesis of polysaccharides (such as alginate, carrageenan, and agar) for bioplastics production [ 15 , 16 ]. In other words, microalgae serve as a sustainable source for the commercial production of biopolymers via cultivation or natural harvest [ 8 ]. Bioplastics are produced by conversion of algal biomass through fermentation, plasticization, blending, and compatibilization processes [ 17 ]. According to a recent survey by European Bioplastics, global bioplastics production is predicted to increase from 2.4 million tons in 2022 to 7.5 million tons by 2026 as a substitute for conventional plastics [ 18 ]. Microalgae-derived plastics are economical, highly recyclable, biocompatible, biodegradable, energy efficient, flexible, have a lesser carbon footprint, and generate no toxic by-products, leading to a more sustainable circular economy [ 13 ]. However, bioplastics are brittle with low melt strength and weak barrier properties. These include bio-polybutylene succinate (bio-PBS), polylactic acid (PLA), polyhydroxybutyrate (PHB), polyurethane (PU), bio-polyethylene (bio-PE), polyhydroxyalkanoates (PHAs), and starch-based, cellulose-based, lipid-based, and protein-based biopolymers [ 19 ] ( Figure 1 ). These biobased plastics are employed in a variety of industrial, agricultural, and biomedical applications [ 20 , 21 , 22 , 23 , 24 ]. The present review provides insights into techniques used by microalgae for the synthesis of biopolymers and bioplastics, while also elucidating strategies for the optimization of microalgae-derived bioplastics for potential applications in industries, biomedicine, and agriculture."
} | 1,936 |
22245575 | null | s2 | 8,568 | {
"abstract": "Domain swapping is a mechanism for forming protein dimers and oligomers with high specificity. It is distinct from other forms of oligomerization in that the binding interface is formed by reciprocal exchange of polypeptide segments. Swapping plays a physiological role in protein-protein recognition, and it can also potentially be exploited as a mechanism for controlled self-assembly. Here, we demonstrate that domain-swapped interfaces can be engineered by inserting one protein into a surface loop of another protein. The key to facilitating a domain swap is to destabilize the protein when it is monomeric but not when it is oligomeric. We achieve this condition by employing the \"mutually exclusive folding\" design to apply conformational stress to the monomeric state. Ubiquitin (Ub) is inserted into one of six surface loops of barnase (Bn). The 38-Å amino-to-carboxy-terminal distance of Ub stresses the Bn monomer, causing it to split at the point of insertion. The 2.2-Å X-ray structure of one insertion variant reveals that strain is relieved by intermolecular folding with an identically unfolded Bn domain, resulting in a domain-swapped polymer. All six constructs oligomerize, suggesting that inserting Ub into each surface loop of Bn results in a similar domain-swapping event. Binding affinity can be tuned by varying the length of the peptide linkers used to join the two proteins, which modulates the extent of stress. Engineered, swapped proteins have the potential to be used to fabricate \"smart\" biomaterials, or as binding modules from which to assemble heterologous, multi-subunit protein complexes."
} | 406 |
37239957 | PMC10218387 | pmc | 8,569 | {
"abstract": "Copiotrophic bacteria that respond rapidly to nutrient availability, particularly high concentrations of carbon sources, play indispensable roles in marine carbon cycling. However, the molecular and metabolic mechanisms governing their response to carbon concentration gradients are not well understood. Here, we focused on a new member of the family Roseobacteraceae isolated from coastal marine biofilms and explored the growth strategy at different carbon concentrations. When cultured in a carbon-rich medium, the bacterium grew to significantly higher cell densities than Ruegeria pomeroyi DSS-3, although there was no difference when cultured in media with reduced carbon. Genomic analysis showed that the bacterium utilized various pathways involved in biofilm formation, amino acid metabolism, and energy production via the oxidation of inorganic sulfur compounds. Transcriptomic analysis indicated that 28.4% of genes were regulated by carbon concentration, with increased carbon concentration inducing the expression of key enzymes in the EMP, ED, PP, and TCA cycles, genes responsible for the transformation of amino acids into TCA intermediates, as well as the sox genes for thiosulfate oxidation. Metabolomics showed that amino acid metabolism was enhanced and preferred in the presence of a high carbon concentration. Mutation of the sox genes decreased cell proton motive force when grown with amino acids and thiosulfate. In conclusion, we propose that copiotrophy in this Roseobacteraceae bacterium can be supported by amino acid metabolism and thiosulfate oxidation.",
"conclusion": "5. Conclusions Together, the present study has clarified the mechanisms governing copiotrophy in a marine Roseobacteraceae bacterium. We showed that copiotrophy in this bacterium is the result of amino acid metabolism and the regulation of functionally relevant genes. These findings contribute to the understanding of marine bacterial lifestyles and provide an explanation of bacterial adaptation in an ecological framework. Future perspectives would include a further study to illuminate the gene regulatory pathways or networks that control the response to the amino acid concentration, and a broad study to overview similar mechanisms among marine bacteria.",
"introduction": "1. Introduction Carbon is essential for the survival of microorganisms. In marine environments, the carbon concentration plays a major role in determining both the growth and niche distribution of bacteria. This is especially true for chemoheterotrophic bacteria that obtain carbon entirely from organic compounds. Marine heterotrophic bacteria can be divided into oligotrophs and copiotrophs according to their nutritional strategy [ 1 ]. Oligotrophic bacteria grow slowly in all types of carbon sources, with high concentrations of organic materials often adversely affecting their growth, whereas copiotrophic bacteria respond rapidly to nutrient availability and can grow well in media with high concentrations of organic matter [ 2 ]. Moreover, copiotrophic bacteria are genetically more capable of sensing, transducing, and integrating extracellular stimuli [ 1 , 2 ]. These characteristics may be key to their ability to fine-tune and respond rapidly to changes in environmental conditions, such as a sudden influx or depletion of nutrients [ 1 ], and allow for them to degrade and utilize high-molecular weight particulate organic materials in the marine ecosystem [ 3 ]. However, the molecular and metabolic mechanisms governing the response of copiotrophic bacteria to changes in carbon source concentrations are not well understood. Bacterial members of the family Roseobacteraceae, previously known as the Roseobacter group and belonging to the family Rhodobacteraceae [ 4 ], are a phylogenetically uniformed and physiologically heterogeneous group within the class Alphaproteobacteria [ 5 , 6 , 7 ]. They are widely distributed in the ocean, ranging from offshore waters to the open ocean, from surface waters to the deep sea, and from the tropics to the polar regions [ 6 , 7 , 8 ], indicating that Roseobacteraceae have adapted to a variety of carbon sources and concentrations. Thus, an understanding of carbon metabolism and trophic strategies in Roseobacteraceae bacteria is of particular significance. Ruegeria pomeroyi DSS-3, previously known as Silicibacter pomeroyi DSS-3 [ 9 ], is the first Roseobacteraceae bacterium with a complete genome sequence [ 10 ] and is used as a model microorganism for the study of the eco-physiological strategies of heterotrophic and free-living bacteria [ 11 ]. Moreover, DSS-3 has been defined as a moderate copiotroph based on genomic and preliminary physiological evidence [ 12 ]. In addition to free-living lifestyles, several Roseobacteraceae lineages prefer living in biofilms [ 13 ]. In coastal marine environments, biofilms can develop on a variety of substrates, including natural stones, sedimental particles, and various man-made materials [ 14 , 15 ], and they often possess great microbial diversity [ 16 , 17 ]. In a recent study [ 18 ], we systematically explored the diversity and thiosulfate metabolism features of Roseobacteraceae strains in marine biofilms and highlighted their important roles in biogeochemical cycles. On the other hand, coastal marine biofilms are believed to be dominated by copiotrophic bacteria, attributed to the relatively higher concentrations of organic carbon in these waters compared with the open ocean [ 19 , 20 ]. However, due to the complexity and high species diversity of marine biofilms, biofilm-associated bacteria that perform copiotrophic lifestyles have not been well-studied. In the present study, we explored the copiotrophic strategy of marine-biofilm-associated Roseobacteraceae bacterium, focusing on one fast-growing strain isolated from biofilms on coastal stone surfaces. The growth and biofilm formation of this bacterium were examined in the presence of different carbon source concentrations, using Ruegeria pomeroyi DSS-3 as a reference strain. We then adopted genomic, transcriptomic, and metabolomic analyses, as well as physiological and biochemical experiments, to identify specific genes and pathways that contribute to the response of this bacterium to variations in carbon source concentrations.",
"discussion": "3. Discussion In the present study, we discovered the copiotrophic characteristics of a Roseobacteraceae bacterium derived from coastal marine biofilms. By comparing Roseobacteraceae sp. M382 with R. pomeroyi DSS-3, we identified it as a copiotroph. M382 probably represents a new genus that is phylogenetically close to Ruegeria , indicating that physiological comparisons between M382 and DSS-3, a known copiotroph, are reasonable. It is apparent that the carbon concentration in the media has a profound effect on the growth of M382. M382 grew faster and accumulated more biomass than DSS-3 when grown in the presence of a high carbon concentration. Moreover, complete genome sequencing indicated that M382 has a larger genome than DSS-3, and it is known that copiotrophs often have large genomes as they possess more functional genes for transporting or metabolizing carbon sources, reflecting a high degree of flexibility in their response to environmental circumstances [ 29 , 30 ]. In addition, a comparison between M382 and its four close relatives revealed unique genes belonging to transcription factors, also in congruent with the notion that diverse regulatory genes are often detected in the genomes of copiotrophs [ 29 , 30 ]. Multi-omics analyses, including genomics, transcriptomics, and metabolomics, demonstrated that the response of M382 to carbon concentration is largely dependent on amino acid metabolism. Genomic comparisons revealed the presence of many unique genes that were related to the metabolism of various amino acids as well as genes that contributed to biofilm formation. The presence of these unique genes may be the result of the biofilm lifestyle, as marine biofilms have been suggested to be a hotspot for the bacterial acquisition of useful genes [ 31 , 32 ]. These results also implied that amino acid metabolism is adopted by M382. Transcriptomics indicated significant alterations in the expression of 28.4% of genes in response to the environmental carbon concentration, and a number of these genes were responsible for amino acid metabolism. These results highlighted the importance of amino acids in supporting copiotrophy in M382. Moreover, the gene transcription profiles at the high carbon concentration suggested the likelihood of the TCA cycle being directly driven by amino acids, rather than by pyruvate from the EMP, ED, or PP routes. This notion is based on the up-regulation of metabolic pathways related to glutamate, aspartate, and alanine, which can be transformed into 2-oxoglutaric acid, oxaloacetate, and pyruvate, respectively, and enter the TCA cycle. In contrast, genes from the EMP, ED, and PP pathways were observed to be down-regulated at a high carbon concentration. Consistently, the metabolomics results showed that M382 accumulates higher amino acid contents when grown in high concentrations of mixed carbon sources, suggesting the selective consumption of amino acids from the mixed carbon sources. Notably, glutamate, aspartate, and alanine were the top three amino acids ranked by concentration, in line with the transcriptomics results showing the up-regulation of genes related to the metabolism of these three amino acids. In addition, a stronger influence of amino acid concentration on bacterial growth than the saccharide concentration was observed, which is also consistent with the notion that the copiotrophic growth of M382 can be supported by amino acid metabolism. The demonstrated correlation between thiosulfate oxidation, amino acid metabolism, and energy production further supports the role of amino acids in supporting copiotrophy in M382. It is well known that in autotrophic bacteria, especially bacteria living in deep-sea environments, thiosulfate oxidation provides a large proportion of the energy requirements [ 33 ]. Although direct evidence is lacking, it is possible that thiosulfate oxidation is also associated with energy production in heterotrophic bacteria. We addressed this question by measuring the PMF of the wild-type M382 strain as well as two sox mutants. PMF is an electrochemical proton gradient across the cytoplasmic membrane that drives vital processes in cells such as ATP synthesis and the transport of a wide range of substrates [ 34 ]. The higher PMF values observed for the wild-type strain in higher amino acid concentrations point to an association between the metabolism of these carbon and sulfur elements. Moreover, these findings are consistent with the transcriptomics results showing up-regulation of the sox gene cluster by high carbon source concentration."
} | 2,721 |
32211602 | PMC7083766 | pmc | 8,570 | {
"abstract": "In an attempt to mimic the outstanding mechanical properties of wood and bone, a 3D heterogeneous chemistry approach has been used in a biomorphic transformation process (in which sintering is avoided) to fabricate ceramics from rattan wood, preserving its hierarchical fibrous microstructure. The resulting material (called biomorphic apatite [BA] henceforth) possesses a highly bioactive composition and is characterised by a multiscale hierarchical pore structure, based on nanotwinned hydroxyapatite lamellae, which is shown to display a lacunar fractal nature. The mechanical properties of BA are found to be exceptional (when compared with usual porous hydroxyapatite and other ceramics obtained from wood through sintering) and unique as they occupy a zone in the Ashby map previously free from ceramics, but not far from wood and bone. Mechanical tests show the following: (i) the strength in tension may exceed that in compression, (ii) failure in compression involves complex exfoliation patterns, thus resulting in high toughness, (iii) unlike in sintered porous hydroxyapatite, fracture does not occur ‘instantaneously,’ but its growth may be observed, and it exhibits tortuous patterns that follow the original fibrillar structure of wood, thus yielding outstanding toughness, (iv) the anisotropy of the elastic stiffness and strength show unprecedented values when situations of stresses parallel and orthogonal to the main channels are compared. Despite being a ceramic material, BA displays a mechanical behavior similar on the one hand to the ligneous material from which it was produced (therefore behaving as a ‘ceramic with the signature of wood’) and on the other hand to the cortical/spongy osseous complex constituting the structure of compact bone.",
"conclusion": "5 Conclusions BA, a ceramic obtained through a biomorphic transformation process from rattan wood, presents a nanostructure and microstructure that retains elements inherited from its parent material, the ligneous essence. This structure has been shown to generate a lacunar fractal porosity, which yields the outstanding mechanical properties that we have measured for BA. In particular, BA shows a damage and defect tolerance, strength, and elastic stiffness superior to its competitor material, a porous sintered HA, but also to other biomorphic HAs obtained from different types of woods that are present in the literature [ 7 , 8 , 10 ]. Moreover, BA is transversely isotropic so that it can be effectively used in all the situations (for instance, bone replacement) where the ceramic piece is subjected to a load having a privileged direction. The fact that BA is produced without sintering and its superior mechanical performance suggest use of this material for several challenging biotechnologies.",
"introduction": "1 Introduction Wood and bone are natural materials displaying a similar and exceptionally performing mechanical behavior, which is rooted in their highly hierarchical and fibrous microstructure [ 1 , 2 ]. Therefore, mimicking of this microstructure to produce superior materials has been the objective of a massive research investment [ [3] , [4] , [5] , [6] , [7] , [8] , [9] , [10] , [11] , [12] , [13] , [14] , [15] , [16] , [17] , [18] , [19] , [20] , [21] , [22] , [23] , [24] , [25] , [26] ] (see Fig. 23 , Fig. 24 ). In this vein, ceramics implementing a microstructure and nanostructure almost identical to wood can be easily believed to exhibit unprecedented mechanical performances and also to be ideal candidates for bone replacement. Following these ideas, biotemplating processes to fabricate 3D functional ceramics (such as apatites) from natural wood have been proposed, all terminating with sintering [ 5 , 7 , 8 , 10 ]. Recently, Tampieri et al. [ 27 ] developed a new procedure 1 to chemically transform rattan wood into a biomimetic hierarchically structured hydroxyapatite (HA). Unlike in previous approaches where the wood was infiltrated with HA slurries and finally sintered to eliminate the organic component and to consolidate the final ceramic [ 8 , 10 ], the new procedure can directly transform wood pieces into large HA scaffolds, preserving the original multiscale structure through a heterogeneous reaction under supercritical conditions directly in the 3D state, without adopting any sintering process. This is an important advantage because high-temperature treatments can represent a serious drawback for the scaffold bioactivity. 2 The resulting ceramic material (referred henceforth as ‘biomorphic apatite’ [BA]) not only maintains a highly bioactive composition and a multiscale pore hierarchy almost identical to those of the parent wood (thanks to a careful control of the reaction kinetics that prevents critical deformations at all scales and to the absence of sintering processes) but also shows a nanosize structure and unprecedented mechanical properties owing to great cohesion of the neo-formed nanocrystals [ 27 ]. BA represents a significant improvement in the development of 3D inorganic devices with a complex microstructure and multiscale details, which are relevant for smart functionality, which is still an open challenge owing to the ineffectiveness of the current ceramic fabrication processes. The objective of the present study is the systematic investigation of the mechanical properties of BA, which are found to be similar to the mechanical properties of both wood and bone, so that BA becomes promising for bone regeneration, especially for load-bearing regions. A morphological investigation is reported, which shows that the porosity of BA displays a fractal nature of the lacunar type, a feature that may justify its excellent damage tolerance. In fact, the fractality of the porosity explains our experimental observation that the fracture is not abrupt and straight (as usual for ceramics) but evidences growth and tortuosity (as usual for bone). The results from an experimental campaign are presented, based on multiple mechanical tests: uniaxial compression (in both a standard configuration and in situ by scanning electron microscopy), three-point bending, ring tests, and ultrasound evaluations. It is shown that BA performs better than the standard HA [ 31 , 32 ] in several mechanical characteristics: elastic stiffness, strength, damage tolerance before failure, and related toughness. Moreover, unlike in HA and similar to wood and bone, BA is found to exhibit a strength in tension often superior to that in compression, to evidence a transversely isotropic behavior, which permits optimization of stiffness in the direction of loading, a feature particularly useful for bone replacement. Our experiments allow the placement of BA in the Ashby charts, in terms of Young modulus vs strength or vs porosity, Fig. 1 . Fig. 1 Ashby charts reporting Young modulus vs strength (upper part) and vs porosity (lower part) for biomorphic apatite, loaded parallel (BA / / ) and perpendicular (BA ⊥ ) to the microtubule structure (‘grain’ in the following), for rattan wood (from which BA was obtained), and for bones. BA, biomorphic apatite; HA, hydroxyapatite. Fig. 1 The Ashby charts reveal that BA occupies a virgin zone for ceramics and displays a similarity to the ligneous material from which it was ‘born’ and to several different bones, which evidences that the material is particularly suited for several biotechnologies. It has to be finally mentioned that the results reported in the present study indicate that outstanding mechanical properties could be obtained for materials different from BA but obtained with a process similar to that used for BA, which can therefore be used as a guide for the fabrication of a new generation of inorganic materials with significant improvement in structural performance. This perspective is encouraged by previous results obtained with the chemical transformation of natural woods into various oxide (for instance, Al 2 O 3 , ZrO 2 , TiO 2 , and MnO) [ [33] , [34] , [35] , [36] , [37] ], and non-oxide (for instance, SiC, TiC, and ZrC) [ [38] , [39] , [40] , [41] , [42] ] ceramics, which are particularly relevant for structural applications."
} | 2,046 |
39099317 | PMC11328171 | pmc | 8,571 | {
"abstract": "The rise of soft\nrobotics in recent years has motivated significant\ndevelopments in smart materials (and vice versa), as these materials\nallow for more compact robotic designs thanks to the embodied intelligence\nthat they provide. Hydrogels have long been postulated as one of the\npotential candidates to be used in soft robotics due to their softness,\nelasticity, and smart properties that can be tuned with nanomaterials.\nHowever, nowadays they represent only a small percentage of the materials\nused in the field. In this perspective, the drawbacks that have hindered\ntheir utilization so far are analyzed as well as the current state\nof hydrogel-based soft actuators, sensors, and manufacturing possibilities.\nThe future improvements that need to be made to achieve a real application\nof hydrogels in soft robotics are also discussed.",
"conclusion": "Conclusions and Outlook Despite the promising start of electroactive hydrogels in the 90s\nand their early stagnation, hydrogels have once again become part\nof the soft robotics materials palette. Even so, their presence in\nsoft robotic actuators is not significant in comparison to other materials.\nSo far, pneumatic and tendon-driven actuators are still the main option\nin the field, and although smart hydrogels can provide additional\ncapabilities to these systems, smart properties are not always required\nin today’s prototypes. Nonetheless, in search of more\ncompact, integral, and untethered\nsoft robots, elements such as air compressors or motors to pull tendons\nshould be removed. It is in this path that hydrogels must play an\nimportant role in the upcoming years. The smart features that they\nexhibit, the additional properties they offer, such as the self-x\nproperties (self-healing, self-sensing, etc.) or the biomimetism,\nare perfectly suited to the bioinspiration trend, and the huge list\nof potential existing demonstrators makes these materials an appealing\noption to evolve in the soft robotics field, despite still being in\na low level of practical applicability (refer to Figure 8 for a schematic summary of\nthe challenges and opportunities related to hydrogels). Figure 8 Challenges\nand opportunities of hydrogels in soft robotics. The great leap of hydrogels in soft robotics must be achieved through\nan improvement in smart responses and, more importantly, mechanical\nstrength and consistency. Current hydrogel-based actuators whose working\nprinciple is based on smart properties do not exhibit enough force\nfor many applications, so, as commented on in the text, the stiffness\nmodulation to achieve actuators with enough strength is a crucial\nfactor. The chemical formulations, the mechanisms to actively regulate\nthe stiffness, and the addition of nanomaterials play a key role in\nthe future of smart actuators based on hydrogels. 44 However, there is also another way to approach this fact.\nPerhaps we cannot expect these materials to be used to build the kind\nof classic robot that everyone has in mind. We may have to be more\ncreative and think of capabilities that are not possible with the\nrigid materials used so far, robots that perform functions that we\nhave not yet imagined. On the other hand, hydrogels should overcome\nthe evaporation problem.\nWhen working in air conditions, hydrogels exchange water with the\nenvironment. Depending on the ambient humidity, their swelling can\nbe different (drier in a dry environment and more swollen in a humid\nenvironment), which entails changes in their behavior. This situation,\nwhich in some applications could serve as a humidity sensor, is undesirable\nin other cases. Solutions to keep the swelling constant can be found\nat the physical level, like coating the hydrogel, or at the material\nlevel, through the hydrogel’s formulation or adding nanomaterials\nthat prevent the evaporation. 15 , 17 The solution\nto these issues must be accompanied by a series of\nimprovements in the field of soft robotics that affect not only hydrogels.\nAn example is more refined 3D printing techniques to produce complex\nshapes with good definition and including multiple materials, 40 or the development of truly stretchable electrodes\nwith great conductivity, which is something that has been studied\nfor years without getting an outstanding solution that works for all\napplications. Stretchable electrodes with great conductivity would\nbe of great help in the development of electroactive hydrogel-based\nactuators as well as improve the efficiency of hydrogel-based sensors\nand simplify its instrumentation. Beyond the classical robotics\napplications, such as manipulator\narms or grippers, hydrogels must be considered for many robotics-related\napplications thanks to the properties they can offer. For instance,\nbiocompatibility and biomimetism (i.e., the resemblance to biological\ntissues) make them ideal prospects for prosthetics or wearables. That\nis the real advantage of hydrogels over other materials: the vast\nvariety of features that one material can have all-in-one. And that\nis the reason why they are utilized in different fields beyond robotics,\nlike medicine (drug delivery, cell culture, tissue engineering), agriculture,\nor body care and hygienics (diapers, contact lenses, etc.), among\nothers. This versatility favors their insertion not only in known\nissues but also in future unknown problems. Another advantage\nof hydrogels is that they can be biodegradable,\npromoting environmental sustainability. Furthermore, some hydrogels,\nas many other polymeric materials, can be decomposed on purpose to\nreutilize the raw material to generate other hydrogel pieces, reducing\nmaterial wastes and favoring the circular economy. 45 All in all, considering the whole view, hydrogels\nare ideal candidates\nto take over soft robotics in the upcoming years due to the large\nlist of properties they can exhibit (e.g., response to stimuli, self-healing,\ntransparency, biocompatibility, biomimetism, etc.), which are aligned\nwith bioinspiration in the search for a living organism-like robot.\nTheir future in the field is guaranteed."
} | 1,508 |
39246340 | PMC11376081 | pmc | 8,575 | {
"abstract": "Relaxor ferroelectrics are well-known for their high dielectric constants, low dielectric losses, and excellent electromechanical properties, making them valuable for various electronic devices. Despite recent efforts to enhance the durability of ferroelectrics through chemical cross-linking, achieving elasticity in relaxor ferroelectric materials remains a significant challenge. These materials inherently possess traits such as low crystallinity and small crystal size, while chemical crosslinking tends to diminish polymer crystallinity considerably. Thus, a key obstacle to making relaxor ferroelectric polymers elastic lies in safeguarding their crystalline regions from the effects of slight crosslinking. To tackle this issue, we selected P(VDF-CTFE-DB) with highly reactive 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 C double bonds as crosslinking sites, reducing the amount of cross-linking agents added and thereby lessening their impact on crystallinity. Through peroxide crosslinking, we transformed linear P(VDF-CTFE-DB) into a network structure, successfully producing a resilient relaxor ferroelectric material with maintained polarization intensity for ferroelectricity. Notably, this elastic relaxor ferroelectric was synthesized at relatively low temperatures, exhibiting a remarkable dielectric constant, superior resilience, fatigue resistance, and a stable ferroelectric response even under strains of up to 80%. Our approach paves the way for developing low-cost, high-dielectric-constant elastomers suitable for wearable electronics and related applications.",
"conclusion": "Conclusions Through a successful free-radical crosslinking reaction at relatively low temperatures, we have synthesized relaxor ferroelectric materials exhibiting high elastic recovery while retaining their ferroelectricity. This material features a relatively high dielectric constant (∼22 at 100 Hz) and a low modulus (∼10 MPa), which is the highest value among all reported intrinsic elastomers. In addition, the superior resilience and fatigue resistance of our elastomer is better than that of commercial fluorubbers. It demonstrates outstanding stability in organic solvents, strong acids, and strong bases. Even under strains as high as 80%, it maintains a stable ferroelectric response. Our strategy provides a novel approach to the elastification of relaxor ferroelectric materials, offering a new elastomer for applications in the field of flexible electronics, such as soft robotics and elastic electronic skin.",
"introduction": "Introduction Relaxor ferroelectrics (RFEs) possess vast potentials in contemporary electronic devices, particularly within the realm of flexible electronics, designated as the preferred materials for numerous electronic components, including acoustic sensors, solid-state coolers, transducers, and actuators. This preference stems from their exceptional properties, including high dielectric constant, low hysteresis, and superior electromechanical and pyroelectric properties. 1–5 Therefore, numerous efforts have been made to explore new methods, such as ‘ferroelectric chemistry’ to advance the synthesis of novel ferroelectrics. 6–8 Classic relaxor ferroelectrics, such as PMN-PT, exhibit diffuse phase transitions and strong frequency-dependent dielectric properties, making them highly suitable for applications in actuators, sensors, and transducers due to their large electro-mechanical coupling and high piezoelectric coefficients. The emergence of wearable electronics has propelled the demand for the elastification of functional materials, 9–12 leading us to develop elastic polymer ferroelectrics capable of delicately balancing crystallinity and resilience by slight crosslinking. 13 During the elastification wave of functional materials, relaxor ferroelectrics have lagged behind. 14,15 Polymer-based relaxor ferroelectric materials, such as the poly(vinylidene fluoride) (PVDF)-based terpolymers or tetrapolymers, have emerged as the premier choice for rendering relaxor ferroelectric materials elastic, due to their solution processability, chemical stability, and high flexibility. 16–19 Currently, significant strides have been made in elucidating the fundamental mechanisms underlying PVDF-based relaxor ferroelectric materials, thereby guiding the discovery of new relaxor ferroelectric organic materials for flexible, scalable, biocompatible sensor and energy applications. 20 The elastification of relaxor ferroelectric polymers holds profound importance and warrants comprehensive investigation, given its pivotal role in wearable or flexible electronics. However, the high cost of PVDF terpolymers or tetrapolymers, combined with their inherent challenges of low crystallinity and small crystal domains, 21 presents a dual challenge involving cost consideration and the intricate balancing between crystallinity and ferroelectricity in the elastification process of relaxor ferroelectric polymers. To address the challenge, we meticulously selected a PVDF-based polymer, P(VDF-CTFE-DB), 22,23 as the starting material for several reasons. Firstly, P(VDF-CTFE) is highly cost-effective compared to other PVDF-based copolymers, with its cost being less than one percent of that of commonly used relaxor ferroelectric polymers like P(VDF-TrFE-CFE) (Table S1 † ). Secondly, the molecular structure of P(VDF-CTFE-DB) incorporates unsaturated –C C– bonds, which not only simplifies the synthesis process but also significantly enhances reactivity and crosslinking efficiency. 24 Lastly, the synthesis of P(VDF-CTFE-DB) from the affordable P(VDF-CTFE) involves only a single-step reaction in a basic environment using triethylamine. 22 This straightforward synthesis, coupled with simple separation, is highly suitable for industrial scale-up. In our previous work, the thermal crosslinking process requires high temperatures of up to 240 °C, 13 posing an additional high risk for CMOS or organic electronic procedures. 25–27 Thus, there is an urgent quest for alternative methods that operate under milder conditions to realize the elastification of ferroelectrics, 28,29 Herein, we intentionally opted for peroxide crosslinking, motivated by three key considerations. Firstly, the crosslinking temperature can be decreased by choosing suitable and active peroxides. 30 Secondly, peroxide crosslinking relies on highly reactive free radicals, ensuring high crosslinking efficiency, thus requiring lower additive amounts for efficient crosslinking. 31,32 Thirdly, the films prepared from peroxide crosslinking usually exhibit high thermal stability and excellent chemical stability due to their dense network structure. 33,34 Bis(tert-butyldioxyisopropyl)benzene (BIPB) was selected as the initiator for the crosslinking due to its multifaceted advantages. This choice facilitates a reduction in additive amounts, along with offering high thermal decomposition activation energy, a slow decomposition rate, and the production of high-quality crosslinked products, minimal release of irritating odors, contributing to environmental friendliness. 35,36 Additionally, triallyl isocyanurate (TAIC) was chosen as the co-crosslinking agent to enhance crosslinking efficiency. As the best co-crosslinking agent, the triazine ring in TAIC imparts strong chemical and thermal stability, reinforcing the crosslinking network and further improving the thermodynamic and chemical stability of the crosslinked samples. 31,32,37 Herein, a relaxor ferroelectric elastomer (crosslinked-P(VDF-CTFE-DB)) was successfully prepared by peroxide crosslinking at low temperatures (160 °C), exhibiting a high elastic recovery, high-dielectric constant, superior chemical and thermal stability, and cost-effectiveness. The resulting relaxor ferroelectric displays a high dielectric constant (approximately 22 at room temperature and 100 Hz), and a broad ferroelectric-to-paraelectric transition temperature range, indicating its capacity to maintain high dielectric constant across various temperature ranges. Moreover, this crosslinked P(VDF-CTFE-DB) film maintains a stable ferroelectric response even under strains up to 80%. In particular, compared to commercial fluorubbers, it exhibits superior resilience and fatigue resistance. By employing the peroxide initiator BIPB and the co-crosslinking agent TAIC, we obtained relaxor ferroelectric elastomers with high resilience and high dielectric constants at room temperature under a mild crosslinking condition.",
"discussion": "Results and discussion Preparation of elastic polymer RFEs via peroxide crosslinking The polymer P(VDF-CTFE-DB) was synthesized following the reported method, 22 and its 1 H NMR spectrum in acetone-d 6 is shown in Fig. S1. † The pristine P(VDF-CTFE-DB) was dissolved in acetone, with a certain ratio of BIPB and TAIC. The resulting solution was then either cast or spin-coated onto substrates to form thick and thin films, respectively. After natural evaporation of the solvent in a fume hood, the blended P(VDF-CTFE-DB) was obtained with any residual solvent completely removed under vacuum. Subsequently, thermal crosslinking was conducted by step-heating in ambient air to gain the crosslinked P(VDF-CTFE-DB). The crosslinking mechanism is illustrated in Fig. 1 . Upon heating, the peroxide bonds in BIPB undergo homolytic cleavage, yielding alkoxyl free radicals. 31,32,38 These radicals subsequently initiate the attack on unsaturated C C bonds within the PVDF-based polymer, generating polymer free radical intermediates. Interaction between these intermediates and the allylic bonds present in the tri-functional TAIC initiates the crosslinking reaction. Notably, TAIC played a crucial role as the co-crosslinking agent, facilitating the formation of the network with higher crosslinking density even at low BIPB feeding amounts. 37 Throughout this crosslinking process, by-products such as acetone, butanol, and methane were released in gaseous form. Fig. 1 The formation mechanism of elastic polymer ferroelectric by peroxide crosslinking. Crosslinking and characterization of elastic RFE films Through differential scanning calorimetry (DSC) analysis of the blended P(VDF-CTFE-DB) films illustrated in Fig. 2a , an exothermic peak corresponding to the crosslinking reaction is observed within the range of 120–215 °C. To ensure thorough crosslinking, thermal treatment at 160 °C for one hour was employed, resulting in complete crosslinking, as confirmed by the absence of an exothermic peak during the first heating cycle of the DSC curve (the inset of Fig. 2a ) for the crosslinked P(VDF-CTFE-DB) film. Subsequently, DSC testing was conducted on crosslinked P(VDF-CTFE-DB) films with varying ratios of BIPB and TAIC, as shown in Fig. S2. † It is observed that the content of the crosslinking initiator within this range had no significant impact on the melting temperature of the crosslinked P(VDF-CTFE-DB) film. Nevertheless, as the ratio of the BIPB and TAIC increased, the crystalline melting enthalpy of the crosslinked P(VDF-CTFE-DB) film consistently decreased. It reaches an extremum at a peroxide initiator ratio of 10%, indicating enhanced crystallinity at this point—a crucial assurance for a favorable ferroelectric relaxor response, as illustrated in Fig. 2b . Fig. 2 Crosslinking characterization of elastic RFE films with the ratios of BIPB and TAIC is 10% and 8%, respectively. (a) DSC curves of the blended P(VDF-CTFE-DB) (‘B-DB’ for short) film indicating in situ crosslinking. (Inset) DSC curves of the crosslinked P(VDF-CTFE-DB) (‘C-DB’ for short). (b) Melting enthalpy of crosslinked P(VDF-CTFE-DB) films with different crosslinking densities. (c and d) FT-IR spectra (c) and XRD patterns (b) of pristine (‘DB’ for short), blended and crosslinked P(VDF-CTFE-DB) films. The crosslinked P(VDF-CTFE-DB) films exhibited excellent thermal and chemical stability. Specifically, the crosslinked P(VDF-CTFE-DB) film demonstrated exceptional thermal stability, with a temperature of decomposition ( T d ) exceeding 400 °C, as evidenced by the thermogravimetric analysis curve (Fig. S3 † ). Additionally, when subjected to various organic solvents such as acetone, cyclohexanone, isophorone, and DMF, the crosslinked P(VDF-CTFE-DB) films displayed stability, as shown in Fig. S4. † After immersion in these solvents for two weeks, minimal swelling was observed, and the gel contents remained around 90%, as presented in Table S2. † Moreover, in subsequent acid–base resistance tests, no obvious color or volume changes were observed in the crosslinked P(VDF-CTFE-DB) films after two weeks of immersion in concentrated sulfuric acid and saturated sodium hydroxide aqueous solution, as depicted in Fig. S5. † These findings highlight the enhanced stability of electronic devices fabricated using crosslinked P(VDF-CTFE-DB) films, thereby expanding their potential applications. Fourier-transform infrared spectroscopy (FT-IR) analysis revealed structural changes in the films pre- and post-crosslinking, as shown in Fig. 2c . In the blended P(VDF-CTFE-DB) film, the disappearance of the C C double bond signal of the crosslinking agent TAIC at ∼1697 cm −1 after crosslinking suggests its extensive involvement during the crosslinking reaction. Notably, the C C double bonds within the P(VDF-CTFE-DB) chain did not fully participate in the crosslinking process, as evidenced by the persistent presence of C C double bonds at 1720 cm −1 (Fig. S6a † ). After crosslinking, there was a notable increase in the content of the α-phase (610 cm −1 ) and β-phase (1060 cm −1 ), particularly in the α-phase, as observed in Fig. S6b–S6d. † This finding was further confirmed by X-ray diffraction (XRD) analysis ( Fig. 2d ), which demonstrated an increased content of α-phase (18.9°) and β-phase (19.2°) after crosslinking. 39 Further insights were gained into the significant influence of the content of BIPB and TAIC on the mechanical properties of the crosslinked P(VDF-CTFE-DB) films. Stress–strain tests were conducted on crosslinked P(VDF-CTFE-DB) films with varying ratios of the peroxides, as depicted in Fig. 3 . When comparing to the pristine P(VDF-CTFE-DB) film lacking a crosslinking initiator (Fig. S7 † ), the elongation at break sharply decreased from ∼1300% for the pristine P(VDF-CTFE-DB) to ∼340% for the crosslinked P(VDF-CTFE-DB) with a BIPB ratio of 2%. As the content of the BIPB and TAIC increased, a decreasing trend in the elongation at break was observed, alongside an increasing trend in modulus, as shown in Fig. 3b . Furthermore, cyclic stress–strain tests were conducted on crosslinked P(VDF-CTFE-DB) films with initiator ratios of 8% and 10% (refer to Fig. S8 † ), indicating superior cyclic performance for the films with a 10% BIPB ratio. Based on the comprehensive results of mechanical properties and crystalline melting enthalpy, the 10% initiator ratio was deemed more suitable for achieving a balance between elasticity and crystallinity, thus using it as the preferred ratio for all subsequent experiments. Fig. 3 Mechanical properties of elastic RFE films. (a) Stress–strain curves of crosslinked P(VDF-CTFE-DB) films at various crosslinking initiator ratios. (b) Modulus and elongation of crosslinked P(VDF-CTFE-DB) films at various crosslinking initiator ratios. (c) Cyclic stress–strain curves of crosslinked P(VDF-CTFE-DB) films with a crosslinking initiator ratio of 10% and pristine P(VDF-CTFE-DB) films (inset) under different strains (the first cycle was omitted owing to clamp sliding and the X -axis is shifted for clarity), indicating the excellent resilience of the crosslinked P(VDF-CTFE-DB) films compared with the unrecoverable deformation of the pristine P(VDF-CTFE-DB) films. (d) The fatigue resistance of crosslinked P(VDF-CTFE-DB) film with a crosslinking density of 10% under 50% strain, compared to that of a commercial fluorubber P(VDF-HFP) (DAI-EL, G-801). Further cyclic stress–strain testing was performed, comprising 5 cycles (depicted in Fig. 3c ) and an extended 3000 cycles (illustrated in Fig. 3d ). The results indicated that the recovery rate of the film remained consistently above 95% under strains ranging from 40% to 80% during cyclic loading. Compared to commercial fluorubber, the crosslinked P(VDF-CTFE-DB) film exhibited superior fatigue resistance and higher recovery ratios. Remarkably, even after 100 cycles, the crosslinked P(VDF-CTFE-DB) film promptly maintained stable elastic recovery. To decipher the underlying mechanism behind the film's elasticity, stress–temperature performance testing was conducted on the crosslinked P(VDF-CTFE-DB) films. The elasticity of the crosslinked P(VDF-CTFE-DB) film, as determined by varying the temperature under different strains (Fig. S9 † ), was attributed to entropy elasticity rather than energy elasticity. 13 This inherence arises from the transformation of macromolecular chain conformations within the crosslinked P(VDF-CTFE-DB) film from coil-like to rod-like shape under external forces, resulting in a decrease in the entropy change. The rod-like state of the system becomes unstable, and upon removal of the external force, due to thermal motion, the molecular chains spontaneously tend toward an increase of the entropy change in the system. Consequently, the molecular chains revert from the rod-like shape to the coil-like shape, exhibiting excellent elastic recovery properties. 40–43 These results confirm that the intrinsic elasticity of relaxor ferroelectric polymers is achieved through peroxide initiation at a relatively low temperature. Relaxor ferroelectricity of crosslinked P(VDF-CTFE-DB) The ferroelectric relaxor behavior of crosslinked P(VDF-CTFE-DB) is characterized using dielectric-temperature curves, polarization-electric field curves, and piezoresponse force microscopy (PFM) testing, as illustrated in Fig. 4 . Fig. 4 Ferroelectric properties of crosslinked P(VDF-CTFE-DB) films. (a) Temperature-dependent ε and dielectric loss of crosslinked P(VDF-TrFE-DB) films at different frequencies. (b and c) P – E loops of Au/C–P(VDF-CTFE-DB)/Au under different electric fields at 1000 Hz (b) and different frequencies (c). (d) Phase-voltage hysteresis and amplitude–voltage butterfly loop. (e and f) phase (e) and amplitude (f) mapping of PFM. The dielectric–temperature curve of the crosslinked P(VDF-CTFE-DB) film reveals distinct relaxor characteristics with a broader Curie transition temperature ( T c ) range. As the test frequency increases, the T c shifts towards higher temperatures, enabling the crosslinked P(VDF-CTFE-DB) film to possess a higher dielectric constant near room temperature. Notably, both the T c and the peak temperature of dielectric loss shift towards higher temperatures with increasing frequency, demonstrating significant frequency dependence. Compared to pristine P(VDF-CTFE-DB) (Fig. S10 † ), the crosslinked film exhibits a higher dielectric constant and a wider range of low dielectric loss temperatures. At room temperature and under a frequency of 100 Hz, the dielectric constant reaches approximately 22, with a dielectric loss tangent below 0.1 ( Fig. 4a ). The Vogel–Fulcher relationship is employed to describe the dynamics of thermally activated dipoles and the freezing behavior of relaxor ferroelectrics. It reflects the freezing behavior caused by the interaction between dipoles. The Vogel–Fulcher relation as shown in eqn (1) (ref. 44 ) 1 where E a is the activation energy, f is the probing frequency k B is the Boltzmann constant (8.617 × 10 −5 eV k −1 ), f 0 is the relaxation frequency of dipoles, T f is the freezing temperature, and T m is the temperature corresponding to the maximum dielectric constant. Vogel–Fulcher equation fits of our permittivity data measured for frequencies ranging between 100 Hz and 1 MHz are given in Fig. S11. † They show the predicted linear relation between the reduced frequency [ln( f / f 0 )] −1 and T max . All fitting parameters are consistent with the Vogel–Fulcher law. Intrinsic elastomers with such high dielectric constant and low dielectric loss are the key materials for soft robots and wearable devices, serving as elastic sensors, energy storage units, and actuators. The P – E loops of crosslinked P(VDF-CTFE-DB) were obtained using a sandwich structure device (Au/C-DB/Au/Si). As shown in Fig. 4b , the loops exhibited a low rectangularity, appearing slender, which is a typical feature of a relaxor ferroelectric material. As the applied electric field increases, the initial hysteresis does not gradually expand but maintains a slender shape, characterized by a significantly large ratio between saturated polarization ( P max ) and remanent polarization ( P r ), resulting in lower rectangularity. At 1 kHz and 330 MV m −1 , the P max and P r of the crosslinked P(VDF-CTFE-DB) film are 6.23 and 0.87 μC cm −2 , respectively. In comparison, the pristine P(VDF-CTFE-DB) exhibits P max and P r values of 3.32 and 0.52 μC cm −2 , respectively (Fig. S12a † ). The coercive field ( E c ) of ∼36 MV m −1 remains unchanged before and after crosslinking. However, as the applied electric field continues to increase, both P max and P r continue to grow, exhibiting a saturating trend. Eventually, at a high electric field of 880 MV m −1 , P max and P r reach 14 and 4.5 μC cm −2 , respectively, with a coercive field E c of approximately 177 MV m −1 (Fig. S13 † ). This demonstrates a pronounced saturation trend, highlighting the saturated characteristics of relaxor ferroelectric material under a strong electric field. The P r of the crosslinked P(VDF-CTFE-DB) increases from 0.51 to 3.23 μC cm −2 across test frequencies ranging from 10 kHz to 10 Hz ( Fig. 4c ). Compared to pristine P(VDF-CTFE-DB) (Fig. S10b † ), the crosslinked P(VDF-CTFE-DB) film exhibits more pronounced relaxor behavior due to a more condensed network introduced by peroxide crosslinking. The introduction of C C double bonds in the ferroelectric polymer reduces spatial hindrance for ferroelectric phase formation and promotes the generation of the ferroelectric phase. This enhancement is partly attributed to the increase in both the non-polar α-phase and the polar β-phase. In addition, the enhanced polarization effect at the interface resulting from the crosslinked network structure also contributes to polarization. However, the polarization at the interface cannot persist after the removal of the electric field, leading to a significant increase in saturated polarization while P r undergoes minimal changes, thereby enhancing the relaxor behavior of the crosslinked P(VDF-CTFE-DB). We utilized PFM to investigate the phase and amplitude changes of crosslinked P(VDF-CTFE-DB) under the stimuli of positive and negative electric fields, 45 confirming the excellent piezoelectric properties of the crosslinked P(VDF-CTFE-DB) film ( Fig. 4d–f ). A 250 nm-thick film was spin-coated on a Pt/Si substrate, and after crosslinking, PFM electrically conductive probes were utilized for ferroelectric property testing. Hysteresis and butterfly curves obtained from a single scan demonstrate the complete switching of ferroelectric domains under the influence of the electric field ( Fig. 4d ). By applying a −10 V bias to a 5 × 5 μm 2 region, maintaining the same center, and subsequently applying a +10 V bias to a 3 × 3 μm 2 region, phase and amplitude maps of a “box-in-box” pattern were obtained ( Fig. 4e and f ), suggesting that the polarity of the ferroelectric domains in the crystalline can be reversibly switched by applied field in a zone rather than only a spot. Additionally, PFM reveals a piezoelectric coefficient of 10.7 pm V −1 (Fig. S14 † ), which is a typical value for PVDF-based ferroelectric polymers. Relaxor ferroelectricity response of crosslinked P(VDF-CTFE-DB) under strains A fully elastic capacitor device, utilizing liquid metal (Ga, gallium) as elastic electrodes, was fabricated using a sacrificial-layer microfabrication method to investigate the relaxor ferroelectric response of the crosslinked P(VDF-CTFE-DB) film under various applied strains and frequencies. 16 The schematic structure of the elastic device is illustrated in Fig. 5a . The elastic device was affixed in a custom-made single-shaft tensile clamp and gradually stretched to 80% strain, as depicted in Fig. 5b . The test results are summarized in Fig. 5c,d . Fig. 5 FE response of elastic RFEs under strains. (a and b) Schematic structure of elastic device (a) in the stretching set-up under 0 and 80% strain (b). (c) P – E loops at 1 kHz under strain from 0 to 80%. (d) P r , P max , and E c under different strains. The P – E loops of this fully elastic device (Fig. S15 and 16 † ) without strain are similar to that of the rigid device with Au electrodes, and the P r remains almost constant throughout the stretching process (Fig. S17–S24 † ). Although there is a slight fluctuation in the P max during stretching, the coercive electric field remains nearly constant within the 0–40% strain range. However, it begins to fluctuate when the strains reach 40–60% and remain stable again in the 60–80% strain range. These observations suggest that the RFE response of the elastic relaxor ferroelectrics is insensitive to stress and frequency. Intrinsic elastomer with high dielectric constants for wearable electronics High-dielectric-constant elastomers are crucial components in emerging wearable electronics, such as non-volatile memory, energy-storage devices and fully elastic circuits. 46,47 The elastomers for wearable electronics will undergo large strains (the maximum up to 30–50%) at high frequencies (0.1–100 Hz), therefore, they require high resilience (elastic recovery, fatigue resistance), excellent electrical performance (high dielectric constant and low loss, high break-down electric field), and efficient solution-processability. 48 In addition, cost-effectiveness is also a primary consideration for materials used in wearable electronics, given their status as daily consumables with a massive demand for consumption. However, the properties of current intrinsic elastomers fall short of the above requirements. For example, the dielectric constant of the most commonly used PDMS elastomer in wearable devices is just 2.8 at 1 kHz, and even the best dielectric elastomers have a range of 3–8. 49 Therefore, it is urgent to develop intrinsic elastomers with high dielectric constant and low loss, robust elastic properties at low cost. Compared to PDMS and other reported intrinsic elastomers, our crosslinked P(VDF-CTFE-DB) exhibits a dielectric constant of ∼20 at 1 kHz, which is the highest value among all reported intrinsic elastomers. The resilience of crosslinked P(VDF-CTFE-DB) is also as excellent as PDMS and surpasses most reported dielectric elastomers. Even compared with commercial PVDF ternary polymer materials, our material is not only cost-effective but also exhibits a low modulus and high resilience (Fig. S25 † ). One of the most notable advantages of our cross-linked P(VDF-CTFE-DB) is its cost, which is less than 1% of that of P(VDF-TrFE) and PVDF-based terpolymers. Above all, the outstanding performance of our crosslinked P(VDF-CTFE-DB) films significantly broadens the potential application scenarios of organic relaxor ferroelectric materials."
} | 6,967 |
36627716 | PMC9832610 | pmc | 8,576 | {
"abstract": "Background Bacillus subtilis is generally regarded as a ubiquitous facultative anaerobe. Oxygen is the major electron acceptor of B. subtilis , and when oxygen is absent, B. subtilis can donate electrons to nitrate or perform fermentation. An anode electrode can also be used by microorganisms as the electron sink in systems called anodic electro-fermentation. The facultative anaerobic character of B. subtilis makes it an excellent candidate to explore with different electron acceptors, such as an anode. This study aimed to optimise industrial aerobic bioprocesses using alternative electron acceptors. In particular, different end product spectrum of B. subtilis with various electron acceptors, including anode from the electro-fermentation system, was investigated. Results B. subtilis was grown using three electron acceptors, i.e. oxygen, nitrate and anode (poised at a potential of 0.7 V vs. standard hydrogen electrode). The results showed oxygen had a crucial role for cells to remain metabolically active. When nitrate or anode was applied as the sole electron acceptor anaerobically, immediate cell lysis and limited glucose consumption were observed. In anode-assisted electro-fermentation with a limited aeration rate, acetoin, as the main end product showed the highest yield of 0.78 ± 0.04 mol product /mol glucose , two-fold higher than without poised potential (0.39 ± 0.08 mol product /mol glucose ). Conclusions Oxygen controls B. subtilis biomass growth, alternative electron acceptors utilisation and metabolites formation. Limited oxygen/air supply enabled the bacteria to donate excess electrons to nitrate or anode, leading to steered product spectrum. The anode-assisted electro-fermentation showed its potential to boost acetoin production for future industrial biotechnology applications. Supplementary Information The online version contains supplementary material available at 10.1186/s13068-022-02253-4.",
"conclusion": "Conclusions The biomass production, alternative electron acceptors utilisation and end product spectrum of the B. subtilis strain differed under different oxygen supplies. Both nitrate and anode-assisted respiration of B. subtilis occurred under oxygen-limited conditions, while fast cell lysis was observed in strictly anaerobic environments. Overall, depending on the electron acceptor the biochemical production shifted from acetate in aerobic conditions to acetate and acetoin under nitrate respiration, and further to mostly acetoin in BESs. Under limited aeration, poised anode potential at 0.7 V showed steered product spectrum towards acetoin production (17.8 ± 0.6 mM at applied potential vs. 9.7 ± 2.0 mM at open circuit), with lactate and 2,3-butanediol as intermediate products. In such anode-assisted systems, a high acetoin yield of 0.7 mol product /mol glucose was achieved with more balanced energy distribution between biomass production and end product formation compared to systems without poised anode potential or aerobic shake flask experiments. This research highlighted the potential of anode-assisted electro-fermentation to connect an industrial aerobic microorganism with bioelectrochemical systems with reduced oxygen dependency for biochemical production.",
"discussion": "Discussion The role of oxygen for B. subtilis growth and end product spectrum In this study, the absence of oxygen restricted the metabolic activity of B. subtilis and resulted in abrupt cell lysis, although many studies have shown B. subtilis can grow without oxygen using alternative electron acceptors, including nitrate [ 13 , 32 , 33 ]. Among all the tested conditions, the highest cell density ( OD 600 = 5.1) was measured in aerobic shake flasks with an average doubling time of 1.6 h. A typical oxygen volumetric mass transfer coefficient (k L a) in similar aerobic shake flasks and growth conditions is ranging from 160 to 226 h –1 [ 34 ], nearly five times higher compared to the estimated 36 h –1 in H-type reactors used with moderate aeration in anode-assisted electro-fermentation in this study [ 35 ]. When nitrate or anode was used as the alternative electron acceptor by B. subtilis , lower glucose consumption rates (< 0.7 mM/h) and cell densities ( OD 600 < 2) compared to aerobic shake flasks (2.3 mM/h) were observed, indicating that limited energy was available with restricted biomass formation when using the alternative electron acceptors. Oxygen as the electron acceptor facilitates sugar metabolism, energy conservation and nicotinamide cofactors, e.g. nicotinamide adenine dinucleotide (NAD + ), regeneration in many microorganisms [ 2 ]. In B. subtilis , reduced dissolved oxygen concentration during cultivations has also earlier shown enhanced fermentation products with 50% less biomass formed [ 24 ]. According to the results, low oxygen supply enabled B. subtilis to donate electrons to alternative electron acceptors. The results suggested that low concentrations of oxygen were essential for cells to activate nitrate and anode-assisted respiration. Oxygen might not only function as electron acceptor but could have also acted as the transcriptional signal chemical to enable the anaerobic pathway [ 35 ]. For example, oxygen limitation induces the transcription of anaerobic genes and regulators for nitrate respiration in B. subtilis [ 36 ]. For anode-assisted respiration, oxygen may have played a more indispensable role as electron acceptor, since constant supply of oxygen was needed. At the energy level, nitrate respiration releases more free energy than fermentation: ΔG 0 = − 2870 kJ/mol glucose for oxygen, − 858 kJ/mol glucose for nitrate and − 218 kJ/mol glucose for fermentation, respectively [ 2 , 37 , 38 ]. Cell growth ( OD 600 = 1.7) was also observed under the microaerobic condition with consumption of nitrate in contrast to fast cell lysis under anaerobic conditions in this study. However, the fundamental understanding of nitrate respiration in B. subtilis under the microaerobic condition remains to be further examined. Different aeration schemes affected the end product spectrum of B. subtilis . In aerobic shake flasks, acetate was the main end product (15.2 mM), whereas in BES experiments with both aeration schemes one-fourth of acetate concentration was measured. In B. subtilis , acetate is typically formed from acetyl-CoA via a two-step reaction encoded by pta and ackA and the formation of acetate during aerobic respiration is explained by metabolic overflow, which activates fermentative pathways to allow faster ATP production per unit cell membrane area [ 39 – 41 ]. During the nitrate respiration, besides acetate formation, other fermentation products, including acetoin and 2,3-butanediol, also started to appear. An increase in overall fermentative activities can be expected when oxygen is not available for the cofactors’ regeneration. In B. subtilis , increased NADH-coupled products secretion, i.e. lactate and 2,3-butanediol (up to 16.3 mM and 12.3 mM, respectively), was observed under the oxygen-limited conditions in BESs. However, lactate was not excreted in nitrate-added serum flasks. It has been reported that under nitrate respiration fermentation is less preferred by B. subtilis in terms of cofactors regeneration [ 36 ] and that the presence of nitrate under oxygen-limited conditions can drastically decrease fermentative enzyme activities, including lactate dehydrogenase [ 42 ]. Anode-assisted electro-fermentation with B. subtilis Previous anodic electro-fermentation studies have demonstrated that the combination of poised anode potential (0.7 V vs. SHE) and K 3 [Fe(CN) 6 ] as the redox mediator enhanced the anaerobic metabolic activities of Pseudomonas putida F1 and Corynebacterium glutamicum lysC [ 25 , 26 , 43 ]. Under anode respiration in this study, B. subtilis cells were unable to maintain metabolic activity showing incomplete glucose oxidation. In contrast to the shake flasks, anaerobic BESs had almost zero glucose consumption, no planktonic cell growth and limited end product formation. Moderate aeration resulted in at least three-fold more charge transferred to the anode compared to anaerobic conditions. However, under moderate aeration, the anode did not show significant impacts on changing the product spectrum and oxygen was likely still acting as the main electron acceptor. With moderate aeration, the anode could not function as an electron sink for cells to uninterruptedly balance the redox per se causing almost identical end product concentrations with or without poised anode potential. When the oxygen supply was switched to limited aeration in BESs, cells were forced to seek alternative electron acceptors and an increased number of surplus electrons (926.2 ± 22.7 C, limited aeration) with enhanced acetoin formation were observed. Chen et al. [ 44 ] suggested that NAD(H) + played an important role during the extracellular electron transfer of B. subtilis , which likely explained the different NADH-coupled metabolites showed in this study: as the total amount of electrons shuttled to the anode increased over time, cofactor levels might have been altered by the anode, in which case less NADH-coupled metabolites are needed to balance the redox state (e.g. lactate and 2,3-butanediol). In B. subtilis, NAD + regeneration is typically mediated by cytoplasmic lactate dehydrogenase [ 45 ]. One hypothesis is that in anode-assisted systems under limited aeration, NAD + was regenerated in several ways; hence, the lower lactate concentration was observed under the poised potential. However, anode did not significantly influence other fermentation steps which involved NAD + regeneration, e.g. production of 2,3-butanediol from acetoin, as similar levels of 2,3-butanediol productivities were observed under limited aeration with poised potential (0.11 ± 0.02 mmol/L/h) and in open circuit (0.09 ± 0.03 mmol/L/h). On the other hand, if the anode helped to increase the cofactor levels, such as NAD + , the conversion of lactate to pyruvate might also have been accelerated. Pyruvate can either re-enter the TCA cycle to potentially enhance the acetoin production or accumulate in cells to trigger the acetoin excretion as a strategy for cells to maintain the intercellular pH during the stationary phase [ 45 ]. Alternatively, the applied anode potential may also change the oxidation–reduction potential (ORP) of the fermentation broth or the intracellular redox state, the principle of which is still to be discovered [ 47 ]. Higher acetoin yield with anode-assisted electro-fermentation systems Adaptation of B. subtilis to alternative electron acceptors revealed future perspectives of using anode-assisted electro-fermentation under oxygen-limited conditions for enhanced acetoin production. First, among all tested conditions (Table 1 ), the high yield of acetoin (0.71–0.78 mol product /mol glucose ) achieved under oxygen-limited conditions (Fig. 5 ) is comparable to the acetoin production of metabolically engineered B. subtilis strains (from 0.62 to 0.77 mol product /mol glucose [ 48 – 50 ]). Although the highest production rate of acetoin (0.3 ± 0.1 mM/h) was found under nitrate respiration, a lower yield (0.3 ± 0.01 mol product /mol glucose ) indicated the limitation for acetoin production under the anaerobic respiration. The results suggested that aerobic acetoin bioprocesses can be optimised with a very low air supply rate along with poised anode potential. A low aeration rate can bring financial benefits regarding operational costs and fewer foaming-related issues for industrial applications [ 2 ]. Furthermore, another industrially related strain, P . putida KT2440, was reported to improve glycolipid surfactants production under similar oxygen-limited conditions with poised anode potential [ 51 ], highlighting the potential of anodic electro-fermentation to decrease the oxygen requirements of aerobic bioprocesses. Fig. 5 Maximum production yields (x-axis) associated with production rates (y-axis) under different conditions tested with oxygen, nitrate and anode as electron acceptors. In contrast to acetoin (neon green) and acetate (metallic violet), intermediate products 2,3-butanediol (crimson red) and lactate (light azure) concentrations declined during the experiments; therefore, only the highest production rates are shown"
} | 3,101 |
37435552 | PMC10331310 | pmc | 8,577 | {
"abstract": "The dynamic adhesive systems in nature have served as inspirations for the development of intelligent adhesive surfaces. However, the mechanisms underlying the rapid controllable contact adhesion observed in biological systems have never been adequately explained. Here, the control principle for the unfolding adhesive footpads (alterable contact area) of honeybees is investigated. The footpads can passively unfold, even without neuro-muscular reflexes, in response to specific dragging activity (generating shear force) toward their bodies. This passive unfolding is attributed to the structural features of the soft footpads, which cooperate closely with shear force. Then, the hierarchical structures supported by numerous branching fibers were observed and analyzed. Experimental and theoretical findings demonstrated that shear force can decrease fibril angles with respect to the shear direction, which consequently induces the rotation of the interim contact area of the footpads and achieves their passive unfolding. Furthermore, the decrease in fibril angles can lead to an increase in the liquid pressure within the footpads, and subsequently enhance their unfolding. This study presents a novel approach for passively controlling the contact areas in adhesive systems, which can be applied to develop various bioinspired switchable adhesive surfaces.",
"conclusion": "4 Conclusion The exceptional dynamic adhesion of many insects is closely related to the variable morphologies of their unfolded arolium, which are rapidly controlled by their shear actions. In this study, the passive controllable mechanisms of the unfolding honeybee arolium (i.e. contact area) were investigated, allowing honeybees to respond to sudden environmental disturbances. Experimental results ( Fig. 1 ) showed that the unfolding arolium (lateral extension) and the increasing contact area, are passively controlled by shear force. Then the mechanical properties of the arolium, especially including its inner branching fibers, were investigated ( Fig. 3 ). Theoretical model and experiments demonstrated that shear force can lead to a decrease in fibril angles, and then result in the torsion of the arolium's cuticles, which forming a new contact region and achieve the unfolding of the arolium. Additionally, the alterable pressure within the liquid-filled arolium was found to enhance its unfolding. The experimental preload to the inner pressure verified its effects on the unfolding of the honeybee arolium. Besides, the increasing pressure resulting from the decreasing height (decreasing fibril angles) controlled by the shear force, was theoretically revealed using the Bernoulli equation and simulated through Finite Element Method (FEM) ( Fig. 5 , Fig. 6 ). The unfolding of arolium caused by the increased inner pressure is coincident with the observed lateral extension ( Fig. 1 ). Notably, this passive controllable mechanism of the contact area in the honeybee arolium offers new perspectives for high dynamic adhesion, and can be applied to bionic switchable applications, such as intelligent transports and climbing robots.",
"introduction": "1 Introduction Over the past decade, natural adhesive systems have garnered significant attention due to their strong and reversible adhesion on nearly all known surfaces. Examples include the hierarchical setae of the geckos, which generate van der Waals force to support its weight [ 1 ], the hexagonal patterns of the tree frogs, which form capillary force on humid surfaces [ 2 ], and the suckers of the octopus and leeches, which adsorb on the dry and wet substrates with negative pressure [ 3 , 4 ]. Inspired by these adhesive systems, various artificial adhesive surfaces have been employed in smart devices, such as intelligent grippers [ 5 ], wearable electronic skins [ 6 ], and wound dressings [ 7 , 8 ]. Despite their rapid development, artificial adhesive surfaces suffer from long switching times between attachment and detachment, which range from seconds to minutes [ [9] , [10] , [11] ]. Whereas, the adhesive systems of many insects, such as ants and honeybees, exhibit rapid switching due to their highly rapid stride frequencies [ 12 , 13 ]. These dynamic biological adhesive systems can be attributed to precise control of adhesive strength and contact area. Some insects, such as ants [ 14 ], can sometimes rapidly change and control their contact areas in less than a millisecond. Therefore, gaining further insight into the mechanisms that regulate contact areas in insect adhesive systems is highly beneficial to improve the rapid control of artificial adhesives. The alterable contact area of footpads can be passively adjusted by shear force during attachment and detachment when pressed on the ground. For example, the hairy adhesive pads of certain organisms such as geckos [ 8 ], spiders [ 15 ], and flies [ 16 ] possess numerous spatulated setae [ 13 ], which can be translated into full contact with a substrate by a shear force toward their bodies [ 13 , 17 ]. Other smooth adhesive pads evolve different strategies in response to shear force. Stick insects possess a fibrous inner structure in the procuticle of their footpads [ 18 ] that produces a lateral expansion of the contact area with hydraulic function in the pads [ 18 , 19 ], while the drag activity (generating shear force) reduces the fibrous angle between the fibers and the substrate. Similarly, the dense array of nanopillars in the toe pads of tree frogs improves adhesion and friction [ 20 , 21 ], which is likely proportionate to the increment of the contact area resulting from the changing angles between the pillars and the pad's surface by shear force [ 20 , 22 ]. Specifically, some insects possess a smooth footpad that can switch between folded and unfolded morphologies [ 13 ]. The adhesive pads of ants and bees have been observed to rapidly unfold in response to imposed shear force, and a hypothetical pulling model has been established to describe this function [ 23 ]. This passive unfolding retraction in ants can sometimes occur within less than a millisecond [ 14 ], which is much shorter than the neuronally controlled reflexes in insects [ 24 ]. Although these regulatory mechanisms of contact area via shear forces have been widely observed in different adhesive systems, especially the folding adhesive pads, corresponding physical models have not yet been fully developed and verified. Arolium is an adhesive footpad of honeybees characterized by its rapid unfolding and folding capacity. In this study, the rapid evolution processes of the inner structures in the honeybee arolium were investigated and the inherent relationships between the contact area and shear force were explored. Firstly, the structures and deformations of the honeybee arolium were observed during attachment and detachment, and the collaboration of the contact area and lateral expansion of unfolded arolium controlled by shear force was investigated. Secondly, the morphologies of the outer and inner structures of arolium were characterized, and the unfolding and the folding mechanisms of arolium were theoretically analyzed. Finally, Finite Element Method (FEM) models were established to verify the theoretical and experimental analyses.",
"discussion": "3 Results and discussion 3.1 Evolution process of contact area passively controlled by dragging movement Honeybees exhibit preeminent capacities to easily traverse smooth artificial surfaces or flower petals for drinking and foraging owing to their highly dynamic wet adhesion. These superior capacities including rapid attachment and detachment occur millions of times during their lifetime. Fig. 1 A illustrates the experimental setup for observing the arolium morphologies and measuring the adhesive forces of honeybees during adhesive periods. As shown in Fig. 1 B, the foldable honeybee arolium constantly changed its contact area (folding and unfolding) during its attachment and detachment process ( Video S1 ). Due to this rapid evolution process, the adhesive force of honeybees (including normal and shear adhesion) exhibited highly dynamic performances ( Fig. 1 C), reflecting an adhesive period less than 1 s though the movement of the experimental honeybees was restricted ( Fig. 1 A). This is because the adhesive forces can be attributed to the product of adhesive strength and changing contact area. Besides, the negative value of normal adhesion indicated the preload during attachment ( Fig. 1 C), which is corresponding with the pre-contact observed in the changing contact area ( Fig. 1 C and D). Fig. 1 The rapid folding and unfolding of arolium passively controlled by shear force. (A) Schematic of the configuration used to measure the arolium's adhesion and observe the contact area with restricted movement. Rapid changes of the morphologies (Bottom) (B) and adhesion of a living honeybee arolium (C) during attachment and detachment. The scale bar is 100 μm. (D) Unfolding arolium of an anatomical leg switched by the artificial dragging movement (Lateral). The scale bar is 100 μm. Simplified contact area (E) (length × width), structural size (F) (length and width), and aspect ratio (G) (width/length) of the honeybee arolium after the shear dragging movement. Fig. 1 The folding and unfolding of the honeybee arolium were controlled by the dragging movement of the tarsus ( Fig. 1 B and D), generating shear force parallel to the dragging direction. After the pre-contact, when the dragging movement toward the proximal direction happened, the foldable arolium simultaneously unfolded over a period of approximately 350 ms. In contrast, when the aforementioned dragging operation ceased or a contrary movement started, the arolium folded rapidly within about 300 ms. Besides, the pre-contact and newly formed contact areas were fixed on the substrate owning to their adhesion. To further ascertain the effects of shear force on the unfolding arolium rather than the neural operation, a freshly amputated leg with honeybee arolium was observed at the lateral view when the dragging movement operated. Similarly, the arolium unfolded passively (without any muscular action) in response to the shear force controlled by the artificial dragging movement ( Fig. 1 D) ( Video S2 ). Furthermore, the changes in the contact area of the honeybees are closely related to the lateral extension of their arolium ( Fig. 1 E–G). The increase in contact area ( Fig. 1 E) mainly resulted from the significant increase in the lateral width of the arolium during unfolding, whereas the longitudinal length remained generally constant ( Fig. 1 F). Meanwhile, the aspect ratio (width/length) rose from 0.78 to 1.45 ( Fig. 1 G). These results indicate that the increases in the contact area of honeybees were primarily induced by the lateral expansion of the arolium, which was passively (i.e. without any muscular action) controlled by the shear force toward the proximal direction. 3.2 Morphological observation of honeybee arolium The mechanism underlying the passive unfolding of the honeybee arolium was investigated by first focusing on its morphology and material properties ( Fig. 2 ). The honeybee's adhesive system mainly consisted of the arolium, arcus, and manubrium ( Fig. 2 A). The distributed elastic arcus ( Fig. 2 A) acts as the primary internal supports for the soft arolium and contributes to the rapid detachment via the release of elastic energy stored from its deformation (accompanied with the unfolding arolium) caused by the shear force. Besides, the bilateral edges of the fully unfolding arolium remain folded (with an included angle close to 180 ° in Fig. 2 B) rather than entirely spreading out on the substrate. Then, the thin-film peeling theory was used to explain the functions of the folding edge of the arolium ( Fig. 2 B and C). Based on the model of Kendall [ 27 ], the peeling force ( F p ) can be estimated as (1) F p = E t b 1 − cos θ , where θ is the peeling angle (included angle), b is the width of the elastic membrane, and E t is the experimental adhesive energy required to fracture per unit area which can be formulated in terms of the work of adhesion ( W ) [ 28 , 29 ]. Therefore, the larger peeling angle needs a smaller peeling force, resulting in a lower requirement of the strain energy (arcus and arolium) stored from the shear force, and then promotes the folding of arolium during detachment. Moreover, the arolium has been proven to process rotatable architecture around a certain joint ( Fig. 2 A), which is controlled by the synergistic effects of the arcus and the inner shrinkable tendon (blue schematic in Fig. 2 A) [ 30 , 31 ]. However, the neuro-muscular activity still cannot explain the passive lateral extension of the arolium. Material properties are another important feature that influences the folded states of the honeybee arolium. As shown in Fig. 2 D and E, resilin autofluorescence (blue region) is evenly distributed throughout the arolium. Resilin, a protein with high resilience that has been already found in arthropod joints and deforming cuticles [ 32 , 33 ], exhibits remarkable resistance to mechanical fatigue under repetitive loading conditions. Additionally, the arcus presents an arched black region while other regions of the arolium are nearly transparent ( Fig. 2 A and B). This factor suggests that the elastic modulus of the arcus is much higher than other regions of the arolium, which is a reasonable conclusion based on its supportive role in the arolium. Although the aforesaid construct the basics (such as the high resilience of arolium) of the folding and unfolding arolium during attachment and detachment, they are still insufficient to fully elucidate the unfolding process of arolium passively controlled by shear force. Fig. 2 Microscopy images of structural morphologies of honeybee arolium. Lateral view of folded arolium (A) via stereoscope and rotating top view of unfolded arolium (B) via Cryogenic scanning electron microscopy (Cryo-SEM). (C) Schematic of the peeling theory, indicating the functions of the folded edge of the arolium. Overlay of a bright-field micrograph (Bottom) (D) and a wide-field fluorescence micrograph (Bottom) (E) showing the resilin autofluorescence in arolium. int: inner tendon, pl: planta, tar: tarsus, cl: claw, arc: arcus, mn: manubrium, ar: arolium. Fig. 2 3.3 Different morphologies of branching fibers within arolium responding to shear force The fibrous structures within the arolium have been shown to play a crucial role in controlling the contact area of insects in response to shear forces, such as the arolium of stick insects which lacks folding capacities [ 18 ]. Therefore, to comprehend the folding and unfolding mechanism of the honeybee arolium, its inner 3D structures were investigated ( Fig. 3 ). Various focused ion beam (FIB) partial cuttings of the honeybee arolium were observed to verify the accurate orientation of fibers. These fibers, which filled the outer cuticle layer of the arolium and were connected via the smaller branched fibers ( Fig. 3 C–E), were spatially orientated toward the shear direction (X-axis) ( Fig. 3 A and B). For the first feature of the fibers projecting on the X-Z plane, the vertical fibril angles ( θ x z ) between the fibers and epicuticle were approximately 29°–57° ( Fig. 3 B), which were obtained through a geometric transform of the observed angles ( Fig. 3 C–H) considering a sloping angle of 40° during FIB observation. Besides, these angles increased for fibers located further away from the symmetric axis of the arolium, ranging from 31.3° to 44.3° in Fig. 3 C–E and 19.7°–36.1° in Fig. 3 F–H. As arolium unfolded, θ xz descreased compared to the nearly upright fibers ( θ x z ≈ 70 − 85 ° ) of the folded arolium [ 26 , 34 ]. Fig. 3 Scanning electron microscopy images of the branching fibrous structure within the unfolded arolium of a honeybee via focused ion beam (FIB). (A) Bottom view of unfolded arolium with the schematic direction of different fibers. (B) 3D schematic of a single fiber showing the spatial angle compared to the drag direction. (C–H) Fibers slanted toward the contact cuticle with the observing view rotated 40° around the axis (X). (C-E) and (D-F) in response to the location further away from the symmetric axis. Fig. 3 Another feature of the fibers in the honeybee arolium has not been observed in other insects before. Projecting on the X–Y plane, the horizontal fibril angles ( θ x y ) between fibers and the shear direction (X-axis) are illustrated in Fig. 3 B. Angle θ x y increased gradually for fibers located further away from the symmetric axis of the arolium ( Fig. 3 B), which was indirectly supported by the existence and increasing density of fibrous fractures ( Fig. 3C–E and Fig. 3 F–H). For example, the number of fibrous fractures in Fig. 3 E is greater than that in Fig. 3 C (about 10 > 4, green region in Fig. 3 C and E), which means the increasing density of fibrous fractures, and then verifying that larger fibril angles θ x y is corresponding to the location further away from the symmetric axis. 3.4 Evolution process of fibers and contact area controlled by shear force The effects of shear force in controlling the contact area of honeybees were analyzed by using theoretical models to reveal the evolution process of the inner hierarchical structures of the arolium ( Fig. 4 ). Fig. 4 A shows the various forces acting on the arolium including the capillary and friction force. Concretely, the capillary force ( F c a p ) can be expressed as the summation of the Young-Laplace pressure effect and surface tension [ 35 ] as represented by Equation (2) , (2) F cap = Δ P π R 2 + 2 π γ R sin θ , where γ is the surface tension of the liquid secretion, θ is the contact angle between the contact substrate and the liquid secretion, and Δ P is the Young-Laplace pressure which can be estimated as [ 35 ] (3) Δ P = γ ( 1 R + 1 r ) , where R and r are the two meniscus radiuses of the liquid bridge. Then, based on the thinner liquid film, the friction ( F f ) (with a friction coefficient, μ ) can be estimated as [ 36 ] (4) F f = μ F cap Fig. 4 The theoretical models for the passive increase of contact area in honeybee hierarchical arolium. (A) The various forces in the arolium during attachment and detachment. (B–D) Different performances of the partial contact area with deformable fibers (C) as a response to shear force. (E–F) FEA simulations of fibers and hierarchical structures to verify the evolution process of fibers and contact area controlled by shear force. Fig. 4 These two forces contribute to resisting shear force during attachment and detachment and peculiarly exhibit the pre-fixation as the basics of the passive unfolding during arolium. Considering the elastic and hierarchical structures of arolium ( Fig. 2 , Fig. 3 C), the shear force cannot be directly transmitted to the contact interface, which subsequently affects the arolium's unfolding. Then, micro-units in the interim region between the contact and non-contact regions ( Fig. 4 B-D) were selected to future analyze the evolution process of fibers controlled by shear force, which are more prone to bending rather than stretching due to their greater slenderness ratio (length/diameter >20) ( Fig. 3 C). Firstly, the rapid response of a single fiber to the internal shear force was analyzed ( Fig. 4 C). Due to the smaller dragging angle ( θ d ) ( Fig. 1 , Fig. 4 A) during arolium unfolding, the shear component force on a fiber [ F s f ( θ x y , θ x z ) ] is much greater than the normal component force [ F n f ( θ x y , θ x z ) ] transformed from the dragging action, which leads to the decrease of fibril angles ( θ x y and θ x z in Fig. 4 C) with a fixed boundary resulting from the capillary and friction force. These rapid changes in fibers are coincident with a previous study about arolium in insects [ 18 ]. Consequently, the evolution processes of a single fiber were extended to the interim region of arolium ( Fig. 4 B and D). Due to the flexible fibers ( Fig. 4 C), the micro-units (including fibers, upper and contact cuticles) exhibit the consequential torsional deformation driven by shear force ( Fig. 4 D). For a square micro-unit, the resultant forces ( F s m , F n m ) (supported by four fibers) can be expressed as (5) F s m = ∑ F s f ( θ x y , θ x z ) , (6) F n m = ∑ F n f ( θ x y , θ x z ) . The fixed foundation of the contact cuticle and fibers, resulting from capillary force and friction, can induce the torsion of the left unit (contact region) towards the lower right corner when cooperating with the component forces of F s m a n d F n m (red schematic forces in Fig. 4 D) perpendicular to the fibers. Considering the interconnections of fibers by smaller branched fibers, the right unit (non-contact region) will be subsequently propelled in the right direction ( Fig. 4 D). Then the torsion of the right unit induces a decrease in the non-contact angle ( θ n a ), resulting in the generation of a new contact region when θ n a approaches zero or the generation of a new capillary force. Thus, the arolium's unfolding process is ultimately achieved and induced by shear force. To prove the theoretical torsions, a Finite Element Method (FEM) model was developed, using simulation software (Creo 5.0, PTC, USA), with the boundary conditions and applied forces as depicted in Fig. 4 D. Under these forces, the micro-units exhibit the torsions on the X–Y and Y-Z planes ( Fig. 4 E and F), particularly the torsion of the right unit (non-contact region) towards the lower right direction, which aligns with the mechanical prediction ( Fig. 4 D). Besides, the decrease in total height responding to the shear force ( Fig. 4 F) indirectly demonstrates the reduction in the fibril angle ( θ x z ) ( Fig. 4 C). Along with the decrease in fibril angle ( θ x y ) ( Fig. 4 F), the observed decreases in the fibril angles ( Fig. 3 ) induced by shear force were verified. Moreover, the decrease in the non-contact angle ( θ n a ) is observed ( Fig. 4 F), indicating the unfolding of the arolium. Notably, this passively controllable mechanism of the contact area is an ingenious energy-saving system that enables honeybees to transform gravity into shear force when attaching to vertical surfaces. 3.5 Effects of the inner pressure on the unfoldment of honeybee arolium The decreases in the fibril angles possess the potential to influence the pressure of the liquid surrounding the inner fibers, thereby enhancing the unfolding process of the honeybee arolium. Fig. 5 A illustrates the experimental setup to investigate the effects of the alterable inner pressure of arolium. Primarily, the soft arolia of insects are liquid-filled materials without any inner air [ 37 ], and the inner liquid is incompressible under normal pressure and temperature. To indirectly increase the inner pressure of arolium, a tiny volume of deionized water was gradually injected into the tarsal segments of honeybees ( Fig. 5 A and B). Meanwhile, the freshly removed arolium unfolded in response to this injecting process ( Fig. 5 B). The two observed unfolded states indicate the enhancement of the increasing inner pressure on the unfolding of the honeybee arolium ( Fig. 5 B). This result is consistent with a previous study that demonstrated the arolium inflation through increased pressure in ants ( O. smaragdina ) [ 23 ]. Fig. 5 Schematic of the effects of the inner pressure on the unfolding of the honeybee arolium. (A) The experimental setup to investigate the connection between the inner pressure and the arolium's unfolding. (B) Experimental results showed the lateral unfolding of the arolium caused by the increase in the inner pressure. (C) The changes in the liquid form within the arolium respond to the changing fibril angles and shear force. (D) A simplified model shows the increase in the inner pressure resulting from the change in the liquid form. Fig. 5 To investigate the effect of the alterable fibril angles on the inner pressure, a simplified partial model of the honeybee arolium was established, comprising bionic fibers and inner liquid ( Fig. 3 , Fig. 4 C). The decreases in the fibril angles ( θ x y and θ x z ) controlled by the shear force resulted in a decrease in the model's height ( Fig. 4 F), which subsequently leads to a change in the liquid form considering the constant volume of liquid ( Fig. 5 C). Here, the Bernoulli equation [ 38 ] was employed to estimate the pressure difference between the two states ( Fig. 5 D) as (7) P 0 + 1 2 ρ V 0 2 + ρ g h 0 = P 1 + 1 2 ρ V 1 2 + ρ g h 1 = C where P is the inner pressure, ρ is the liquid density, V is the local velocity, g is the gravity acceleration, h is the vertical height, C is a constant, and subscripts 0 and 1 represent the different states. Considering the steady states of the initial and final state, the local velocity ( V ) can be set to zero. Then the decrease in height ( h ) can lead to an increase of the inner pressure ( Δ P i n = ρ g ( h 0 − h 1 ) ), which likely promote the unfolding of the arolium. However, generalizing this function to the unfolding of the honeybee arolium necessitates assuming a constant volume of the inner liquid within the arolium. In fact, this assumption is supported by the slow volume change of the liquid as it permeates through the epidermal cells on the honeybee arolium [ 17 ]. Subsequently, to model the liquid-structure interactions, including the liquid's pressure and the deformation induced by shear force, the coupled Euler-Lagrange (CEL) approach was adopted ( Fig. 6 A) [ 39 ]. The liquid model, discretized utilizing the EC3D8R element in the Eulerian domain, was carried out using the Explicit method in the simulation software (Abaqus 2021, Dassault Systemes, France) with defined boundary conditions and applied loads ( Fig. 6 A). Fig. 6 A illustrates the evolution process of the liquid pressure and the mechanical deformations induced by shear force. During the loading process, the peak liquid pressure rapidly rises and falls, eventually achieving a steady pressure greater than the initial pressure ( Fig. 6 B). This rapid rise in peak pressure is driven by the sudden mechanical deformations induced by shear force, and is subsequently neutralized by other liquids. However, the stable increased pressure (steady state), corresponding to the decrease in height and fibril angle caused by shear force ( Fig. 6 B), aligns with the predication of Eq. (7) ( Fig. 5 D). In conclusion, the increase in liquid pressure is proved to be induced by shear force. Considering the cantilever beam-like structures of the honeybee arolium, characterized by bilateral bends and intermediate fixation ( Fig. 2 , Fig. 6 C), the increased pressure induced by shea force can further enhance the increase in the lateral width (lateral unfolding) rather than the proximal–distal length. This observation is unified with the observed unfolding of the honeybee arolium under the experimentally applied pressure ( Fig. 5 B). Fig. 6 FEA simulations for the increasing inner pressure caused by shear force. (A) The evolution process of the liquid pressure and the mechanical deformations. (B) The changes in the liquid pressure and inner fibril angle. (C) Unfolding arolium caused by the increase of inner pressure. Fig. 6"
} | 6,864 |
37000872 | PMC10065435 | pmc | 8,578 | {
"abstract": "Ongoing climate change is driving the search for renewable and carbon-neutral alternatives to fossil fuels. Photocatalytic conversion of fatty acids to hydrocarbons by fatty acid photodecarboxylase (FAP) represents a promising route to green fuels. However, the alleged low activity of FAP on C2 to C12 fatty acids seemed to preclude the use for synthesis of gasoline-range hydrocarbons. Here, we reveal that Chlorella variabilis FAP ( Cv FAP) can convert n -octanoic acid in vitro four times faster than n -hexadecanoic acid, its best substrate reported to date. In vivo, this translates into a Cv FAP-based production rate over 10-fold higher for n -heptane than for n -pentadecane. Time-resolved spectroscopy and molecular modeling demonstrate that Cv FAP’s high catalytic activity on n -octanoic acid is, in part, due to an autocatalytic effect of its n -heptane product, which fills the rest of the binding pocket. These results represent an important step toward a bio-based and light-driven production of gasoline-like hydrocarbons.",
"introduction": "INTRODUCTION Fatty acid photodecarboxylase (FAP; EC 4.1.1.106), a member of the glucose-methanol-choline oxidoreductase family, is an algae-specific enzyme harboring a flavin adenine dinucleotide (FAD) cofactor ( 1 ). FAP is one of the few known natural photoenzymes besides photosynthetic reactions centers (which are protein complexes), DNA-repairing enzymes photolyases ( 2 ), and light-dependent protochlorophyllide oxidoreductases ( 3 ). Despite the short time since the discovery of FAP in the green microalga Chlorella variabilis by Sorigué et al. ( 1 ), numerous groups have already explored the potential applications of this enzyme in biocatalytic processes. C. variabilis FAP ( Cv FAP) represents a new, attractive, light-driven, and redox-neutral means for the production of n -alkanes and n -alkenes as a basis for fuels, chemistry, and cosmetics ( 4 – 7 ). Cv FAP also appears to be a very promising tool for the photocatalytic synthesis of specialty chemicals such as deuterated hydrocarbons ( 8 ), enantiomerically pure α-amino acids, α-hydroxy acids ( 9 ), secondary fatty alcohols ( 10 ), and aliphatic amines and esters ( 11 ). The natural substrates for the photoproduction of hydrocarbons in the green algae C. variabilis and Chlamydomonas reinhardtii were previously shown to be C16 to C18 linear fatty acids (FAs) ( 12 ). Following the discovery of FAP, an initial in vitro characterization of purified Cv FAP and its Chlamydomonas homolog indicated that these enzymes exhibited higher affinity for longer (C16 to C18) FAs compared to C12 to C14 FAs ( 1 ). It was also found that the crystal structure of Cv FAP expressed in the heterologous host Escherichia coli , which contains C10 to C18 FAs, had two “native” (likely unsaturated) C18 FA substrates—one at the active site and another one stabilized at the surface of the protein close to the entrance to the tunnel leading to the active site ( 13 ). In addition, expression in E. coli of four other FAPs chosen in different algal groups among almost 200 putative FAPs identified in algae genomes or metagenomic data showed that they were preferentially performing photodecarboxylation of the endogenous C16 to C18 FAs rather than C10 to C14 FAs ( 14 ). Use of E. coli cell-free extracts expressing the enzyme confirmed the preference of Cv FAP for C16 to C18 FAs over C12 to C14 FAs ( 4 ) and showed that it had low activities on C2 to C6 FAs ( 15 ). Together, all these data lead to the view that Cv FAP was adapted to act on C16 to C18 substrates and was not likely to be an efficient biological catalyst to produce gasoline-range (C5 to C11) hydrocarbons. Here, we reveal that, under the right conditions, Cv FAP can be highly active on C8 to C10 medium-chain FAs in vitro. We also show that the high activity of Cv FAP on C8 to C10 FAs can be attributed in part to an unexpected autocatalytic effect, for which we provide spectroscopic evidence and molecular modeling support. Last, a bioconversion experiment using bacterial cultures expressing Cv FAP provides evidence that this biocatalyst can be used to produce medium-chain hydrocarbons in a much more efficient way than long-chain hydrocarbons.",
"discussion": "DISCUSSION Screening the quantum yield of medium-chain (C7 to C12) FAs photodecarboxylation by Cv FAP using time-resolved spectroscopy, we observed an autocatalytic effect, whereby the initially formed n -alkane products enhanced the decarboxylation of further FA substrates by mimicking the missing part of the long chain, for which the FA binding site of FAP is adapted. This autocatalytic effect was observed for C7 to C10 FAs but it was most pronounced for the C8 FA substrate. These observations inspired us to test the effects of C5 to C12 n -alkanes added as cocatalysts to C7 to C12 FA substrates. Our results show that n -alkanes can serve as cocatalysts and decarboxylation quantum yields previously achievable only for long-chain FAs can now be obtained also for medium-chain (C7 to C12) FAs, provided that the right combination of FA and n -alkane is chosen. The highest quantum yields (60 to 70%) were obtained when the total number of carbon atoms of the FA substrate and the n -alkane cocatalyst was 16 ± 1, in line with the most pronounced autocatalytic effects observed for C8 FA and with C16 FA being one of the best reported substrates for Cv FAP ( 1 ) [note that the ideal length of ~16 total carbon atoms was also suggested by Zhang et al. ( 15 ), who attempted to enhance Cv FAP enzymatic activity under continuous light on short (C1 to C6) substrates by adding C7 to C17 n -alkanes to crude cell extracts from E. coli expressing Cv FAP]. The outcomes of our MD simulations suggest that the auto- and/or cocatalytic effect of the n -alkanes can be attributed to a steric hindrance that favors the positioning of the medium-chain FAs close enough to the FAD cofactor so that the initial forward electron transfer from the substrate to FAD can occur upon photoexcitation of the latter. The simulations also establish that the conformation of C8 FA and C10 or C7 alkanes closely resembles that observed in the MD simulations with the corresponding long-chain FAs (C18 or C15, respectively). Our in vitro assay demonstrates that the autocatalytic effect (observed in the spectroscopic experiments and rationalized by MD simulations) together with the optimum pH and substrate concentration can considerably enhance the photoconversion of a medium-chain FA (C8 FA) by Cv FAP under continuous illumination. The chemical yield of C8 FA decarboxylation exceeding four times that of C16 FA is the best in vitro performance of wild-type Cv FAP on a medium-chain substrate reported so far. The difference in the optimum pH obtained for C8 and C16 FAs (6.0 and 8.5, respectively) is consistent with experimental data on the p K a of C8 to C16 saturated FA salts organized in films ( 18 ), as well as quantum chemistry calculations, which show that the lengthening of an FA carbon chain by one methylene group leads to an increase of 0.43 units in the surface p K a at the air-water interface ( 19 ). Last, our in vivo assay (photodecarboxylation of 1- 13 C–labeled C8 and C16 FAs inside the living E. coli bacteria) shows that the chemical yield of medium-chain n -alkanes by FAP under continuous irradiation can greatly exceed that of long-chain ones, by a factor of ~13 for the case of C7 and C15 n -alkanes. The autocatalytic effect certainly contributes to the efficient decarboxylation of medium-chain FAs, but autocatalysis alone is not sufficient to explain the magnitude of this difference. Given that the permeability of the cell membranes seems to be higher for long substrates (such as C16 FA) and that the kinetics of the initial photochemical steps are also similar (for both C8 and C16 FAs), we suggest that the high chemical yield of medium-chain (C8) FA decarboxylation observed both in vitro and in vivo is most likely due to an acceleration of the “dark” steps of the photocycle, namely, the exchange of the product for a new substrate. Faster replacement of the product with a new substrate (in the case of medium-chain FAs) may also have a positive impact on the (photo)stability of the enzyme ( 20 ). In conclusion, the present work unveils an autocatalytic effect of C7 to C10 FAs decarboxylation products and documents the cocatalytic effects of C5 to C12 n -alkanes on the decarboxylation of medium-chain (C7 to C12) FAs by Cv FAP. It further reveals an unexpectedly high in vivo activity of Cv FAP on C8 FA acid, which is over 10-fold higher than that on C16 FA, the best FAP substrate identified before our current study. These results should stimulate applied research on the use of Cv FAP to convert medium-chain substrates and guide future studies aiming at improving Cv FAP for high-yield production of gasoline-like hydrocarbons."
} | 2,238 |
34668280 | null | s2 | 8,580 | {
"abstract": "Remote coral reefs are thought to be more resilient to climate change due to their isolation from local stressors like fishing and pollution. We tested this hypothesis by measuring the relationship between local human influence and coral community resilience. Surprisingly, we found no relationship between human influence and resistance to disturbance and some evidence that areas with greater human development may recover from disturbance faster than their more isolated counterparts. Our results suggest remote coral reefs are imperiled by climate change, like so many other geographically isolated ecosystems, and are unlikely to serve as effective biodiversity arks. Only drastic and rapid cuts in greenhouse gas emissions will ensure coral survival. Our results also indicate that some reefs close to large human populations were relatively resilient. Focusing research and conservation resources on these more accessible locations has the potential to provide new insights and maximize conservation outcomes."
} | 254 |
34668280 | null | s2 | 8,581 | {
"abstract": "Remote coral reefs are thought to be more resilient to climate change due to their isolation from local stressors like fishing and pollution. We tested this hypothesis by measuring the relationship between local human influence and coral community resilience. Surprisingly, we found no relationship between human influence and resistance to disturbance and some evidence that areas with greater human development may recover from disturbance faster than their more isolated counterparts. Our results suggest remote coral reefs are imperiled by climate change, like so many other geographically isolated ecosystems, and are unlikely to serve as effective biodiversity arks. Only drastic and rapid cuts in greenhouse gas emissions will ensure coral survival. Our results also indicate that some reefs close to large human populations were relatively resilient. Focusing research and conservation resources on these more accessible locations has the potential to provide new insights and maximize conservation outcomes."
} | 254 |
23514423 | null | s2 | 8,582 | {
"abstract": "Neuroscience is at a crossroads. Great effort is being invested into deciphering specific neural interactions and circuits. At the same time, there exist few general theories or principles that explain brain function. We attribute this disparity, in part, to limitations in current methodologies. Traditional neurophysiological approaches record the activities of one neuron or a few neurons at a time. Neurochemical approaches focus on single neurotransmitters. Yet, there is an increasing realization that neural circuits operate at emergent levels, where the interactions between hundreds or thousands of neurons, utilizing multiple chemical transmitters, generate functional states. Brains function at the nanoscale, so tools to study brains must ultimately operate at this scale, as well. Nanoscience and nanotechnology are poised to provide a rich toolkit of novel methods to explore brain function by enabling simultaneous measurement and manipulation of activity of thousands or even millions of neurons. We and others refer to this goal as the Brain Activity Mapping Project. In this Nano Focus, we discuss how recent developments in nanoscale analysis tools and in the design and synthesis of nanomaterials have generated optical, electrical, and chemical methods that can readily be adapted for use in neuroscience. These approaches represent exciting areas of technical development and research. Moreover, unique opportunities exist for nanoscientists, nanotechnologists, and other physical scientists and engineers to contribute to tackling the challenging problems involved in understanding the fundamentals of brain function."
} | 410 |
21993758 | PMC3192504 | pmc | 8,583 | {
"abstract": "DNA molecules provide what is probably the most iconic example of self-replication—the ability of a system to replicate, or make copies of, itself. In living cells the process is mediated by enzymes and occurs autonomously, with the number of replicas increasing exponentially over time without the need for external manipulation. Self-replication has also been implemented with synthetic systems, including RNA enzymes designed to undergo self-sustained exponential amplification 1 - 5 . An exciting next step would be to use self-replication in materials fabrication, which requires robust and general systems capable of copying and amplifying functional materials or structures. Here we report a first development in this direction, using DNA tile motifs that can recognize and bind complementary tiles in a pre-programmed fashion. We first design tile motifs so they form a seven-tile seed sequence, then use the seeds to instruct the formation of a first generation of complementary seven-tile daughter sequences, and finally use the daughters to instruct the formation of seven-tile granddaughter sequences that are identical to the initial seed sequences. Considering that DNA is a functional material that can organize itself and other molecules into useful structures 6 - 13 , our findings raise the tantalizing prospect that we may one day be able to realize self-replicating materials with various patterns or useful functions."
} | 359 |
28731470 | PMC5702722 | pmc | 8,584 | {
"abstract": "Similar to plant growth, soil carbon (C) cycling is constrained by the availability of nitrogen (N) and phosphorus (P). We hypothesized that stoichiometric control over soil microbial C cycling may be shaped by functional guilds with distinct nutrient substrate preferences. Across a series of rice fields spanning 5–25% soil C (N:P from 1:12 to 1:70), C turnover was best correlated with P availability and increased with experimental N addition only in lower C (mineral) soils with N:P⩽16. Microbial community membership also varied with soil stoichiometry but not with N addition. Shotgun metagenome data revealed changes in community functions with increasing C turnover, including a shift from aromatic C to carbohydrate utilization accompanied by lower N uptake and P scavenging. Similar patterns of C, N and P acquisition, along with higher ribosomal RNA operon copy numbers, distinguished that microbial taxa positively correlated with C turnover. Considering such tradeoffs in genomic resource allocation patterns among taxa strengthened correlations between microbial community composition and C cycling, suggesting simplified guilds amenable to ecosystem modeling. Our results suggest that patterns of soil C turnover may reflect community-dependent metabolic shifts driven by resource allocation strategies, analogous to growth rate–stoichiometry coupling in animal and plant communities.",
"conclusion": "Conclusions We observed that C turnover across a large soil C:N:P gradient is in part mediated by P availability, reflecting differences in microbial P metabolism related to growth rates and nutrient use. Interactions among ecosystem C, N and P cycling were mirrored by community metabolic potential and appear related to shifts in community structure arising from trait-based environmental filtering of organisms by soil nutrient availability. Our findings suggest that stoichiometric resource allocation and ribosome copy number are key traits that underpin patterns of microbial community structure and function in some soils and may mediate ecosystem C cycling under multiple nutrient limitation. Grouping microbes by their resource allocation patterns improved relationships with soil nutrients and C turnover in the rice paddy soils studied, and the tradeoffs we observed among traits might help to simplify future efforts to assimilate microbial sequence data into ecosystem models.",
"introduction": "Introduction Although microbial communities are critical to carbon (C) flow in the biosphere, ecosystem models have only recently begun to simulate variation in their metabolism in soils ( Li et al. , 2014 ; Wang et al. , 2015 ; Weider et al. , 2015 ). Increasingly powerful and available data on microbial community structure and function might help to better inform these efforts ( McGuire and Treseder, 2010 ; Singh et al. , 2010 ; Schimel and Schaeffer, 2012 ; Graham et al. , 2016 ). Yet, compared with taxonomically constrained processes such as ammonia or methane oxidation ( Bouskill et al. , 2012 ; Ho et al. , 2013 ), delineation of functional groups for decomposition in soils is challenging due to the broad distribution and functional redundancy of the relevant traits ( Allison and Martiny, 2008 ; Schimel and Schaeffer, 2012 ; Berlemont and Martiny, 2013 ; Martiny et al. , 2015 ). However, consideration of interactions between C and nutrient cycling by soil microbes may help clarify ecologically relevant functional guilds. For example, consistent shifts in terrestrial soil C turnover and retention with mineral N addition ( Craine et al. , 2007 ; Treseder, 2008 ) appear mediated in part by increased abundance of microbes utilizing simple C substrates, at the expense of those mining complex aromatic C and organic N ( Treseder et al. , 2011 ; Fierer et al. , 2012 ; Ramirez et al. , 2012 ; Cederlund et al. , 2014 ; Amend et al. , 2015 ). Phosphorus (P) availability can also contribute to regulation of soil C cycling, and the underlying traits for microbial P cycling could analogously be linked with traits for C and N acquisition. Although less commonly studied than soil C and N interactions, P availability may affect land C sink strength at ecosystem and global scales ( Wang et al. , 2007 ; Vitousek et al. , 2010 ; Goll et al. , 2012 ) and contribute to the regulation of soil C turnover rates even in nominally N-limited habitats, such as grasslands, temperate forests and leaf litter ( Bradford et al. , 2008 ; Manzoni et al. , 2010 ; Strickland et al. , 2010 ; Fisk et al. , 2015 ). Soil P availability may not only affect the biomass of soil microbes ( Griffiths et al. , 2012 ; Zhang et al. , 2015 ) but critically might also control community-scale rates of metabolism ( Strickland et al. , 2010 ; Spohn and Chodak, 2015 ), which could reflect underlying stoichiometric constraints at the scale of individual cells ( Hartman and Richardson, 2013 ). As posited by the growth rate hypothesis ( Elser et al. , 1996 , 2000 , 2003 ), cellular growth rates are linked to biomass N:P ratios by the high P demands of ribosomal RNA, which determines in part the rate of synthesis of N-rich proteins. Differentiation of organism C, N and P demands based on growth rate variation forms the foundation of Ecological Stoichiometry theory ( Sterner and Elser, 2002 ; Vrede et al. , 2004 ; Allen and Gillooly, 2009 ), enabling community shifts under different nutrient regimes to be connected with predictable alterations in ecosystem C cycling, particularly in aquatic ecosystems ( Sterner and Elser, 2002 ; Weber and Deutsch, 2010 ; Follows and Dutkiewicz, 2011 ; Hessen et al. , 2013 ; Mock et al. , 2015 ). Development of a parallel framework to link stoichiometric regulation of microbial metabolism to soil C cycling at the community and ecosystem scales is highly desirable ( Hall et al. , 2011 ; Sistla and Schimel, 2012 ; Zechmeister-Boltenstern et al. , 2015 ), especially given the integration of stoichiometric regulation of primary producers and decomposition into current terrestrial ecosystem models ( Yang et al. , 2014 ; Reed et al. , 2015 ). However, the inter-relationships between soil stoichiometry, microbial communities and soil C cycling are not currently well understood. In culture, microbial growth rates are linked with cell N:P stoichiometry and ribosomal RNA content or gene copy number ( Makino and Cotner, 2004 ; Karpinets et al. , 2006 ; Keiblinger et al. , 2010 ; Vieira-Silva and Rocha, 2010 ; Franklin et al. , 2011 ), which can vary among bacterial lineages ( Mouginot et al. , 2014 ; Roller et al. , 2016 ). Separately, variation or manipulation of P availability and stoichiometry in soils has been associated with shifts in microbial community composition ( Güsewell and Gessner, 2009 ; Fanin et al. , 2013 ; Leff et al. , 2015 ; Spohn et al. , 2015 ), without considering relationships to soil C cycling or metabolic differences among responsive microbes. We postulated that ecosystem-scale relationships between soil C cycling and P availability may arise due to changes in microbial metabolism, which reflect a dependence of growth rates on P availability, and are underpinned by shifts in microbial communities based on differences in their nutrient utilization potential. We combined plot-scale manipulations and measurements of C cycling across a broad soil C:N:P gradient with high-throughput sequencing techniques to address specific hypotheses at different scales of ecological organization, including: (1) Soil C cycling rates are associated with P availability and stoichiometry at the ecosystem scale; (2) At the community scale, taxonomy and function of microbes are associated with soil P stoichiometry; (3) At the individual scale, genomic features and potential utilization of C, N and P substrates differ characteristically among taxa, and (4) Across these scales, rates of soil C cycling are linked with the abundance of microbial groups defined by differences in stoichiometric resource allocation. To enable evaluation of the latter two hypotheses and interrogate relationships across scales, we inferred the stoichiometric utilization potential of individual microbes in situ , using 16S rRNA and shotgun metagenome sequencing in combination with inference from existing reference genomes ( Langille et al. , 2013 ).",
"discussion": "Discussion Stoichiometric nutrient limitation of soil C cycling and metabolism In the wetland soils studied, C cycling rates appeared dependent on the supply of soil C, N and P, yet constraints imposed by P availability featured most prominently. Non-linear patterns of soil CO 2 fluxes reflected declining soil C turnover rates with increasing organic matter ( Figures 1a and b ). One explanation for this decline in soil C turnover is the lower availability of P relative to N as soil C accumulates ( Figure 1c ), consistent with global-scale patterns ( Hartman and Richardson, 2013 ). In agreement with our first hypothesis, soil C turnover was most closely associated with P availability, including N:P ratios and inorganic P ( Table 1 , Figures 1d and e ). Moreover, we observed a breakpoint in N fertilization effects roughly corresponding to the Redfield ratio of N:P=16:1 ( Figure 1d ), which delineates a shift from N to P limitation ( Redfield, 1958 ; Sterner and Elser, 2002 ; Cleveland and Liptzin, 2007 ). Similar to interacting effects of N and P additions on soil C cycling in upland soils ( Nottingham et al. , 2015 ; Poeplau et al. , 2016 ), our results suggest N fertilization altered C cycling only where N was limiting relative to P. Importantly, these results are opposite of those expected from C:N ratios alone, which would predict N fertilization effects only in high C:N organic soils. Alternately, declining C turnover at higher soil C may reflect accumulation of aromatic compounds, which are both resistant to degradation and hypothesized to directly inhibit decomposition ( Melillo et al. , 1982 ; Freeman et al. , 2001 ). Accordingly, we found aromatic and carboxylic compounds were more abundant in higher C organic soils ( Figure 1f ). These shifts were mirrored by changes in microbial C substrate utilization potential in our metagenome data, which revealed more carbohydrate-degrading genes in mineral soils and greater potential utilization of complex C forms in organic soils ( Figure 2 , Supplementary Figure S6 ). Notably, aromatic compounds were most closely associated with increasing C:N, C:P and N:P ratios, opposite of inorganic P availability, while increasing soil C content was linked with greater abundance of both carbohydrates and aromatics ( Supplementary Figure S2 ). Despite its effects on C turnover, N fertilization did not significantly alter microbial metabolic potential or community structure in our rice soils ( Figure 2 , Supplementary Figure S7a and d ), in contrast to upland soils where N fertilization consistently reduces soil C turnover and alters microbial communities ( Ramirez et al. , 2010 ; Fierer et al. , 2011 ; Ramirez et al. , 2012 ). Both nitrous oxide fluxes ( Ye et al. , 2016 ) and gene abundances for N dissimilation were higher in our mineral soils ( Figure 2 ), which could suggest excess N availability consistent with their lower C:N ratios. However, N addition increased C turnover only in mineral soils ( Figure 1 ) and did not alter in situ N 2 O fluxes in any of our soils ( Ye et al. , 2016 ). This decoupling of denitrification from N supply could indicate that denitrifiers are instead limited by available C, as suggested by correlated N dissimilation and carbohydrate genes across the gradient ( Figure 2 , Supplementary Figure S6 ). P fertilization effects on decomposition also appear to be contingent on the availability of labile C ( Fisk et al. , 2015 ; Nottingham et al. , 2015 ; Poeplau et al. , 2016 ). The strong relationship we observed between inorganic P availability and C turnover ( Figure 1 ) in our rice soils resembles coupling of available P with microbial metabolic rates across land use and global habitat types, including wetlands ( Strickland et al. , 2010 ; Hartman and Richardson, 2013 ). We hypothesized that this variation in C cycling and metabolism arises from underlying shifts in microbial community structure and function, and that coupling of ecosystem biogeochemical cycles reflects metabolic coupling of element use within and among microbial taxa. Microbial communities and traits linked with stoichiometry and C cycling Microbial community structure varied with several factors along the soil gradient, including strong independent relationships with soil P stoichiometry ( Table 1 ), in agreement with our second hypothesis. The abundances of several microbial groups were correlated with soil C turnover, including members of the Actinobacteria , Firmicutes , Chloroflexi , some clades of the Acidobacteria and narrower groups of Proteobacteria and Bacteroidetes ( Figure 3 , Supplementary Figure S7b–c , Supplementary Table S7 ). Many of these lineages have been previously associated with higher rates of soil C cycling or microbial metabolism ( Fierer et al. , 2007 ; Goldfarb et al. , 2011 ; Fierer et al. , 2012 ; Ramirez et al. , 2012 ) potentially reflecting their greater carbohydrate utilization potential ( Berlemont and Martiny, 2013 ; Berlemont et al. , 2014 ). In our soils, microbial taxa correlated with C turnover rates were enriched in imputed carbohydrate utilization genes while harboring fewer genes for P scavenging ( Figure 3 ), consistent with biogeochemical and metabolic shifts observed along the gradient ( Figures 1 and 2 ). Organisms that were highly correlated with soil C turnover also had greater imputed ribosome gene copy numbers ( Figure 3 ), which are associated with increased growth rates among cultured microbes ( Klappenbach et al. , 2000 ; Shrestha et al. , 2007 ; Vieira-Silva and Rocha, 2010 ; Roller et al. , 2016 ). This key trait also integrates growth rates with cellular demand for N and P ( Elser et al. , 1996 , 2003 ), underlying functional differentiation and trophic status in higher organisms ( Sterner and Elser, 2002 ; Wright et al. , 2004 ; Arrigo, 2005 ; Hessen et al. , 2013 ). Trait differentiation underlies community assembly and C cycling Tradeoffs in allocation to resource acquisition and growth contribute to metabolic specialization and niche partitioning among organisms ( Tilman, 1990 ; Johnson et al. , 2012 ; Edwards et al. , 2013a ; Litchman et al. , 2015a , b ; Díaz et al. , 2016 ). In our data, organisms inferred to use proportionately more carbohydrates had less allocation to P scavenging ( Figure 4 , Supplementary Figure S8 , Supplementary Table S9 ), consistent with their opposing patterns among OTUs positively or negatively correlated with C turnover ( Figure 3 ), and with shifts in ecosystem C and P cycling in soil biogeochemical and metagenomic data ( Figures 1 and 2 , Supplementary Figure S6 ). Allocation to use of C, N and P substrates differed among high-level taxonomic groups ( Figure 4 , Supplementary Figure S8 , Table 3 ), in agreement with our third hypothesis, that genomic features and potential utilization of C, N and P substrates would differ characteristically among taxa. Taxonomic patterns in carbohydrate use and P scavenging in our data also appeared broadly similar to those observed in comparative genomic and metagenomic studies ( Berlemont and Martiny, 2013 ; Berlemont et al. , 2014 ; Chai et al. , 2014 ). Deeply rooted partitioning of nutrient allocation could help account for consistent biogeographic patterns in microbial communities ( Phillipot et al. , 2010 ; Ramirez et al. , 2014 ) and their responses to fertilization at high taxonomic ranks ( Leff et al. , 2015 ; Spohn et al. , 2015 ). Higher taxonomic ranks also accounted for most of the covariation of microbial communities with chemistry in our soils ( Supplementary Table S10 ), consistent with results compared across terrestrial and aquatic ecosystems ( Lu et al. , 2016 ). Differences in resource allocation strategies may underlie patterns of community assembly across environmental gradients ( Weber and Deutsch, 2010 ; Edwards et al. , 2013a , b ; van Bodegom et al. , 2014 ; Litchman et al. , 2015a , b ). Our CATS models ( Warton et al. , 2015 ) revealed interactions between soil nutrient availability and genomic resource allocation ( Figure 4d ) that suggest differences in nutrient utilization shape community composition. Genomic allocation to carbohydrate use and N dissimilation both had positive interactions with nitrate concentrations across taxonomic ranks ( Figure 4 , Supplementary Figure S9 ), suggesting again that higher denitrification rates ( Ye et al. , 2016 ) and genetic potential in our mineral soils ( Figure 2 ) were associated with greater carbohydrate use, analogous to coupled mineral N and carbohydrate use in upland soils ( Ramirez et al. , 2012 ; Amend et al. , 2015 ). Although P scavenging allocation was less clearly related to soil nutrient availability, ribosome copy numbers had significant positive interactions with inorganic P concentrations at each taxonomic rank tested ( Figure 4d , Supplementary Figure S9 ). This result could suggest that inorganic P availability limits the abundance of high ribosome copy number organisms, whose high growth rates and low C use efficiency ( Roller et al. , 2016 ) may in part account for greater C turnover in our P-rich mineral soils. This mechanism, if observed elsewhere, could in part account for large-scale relationships between inorganic P availability and microbial metabolic rates across soils ( Strickland et al. , 2010 ; Hartman and Richardson, 2013 ). Incorporation of functional trait variation can enable community-driven metabolic modeling of nutrient-mediated changes in ecosystem C cycling ( Weber and Deutsch, 2010 ; Bouskill et al. , 2012 ; Litchman et al. , 2015a ; Martiny et al. , 2015 ), which may be simplified by accounting for tradeoffs among traits ( Edwards et al. , 2012 ; Johnson et al. , 2012 ; Litchman et al. , 2015b ). We tested simplified representations of microbial communities by deriving de novo functional ‘guilds’ of microbes based on similar patterns of imputed C, N and P use. Accounting for differences in carbohydrate allocation ( C carb:arom ) and ribosome copy numbers among these guilds yielded stronger relationships with soil C cycling in our soils than phylogenetic relationships alone ( Table 4 ), while using fewer free parameters ( Supplementary Table S11 ). This finding supports our fourth hypothesis that rates of soil C cycling are linked with the abundance of microbial groups defined by differences in stoichiometric resource allocation, although this awaits confirmation in other ecosystems."
} | 4,734 |
27242064 | null | s2 | 8,585 | {
"abstract": "In this paper we demonstrate that 3D printing with a digital light processor stereolithographic (DLP-SLA) 3D printer can be used to create high density microfluidic devices with active components such as valves and pumps. Leveraging our previous work on optical formulation of inexpensive resins (RSC Adv., 2015, 5, 106621), we demonstrate valves with only 10% of the volume of our original 3D printed valves (Biomicrofluidics, 2015, 9, 016501), which were already the smallest that have been reported. Moreover, we show that incorporation of a thermal initiator in the resin formulation along with a post-print bake can dramatically improve the durability of 3D printed valves up to 1 million actuations. Using two valves and a valve-like displacement chamber (DC), we also create compact 3D printed pumps. With 5-phase actuation and a 15 ms phase interval, we obtain pump flow rates as high as 40 μL min(-1). We also characterize maximum pump back pressure (i.e., maximum pressure the pump can work against), maximum flow rate (flow rate when there is zero back pressure), and flow rate as a function of the height of the pump outlet. We further demonstrate combining 5 valves and one DC to create a 3-to-2 multiplexer with integrated pump. In addition to serial multiplexing, we also show that the device can operate as a mixer. Importantly, we illustrate the rapid fabrication and test cycles that 3D printing makes possible by implementing a new multiplexer design to improve mixing, and fabricate and test it within one day."
} | 382 |
35517309 | PMC9054928 | pmc | 8,586 | {
"abstract": "In this study, two alginate-based hydrogels with good mechanical strength, toughness and resilience were synthesized by hydrophobic interaction and coordination bonding. Sodium alginate/poly(acrylamide) semi-interpenetrating network (NaAlg/PAM semi-IPN) hydrogels were first synthesized through the micelle copolymerization of acrylamide and stearyl methacrylate in the presence of sodium alginate, then calcium alginate/poly(acrylamide) double network (CaAlg/PAM DN) hydrogels were prepared by immersing the as-prepared NaAlg/PAM semi-IPN hydrogels in a CaCl 2 solution. FT-IR and XPS results revealed NaAlg/PAM semi-IPN hydrogels and CaAlg/PAM DN hydrogels were successfully synthesized through non-covalent interactions. The tensile strength of CaAlg/PAM DN hydrogels could reach 733.6 kPa, and their compressive strengths at 80% strain are significantly higher than those of the corresponding NaAlg/PAM semi-IPN hydrogels, which is attributed to the alginate network crosslinked by Ca 2+ . The dual physically crosslinked CaAlg/PAM DN hydrogels can achieve fast self-recovery, and good fatigue resistance, which is mainly assigned to energy dissipation through dynamic reversible non-covalent interactions in both networks. The self-healing ability, swelling behavior and morphology of the synthesized alginate-based hydrogels were also evaluated. This study offers a new avenue to design and construct hydrogels with high mechanical strength, high toughness and fast self-recovery properties, which broadens the current research and application of hydrogels.",
"conclusion": "4. Conclusion We have successfully proposed a novel strategy to synthesize fully physically cross-linked alginate-based hydrogels with high mechanical strength, toughness and self-recovery capability, which originates from the dynamically reversible non-covalent interactions. These dynamically reversible non-covalent interactions act as sacrificial bonds to dissipate energy during the deformation process. Mechanical properties of alginate-based hydrogels could be regulated through their network structure and composition. CaAlg/PAM DN hydrogels exhibit high toughness with respect to PAM single hydrogel and NaAlg/PAM semi-IPN hydrogel, which may be attributed to the different network structures and the unzipping of coordination interactions between carboxyl anion of alginate and Ca 2+ . The synthesized alginate-based hydrogels exhibit good mechanical properties, and the CaAlg/PAM DN hydrogels also show fast self-recovery and good fatigue resistance properties. Swelling experiment further demonstrates the formation of coordination interactions between carboxyl anion of alginate and Ca 2+ . This strategy enriches the exploration of alginate-based hydrogels based on non-covalent interactions, and would expand their applications.",
"introduction": "1. Introduction Sodium alginate, which could be obtained mainly from marine brown algae belonging to the Phaeophyceae , is a naturally occurring polysaccharides. 1,2 It is a polyanionic linear copolymer of 1,4-linked-α- l -guluronic acid and β- d -mannuronic acid residues in varying proportions, and is considered as a unique biocompatible, biodegradable and non-toxic polymer. 3 Sodium alginate-based materials have been extensively studied and used in biomedical applications, such as drug delivery, 4 tissue engineering 5 and wound healing. 6 Various material forms based on sodium alginate, such as films, fibers, microspheres and hydrogels, have been designed to meet the needs of the applications. 7 Among these material forms, hydrogels are known for their similarity to the natural extracellular matrix and are being increasingly used in the field of tissue engineering. 8,9 Hydrogels possess the ability to absorb and retain large volumes of water or biological fluids, and don't dissolve in the solvent. 10 Due to their soft and rubbery consistency, water uptake capacity, biocompatibility and similar properties to human tissue, hydrogels are extremely suitable as tissue engineering materials, such as contact lenses, engineering scaffolds and biosensors. 11,12 However, traditional hydrogels exhibit brittleness and inefficient energy consumption, which originates from their low resistance to crack propagation due to the lack of an efficient energy dissipation mechanism in the gel network. 13,14 This leads to poor mechanical performances of hydrogels, and limits their applications. 15 Therefore, imparting excellent mechanical properties to the hydrogels is an urgent problem that must be solved. To obtain a hydrogel with excellent mechanical properties, it is necessary to increase the total viscoelastic dissipation along the hydrogel by introducing a dissipation mechanism at the molecular level. 13,15 Recently, many techniques for synthesizing hydrogels with excellent mechanical properties have been proposed, including interpenetrating network hydrogels, 16 double network hydrogels, 17,18 nanocomposite hydrogels 19,20 and topological hydrogels. 21 Among them, double network hydrogels have demonstrated their excellent mechanical properties. In the double network hydrogels, both networks with contrasting structures are separately cross-linked, and the interpenetration of two networks makes the hydrogels both tough and soft. 22 The toughening mechanism for double network hydrogels is mainly based on sacrificial bonds that break from the first network to effectively dissipate energy and protect the second network, thereby sustain stress and store elastic energy. So the mechanical properties of hydrogels are reinforced. But the fracture of the first network leads to the cleavage of irreversible and permanent bond, so it is difficult to repair and recover the hydrogel from damage and fatigue. 23 The self-healing and self-recovery properties of hydrogels are also especially important for expanding their applications, such as cartilage, tendon, muscle, and blood vessel. The introduction of the reversible non-covalent interaction in place of the sacrificial covalent bond in the first network is considered to be an important way to improve the self-healing and fatigue properties of double network hydrogels caused by permanent bond cleavage. Various non-covalent interactions, such as hydrogen bond, 24 metal-coordination 25,26 and hydrophobic interaction, 27,28 have been incorporated into the double network hydrogels, resulting in a range of attractive properties, such as high strength, fatigue resistance, self-healing, self-recovery, shape memory, and remoldability/recyclability/reusability. A fully, physically crosslinking double network hydrogels were synthesized using a hydrogen bond-associated agar gel as the first network and a hydrophobically associated polyacrylamide gel as the second network, and found that the synthesized hydrogels not only exhibit excellent mechanical strength and high toughness, but also have rapid self-recovery, remarkable fatigue resistance, and notable self-healing performances. 22 Yuan et al. 27 reported a dual physically cross-linked polyacrylamide/xanthan gum double network hydrogels. The results revealed that the prepared hydrogels possess fracture stresses high as 3.64 MPa and compressive stresses at 99% strain of more than 50 MPa. In addition, the introduction of the physically crosslinked structure also gives the hydrogels excellent fatigue resistance and self-healing properties. In this study, alginate-based hydrogels with semi-interpenetrating structure were first synthesized by hydrophobic interaction in the presence of sodium alginate. Then the synthesized hydrogels were soaked in an aqueous solution of CaCl 2 , causing that α- l -guluronic acids (G) residues of sodium alginate could couple with Ca 2+ to form metal-coordinate bond. This acts as a physical cross-linking point, leading to the formation of alginate network. 17,29 So alginate-based hydrogels with double network structure were constructed by hydrophobic interaction and metal-coordination. The mechanical strength, fatigue resistance, self-recovery, self-healing properties and pH-sensitivity of two different structures of alginate-based hydrogels were discussed. As a result, alginate-based hydrogels with double network structure showed excellent mechanical strength, fatigue resistance and self-recovery properties, while alginate-based hydrogels with semi-interpenetrating network structure exhibited excellent self-healing properties. Therefore, the synthesized alginate-based hydrogel with excellent properties has potential application value in many fields.",
"discussion": "3. Results and discussion 3.1 Synthesis and characterization of alginate-based hydrogels \n Scheme 1 displays the schematic diagram for the synthesis of alginate-based hydrogels. As shown in Scheme 1 , SMA is first dissolved in the SDS micelles to form polymerizable micelles, which are hydrophobic micelles. 27 Then sodium alginate is added, and stirred to completely dissolve. After that, poly(acrylamide)-based network is constructed by the micelle copolymerization of acrylamide and stearyl methacrylate in the presence of sodium alginate, where the formed polymerizable micelles act as the crosslinking points. Therefore, NaAlg/PAM hydrogels with semi-interpenetrating network structure are synthesized. The synthesized NaAlg/PAM semi-IPN hydrogels are further immersed in an aqueous solution of CaCl 2 . A ionic bonding cross-linked network is formed through the metal-coordination between carboxyl groups of alginate chain and Ca 2+ ions. 17,29 Because the PAM network intersects with the alginate network each other, the non-covalent interactions, such as van der Waals force and hydrogen bonding, could be formed between PAM and alginate. Therefore, CaAlg/PAM double network hydrogels are synthesized. Scheme 1 Schematic diagram of NaAlg/PAM semi-interpenetrating network hydrogels and CaAlg/PAM double network hydrogels. \n Fig. 1(a) shows FT-IR spectra of sodium alginate, PAM, NaAlg/PAM and CaAlg/PAM hydrogels. In the spectrum of sodium alginate, the broad peak at 3400 cm −1 is due to OH stretching vibration. The obvious absorption peaks at 1612 cm −1 and 1415 cm −1 are attributed to asymmetric and symmetric stretching vibrations of –COO − group of alginate, respectively. 30 The peak at 1030 cm −1 is assigned to the C–O stretching vibration of polysaccharide structure. 30,31 The spectrum of PAM clearly shows the stretching vibration of N–H at 3357 and 3194 cm −1 , C \n \n\n<svg xmlns=\"http://www.w3.org/2000/svg\" version=\"1.0\" width=\"13.200000pt\" height=\"16.000000pt\" viewBox=\"0 0 13.200000 16.000000\" preserveAspectRatio=\"xMidYMid meet\"><metadata>\nCreated by potrace 1.16, written by Peter Selinger 2001-2019\n</metadata><g transform=\"translate(1.000000,15.000000) scale(0.017500,-0.017500)\" fill=\"currentColor\" stroke=\"none\"><path d=\"M0 440 l0 -40 320 0 320 0 0 40 0 40 -320 0 -320 0 0 -40z M0 280 l0 -40 320 0 320 0 0 40 0 40 -320 0 -320 0 0 -40z\"/></g></svg>\n\n O stretching at 1661 cm −1 , and N–H deformation at 1611 cm −1 . 27 In the spectra of NaAlg/PAM and CaAlg/PAM hydrogels, the characteristic peaks of both alginate and polyacrylamide could be observed, and no new distinct peaks are observed. Compared with sodium alginate, the wavenumber for asymmetric –COO − stretching of alginate in the hydrogels decreases, which may be attributed to the electrostatic interaction between metal ions and carboxy anions, hydrogen bonding between alginate and polyacrylamide. 32,33 Especially for CaAlg/PAM hydrogel, the intensity of the peaks corresponded to O–H, asymmetric –COO − stretching and symmetric C–O stretching in C–O–C structure decrease. The possible reason is that the alginate chains are cross-linked to form a network through the strong electrostatic attraction between alginate chains and Ca 2+ ions. 32,33 These results indicate that alginate-based hydrogels with different structures can be synthesized through physically cross-linking interactions. Fig. 1 FT-IR (a) and XPS wide-scan (b) spectra of SA, PAM, NaAlg/PAM and CaAlg/PAM hydrogels; C 1s (c) and O 1s (d) level spectra of SA, PAM, NaAlg/PAM and CaAlg/PAM hydrogels. More detailed structure analysis about NaAlg/PAM semi-IPN hydrogels and CaAlg/PAM double network hydrogels was obtained by XPS. As displayed in Fig. 1(b) , C, O, N and Na elements are presented on the surface of the NaAlg/PAM hydrogel. While Na element is not detected on the surface of CaAlg/PAM hydrogel, but there presents Ca element. This suggests that the ion-exchange reaction occurs when NaAlg/PAM hydrogel is soaked in CaCl 2 solution. The high-resolution C 1s and O 1s spectra are shown in Fig. 1(c) and (d) , respectively. For SA, the C–C/C–H, C–O, and O–C O peaks appeared at 284.8, 286.4 and 288.0 eV, respectively. 34 In the O 1s spectrum of SA, the peaks situated at 531.5 and 532.9 eV indicated the presence of O–C O and C–OH/O–C–O, 35 which is consistent with the C 1s spectrum of SA. The peak at 535.8 eV is attributed to the water absorbed by SA. The peaks at 284.8 and 287.7 eV in the C 1s spectrum of PAM are assigned to C–H/C–C and C–N, respectively. The O 1s spectrum of PAM shows a peak at 532.1 eV, corresponded to N–C O. 36,37 For NaAlg/PAM semi-IPN hydrogel, the C 1s spectrum has four distinctive chemical states: at 284.8 eV related to C–H/C–C, at 285.4 eV for C–O, at 286.4 eV for C–N and 288.8 eV corresponded to O–C O. The corresponding O 1s spectrum exhibits three peaks at 531.8, 532.5 and 533.4 eV, which are attributed to N–C O, C–O and O–C O, respectively. It is observed that the chemical states of C and O in NaAlg/PAM semi-IPN hydrogel contain the chemical states of SA and PAM, and no new chemical states are formed. This reveals that NaAlg/PAM semi-IPN hydrogels are synthesized by the non-covalent interactions. It is clear that the C 1s and O 1s spectra of CaAlg/PAM DN hydrogel are similar to those of NaAlg/PAM semi-IPN hydrogel. However, this is significant that the peak intensity at 288.8 eV corresponded to O–C O reduces, implying the strong coordination between –COO − and Ca 2+ . This indicates that CaAlg/PAM DN hydrogels are prepared by immersing NaAlg/PAM semi-IPN hydrogels in CaCl 2 solution. These results are consistent with the FT-IR result, further confirming that the NaAlg/PAM semi-IPN hydrogel and CaAlg/PAM DN hydrogel based on non-covalent interactions are successfully prepared. The synthesized NaAlg/PAM semi-IPN hydrogels and CaAlg/PAM DN hydrogels present an excellent performance in ductility and mechanical properties, as displayed in Fig. S1 † and 2 . It is clear in Fig. 2(a) and (b) that the CaAlg/PAM DN hydrogel sample (60 mm × 10 mm) is easily stretched to 200% of its original length, and the central notched ( Φ 4 mm) hydrogel sample also remains stable without crack propagation. It can be observed that the sample can be restored to the original length. A rod sample of 8 mm diameter not only can be able to a certain amount of pressure, but also can resist to slicing with a knife, as displayed in Fig. 2(c) and (d) . When the external force is removed, the sample can return to its original state. The sample film can lift a block of 100 g without breaking (as Fig. 2(e) ). The synthesized NaAlg/PAM semi-IPN hydrogels can also go through a similar process without damaging, as displayed in Fig. S1. † Therefore, the synthesized hydrogels display stretchability, flexibility, and self-recovery properties. Next, we would further analyze the performance of the synthesized hydrogels through tensile test, compression test, loading–unloading cycle test and rheological test. Fig. 2 The photographs of CaAlg/PAM DN hydrogel demonstrating the excellent mechanical behaviors: under stretching (a) without and (b) with central notch, compressing (c), slicing with a knife (d) holding a weight of 100 g (e). 3.2 Mechanical properties Mechanical properties of the synthesized hydrogels were evaluated by tensile test and compression test, and the obtained results are shown in Fig. 3 and Table S2. † The previous literature has reported that the mechanical properties of hydrogels are closely related to their water content, their tensile strength decreases as the water content increases. 38 Therefore, the water content of the synthesized hydrogels was measured by gravimetric method, the obtained results are displayed in Table S1. † It is clear that the water content of NaAlg/PAM semi-IPN hydrogels is 39.9–57.5%, while CaAlg/PAM DN hydrogels have a higher water content (67.0–73.4%). This is because the prepared NaAlg/PAM semi-IPN hydrogels don't reach the swelling equilibrium, so it would further swell at the CaCl 2 solution. Therefore, the water content of CaAlg/PAM DN hydrogels is higher than that of the corresponding NaAlg/PAM semi-IPN hydrogels. As shown in Fig. 3 and Table S2, † tensile strength of CaAlg/PAM DN hydrogels with higher water content is larger than that of the corresponding NaAlg/PAM semi-IPN hydrogels, while their elongation at break is significantly less than that of the corresponding NaAlg/PAM semi-IPN hydrogels, which is mainly due to the different network structures. Because the interaction between Ca 2+ and –COO − in alginate chains increases the cross-linking density of polymer network, so the CaAlg/PAM DN hydrogels cross-linked by Ca 2+ exhibit higher mechanical properties. 39 As listed Table S2, † with increasing of the SMA concentration from 1% to 4%, the tensile strength of NaAlg/PAM semi-IPN hydrogels increases from 202.7 kPa to 529.7 kPa and the elongation at break also increases from 18.5 mm mm −1 to 33.6 mm mm −1 . Fig. 3(a) shows the tensile stress–strain curves of CaAlg/PAM DN hydrogels with various SMA concentrations, and the tensile strength and elongation at break obtained from Fig. 3(a) are shown in Fig. 3(b) . CaAlg/PAM DN hydrogels with various SMA concentrations exhibit excellent tensile performances (575.1–733.6 kPa tensile strength and 11.5–17.6 mm mm −1 elongation at break). The tensile strength and elongation at break of CaAlg/PAM DN hydrogels first increase and then decrease as the SMA concentration increases. For CaAlg/PAM DN hydrogel with 2% of SMA, its tensile strength could reach 733.6 kPa, while the elongation at break of the hydrogel with 3% of SMA is largest (17.6 mm mm −1 ). The increased SMA concentration results in the increasing of cross-linking density of the PAM network. Because the higher the density of cross-linking is, the more rigid network is, and the smaller the elasticity is. 40 Therefore, with increasing of the SMA concentration, the tensile strength and elongation at break first increase and then decrease. Compressive stress–strain curves of CaAlg/PAM DN hydrogels with various SMA concentrations are displayed in Fig. 3(c) . It can be observed that the compressive strength at strain 80% first increases and then decreases as the SMA concentration increases, and the compressive strength of CaAlg/PAM DN hydrogel with 2% of SMA could reach 1.354 MPa. However, with increasing of the SMA concentration from 1% to 4%, the compressive strength of NaAlg/PAM semi-IPN hydrogels decreases from 0.528 MPa to 0.396 MPa. Fig. 3 (a and d) Tensile stress–strain curves of CaAlg/PAM DN hydrogels; (b and e) tensile strength and elongation at break of CaAlg/PAM DN hydrogels; (c and f) compressive stress–strain curves of CaAlg/PAM DN hydrogels: (a, b and c) hydrogels with various SMA concentrations; (d, e and f) hydrogels with various SA concentrations. The influence of the sodium alginate concentration on mechanical properties of prepared hydrogels was also investigated. As listed in Table S2, † with increasing of the SA concentration, the tensile strength of NaAlg/PAM semi-IPN hydrogels first increases and then decreases. Tensile strength of NaAlg/PAM semi-IPN hydrogel with 7.5% of SA could reach 678.2 kPa. And NaAlg/PAM hydrogels exhibit larger elongation at break (24.3–38.2 mm mm −1 ). Fig. 3(d) shows the tensile stress–strain curves of CaAlg/PAM DN hydrogels with various SA concentrations, and the obtained results from Fig. 3(d) are displayed in Fig. 3(e) . The tensile strength of CaAlg/PAM DN hydrogels with different SA concentrations is 492.6–733.6 kPa, their elongation at break is 9.9–17.1 mm mm −1 . It is observed that with increasing of the SA concentration, tensile strength and elongation at break first increase and then decrease. The carboxyl anion of alginate can be coupled with Ca 2+ , so the second network based on alginate is formed. As the concentration of SA increases, the second network becomes much denser. The previous literatures 39,41 have reported that because alginate chains could interpenetrate with covalently cross-linked PAM network, the alginate network ionically cross-linked by multivalent cationic could be zipped. For alginate network with lower cross-linking density, the loose alginate network is prone to deformation during stretching, improving the tensile strength and elongation at break. For alginate network with higher cross-linking density, the dense alginate network results in a decrease in the elasticity of the materials, which reduces both tensile strength and elongation at break. As listed in Table S2, † with increasing of the SA concentration, the compressive strength of NaAlg/PAM semi-IPN hydrogels at strain 80% first increases and then decreases. The compressive strength of NaAlg/PAM semi-IPN hydrogel with 5.0% of SA could reach 0.516 MPa. Fig. 3(f) displays the compressive stress–strain curves of CaAlg/PAM DN hydrogels with different SA concentrations. It is clear that the compressive strength of the DN hydrogels first decreases and then increases as the SA concentration increases. For CaAlg/PAM DN hydrogel with 10% of SA, its compressive strength at strain 80% could reach 1.985 MPa. The mechanical properties of alginate-based hydrogels with similar systems are summarized in Table S3. † It is clear that compared with alginate-based hydrogels reported in the literatures, the synthesized alginate-based hydrogels exhibit excellent mechanical properties. On the basis of the above results, we found that the mechanical properties of the synthesized alginate-based hydrogels are not only related to their composition, but also to their network structure. For NaAlg/PAM hydrogel, the interpenetration of alginate chains and PAM network forms a semi-interpenetrating network structure. The increased SMA concentration enhances the cross-link density of the PAM network, and the increased SA could increase the entanglement of alginate and PAM network. So the cross-linking density of alginate-based hydrogel could be improved as the SMA or SA concentrations increase, resulting in the formation of more robust and flexible network structure. This could lead to the increasing of the strength and stretchability of the hydrogels. 42 For CaAlg/PAM hydrogel, alginate chains were cross-linked and formed interpenetrated ionic network due to strong electrostatic attraction between positively charged Ca 2+ ions and negatively charged alginate chains. So the CaAlg/PAM hydrogel exhibits double network structure. With increasing of the concentrations of SMA and SA, the cross-linking densities of the first network and the second network increase, respectively. At lower cross-linking density, the increased cross-linking density could obviously improve the strength and stretchability of the materials. At higher cross-linking density, the increased cross-linking point makes the hydrogel network more denser, so that the polymer chain is not easy to move when the sample is deformed, which could cause the sample to break. Therefore, to construct a double network hydrogel with excellent mechanical properties, when one network has a high cross-linking density, the other must be a loose network. When the synthesized DN hydrogel is deformed under stress, the loose network ruptures to efficiently dissipate energy, protect the soft and ductile network from crack propagation, thereby improving the mechanical properties of the hydrogels. 43,44 3.3 Hysteresis and self-recovery The loading–unloading test was used to evaluate the energy dissipation of hydrogels with different structures. Fig. 4(a) displays the loading–unloading curves of PAM, NaAlg/PAM and CaAlg/PAM hydrogels at strain 600%. For PAM and NaAlg/PAM hydrogels, there is no hysteresis loop on the loading–unloading curves, revealing that they show a typical rubber elastic behavior. 20 The loading–unloading curve of CaAlg/PAM DN hydrogel shows an obvious hysteresis loop, indicating CaAlg/PAM DN hydrogel dissipates energy effectively. The dissipated energies calculated by Fig. 4(a) are shown in Fig. 4(b) . For CaAlg/PAM DN hydrogel, its dissipated energy could reach 540.7 kJ m −3 , which is greater than that of PAM (62.1 kJ m −3 ) and NaAlg/PAM semi-IPN hydrogels (134.1 kJ m −3 ). These results reveal that CaAlg/PAM DN hydrogel has more effective energy dissipation pathways with respect to PAM and NaAlg/PAM semi-IPN hydrogels, which may be ascribed to the interaction between Ca 2+ and –COO − of alginate chains. When the external loading is applied on the CaAlg/PAM hydrogel, the ionic coordination bonds serve as reversible sacrificial bonds and are cracked to effectively dissipate energy. 43 Furthermore, the cycle tensile loading–unloading tests of CaAlg/PAM DN hydrogel at different strains during loading–unloading cycles were measured, and the obtained results are shown in Fig. 4(c) and (d) . It is evident that the area of loop increases with the increasing strain (as Fig. 4(c) ); i.e. , the dissipated energy increases from 73.0 kJ m −3 at strain 200% to 655.7 kJ m −3 at strain 1200% ( Fig. 4(d) ). This further confirms that CaAlg/PAM DN hydrogel has high energy dissipation. However, the reloading curve of CaAlg/PAM DN hydrogel is close, but does not completely overlap with the previous loading curve. The similar phenomenon has been reported in agar/PAMAAc-Fe 3+ double network hydrogels. 25 The possible reason is that the previous loading has induced a change in the gel network, and a portion of dynamic reversible non-covalent interactions can be restored before the next loading. Fig. 4 (a) Loading–unloading curves and (b) the corresponding toughness of NaAlg 0%/PAM 2% hydrogel, NaAlg 5.0%/PAM 2% hydrogel and CaAlg 5.0%/PAM 2% hydrogel; (c) cycle loading–unloading curves and (d) the corresponding toughness of CaAlg 5.0%/PAM 2% hydrogel at various strains. Self-recovery capacity of CaAlg/PAM DN hydrogels was evaluated by the cycle loading–unloading tests, and the obtained results are shown in Fig. 5(a) and (b) . As shown in Fig. 5(a) , no resting time is applied between two consecutive loadings, the area of hysteresis loop significantly reduces. With increasing of the resting time, the area of hysteresis loop increases. It is observed from Fig. 5(b) that toughness recovery ratio increases as the resting time increases. After 30 min of resting, CaAlg/PAM DN hydrogel can recover ∼60% of toughness. With increasing of the resting time from 0 min to 30 min, the residual strain decreases from 38.6% to 13.7%. Furthermore, the fatigue resistance of CaAlg/PAM DN hydrogels was analyzed by the ten cycles loading–unloading at strain of 600%. As shown in Fig. 5(c) , the hysteresis loop and stress of CaAlg/PAM DN hydrogels display an obvious decrease with respect to original cycle, and these figures further decrease as the increase of number of cycle. This indicates that the hydrogel network has been destroyed during loading–unloading process, and it could not be fully recovered quickly. The sample subjected to 10 cycles of loading–unloading was allowed to recover at room temperature for 24 hours, and then the sample was measured again by the ten loading–unloading cycles at strain of 600% (as Fig. 5(d) ). It is observed that maximum stress of the recovered sample is slightly larger than its original stress. This reveals that the destroyed network can be formed again, which is mainly attributed to the dynamic reversible non-covalent interaction. And the reformed network could achieve the optimal reorganization with relieving internal stress of the hydrogel. 45 These demonstrate that CaAlg/PAM DN hydrogels exhibit the excellent fatigue resistance. Fig. 5 (a) Cyclic loading–unloading curves and (b) toughness recovery ratio and residual strain of CaAlg 5.0%/PAM 2% DN hydrogel at different resting time; the ten successive tensile loading–unloading curves of the (c) as-prepared samples and (d) recovered sample after 24 h resting time. 3.4 Rheological behavior \n Fig. 6(a) shows the results of strain sweep for NaAlg 5.0%/PAM 2% semi-IPN hydrogel and CaAlg 5.0%/PAM 2% DN hydrogel. It is clear that the modulus values of NaAlg 5.0%/PAM 2% hydrogel and CaAlg 5.0%/PAM 2% hydrogel have no dependence on the strain when the strain is less than 5%. With further increase of strain, the G ′ value decreases. This indicates that the linear viscoelastic domain for CaAlg 5.0%/PAM 2% hydrogel is larger than that of NaAlg 5.0%/PAM 2% hydrogel. It can be observed that the G ′ value of CaAlg 5.0%/PAM 2% hydrogel in the linear viscoelastic domain is higher than that of NaAlg 5.0%/PAM 2% hydrogel. This may be closely related to cross-linking density of hydrogel network. The G ′′ values of hydrogels display a obvious peak prior to the final decrease, revealing an increasing use of deformation energy to deform subdomains of network structure before the inner structure finally breaks. 46 Fig. 6 Dynamic rheological behaviors of NaAlg 5.0%/PAM 2% and CaAlg 5.0%/PAM 2% hydrogels: (a) strain sweep by frequency of 10 rad s −1 at 25 °C; (b) oscillatory frequency sweeps by 1.0% strain at 25 °C; (c) cyclic continuous step strain measurements in which the strain was switched from 1% strain for 100 s to 100% strain for 100 s; (d) cyclic continuous step strain measurements in which the strain was switched from 1% strain for 100 s to various larger strains (100%, 200%, 300% and 400%) for 100 s. \n Fig. 6(b) displays the frequency dependence of G ′ and G ′′ for NaAlg 5.0%/PAM 2% semi-IPN hydrogel and CaAlg 5.0%/PAM 2% DN hydrogel. It is observed that the G ′ over the measured frequency range is always larger than the G ′′, and both of them have poor frequency dependence. This reveals that the NaAlg/PAM semi-IPN hydrogel and CaAlg/PAM double network hydrogel are prepared by physical interactions. Similar result has been reported by Yang et al. 47 It is also clear that the G ′ value of CaAlg 5.0%/PAM 2% hydrogel is higher than that of NaAlg 5.0%/PAM 2% semi-IPN hydrogel, which is mainly due to the second network formed by Ca 2+ crosslinking of alginate. The loss factor (tan δ = G ′′/ G ′) represents the lost energy to storage energy during deformation. Compared to NaAlg 5.0%/PAM 2% semi-IPN hydrogel, the tan δ value of CaAlg 5.0%/PAM 2% double network hydrogel is much lower (about 0.1), revealing that the formation of network structure leads to the increasing of elastic property more obvious than that of viscous property. 48 And the tan δ value of CaAlg 5.0%/PAM 2% hydrogel is nearly stable and almost independent of frequency. The changes in moduli of NaAlg 5.0%/PAM 2% semi-IPN hydrogel and CaAlg-5.0%/PAM-2% DN hydrogel with the shear strain jump between 1% and 100% are shown in Fig. S2(a) † and 6(c) . As displayed in Fig. 6(c) , the initial G ′ and G ′′ values of CaAlg 5.0%/PAM 2% hydrogel at 1% strain are 15.7 and 1.9 kPa, respectively. The G ′ value decreases to 5.3 kPa and the G ′′ value increases to 4.8 kPa, and the both almost completely return to their original values upon the strain returning to 1% over all the cycles. NaAlg 5.0%/PAM 2% semi-IPN hydrogel also shows a similar phenomenon. This indicates that the synthesized samples show stable hydrogel properties under low shear strain value. During the abrupt increase in the strain up to larger strain (as Fig. S2(b) † and 6(d) ), the G ′ value is lower than G ′′, revealing that the viscous regime dominates over elastic. Upon reversal of the strain, G ′ and G ′′ could completely return to their original values, demonstrating the reformation of the gel network. Similar result has been reported by Sadiys Anjum et al. 49 As shown in Fig. 6(d) , the cycle is repeated with larger strain for up to 400%, the elastic state suddenly retracts without any significant loss in modulus. These results reveal that the synthesized hydrogel exhibits good self-recovery ability, which is consistent with the result of cycle loading–unloading tensile test. 3.5 Self-healing properties Self-healing properties of materials are important for improving the recycling and extending application of materials. The self-healing properties of the materials benefit from dynamic and reversible bonds, such as hydrogen bonding, metal-coordination, and hydrophobic interaction. 50,51 In this study, NaAlg/PAM semi-IPN hydrogels and CaAlg/PAM DN hydrogels were constructed through hydrophobic interaction and metal-coordination. So the self-healing properties of the synthesized hydrogels were evaluated. Fig. 7 shows the self-healing properties of NaAlg 5.0%/PAM 2% semi-IPN hydrogel. As displayed in Fig. 7(a) , NaAlg 5.0%/PAM 2% semi-IPN hydrogel was first cut into two pieces and one piece was dyed with methylene blue, and then the cut surfaces were kept in contact at 40 °C for 24 h. The healed sample can be also withstood bending and stretching deformation, as Fig. 7(b) . These results reveal that there occurs the migration of the polymer chain and the reconstruction of dynamic reversible hydrophobic interaction at the interface of the hydrogel, which could give the materials self-healing properties. The identical result was acquired by an optical microscopy, as presented in Fig. 7(c) . It is clear that the crack of the sample becomes smaller as the self-healing time prolongs, and almost disappears at 48 h. Tensile stress–strain curves of samples subjected to different healing time are shown in Fig. 7(d) , and the healing efficiencies obtained from Fig. 7(d) are presented in Fig. 7(e) . It is observed that the healing efficiency increases as the healing time prolongs. For the sample healed for 48 h, its stress and elongation at break are 231.9 kPa and 4.06 mm mm −1 , respectively, showing a higher self-healing ability (79.5% HE of stress, 16.7% HE of strain). However, CaAlg/PAM DN hydrogels exhibited very low self-healing efficiency, so the stress–strain curves of the samples could not be obtained. Because of the higher binding strength of Ca 2+ and –COO − of alginate, thereby the reversibility of dynamic non-covalent association is greatly reduced. This reduces the ability to achieve self-healing through weak non-covalent bond interactions. Therefore, we should consider how to improve the self-healing efficiency of CaAlg/PAM DN hydrogels in future research. Fig. 7 Self-healing properties of NaAlg 5.0%/PAM 2% hydrogel: (a) digital photographs of the process to prepare healed sample; (b) digital photographs of healed sample that can withstand different deformations; (c) optical microscopy images of the sample after being healed for various time; (d) typical stress–strain curves of healed hydrogels; (e) healing efficiency of healed hydrogels. 3.6 Swelling behavior and surface morphology \n Fig. 8 shows swelling behavior of NaAlg 5.0%/PAM 2% semi-IPN hydrogel and CaAlg 5.0%/PAM 2% DN hydrogel in pH = 7.4 buffer solution. It is clear in Fig. 8(a) that the swelling ratio of NaAlg/PAM semi-IPN hydrogel is larger than that of CaAlg/PAM DN hydrogel. This further confirms that CaAlg/PAM DN hydrogels have a higher cross-linking density, which is attributed to the metal-coordination between carboxyl anion of alginate and Ca 2+ . It is observed in Fig. 8(b) that the volume of the sample increases significantly as the pH value of the solution increases, which may be attributed to ionization of hydrophilic COOH groups of hydrogels. 52 The influence of the composition on equilibrium swelling ratio of the synthesized hydrogels at various pH buffer solutions (1.5, 7.4 and 12.0) are shown in Fig. 8(c), (d) and S3. † As shown in Fig. 8(c) and S3(a) † , with increasing of the SMA concentration, the equilibrium swelling ratio at similar pH buffer solution decreases. The possible reason is that the increased SMA leads to an increase in the cross-linking density of the PAM network, which limits the swelling of the hydrogel. For CaAlg/PAM DN hydrogels with various SA concentrations (as Fig. 8(d) ), its equilibrium swelling ratio in pH 1.5 and 7.4 buffer solutions slight increases as the SA concentration increases from 0% to 5.0%. The possible reason is that the introduction of SA may initially interfere with the PAM network, resulting in the decreasing of physical interactions in the hydrogel. With further increasing of the SA concentration, the equilibrium swelling ratio decreases, which may be assigned to filling up of the void spaces of the network chains by alginate. The similar result has been reported by Samanta, et al. 53 In the pH 12.0 buffer solution, the swelling ratio of CaAlg/PAM hydrogels has been increasing as the SA concentration decreases. The physical interactions between the polymer chains in higher pH solution are significantly weakened, so the swelling ratio of the sample is determined by its crosslink density. With increasing of the SA concentration, the alginate network cross-linked by Ca 2+ is more denser, resulting in a low swelling ratio. However, NaAlg/PAM hydrogels with various SA concentrations exhibit a different phenomenon from CaAlg/PAM hydrogels, as displayed in Fig. S3(b). † This may be related to their different material structures. Fig. 8 (a) Swelling behaviors of NaAlg/PAM and CaAlg/PAM hydrogels in pH = 7.4 buffer solution; (b) the mechanism of pH-sensitivity swelling behavior of CaAlg/PAM hydrogels; (c and d) equilibrium swelling ratios of CaAlg/PAM hydrogels with different compositions in different buffer solutions: (c) hydrogels with various SMA concentrations, (d) hydrogels with various SA concentrations. SEM images of hydrogels that reached equilibrium swelling in pH 7.4 buffer solution are displayed in Fig. 9 . As displayed in Fig. 9(a) and (b) , NaAlg 5.0%/PAM 4% and NaAlg 10.0%/PAM 2% semi-IPN hydrogels show a markedly porous structure and intersected pore channels. However, the corresponding CaAlg/PAM DN hydrogels show a relatively dense surface, as shown in Fig. 9(e) and (f) . This may be mainly attributed to the formation of Ca 2+ ion cross-linked alginate network. It is clear that the SEM image of CaAlg 5.0%/PAM 4% DN hydrogel presents a honeycomb-like surface, and the pore size clearly reduces with respect to NaAlg 5.0%/PAM 4.0% semi-IPN hydrogel. Compared with NaAlg 10.0%/PAM 2% semi-IPN hydrogel, the pore density of CaAlg 10.0%/PAM 2% DN hydrogel is significantly reduced. This is because CaAlg/PAM DN hydrogels have a higher cross-linking density with respect to the corresponding NaAlg/PAM semi-IPN hydrogels. The results are consistent with the results of swelling test and mechanical performances. Comparing CaAlg 5.0%/PAM 2% ( Fig. 9(c) ) and CaAlg 5.0%/PAM 4% ( Fig. 9(e) ) DN hydrogels, it can be found that the sample with higher SMA concentration presents larger pore density. The increased SMA concentration increases the cross-linking density of PAM network, limiting the swelling of the hydrogel. This could result in the reduced pore size and increased pore density. With increasing of the SA concentration, the pore density of the hydrogel decreases, as displayed in Fig. 9(c), (d) and (f) . The possible reason is that the increased SA concentration leads to a denser alginate network cross-linked with Ca 2+ ions, so that the hydrogel is less swellable, resulting in a decrease in pore density. These results reveal that the surface morphology of the prepared alginate-based hydrogel can be adjusted by the network structure and composition of the hydrogel. Fig. 9 SEM images of alginate-based hydrogels: (a) NaAlg 5.0%/PAM 4%; (b) NaAlg 10.0%/PAM 2%; (c) CaAlg 5.0%/PAM 2%; (d) CaAlg 0%/PAM 2%; (e) CaAlg 5.0%/PAM 4%; (f) CaAlg 10.0%/PAM 2%."
} | 10,087 |
35310623 | PMC8924478 | pmc | 8,587 | {
"abstract": "Algae (including eukaryotic microalgae and cyanobacteria) have been genetically engineered to convert light and carbon dioxide to many industrially and commercially relevant chemicals including biofuels, materials, and nutritional products. At industrial scale, genetically engineered algae may be cultivated outdoors in open ponds or in closed photobioreactors. In either case, industry would need to address a potential risk of the release of the engineered algae into the natural environment, resulting in potential negative impacts to the environment. Genetic biocontainment strategies are therefore under development to reduce the probability that these engineered bacteria can survive outside of the laboratory or industrial setting. These include active strategies that aim to kill the escaped cells by expression of toxic proteins, and passive strategies that use knockouts of native genes to reduce fitness outside of the controlled environment of labs and industrial cultivation systems. Several biocontainment strategies have demonstrated escape frequencies below detection limits. However, they have typically done so in carefully controlled experiments which may fail to capture mechanisms of escape that may arise in the more complex natural environment. The selection of biocontainment strategies that can effectively kill cells outside the lab, while maintaining maximum productivity inside the lab and without the need for relatively expensive chemicals will benefit from further attention.",
"introduction": "Introduction Genetic modification of algae, including eukaryotic microalgae and cyanobacteria, is expected to facilitate direct conversion of light energy and inorganic carbon to a wide variety of valuable chemicals ( Angermayr et al., 2015 ; Gomaa et al., 2016 ; Santos-Merino et al., 2019 ; Arora et al., 2020 ). As with other genetically modified organisms (GMOs), the environmental risk of large-scale cultivation must be assessed, and appropriate measures must be taken to mitigate those risks. Previously, Henley et al. (2013) reported a risk assessment for genetically engineered microalgae ( Henley et al., 2013 ), finding that risks to human health, the environment, and the economy, were generally low, but not zero. Given that genetically engineered algae may be grown outdoors, possibly in open ponds, they determined that the potential for these cells to escape into the environment is elevated beyond that of typical industrial microbial cultivation. Henley et al. (2013) therefore, recommended the development of biocontainment strategies which reduce growth fitness in the natural environment, that are conditionally lethal to the cells when they are not in the lab or industrial setting, and that have reduced capability to transfer genetic material to other organisms. Since that report, many new genetic biocontainment strategies have been developed for microalgae and other industrially relevant microorganisms which achieve one or more of those aims (reviewed by Lee et al., 2018 ; Wang and Zhang, 2019 ; Kim and Lee, 2020 ; Arnolds et al., 2021 ; Kallergi et al., 2021 ). In synthetic auxotrophy, cells are modified to make their growth dependent on an unusual or nonnatural nutrient or an unnaturally high concentration of a nutrient. Examples include dependence on unusual phosphorous sources like phosphite ( Motomura et al., 2018 ) and dependence on high concentrations of carbon dioxide ( Clark et al., 2018 ; Lee et al., 2021 ). Further efforts have been made to express toxic proteins, such as nucleases, in the cells in a manner dependent on the conditions outside the lab, typically, the loss of some synthetic signal molecule or an unnatural concentration of a signal molecule. Biocontainment strategies have been collected in the Biocontainment Finder on the Standardsinsynbio.eu website. This review will summarize the rationale for designing genetically encoded biocontainment systems and the efforts made thus far to assess their efficacy in genetically engineered algae. First, we discuss the possible escape routes and fates of escaped algae. The regulatory requirements for outdoor cultivation of genetically engineered algae in a few regions are then summarized. Next, an overview of the different types of genetically encoded biocontainment strategies that may be used in algae is provided. We examine whether lab tests, which frequently demonstrate the achievement of meeting the NIH guideline of a 10 −8 cell survival rate ( USA Department of Health and Human Services: National Institutes of Health, 2019 ), are truly representative of what may occur if cultures were released into the natural environment. Finally, we discuss the results of some specific examples of genetically encoded biocontainment found in recent publications and finish by suggesting future directions. How Might GE Algae Escape? What Are the Consequences of Escaped GE Algae? Biological invasions may proceed through different stages of escape, including proliferation, spread, and persistence. Invading organisms often die out, but in some cases may “alter fundamental ecological properties such as the dominant species in a community and an ecosystem’s physical features, nutrient cycling, and plant productivity” ( Mack et al., 2000 ). We focus first on dispersal and how algal cultivation is likely to differ from that of heterotrophs. Heterotrophic microbes are generally grown in fermenters, inside buildings, with little exposure to the environment. In this situation, escape is most likely to occur via discharge of spent growth media with imperfect prior removal of the microbes. Large accidental spills from fermenters or the harvesting equipment may potentially flow out of buildings or greenhouses. At smaller scales, microbes can also hitch a ride on any equipment or workers in contact with the culture. Algae grown at an industrial scale are likely to grow outside, possibly in open ponds, to take advantage of the free energy source provided by sunlight. This direct exposure to the environment presents challenges in terms of the lack of control over conditions such as temperature and light intensity, as well as the significant problem of biological contamination. Competition from natural algae may reduce yields of the desired product and predators may quickly devour the cultivated species ( US DOE EERE BETO, 2021 ). To prevent genetically engineered algae (GE algae) grown outdoors from leaving the ponds, regulatory agencies in the United States and Mexico have required secondary containment, such as earthen berms, around the ponds to prevent spills from leaving the facility. Further, netting has been required to prevent birds, small mammals, and insects including aphids from entering the pond and potentially carrying away algae to another location ( Szyjka et al., 2017 ; González-Morales et al., 2020 ). Regulatory agencies may consider enclosed bioreactors differently from open ponds. However, there is likely a heightened risk of glass or plastic bioreactors breaking if they are located outside, compared with those located indoors. In contrast, regulatory agencies have focused less on whether GE algae can establish themselves in the natural environment once they have escaped the cultivation system. It is generally expected that GE algae are poorly suited for growth in the environment. Some species which are considered model species have been grown continuously in laboratories and may have evolved or acclimated to the favorable environment of the lab where they typically have media much richer than anything found in the environment, they are not exposed to UV light, and where predators and competitors are carefully excluded by researchers. Cells engineered to produce large quantities of valuable products may be further disadvantaged by the metabolic burdens imposed in generating those products. It is not clear, however, that this common conception has been tested by examining the growth of such organisms in natural conditions. In addition, reduced growth rates do not preclude the persistence of escaped cells in the environment. As others have concluded, the risk of GE algae growing in the environment is not zero ( Henley et al., 2013 ). In addition to the potential for growth at some low rate, escaped cells may continue to exist in a state of persistence. Some bacteria can form spores to persist in environmental conditions unfavorable for growth. Even bacteria that cannot form spores may enter a state of low growth to persist in nutrient-limited conditions ( Gray et al., 2019 ). The persistence state may allow cells time to mutate any toxic genes used for biocontainment, and thus escape. The assays commonly used to assess escape frequency, such as growth curves and colony counting, may not capture this mechanism since the extremely slow growth of the persistent state may appear the same as cell death. If GE microbes escape physical containment, establish themselves, and persist in the natural environment, what harm may be done? A report commissioned by the government of Netherlands summarized possible risks of escape ( Enzing et al., 2012 ). GE microbes that escape compete with native species for nutrients. They may continue producing the valuable products they have been engineered to make, in turn impacting microbial community dynamics. In some cases, those products may have some toxicity to other organisms. The engineered organisms may themselves become food for other organisms which may alter the environment in an unpredictable way. GE algae may contribute to worsening eutrophication, or enriched nutrients in natural water bodies, which can lead to the reduction of dissolved oxygen that may follow an algal bloom. Horizontal gene transfer from GE algae to other organisms could also result in the further spread of antibiotic resistance genes as these genes are often used as selection markers for genetic modification. Ideally, all of these possible ecological disruptions should be avoided. Figure 1 summarizes some of the possible regulatory requirements for outdoor growth of GE algae (A) and the two types of genetically encoded biocontainment strategies that have been demonstrated in lab tests for GE algae (B). Figure 1 (A) Present regulatory requirements for outdoor cultivation of genetically engineered algae have focused on the use of netting and secondary containment such as earthen berms to prevent physical escape and monitoring for escape with nearby catch traps. Catch traps are monitored for growth of the GE algae species grown in the main cultivation system using quantitative PCR. (B) Genetically encoded biocontainment is expected to provide another layer of security to prevent growth of GE algae outside the lab or industrial setting. Synthetic auxotrophy creates a growth dependence on an unusual nutrient (dark-circled stars) or an unusually high concentration of a nutrient not likely to be found in nature. Modifications to prevent growth on the usual nutrients (light-circled stars) found in the environment (e.g., knockout of transporters for the usual nutrient—red do not enter symbols) may be necessary to enforce this dependence. Growth can proceed normally in lab or industrial setting when the unusual nutrient is provided in sufficient quantity. In the natural environment, the GE biocontained cells cannot utilize the usual nutrient and cannot grow. Active biocontainment strategies rely on a signal molecule (light-circled star) that is not likely to be found outside the lab/industrial setting to control expression of a toxic protein. Expression of the toxic gene (e.g., a nuclease) is repressed by the signal molecule which can be provided in the lab or industrial setting. Induction by the signal molecule’s absence in the natural environment results in cell death. Recent Field Tests and Regulatory Requirements for Outdoor Growth of GE Algae The legal status of GE algae (and genetically engineered organisms in general) in the United States has been reviewed in 2014. The authors focus on the challenges for regulating modified microbes with one of the most difficult challenges being in the research and development stage, when the hazards of deployment may not be known until the research is completed ( Mandel and Marchant, 2014 ). Some regulations may apply depending on what genetic modifications were made and how the modified organism will be used. The Federal Insecticide, Fungicide, and Rodenticide Act (FIFRA) may apply to microorganisms intended to be used as pesticides. The FDA regulates microbes that alter the nutrition of food (including if the algae are the food/drug/nutrition supplement) and the Federal Food, Drug, and Cosmetics Act may apply depending on the product application. In a few cases in the United States and Mexico, companies have discovered what the environmental regulations entail for pilot-scale plants growing GE algae. In the United States and Mexico, environmental agencies have focused first on reducing the chance that the GE algae can leave the cultivation system. For open ponds, this has meant utilizing secondary containment berms around the ponds, netting to avoid contact between the culture and insects and animals, and monitoring of traps arranged outside of the culture vessel or pond to detect escape via wind carried water droplets. We next review some details of the permitting process for a pilot plant proposed by the algae biotechnology company, Algenol, in the United States as it has been well documented. Algenol previously pursued modification of a proprietary cyanobacteria species, called AB1, to produce ethanol ( Legere et al., 2017 ). In 2009, Algenol obtained a determination from the USDA Biotechnology Regulatory Services of the Animal and Plant Health Inspection Service (APHIS) that the strains they utilized were not pathogenic for plants, animals, or humans, and they were therefore not regulated under the Plant Protection Act. A permit was still needed under APHIS for interstate transport of hybrid algae. Research and development activities conducted inside a structure by Algenol were also exempted from permitting under the Toxic Substances Control Act (TSCA). GE algae grown outdoors generally require TSCA Environmental Release Applications (TERA). Since Algenol proposed to cultivate GE algae in closed photobioreactors, the EPA indicated that the facility would be exempted from the TERA application process because they were considered a “contained structure” under the Code of Federal Regulations (CFR) 40 CFR 725.234. 1 Forty CFR 725.3 simply defines “structure” as “building or vessel which effectively surrounds and encloses the microorganism and includes features designed to restrict the microorganism from leaving.” The Environmental Protection Agency (US) provides specific guidance for TSCA application for algae in the document “Algae Supplement to the Guidance Document: Points to Consider in the Preparation of TSCA Biotechnology Submissions for Microorganisms.” 2 More recent examples of permitted outdoor cultivation of GE algae at facilities operated by Sapphire Energy in California, United States, and StelaGenomics México in Guanjuato, Mexico are referenced in recent publications. The requirements of the TERA for Sapphire Energy included maintaining secondary containment, netting, and monitoring bucket traps outside the main cultivation pond for growth of the GE algae ( Szyjka et al., 2017 ). Similar requirements were made in Mexico under the Biosafety Law of GMOs ( González-Morales et al., 2020 ). A strain of Acutodesmus dimorphus , modified to increase fatty acid synthesis, was grown at a Sapphire Energy pilot plant in California. Traps outside the pond were filled with tap water supplemented with algae growth medium. qPCR and metagenomic analysis were used to monitor the growth in the traps. Very low abundance of the GE strain was found. The wild-type (WT) Acutodesmus dimorphus was relatively high in abundance in the traps leading the authors to speculate that the WT may have arrived from the surrounding environment since it is known to have natural abundance in the vicinity. This work also examined the invasion potential of the GE A. dimorphus compared with the wild type by inoculating water collected from nearby lakes with each strain. Both strains were found to have minimal impacts on the natural diversity of the lakes. Similar monitoring was done by González-Morales et al. (2020) to comply with regulations for biosafety, in Mexico. Tap water supplemented with BG11 media was placed in traps outside the main cultivation site (a shallow pond with a paddlewheel in a racetrack configuration or raceway pond) between 3 and 28 m distant. PCR and RT-qPCR did not amplify genomic DNA from the engineered Synechococcus elongatus PCC 7942 grown in the raceway. In these two cases, transfer of GE algae to the surroundings was minimal. In Netherlands, another algae biotechnology company, Photanol, has operated a pilot plant and demonstration plant in Amsterdam and Delfzijl, which cultivate genetically modified cyanobacteria. The Dutch government requires permits for contained use and introduction into the environment of GMOs. Market applications of such organisms require a third permit. In the European Union, member nations implement the directives of the European Commission (EC). According to a report commissioned by Netherlands Commission on Genetic Modification, EC directives 2009/41/EC and 2001/18/EC require regulation of contained uses, which generally may apply to GMO algae/cyanobacteria grown indoors (2009/41/EC) and those which may be grown outdoors, including those grown in sealed photobioreactors if they are outdoors (2001/18/EC). For both, an environmental risk assessment is required. However, the facilities can be exempted from risk assessment if the cultivation system has a history of safe use under “good industrial large-scale practice” (GILSP), and the particular GMO is composed of a non-pathogenic host, a “safe” vector, and insert, and the resulting GMO has a lower fitness in the environment than the wild-type host organism ( Enzing et al., 2012 ). Overview of Strategies for Genetically Encoded Biocontainment To address the risks associated with outdoor growth of GE algae, several genetically encoded biocontainment systems have been developed. General strategies have included developing synthetic auxotrophy ( Figure 1B ) where cell growth is made dependent on some unusual nutrient, and genetic circuits that can sense a change in environmental conditions that indicates the cell has left the controlled conditions of the lab and induce expression of toxic proteins or suppress expression of essential genes ( Arnolds et al., 2021 ). Synthetic auxotrophy may be conferred by knockout of genes required for nutrient utilization combined with the introduction of genes needed to utilize some unusual nutrient that the cells are unlikely to encounter outside of the lab/industrial setting. One advantage of this approach is that it may be less likely for cells to mutate to regain the ability to utilize more common nutrients. Unlike inducible lethal genes, this does not rely on a signaling pathway in which a mutation in any component may result in failure, and the continued growth of the cells in the environment. One example of this is found in work that has introduced genes needed for phosphite uptake and utilization combined with the knockout of the phosphate transporters ( Motomura et al., 2018 ). A slightly different approach was taken by researchers who knocked out the carbon concentrating mechanism (CCM) genes from a cyanobacterium, resulting in a strain that was dependent on high CO 2 concentrations for growth ( Clark et al., 2018 ). Both studies were able to demonstrate hypothetical compliance with the NIH’s escape frequency guideline of one in 10 8 when the cells were grown in zero phosphite and ambient CO 2 concentrations, respectively. A related strategy for biocontainment may be the utilization of organisms that have already evolved to survive in uncommon environments including extremophile species, such as members of Cyanidiales which can only grow in acidic environments. Biocontainment systems which are dependent on the absence of the unusual nutrients should carefully consider whether alternative nutrients may be available in some environments. For example, organic carbon sources may be utilized by some algae which would reduce the efficacy of the high inorganic carbon requirement of the system described above. Phosphite-dependent strains may be able to take advantage of naturally occurring phosphite. One lake in eastern China was found to have 1.58 μg/kg phosphite near the surface, which represented about 5.51% of the total phosphorous ( Han et al., 2013 ). Analysis of freshwater from samples in six locations in Florida, United States showed that phosphite and hypophosphite frequently represented more than 25% of the dissolved phosphorous ( Pasek et al., 2014 ). The highest concentration of phosphite measured in that report was approximately 0.1 mM. Bacteria capable of utilizing reduced phosphorus are widespread ( Stone and White, 2012 ), suggesting that phosphite and hypophosphite are common in the environment. Active biocontainment strategies are distinguished from passive strategies by the utilization of lethal genes, which are induced by a change in the concentration of a signal molecule that the cells are expected to experience if they were to escape from their normal cultivation media into the natural environment ( Figure 1B ). Inducible promoters are typically required to control expression of lethal genes. However, the options for well-characterized inducible promoters are limited in algae. Some of the best candidate promoters rely on a synthetic molecule for repression, such as the allolactose analog, isopropyl ß-D-1-thiogalactopyranoside (IPTG), and the tetracycline analog anhydrotetracycline (aTc). While these have been effective in controlling gene expression and are unlikely to be encountered in significant concentrations in nature, the addition of these to large-scale cultures may be a significant expense. It is important to choose robust signals for induction of the selected containment system. The signal response (cell death) should be strong when the cell exists outside the lab or industrial setting. In the production setting, expression of lethal genes must be minimized to reduce the loss in productivity that is expected to result. Metal ion-inducible promoters such as PnrsB (Ni 2 + -inducible) have been used frequently for this purpose because they are tightly repressed in the absence of metal ions ( Englund et al., 2016 ). This tight repression is an important property when expressing toxic genes ( Cheah et al., 2013 ). However, with this promoter and others like it, toxic gene expression would be repressed in most natural waters because they have low concentrations of metal ions. It may be possible to invert this signal by using the metal ion-inducible promoter to drive expression of a repressor transcription factor that acts on the promoter of the toxic gene. It is unclear how this might impact the leakiness of the lethal gene expression. Types of Lethal Genes A variety of lethal genes have been utilized in cyanobacteria—both for biocontainment and for counterselection including proteins from toxin-antitoxin systems and phage lysis proteins ( Cheah et al., 2013 ; Čelešnik et al., 2016 ; Zhou et al., 2019 ). An important benefit to using toxins from antitoxin systems is that the antitoxin can be co-expressed at a low level (or induced only in the lab/industrial setting) to prevent leaky expression of the toxic protein from reducing growth rates in the lab. Thousands of toxin-antitoxin systems have been identified, and this remains an active area of research in microbiology ( Page and Peti, 2016 ). Such systems may have diverse roles in their native context. They were hypothesized to be important to plasmid maintenance by the mechanism known as post-segregation killing ( Gerdes et al., 1986 ). In such systems, the antitoxin is encoded on a plasmid and the toxin in the chromosomal DNA. Cells that do not inherit a copy of the plasmid with the antitoxin are killed by the expression of the toxin alone. More recently, they have been suggested to be important in inducing a persistence state in poor environmental conditions ( Gerdes, 2016 ; Page and Peti, 2016 ). Under normal growth conditions, the expression of the antitoxin exceeds the expression level of the toxin. A stressor, such as nutrient starvation or antibiotics exposure, perturbs this balance and the toxin expression exceeds that of the antitoxin, leading to growth arrest which can be reversed when more favorable conditions return. Some work has suggested that persistence is stochastically induced within populations ( Verstraeten et al., 2015 ), which can be beneficial in surviving infrequent, severe, and difficult-to-sense stresses ( Kussell and Leibler, 2005 ). Page and Peti (2016) provide an excellent review of the evidence that at least some toxin-antitoxin systems are used in this way ( Page and Peti, 2016 ). They classified systems according to the mechanism of action of the toxin gene and the mechanism by which the antitoxin can inactivate the toxin. Toxins may halt metabolism by depolarizing membranes, prevent production of new proteins by wholesale degradation of mRNA, or degrading the already existing proteins. The potential reversibility of any of these types of toxins may make them less attractive for biocontainment strategies. Generally, it is not known how long a cell can persist while these toxins are active and this performance parameter has not typically been measured in biocontainment reports. Nucleases that degrade the chromosomal DNA might be more advantageous because the mechanism of killing the cell also degrades the recombinant DNA and may prevent horizontal gene transfer to or from the GMO. Restriction enzymes, Cas9, and nucleases that are used to scavenge nucleotides from the environment are all candidates for this category. Interestingly, EcoRI has been used in a biocontainment module in Escherichia coli despite the fact that E. coli DNA should be protected from EcoRI by the native methylation pattern. Apparently, overexpression of the restriction enzyme can easily overcome the protection offered by the native methylation and cause cell death by DNA degradation ( Chan et al., 2016 ). Some nucleases are used by cells to scavenge nucleotides from the environment. The NucA nuclease from Serratia marcescens is one such protein which has been used for biocontainment ( Balan and Schenberg, 2005 ). This nuclease is activated by disulfide bond formation which does not occur in the reducing environment inside the cell. Only when it is secreted into the oxidizing environment outside the cell does it become active. The organism’s own nucleotides are thus protected ( Benedik and Strych, 1998 ). This mechanism of inactivation may limit the utility of this nuclease in biocontainment strategies. Another nuclease, NucA from Anabaena sp. PCC 7120, is inactivated by dimerization with a specific inhibitor protein, NuiA ( Meiss et al., 2000 ). In this case, by careful selection of promoters, NucA and NuiA can be co-expressed as part of a biocontainment module such that NucA is inactivated by NuiA in the lab, but NucA expression exceeds that of NuiA when the cell escapes into the environment. Since NucA is secreted by Anabaena , it may also be important to identify and remove any signal peptides that target the protein to the extracellular space so it can effectively degrade the DNA and RNA inside the cell. Genetic Instability Mutations within toxic protein coding sequences that may result in inactive protein or reduced toxicity are a serious threat to the efficacy of biocontainment modules. Horizontal gene transfer from other organisms potentially can complement knockouts made for synthetic auxotrophy strategies. The natural competence of some cyanobacteria to uptake DNA increases the probability that this may occur. In some species, the pili gene hfq is essential to natural competence ( Dienst et al., 2008 ), and in one study, the natural competence genes were knocked out in order to maintain the synthetic auxotrophy ( Clark et al., 2018 ). This suggests one method to avoid possible failure of the genetic biocontainment module. Mutation hot spots within toxin gene should be avoided if possible ( Rogozin and Pavlov, 2003 ). Overlapping the coding sequence of the toxin gene with an essential gene has been proposed as a method for selecting against mutations, though the process of designing and testing such intertwined coding sequence presents a formidable challenge ( Blazejewski et al., 2019 ). Toxin genes used in the reports discussed below have generally been chromosomally integrated rather than maintained on replicating plasmids which may easily be lost, especially if there is selective pressure against them. Genes needed to facilitate growth through synthetic auxotrophy may be maintained on replicative plasmids. However, this may facilitate the spread of those genes to possible contaminating species through plasmid transfer. Specific Examples of Genetically Encoded Biocontainment Modules in Cyanobacteria In this section, we discuss the recent successes in developing biocontainment modules in cyanobacteria. Table 1 summarizes these studies. Table 1 A summary of recent reports which have tested biocontainment strategies in cyanobacteria. Strain Type Promoter Induction Genes/proteins Escape frequency Reference Synechocystis sp. PCC6803 Toxin-antitoxin \n PcopM \n Zn 2 + NucA/NuiA (from Anabaena ) “Complete autodestruction upon Zn 2 + induction” \n Čelešnik et al., 2016 \n Synechocystis sp. PCC6803 Toxin-antitoxin \n PnrsB \n Ni 2 + or Co 2 + NucA/NuiA (from Anabaena ) Weak autotoxicity \n Čelešnik et al., 2016 \n Synechocystis sp. PCC6803 Toxin-antitoxin \n PcopB \n Zn 2 + ssr1114/slr0664 Weak autotoxicity \n Čelešnik et al., 2016 \n Synechocystis sp. PCC6803 Toxin-antitoxin \n PcopB \n Zn 2 + slr6101/slr6100 Weak autotoxicity \n Čelešnik et al., 2016 \n Synechocystis sp. PCC6803 Toxin-antitoxin \n PrnpB \n Constitutive (antitoxin induced by Zn 2 + ) ssr1114/slr0664 Weak autotoxicity on metal ion withdrawal (antitoxin expressed using P copB ) \n Čelešnik et al., 2016 \n Synechococcus sp. PCC7002 Synthetic auxotrophy High CO 2 dependence (CCM deletion) <1 × 10 −9 /CFU \n Clark et al., 2018 \n Synechococcus elongatus PCC7942 Synthetic auxotrophy Phosphite dependence Below detection limit over 28 days (3.6 × 10 −11 per CFU) \n Motomura et al., 2018 \n Synechococcus sp. PCC7002 Growth on melamine \n Pc223 \n Constitutive Synthetic melamine utilization operon Could be converted to synthetic auxotrophy strategy if ammonia and nitrate uptake inhibited \n Selão et al., 2019 \n Synechococcus sp. PCC7002 Growth on phosphite \n PpsbA (A. hybridus) \n Constitutive PtxD from Pseudomonas stutzeri WM88 Could be converted to synthetic auxotrophy strategy if phosphate uptake inhibited \n Selão et al., 2019 \n Synechococcus elongatus PCC7942 Toxin-antitoxin \n PisiAB \n Reduced iron availability SepA2/SepT2 <1 × 10 −9 /CFU \n Zhou et al., 2019 \n Synechococcus elongatus PCC7942 Toxin-antitoxin \n PisiAB \n Reduced iron availability SepA1/SepT1 Weak autotoxicity \n Zhou et al., 2019 \n Synechococcus elongatus PCC7942 Toxin-antitoxin \n PisiAB \n Reduced iron availability slr6101/slr6100 Weak autotoxicity \n Zhou et al., 2019 \n Synechococcus elongatus PCC7942 Toxin-antitoxin \n PisiAB \n Reduced iron availability ssr1114/slr0664 Limited growth \n Zhou et al., 2019 \n Synechococcus elongatus PCC7942 Membrane disruption \n PisiAB \n Reduced iron availability P22 phage holin-endolysin Slightly reduced growth in induction media \n Zhou et al., 2019 \n Synechococcus elongatus PCC7942 Toxin-antitoxin \n PisiAB \n Reduced iron availability NucA/NuiA (from Anabaena ) Growth arrest after 24 h \n Zhou et al., 2019 \n Synechococcus elongatus UTEX2973 Toxin-antitoxin \n PisiAB \n Reduced iron availability SepA2/SepT2 <1 × 10 −9 /CFU \n Zhou et al., 2019 \n Synechococcus elongatus UTEX2973 Toxin-antitoxin \n PisiAB \n Reduced iron availability SepA1/SepT1 Weak autotoxicity \n Zhou et al., 2019 \n Toxic Proteins Čelešnik et al. (2016) tested several toxin proteins in Synechocystis sp. PCC6803 (S. 6803). They focused on metal ion-inducible promoters including the zinc-inducible PcopB and PcopM , and nickel-inducible PnrsB promoters, which tend to be tightly controlled and not leaky. PnrsB had previously been used by others to control the expression of the toxin mazF gene which could be used as a counterselection marker ( Cheah et al., 2013 ). Toxin proteins included the NucA nuclease from Anabaena , which is used to scavenge nucleic acids from the environment and can degrade single and double-stranded DNA and RNA ( Meiss et al., 1998 ). Along with NucA, two other native toxin-antitoxin systems from S. 6803 were tested with the same three metal ion-inducible promoters. The P copM -NucA combination resulted in growth arrest in liquid culture, loss of viability in a tetrazolium assay (a measure of reducing capacity), and no growth on agar plates when induced by zinc, showing the toxicity of this gene. Strains with PnrsB -driven NucA and PcopB -driven slr0664 (a putative RNase and relative of RelE from E. coli ; Ning et al., 2011 ) were still capable of growing, though at a slower rate. In each strain, the antitoxin was co-expressed to avoid growth defects that may arise due to leaky expression. As the authors point out, metal ion concentrations in natural waters tend to be much lower than the ~4–20 μM concentrations needed to induce these promoters. They, therefore, tested another design which put the antitoxin of the slr0664 toxin under control of the PcopB promoter and the toxin under control of the constitutive PrnpB promoter. This strain showed reduced growth compared to the wild type when grown in standard BG11 and similar growth to wild type when grown in BG11 supplemented with 4 μM zinc ( Čelešnik et al., 2016 ). It has been shown by others that slr0664 is activated by proteolysis of the antitoxin, ssr1114, which may be dependent on growth conditions ( Ning et al., 2011 ), and it is unclear whether this layer of regulation affected the outcome of this experiment. Zhou et al. (2019) took a similar approach in designing biocontainment modules for S. elongatus UTEX 2973 and PCC 7942. They also tested NucA from Anabaena , the holin and endolysin from P22 phage, and a native RNase, SepT2, that is part of a toxin-antitoxin system. The holin and endolysin had been previously used to facilitate nickel-inducible lysis of Synechocystis sp. PCC 6803 using the PnrsB promoter ( Liu and Curtiss, 2009 ). Zhou et al. (2019) first screened several metal ion-inducible promoters and selected a low-iron-inducible promoter, PisiAB from S. 7942, which was expected to be induced by the low-iron content of most water in the natural environment. For SepT2 and NucA, the respective antitoxins were co-expressed under control of the native P psbA2 promoter. The phage lysis proteins reduced growth slightly even when uninduced (iron-replete condition) and minimally reduced growth further when induced by low-iron concentrations. The NucA gene caused growth arrest within 24 h of a shift to low-iron media. The toxin-antitoxin systems ssr1114/slr0664, slr6101/slr6100, and SepA1/SepT1 were found to be bacteriostatic rather than lethal. SepT2 was found to be effective in both S. 2973 and S. 7942 with colony forming units below the detection limit of 10 −9 after 3 days of low-iron media exposure ( Zhou et al., 2019 ). Synthetic Auxotrophy Algae have been engineered to be dependent on chemical species that are in lower abundance in the environment than what is needed to sustain growth ( Figure 1B ). Synthetic auxotrophs that are dependent on high carbon dioxide concentrations, phosphite, and melamine or urea for growth have been examined. Unlike the lethal gene strategies, synthetic auxotrophy may be less likely to be overcome by mutations; while a single point mutation may inactivate a lethal gene, horizontal gene transfer from other organisms may be necessary for synthetic auxotrophs to regain the ability to utilize more common nutrients. Phosphate is the form of phosphorous typically used to grow cyanobacteria. Some organisms can utilize phosphite, and the genes responsible for transport and oxidation of phosphite to phosphate are known. These genes can be introduced to cyanobacteria, and they can be grown in media lacking phosphate, which has been shown to be effective in reducing biological contamination ( González-Morales et al., 2020 ) since not all organisms can utilize phosphite. Selão et al. (2019) completed similar work in creating strains of cyanobacteria that were auxotrophic of phosphite, and melamine or urea for nitrogen sources. To make Synechococcus sp. PCC7942 a synthetic auxotroph of phosphite, the transporter genes HtxBCDE and the phosphite oxidation gene, ptxD , from Pseudomonas stutzeri WM88 were introduced, and the phosphate transporters knocked out. The escape frequency was found to be undetectable over 28 days with a detection limit of 3.6 × 10 −11 per colony forming unit. Although natural water may contain low levels of phosphite, the engineered strain was not able to grow in BG11 media prepared with sterilized fresh water from a natural source (with or without phosphate). The potential for horizontal gene transfer was assessed by growing the engineered strain with the wild-type Synechocystis sp. PCC6803. The co-culture was plated on selective media for the kanamycin resistance cassette, which was used to disrupt the phosphate transporter. All 28 colonies screened were found to be S. 7942 rather than the WT S. 6803 ( Motomura et al., 2018 ). Similar efforts to generate synthetic auxotrophies of nitrogen sources could build on the work of Selão et al. (2019) by knocking out ammonium transporters. Although there are not many examples yet of biocontainment developed for eukaryotic algae, we expect that many of the strategies effective in cyanobacteria may also be effective when used for biocontainment of eukaryotic algae. Growth on phosphite has also been demonstrated in the fast-growing eukaryotic algae, Picochlorum ( Dahlin and Guarnieri, 2022 ). In Chlamydomonas reinhardtii , the chloroplast genome does not use the UGA codon (the opal stop codon); leading Young and Purton (2016) to suggest that it may be utilized for biocontainment. Specifically, they suggested that UGA can be inserted into an essential gene, and the chloroplast transformed to express the corresponding tRNA synthetase that utilizes a nonnatural amino acid. Such a strain would be dependent on external supply of the nonnatural amino acid for growth ( Young and Purton, 2016 ). In a follow-up study, this group expressed the ptxD gene for phosphite utilization, incorporating the UGA codon substitution, and co-expressing a tRNA with its anticodon modified to decode this codon. This design was not intended to render the modified strain dependent on a nonnatural amino acid, but the authors considered this a strategy for reducing the possibility of the ptxD gene being transferred to other organisms, and, thus, another form of biocontainment ( Changko et al., 2020 ). A pair of publications report on creating strains of cyanobacteria which are dependent on high carbon dioxide concentrations for growth by knocking out the genes needed for the CCM. Lee et al. (2021) knocked out the carbon concentrating mechanism in S. elongatus PCC7942. They observed no growth of this strain when grown in photobioreactors with less than 5% CO 2 in the sparge gas. However, even in elevated CO 2 , the strain grew more slowly than wild type and a loss in productivity of the target molecule, farnesene, was also observed. Recovery from these losses was achieved by complementing the strain with a bicarbonate transporter and carbonic anhydrase genes. This strain could still grow, though at a diminished rate, when sparged with air ( Lee et al., 2021 ). Clark et al. (2018) also knocked out the carbon concentrating mechanism genes of Synechococcus sp. PCC7002 to create a strain dependent on high CO 2 for growth. Measuring colony forming units to demonstrate compliance with the NIH escape frequency guideline, they showed that the guideline could be met in ambient air growth. Using co-cultures of the CCM knockout with the wild type, it was shown that the guideline threshold may be exceeded via horizontal gene transfer. Knockout of a gene essential for horizontal gene transfer was effective in reducing the escape frequency down below the guideline threshold. Henley et al. (2013) suggested the potential for horizontal gene transfer from the GE algae to wild organisms, which are carried into the pond—this paper suggests one effective approach to address this risk. While this knockout reduced the frequency of acquisition of the CCM genes in co-culture, it did not appear to change the frequency of gene transfer from the engineered cells to their co-culture partners ( Clark et al., 2018 ). Future Directions Based on studies summarized above, researchers have been quite successful in demonstrating that both toxic genes and synthetic auxotrophy can be effective in reducing escape frequencies to below the NIH guideline. In some cases, growth rates were only reduced, and escape frequencies exceeded the NIH guideline. Researchers have not always examined why some strategies failed. Inspection of the failure mechanism(s) would benefit development of future strategies. There may be many possible explanations. Was the expression level of the toxin protein insufficient to overcome the expression level of the co-expressed antitoxin? Did some cells in the culture mutate to reduce the activity of the toxin? Mechanistic understanding of biocontainment efficacy, and the impact upon strain fitness, will ultimately enable predictive design to concurrently maximize biocontainment and bioproductivity. Given that most tests have relied on counting colony forming units, it may be beneficial to examine whether growth on agar plates in the lab is more, or less, permissive to growth than conditions outside the lab. Fewer stressors may be present in the controlled environment of the lab, but it may be possible for escaped cells to find some ecological niche in the natural environment where the molecule needed to repress a toxin gene can be found. Testing of biocontainment strategies in more realistic environmental conditions would be beneficial to our understanding of the efficacy of genetically encoded biocontainment systems. For example, engineered strains could be grown in media that models the natural environment that the cells are likely to encounter if they physically escape. Growth media could also be developed to demonstrate the escape frequency in a worst-case scenario. For a synthetic auxotrophy strategy, this media would include possible alternative nutrients, or the maximum known environmental concentration of the nutrient for which the algae has been made dependent. A further limitation of present tests of these systems is that they tend to only measure growth or no growth in one condition. It may be possible for cells expressing toxin proteins to persist for some period of time, with the potential for revival. If there is any chance that the kill switch signal can be reversed in the natural environment, the duration that the cells can persist with the switch “on” should be determined. In some cases, it may be possible for escaped cells to experience fluctuations in the environmental signal needed to kill the cell. Another question that may be important to probe is how long must the genetic kill switch be “on” for it to result in complete killing of all cells? Can they recover if the switch is not “on” for long enough? Is it possible for such biocontainment strategies to give rise to persistence in the environment? It is generally assumed that the lab conditions used to test biocontainment strategies are stringent in that the cells are grown without competitors or predators and in media that is richer in nutrients than most natural waters, with light intensity that is not too high or too low and does not include UV radiation. However, the complexity of the natural environment may provide some opportunities for escape that is not represented in such tests. For example, contact with a multitude of other organisms could give the cells the opportunity to obtain the signal molecule which represses the toxic gene or which the synthetic auxotrophy has made the cell dependent on via cross-feeding. Some of the signal molecules may be present in sufficient concentrations in some environments to prevent efficient killing of the cells. Impacts of biocontainment strategies to productivity should also be assessed because they are unlikely to be implemented if they reduce productivity. Further genetic modifications, beyond synthetic auxotrophy and toxic gene strategies, which improve fitness in the cultivated setting but decrease fitness in the natural environment should also be identified to further reduce escape frequencies. The Standards in Synthetic Biology website 3 is currently collecting biocontainment strategies that have been tested and may be applied to GE algae."
} | 11,439 |
38451250 | PMC10999742 | pmc | 8,589 | {
"abstract": "Abstract Cycads are known to host symbiotic cyanobacteria, including Nostocales species, as well as other sympatric bacterial taxa within their specialized coralloid roots. Yet, it is unknown if these bacteria share a phylogenetic origin and/or common genomic functions that allow them to engage in facultative symbiosis with cycad roots. To address this, we obtained metagenomic sequences from 39 coralloid roots sampled from diverse cycad species and origins in Australia and Mexico. Culture-independent shotgun metagenomic sequencing was used to validate sub-community co-cultures as an efficient approach for functional and taxonomic analysis. Our metanalysis shows a host-independent microbiome core consisting of seven bacterial orders with high species diversity within the identified taxa. Moreover, we recovered 43 cyanobacterial metagenome-assembled genomes, and in addition to Nostoc spp., symbiotic cyanobacteria of the genus Aulosira were identified for the first time. Using this robust dataset, we used phylometagenomic analysis to reveal three monophyletic cyanobiont clades, two host-generalist and one cycad-specific that includes Aulosira spp. Although the symbiotic clades have independently arisen, they are enriched in certain functional genes, such as those related to secondary metabolism. Furthermore, the taxonomic composition of associated sympatric bacterial taxa remained constant. Our research quadruples the number of cycad cyanobiont genomes and provides a robust framework to decipher cyanobacterial symbioses, with the potential of improving our understanding of symbiotic communities. This study lays a solid foundation to harness cyanobionts for agriculture and bioprospection, and assist in conservation of critically endangered cycads.",
"introduction": "Introduction Cycads are a group of gymnosperms whose origin has been traced to the Carboniferous [ 1 ] and are currently endangered due to poaching and habitat loss [ 2 ]. These plants can harbour symbiotic Nostocales, known as cyanobionts, inside specialized coralloid roots, a unique symbiotic organ among gymnosperms [ 3 4 ]. These cyanobionts are facultative, recruited from the soil for a transient symbiosis. Recently, it has become apparent that coralloid roots also contain other sympatric bacteria, such as Hypomicrobiales and Caulobacterales [ 5 8 ]. Biological nitrogen fixation (BNF) is believed to be the main function of the coralloid root microbiome, in exchange for carbon sources from the host [ 9 10 ]. Indeed, in nitrogen-poor environments, cycad leaves carry the same nitrogen fractionation signal as their diazotrophic cyanobionts, confirming the plant’s reliance on symbiotic BNF [ 11 ]. Other less studied functions for cycad cyanobionts and associated microbes might also be relevant for the symbiosis, such as diverse biological and ecological functions supported by natural products or specialized metabolites [ 12 14 ]. All known cycad species can develop coralloid roots, but not all Nostocales species are recruited for symbiosis [ 15 ]. Most characterized cycad cyanobionts belong to the family Nostocaceae , with Calothrix species being isolated on rare occasions [ 16 18 ]. Previous morphological [ 4 ] and single-gene studies targetting the 16S rRNA gene or the tRNA-Leu intron [ 16 19 ] show that cycad cyanobionts are phylogenetically related to the facultative cyanobionts from Gunnera [ 20 ], bryophytes [ 12 21 ] and lichens [ 22 23 ]. Whole-genome analyses partially support this pattern, although these studies are limited as they vary in their sampling and the number and composition of presumed symbiotic clades [ 5,21, 24 27 ]. It has been hypothesized that cyanobionts isolated from different hosts with close phylogenetic relationships might have shared symbiosis genes (i.e. genes unique to and conserved in cyanobionts used to form and/or maintain the symbiosis) [ 28 ]. Even though some genes have been experimentally shown to be needed for the symbiosis, such as the hormogonia regulation locus hrmUA [ 29 ], or the genes ntcA, hetR and hetF , which are required for heterocyst development [ 30 ], these are commonly found throughout Nostocales . Thus, the specific genes related to symbiosis remain unknown. Since many cyanobionts show the ability to form symbiotic relationships with different hosts, showing a ‘symbiotic plastic’ behaviour [ 15,25, 31 33 ], the possibility of a symbiotic genomic signature within the context of symbiotic plasticity is also an interesting hypothesis. However, the incomplete phylogenomic framework to date limits comparative analyses aimed to identify common genomic functions related to the mechanisms that underlie cycad and other hosts’ cyanobiont mutualisms. Not surprisingly, the search for symbiotic genes have yielded incongruent results [ 25 27 ]. Warshan and collaborators found a group of 74 genes that were present in the sequenced cyanobiont, but not in the free-living strains. However, a later study on a larger genomic dataset found [ 27 ] that free-living strains also contain these genes. Free-living strains are those found outside of a symbiotic host, although this term does not necessarily imply that the strain is incapable of forming symbiotic relationships. Therefore, it is currently unknown if the phylogenetic relationship among cyanobionts is congruent with specific symbiosis genes or broader symbiotic genomic signatures, such as those found in the diazotrophic symbiosis between Rhizobiales and legumes (e.g. nod genes) [ 34 ]. Just as the putative shared symbiotic genes remain to be identified, it is also unknown if the non-cyanobacterial sympatric bacterial communities of the coralloid root have a role in the symbiotic behaviour of cyanobionts. These communities have been independently identified in lichens, bryophytes and Azolla [ 35 38 ], in addition to coralloid roots [ 5 8 ]. However, even when these microbiomes seem similar, no formal meta-analysis of these datasets has been done, which would be an important first step to test their overall roles during symbiosis, including symbiotic plasticity. The latter is an interesting possibility, as cyanobacteria-associated communities have been shown to interact with the cyanobiont in coralloid roots [ 5 ] and to directly contribute to the Azolla symbiosis by complementary denitrification [ 37 ]. The present study aims to describe cycad-associated microbiomes and the phylogenetic placement of the cyanobiont, in order to identify symbiotic genes or genomic signatures in cyanobacteria associated with cycads and beyond. We began with a taxonomic assessment of the symbiotic communities of coralloid roots and its meta-analysis, coupled with a phylometagenomic reconstruction of the newly generated metagenome assembled genomes (MAGs) of sympatric cyanobionts. These MAGs were added to existing high-quality genomes from cyanobionts and free-living strains (i.e. those obtained from Nostocales isolated outside a symbiotic host) in order to find well-supported phylogenetically related groups of cyanobionts. The genomes in the identified symbiotic clades, in turn, were used for pangenomic comparisons to identify conserved genes specific to the symbiotic genomes. The generated data, which quadruple the number of publicly available cycad cyanobiont genomes, clarifies the phylogenomic distribution of cyanobionts from cycads and other systems and confirms the existence of at least three symbiotic lineages with a relevant functional and phylogenetic signal, or genomic signature, that warrants targeted experimental characterization.",
"discussion": "Discussion Despite the steady increase in the number of cyanobiont genomes in public databases, whether from cycads or other hosts, their phylogenomic analysis has been limited to the analysis of a few selective genomes in each study, resulting in inconsistencies regarding the number of symbiotic clades and the species therein [ 5 8 24 25 , 26 ]. This has hampered the elucidation of both general, and cycad-specific, cyanobacterial symbiotic signals, especially in elucidating cyanobacterial symbiotic adaptations in both aquatic and terrestrial environments [ 81 ]. Furthermore, it is also unknown how big of a role the associated bacterial community may play in the symbiosis process, and whether their composition is dictated by geography [ 8 21 ], cyanobiont selection [ 5 82 ], host selection [ 37 ] or a combination of all these factors. Here, we present a comprehensive phylogenomic analysis of facultative cyanobionts, coupled with the taxonomic meta-analysis of their associated bacterial community, directly from cycad coralloid roots and in culture. We thereby provide a general overview of the diversity found in a variety of environments, clarifying the apparent cladistic distribution of symbiotic Nostocales species. Taxonomic analysis of symbiotic communities in coralloid roots from American [ 5 6 8 ] and Asian [ 7 83 ] cycads has revealed a diverse bacterial community mostly dominated by Cyanobacteria and Pseudomonadota , which have been reported in other symbiotic systems [ 28 ]. Our metagenomes were also dominated by Cyanobacteria and Pseudomonadota, with Bacteroidota or Actinomycetota being the third mot abundant phyla in most of the samples, the former being consistently found in Australian cycads and more sporadically in American cycads ( Fig. 2a ). The ratio between these two phyla, however, was dependent on whether the communities were sequenced directly after biomass extraction from coralloid roots or after co-culturing (Fig. S2). While the relative abundance of Nostocales in culture-independent metagenomes was always >80 %, their abundance in co-culture varied widely ( Fig. 2a ). Given that co-culture conditions are not reflective of the coralloid root inner environment, these changes are in line with previous studies that reported a similar shift in cyanobacterial dominance between the coralloid roots’ endosphere and their surrounding soil [ 6 ]. Interestingly, the taxonomic identity of dominant taxa was also consistent between samples from both extraction methodologies, despite the different nutritional and environmental conditions that communities in co-culture are exposed to compared to the coralloid root, indicating that once formed, symbiotic communities remain stable ( Fig. 1c ). Composition was also similar regardless of sampling location, with Australian cycads having a taxonomic composition comparable to their American counterparts. Even more so, this study confirms that seven bacterial orders, i.e. Nostocales, Hypomicrobiales, Caulobacterales, Sphingomonadales, Burkholderiales, Xanthomonadales and Sphingobacterales , can be found in cycads and other symbiotic systems [ 21,35 38 ] ( Fig. 2b ). This degree of universal conservation implies selection [ 84 ] that might be based in supporting functional roles [ 37 82 ]. Cyanobiont-specific analysis revealed three paraphyletic clades that contained 95 % of all symbiotic species ( Fig. 3 ). All clades included free-living species as well, but, to our surprise, their genomic composition, based on ANI similarity (Fig. S3) and pangenomic analysis ( Figs3c 5a ), was indistinguishable from their symbiotic counterparts, although it should be noted that some of the so-called free living species might also be capable of forming a symbiotic relationship and they just happened to be isolated from a non-symbiotic niche. At first glance, this observation may be in line with the idea that cyanobiont genomes lose symbiosis genes after they are removed from their symbiotic systems and maintained in laboratory conditions [ 27 ], similar to what has been observed in Rhizobium [ 85 ]. However, the MAGs obtained from direct metagenomes were not significantly different from those obtained from co-cultures from the same clade (Fig. S3). Moreover, pangenomic analysis revealed that SYMB-2 and SYMB-3 genomes were different to all other genomes, including those from SYMB-1 ( Fig. 4c ). This indicates that at least two divergent lineages of symbiotic Nostocales have evolved. Even though genomes from SYMB-2 and SYMB-3 are phylogenomically related, the fact that only the latter contains lichen cyanobionts (with the characteristic features of this tripartite association) ( Fig. 3b ) suggests that they might have unique symbiosis mechanisms. Previous efforts aimed at finding shared genomic features in cyanobionts have used genomes that, considering the phylogeny presented here, are not phylogenetically related [ 5 25 ], which may explain why previously reported symbiosis genes have been found outside cyanobionts [ 27 ]. Even though cyanobacteria may use different mechanisms to adopt a symbiotic lifestyle, they still share certain general functional features. These include an increment of the coding genome fraction dedicated to COG functions related to lipid and carbohydrate metabolism ( Fig. 5b ), which suggests that these are necessary for the symbiotic lifestyle, as functional enrichment has been linked to adaptation [ 81 86 ]. This is consistent with previous findings [ 25 78 ] that have found that these functions are necessary for symbiosis. Genomes from SYMB-2 and SYMB-3 were also found to be enriched in genes related to amino acids and secondary metabolism ( Fig. 5b ) which have also been reported to be characteristic of cyanobionts [ 5 25 ]. More than 40 % of the genes assigned to the latter function, namely secondary metabolism, were only found in symbiotic clades, which further emphasizes the uniqueness of natural products produced by symbiotic cyanobacteria, something that has been recently highlighted ( Fig. 5a ) [ 14 ]. Other COG categories with a high proportion of unique genes were those dedicated to intracellular trafficking, motility and signal transduction mechanisms ( Fig. 5a ). The importance of these mechanisms in cyanobiont symbiosis has been previously reported [ 25 27 28 ]. Interestingly, genomes from clade SYMB-1 contained a lower fraction of unique genes from these categories than those from SYMB-2 and 3 ( Fig. 5a ), in line with what was observed in the pangenomic analysis, namely that based on composition, genomes from SYMB-1 are closer to FL-II than SYMB-2 and 3 are to their sister clade FL-XI ( Fig. 3c ). The present study, to the best of our knowledge at the time of writing, has quadrupled the number of available cycad cyanobiont’ genomes, and nearly doubled the overall number of cyanobiont genomes available. Despite this, the cycad-specific clade SYMB-1 is underrepresented ( Fig. 3c ). This study suggests that cyanobionts are not dependent on a specific set of conserved symbiosis genes. Nonetheless, general functional traits are shared by all SYMB genomes. Furthermore, they also share a similar bacterial community. While cyanobionts were clustered by clade rather than geography, taxonomic identity of dominant associated bacteria remained constant. The function of these communities, and the specific mechanisms that different cyanobiont and their hosts use to establish the transition to a symbiotic state, are yet to be elucidated. This study provides researchers with a solid footing to further investigate the establishment of symbiosis, both in the association of cyanobionts with the critically endangered cycads, and other eukaryotic hosts."
} | 3,851 |
39804707 | PMC11783582 | pmc | 8,591 | {
"abstract": "Unprecedented penetration of artificial intelligence\n(AI) algorithms\nhas brought about rapid innovations in electronic hardware, including\nnew memory devices. Nonvolatile memory (NVM) devices offer one such\nattractive alternative with ∼2× density and data retention\nafter powering off. Compute-in-memory (CIM) architectures further\nimprove energy efficiency by fusing the computation operations with\nAI model storage. Electronic characteristics of NVM devices, like\nresistance in the two resistance states, directly affect the circuit\ndesigners’ decisions and result in the varying performance\nof NVM-CIM chips. In this mini review, we assess the bounds on device\nresistances for accuracy and circuit performance to suggest recommendations\nto device engineers for frictionless device–circuit–system\ninteractions. Furthermore, we review challenges in reliably programming\nNVM devices, followed by benchmarking recent NVM-CIM chips. Our literature\nreview and analytical modeling reveal that a high resistance ratio\nand low variability are favored, and the resistance in a low resistance\nstate is bound by accuracy and circuit performance constraints.",
"conclusion": "Conclusion We analyze the constraints on resistance\nof NVM devices from the\npoint of view of accuracy and EDP to identify the bounds. High resistance\nratio ( K = R HRS / R LRS ) is preferred for high accuracy and low\nEDP. However, the absolute value of R LRS has an upper bound for high accuracy readout and a lower bound for\nensuring low EDP. The analysis is independent of the CMOS technology\nnode, and absolute values can be easily calculated by estimating the\nparasitic resistance in the readout path. We also review the write\nchallenges to advocate for the engineering effort for low voltage\nwrite, thermal stability of resistance values, and high endurance\noperation. Our survey of CIM circuits reveals the desirability of\ntechnology scaling for NVM devices for area and energy efficiency.",
"introduction": "Introduction AI workloads require storage of models\nclose to the computation\nunits to avoid data movement to improve both energy efficiency and\nthroughput. 1 , 2 SRAM is conventionally used for on-chip\nstorage because of its CMOS compatibility, reliable operation, and\nscaling to advanced technology nodes. 3 SRAM\nis complemented by DRAM for larger and denser off-chip memory hosted\non a separate die. The difference in fabrication process prevents\nDRAM integration on the CMOS flow, while a larger area of SRAM caused\nby a 6-transistors (6T) bitcell hinders large on-chip storage. Nonvolatile\nmemory (NVM) devices offer advantages of density, like DRAM and CMOS,\ncompatible with on-die integration, like SRAM. 4 − 6 1-Transistor-1-resistor\n(1T1R) bitcells of RRAM and MRAM demonstrate significant high density\ncompared to 6T SRAM bitcells. Their memory operation stems from their\nphysical characteristics like material crystallization in phase change\nmemory (PCM), conductive filament formation in RRAM, and spin alignment\nin MRAM. 7 This gives rise to distinct resistance\nstates called the low resistance state (LRS) and high resistance state\n(HRS). These binary states offer distinguishable currents when sensed\nusing a reading voltage ( V read ) and can\nbe used to store 0/1. NVMs have been shown to be integrated within\nthe CMOS process at various nodes, like 12, 22, and 40 nm for RRAM, 8 − 10 14,\n22, and 18 nm for MRAM, 11 − 13 and 14 and 40 nm for PCM. 14 − 16 Additionally, they retain the stored data even when\nthey are isolated from the power supply, resulting in additional energy\nsavings in standby operations. Typical resistance characteristics\nand operational principles of popular NVMs are shown in Figure 1 a. Figure 1 (a) Nonvolatile memory\n(NVM) devices with their operational principles\nand typical device characteristics. 8 (b)\nEffect of device parameters on application-level performance metrics.\n(c–e) Circuit schematics for conventional von Neumann architecture,\nNVM as on-chip memory, and NVM-based CIM. Nevertheless, energy overheads of fetching and\nstoring the data\nbetween memory and compute still continue if NVMs are used only for\non-chip storage. 10 Figure 1 c shows a conventional Von-Neumann architecture\nwith an AI model stored in the memory with a physically separate compute\nunit carrying out MAC operations. Constant data movement between memory\nand compute causes significant energy and latency overhead. 17 NVM storage improves memory density and may\nreduce transfers to and from DRAM ( Figure 1 d). Compute-in-memory (CIM) takes a more\naggressive approach and merges a part of the MAC operation within\nmemory array to reduce data movement even more for efficiency and\nspeed. Figure 1 e shows\nan example of an analog CIM array with NVM devices like RRAM storing\nthe weights using HRS and LRS as bits 0/1, respectively. The wordline\n(WL) is driven by the input activation through a wordline driver (WLD)\nand the NVM device allows passage of current depending on the resistance\nstate. Bit-wise multiplication happens within the NVM devices, whereas\naccumulation is carried out over the BL/SL. The sum of the currents\nrepresents the multiply and accumulation (MAC) result between activation\nand weight, and it is converted to a digital code for post-MAC processing\nby an analog-to-digital convertor (ADC). Without extra data movement\nand multirow access, the CIM macro usually exhibits high energy efficiency\n(tera-operations/sec/watt) and high compute density (tera-operations/sec/mm 2 ). Numerous material and device candidates have been\nproposed in recent\nyears for NVM-CIM 18 operations with different\nswitching materials and electrodes, 7 , 19 , 20 e.g., resistance in LRS may vary from 700 Ω\nto 930 MΩ for RRAM 20 and 900 Ω\nto 6 MΩ for MRAM. 21 However, the\nresistance ratio ( K = R HRS / R LRS ) remains relatively constant, as\nshown in Figure 1 a.\nDifferent devices provide a large range of resistance values (LRS/HRS),\nwrite characteristics, and endurance performance. 20 MAC operation in CIM is known to be affected by the resistance\nratio ( K = R HRS / R LSR ), read current in LRS ( I LRS = V read / R LRS ), and process-induced variability (σ) 22 , 23 ( Figure 1 b). These\nparameters affect CIM accuracy, energy consumption, and compute latency\nat circuit level, and further, at system level. 24 − 26 Therefore,\nearly identification of device parameter design space given circuit/system\nspecifications helps the material/device researchers to make design\nchoices in the development of these resistive memory devices. In this\nmini review, we survey the recent literature to provide an analytical\nmodel on how device parameters affect circuit designs for CIM readout\nand suggest recommendations to the device and material engineering\ncommunity for seamless device–circuit interactions. We focus\non maintaining accuracy for readout and minimizing the energy-delay\nproduct for the CIM-array to identify bounds on the device parameters.\nOur modeling framework may be useful for early design decisions by\nmaterials, devices, and circuit engineers, while the summary of literature\nshows upcoming challenges and research trends that may push the viability\nof NVM-CIM for commercial applications."
} | 1,803 |
36290352 | PMC9598760 | pmc | 8,592 | {
"abstract": "Simple Summary The humic lake represents a special kind of aquatic ecosystem with high humic substances, low irradiance, and a high potential for greenhouse gas emissions. Despite the special environment and biogeochemical processes in humic lake water, knowledge about the underlying microbial-driven functions remains elusive. Here, we studied the compositions and functional gene structures of microbial communities in a humic lake (HL) and a reference weakly alkaline lake (RAL). We found that the high organic matter content in the HL supported higher gene diversity; and, specifically, the carbon and nitrogen fixations, the degradation of various types of carbon, methane oxidation and methanogenesis, ammonification, denitrification, and assimilatory N reduction might be enhanced more in the HL than in the RAL. By contrast, the humic fractions in the HL might reduce microbial metabolic potential for sulfur oxidation and phosphorus degradation. The potential interactions between different functional microorganisms might be down-regulated provided that there were more easily acquired nutrients in the HL. Overall, our results showed the functional gene “landscape” of microbial communities in the surface water of a humic lake, which helps understand the biogeochemical processes and the remediation of organic matter pollution in lacustrine ecosystems. Abstract Humic lakes (HLs) are special water bodies (high organic matter content, low pH, and low transparency) that are important sources of major greenhouse gases. The knowledge about microbial functional potentials and the interactions among different genes in HL water has been scarcely understood. In this study, we used 16S rRNA gene sequencing and the GeoChip 5.0 to investigate microbial community compositions and functional gene structures in an HL and a reference weakly alkaline lake (RAL). The HL microbial communities showed distinct compositions and functional gene structures than those in the RAL. The functional gene diversity was significantly higher in the HL than in the RAL. Specifically, higher gene relative intensities in carbon and nitrogen fixations, the degradation of various types of carbon, methane oxidation and methanogenesis, ammonification, denitrification, and assimilatory N reduction were observed in the HL samples. By contrast, the metabolic potentials of microorganisms involved in dissimilatory N reduction, phosphorus degradation, and sulfur oxidation were weaker in the HL than in the RAL. Despite higher functional gene diversity, the interaction efficiency among genes (reflected by network geodesic distance and clustering coefficient) might be reduced in the HL. Different functional microbes may develop less interdependent relationships in acquiring nutrients given the high resource availability in the HL. Overall, the enhanced microbial metabolic potentials and less efficient functional interactions might have great consequences on nutrient cycling and greenhouse gas emissions in the HL ecosystem.",
"conclusion": "5. Conclusions Our study showed that the special environmental characteristics in the HL water (high content of organic matter (with the humic substances as the main component), total nitrogen and phosphorus, and relatively low pH) could impact bacterioplankton communities to form unique functional gene compositions and result in disparate ecological processes compared with the reference RAL. The abundant organic matter might shed “priming effects” for the degradation of various carbon types in the HL and increase the metabolic potential of microorganisms to participate in processes such as methanogenesis, nitrogen fixation, and denitrification. These processes may lead to enhanced emissions of greenhouse gases (CO 2 , CH 4 , and N 2 O) in HLs [ 56 ]. However, the relatively low pH and some humic fractions in the HL water may exert inhibitions on the degradation of aromatic organic carbon, sulfur oxidation, and phosphorus degradation in the HL ( Figure 5 ). The high amount of organic matter in the water might also change the interactions among different functional microorganisms, reducing the inter-dependent (or synergistic) relationships within the community to acquire nutrients for their growth. The validity of our results could be further tested using biogeochemical monitoring and other “omic” approaches (such as metatranscriptomics, metaproteomics, and metabolomics).",
"introduction": "1. Introduction The humic lake (HL) is a distinct, dark-colored lake, with poor light transparency, low pH, low oxygen, and high content of humic substances (HSs), usually due to a large input of allochthonous (terrestrially derived) organic matter [ 1 , 2 ]. HSs are complex and heterogeneous mixtures formed by biochemical and chemical reactions during the decay and transformation of biomass, a process known as humification [ 3 ]. They represent one of the most important components of the total carbon on earth and comprise 50–75% of the dissolved organic carbon (DOC) in water [ 4 ]. The high concentration of HLs in the lake had a great role in the carbon and other nutrients’ cycling processes [ 5 ], which were largely driven by planktonic microorganisms [ 1 , 6 ]. Due to the general absence of planktivorous and piscivorous fish, the trophic interactions between microbes and other organisms might be less complex in the HL ecosystem [ 7 ]. The specialized physicochemical and biological conditions have possibly resulted in distinct microbial communities and their functions. The pelagic microorganisms contributed greatly to the primary and secondary production in the HLs [ 8 ], since the enriched organic carbon promoted microbial growth and the dark humic substances hindered the growth and production of aquatic plants [ 1 ]. Evidence has shown that the unique environmental conditions of HLs can result in unique bacterial populations, such as the soil II-III clade of Actinobacteria [ 9 ]. The specialized bacterial populations may conduct differential functions in HLs than in common lakes. There has been emergent evidence about microbial community compositions in HLs [ 9 , 10 ] and the functions of some specific taxa in the degradation of organics in HLs [ 11 , 12 ]. Microbial functional gene compositions provide more direct evidence than taxonomic compositions in inferring the material cycling and functional processes in an ecosystem [ 13 ]. Among technologies used to study microbial functional genes, the GeoChip is a high-throughput microarray-based genomic technology to study various biogeochemical processes and functional activities [ 14 ]. There has been some works using GeoChip to study microbial functional genes in natural lakes and artificial aquatic ecosystems. The carbon, nitrogen, phosphorous, and sulfide cycling functions and the metabolic pathways have been investigated [ 15 , 16 , 17 ]. Gene compositions and functional potentials can be regulated by environmental variables such as pH, DOC, and nitrogen nutrients [ 18 , 19 , 20 ]. For HL lakes, the specific functions such as the degradation of recalcitrant carbon, the utilization of glycolate [ 21 ], and the transformation of organic pollutants [ 22 ] have already been studied. These functions could show elevated activity in HLs and have temporal patterns linked with other organisms in water. However, little is understood about the whole-community functional genes and the potential interactions of symbiotic genes in HLs. Here, we studied a HL (pH 5.16) and a reference common lake (weakly alkaline, pH 8.3, RAL) in Denmark. The GeoChip 5.0 was used to investigate the functional gene differences in the two lakes. The gene association networks were used to infer the potential interactions among different symbiotic functional genes. We hypothesized that (a) the functional gene composition and their functional activities were greatly different between the HL and the common RAL, and (b) the special physicochemical traits in the HL may result in disparate interaction patterns of functional genes than those in the reference common lake. Our results may provide clues for understanding element cycling processes and the amendment of organic matter pollution in lacustrine ecosystems.",
"discussion": "4. Discussion 4.1. Differences in Microbial Functional Gene Compositions and Structures between the HL and RAL In this study, we investigated the functional gene compositions in microbial communities from an HL and a reference natural RAL and found higher diversity and lower homogeneity in the HL compared to the RAL samples ( Figure 2 ). This may be attributed to the high content of DOC in the HL water due to an exceptionally large input of allochthonous organic matter [ 1 , 2 ]. As an important nutrient source that may regulate microbial communities [ 34 ], the higher DOC content might sustain more functional genes to show metabolic activity in HL samples ( Figure 1 A). Compared with the RAL, the HL had more special environmental conditions, e.g., low pH, poor light, low oxygen, but high content of HS ( Table 1 ), which usually represents a harsh environment that filters out the adapted microbial species and functions [ 35 ], meaning lower homogeneity in the HL than that of the RAL ( Figure 2 ). For phylogenetic compositions, the HL and RAL shared the same dominant bacterial phyla, such as Proteobacteria, Actinobacteria, Bacteroidetes, Cyanobacteria, and Verrucomicrobia ( Figure S5A ), which was typical for most lake bacterioplankton communities [ 36 ]. However, at the class level, the two lakes had disparate phylogenetic compositions. Compared with the RAL, the relative abundances of the class Gammaproteobacteria, Bacteroidia, and Bacilli (especially the genus Paenibacillus ) were higher in the HL ( Figure S5B, Table S3 ). The persistent dominance of Gammaproteobacteria and Bacteroidia was also observed in another humic lake [ 9 ]. Interestingly, though the relative abundances of Alphaproteobacteria, Actinobacteria, and Verrucomicrobiae were lower, the relative abundances of some genus within these class, for example, Rhodobacter (Alphaproteobacteria), Microbacteriaceae (Alphaproteobacteria), and Prosthecobacter (Verrucomicrobiaewas) were much higher in the HL ( Figure S5, Table S3 ). Some members of the Rhodobacter and Prosthecobacter had a high capacity for utilizing organic materials [ 37 , 38 ]. 4.2. Differences in Metabolic Potentials between the HL and the RAL Aquatic bacteria play key roles in greenhouse gas emissions in HLs due to their metabolic activities in utilizing the high content of organic carbon in water [ 39 , 40 ]. Genes for degradation of a wide range of carbon, ranging from the labile type (e.g., glucoamylase ) to the recalcitrant type (e.g., ligninase ), showed higher intensity in the HL than in the RAL, indicating that the high amount of carbon in the HL water could induce the “priming effect” for a variety of carbon types. Intriguingly, the carbon fixation gene (i.e., the rubisco gene) also showed a higher intensity in the HL samples, which may be ascribed to the fact that the HL inhabited more Gammaproteobacteria, from which many autotrophic microorganisms use the Calvin cycle for carbon fixation [ 41 ]. Previous studies have reported that HSs could stimulate CH 4 oxidation by acting as an electron shuttle and extracellularly directing electrons to high-valent chemicals [ 42 ]. This stimulation of CH 4 oxidation was verified by our study, in that the metabolic potential of methane oxidation (i.e., pmoA and mmoX ) was significantly higher in the HL than in the RAL. Accordingly, the aerobic MOB from the genus Methylobacter (Gammaproteobacteria) were more abundant in the HL. The elevated methane oxidation may further lead to local oxygen scavenging [ 43 ], resulting in a positive-feedback loop to sustain higher potential methanogenesis ( mcr gene more intensified, Figure 2 ) in anoxic microsites. Our study showed that there was a significant difference in the relative signals of the nifH gene group between the HL and RAL ( Figure 3 C, p < 0.05). The main components of humic (humic acid and fulvic acid) can significantly increase the growth efficiency and nitrogen fixation capacity of N-fixing bacteria [ 44 ]. Denitrification, as a heterotrophic pathway, is an important link between the C and N cycles. Previous studies have revealed that humic acid could promote heterotrophic denitrifying bacteria, such as the Thauera , which could utilize HSs as an electron shuttle to improve denitrification performance, especially for nitrite reduction [ 45 ]. They were also detected in our study, and their nosZ and nirS intensities were significantly enhanced more in the HL than in the RAL (Student t test, both cases, p < 0.05). The relative intensities of most denitrification genes (except the narG and norB ) were higher in the HL than in the RAL, implying that the denitrification processes (though not all steps) were promoted ( Figure 3 , p < 0.01), which may result in the quick removal of N in the water. This may help explain why there was much less difference in NO X − contents, compared with the great difference in TN content, between the two lakes ( Table 1 ). The enhanced nitrogen-fixing and denitrification processes implicated more rapid N cycling in the HL than in the clear RAL. We observed enhanced microbial sulfate reduction potential but lower sulfide oxidation potential in the HL than in the RAL ( Figure 3 D). The high organic carbon availability in HLs was preferential for the heterotrophic sulfate-reducing bacteria [ 46 ], and the high content of humic substances often leads to lower oxygen levels in HL water [ 47 ], which may promote sulfur reduction but inhibit sulfur oxidation. The enhanced sulfate reduction and lower sulfide oxidation suggested more deposition of sulfur (e.g., in the form of sulfide metal) in the HL. We also observed less phosphorus degradation potential in the HL than in the RAL ( Figure 3 D). The humic acid fractions could inhibit phosphatase by binding with some active sites of plant-derived enzymes [ 48 ] and inhibit the activity of plant phytase by forming complexes with the enzyme substrates and having considerable absorption properties [ 49 ]. Our results suggest that similar mechanisms may also be responsible for microbial phosphatase and phytase (i.e., the lower ppx and phytase gene intensities) in the HL water. 4.3. Potential Interactions between Functional Genes The special environmental characteristics of the HL resulted in distinct microbial community structures and also special functional gene network topological traits compared with the RAL. The negative links in the network may represent the competition or difference in metabolic preferences among community members [ 50 , 51 ]. In our study, the HL network had a higher proportion of negative correlations, which implicated that the refractory HSs aggravated the competition in HL microbial communities [ 35 ]. The networks with a shorter path length can transmit environmental fluctuations to the whole network in a shorter time and rapidly change the structure and function of the network [ 17 ]. The higher percentage of positive links in a network also favored quick and broad feedback in the community to environmental perturbations [ 52 ]. The average path distance was longer and the positive edge proportion was smaller in the HL, which indicated that the interactions between functional genes in the HL might be less efficient and respond less quickly to the changes in the environment than those in the RAL. Some studies have shown that the bacterial communities in a humic lake were relatively resilient to extreme weather events [ 53 ]. In addition, there was higher resource availability in the HL water (reflected by the higher contents of organic carbon, total nitrogen, phosphorus, and ferrum) ( Table 1 ), which might reduce the inter-dependent relationships between different functional groups in acquiring nutrients for their growth [ 54 ]. Two of the top five high-degree genes in the HL network were the nosZ gene, highly connecting with the cellobiase , nrfA , and xylA genes, which reflected the enhanced nitrogen-reduction processes, and these processes depended on the degradation of organic carbons in the HL water. The highlighted sox gene in the RAL network reflected that the sulfur oxidation was characterized in the clear water of the RAL, which developed high correlations with the carbon-fixation genes ( rubisco and FTHFS ) and the nosZ gene. Similar gene co-occurrence of the sox and the carbon-fixation and denitrification genes were also found in a movile cave [ 55 ], which suggested that the chemolithoautotrophic sulfur-oxidizing bacteria might play a key role as a primary producer in the studied RAL ecosystem."
} | 4,213 |
26727469 | PMC4699747 | pmc | 8,593 | {
"abstract": "Microalgae have attracted wide attention as one of the most versatile renewable feedstocks for production of biofuel. To develop genetically engineered high lipid yielding algal strains, a thorough understanding of the lipid biosynthetic pathway and the underpinning enzymes is essential. In this work, we have systematically mined the genomes of fifteen diverse algal species belonging to Chlorophyta, Heterokontophyta, Rhodophyta, and Haptophyta, to identify and annotate the putative enzymes of lipid metabolic pathway. Consequently, we have also developed a database, dEMBF (Database of Enzymes of Microalgal Biofuel Feedstock), which catalogues the complete list of identified enzymes along with their computed annotation details including length, hydrophobicity, amino acid composition, subcellular location, gene ontology, KEGG pathway, orthologous group, Pfam domain, intron-exon organization, transmembrane topology, and secondary/tertiary structural data. Furthermore, to facilitate functional and evolutionary study of these enzymes, a collection of built-in applications for BLAST search, motif identification, sequence and phylogenetic analysis have been seamlessly integrated into the database. dEMBF is the first database that brings together all enzymes responsible for lipid synthesis from available algal genomes, and provides an integrative platform for enzyme inquiry and analysis. This database will be extremely useful for algal biofuel research. It can be accessed at http://bbprof.immt.res.in/embf .",
"introduction": "Introduction With the irreversible depletion of petroleum resources, renewable biofuels are sustainable alternative to meet the global energy needs. Microalgae as a rich source of lipid, especially triacylglycerols (TAGs) have emerged as a potential biofuel feedstock due to several distinct advantages over other starch-based or lignocellulosic plant species, such as higher photosynthetic efficiency and higher biomass production rate. Besides, microalgae can be grown on non-arable land using wastewater, thus not competing with agri-resources and mitigating CO 2 emissions efficiently [ 1 , 2 ]. However, to make biofuel production from microalgae a cost-competitive process, the oil content in oleaginous algae needs to be significantly improved through genetic engineering techniques [ 3 , 4 ]. It has been proposed that lipid biosynthesis can be increased by over expressing the rate-limiting enzymes of fatty acid biosynthesis pathway, of which acetyl-CoA carboxylase (ACCase) that catalyzes the first committed step of fatty acid synthesis viz., conversion of acetyl CoA to malonyl CoA plays a pivotal role [ 5 , 6 ]. In addition, overexpression of the acyltransferases enzymes catalyzing the main regulatory steps involved in TAG biosynthesis, widely known as the Kennedy pathway, have also been determined as a potential approach to boost oil accumulation. For instance, overexpression of a type 2 diacylglycerol acyltransferases (DGAT) enzyme in the diatom Phaeodactylum tricornutum resulted in 35% increase in TAG content [ 7 ]. In another study, co-overexpression of multiple genes of the Kennedy pathway including glycerol-3 phosphate acyltransferase (GPAT), lysophosphatidyl acyltransferase (LPAT), phosphatidic acid phosphatase (PAP), diacylglycerol acyltransferase, glycerol-3 phosphate dehydrogenase (GPDH) and phospholipid:diacylglycerol acyltransferase (PDAT) in Chlorella minutissima resulted in a two-fold increase of TAG content [ 8 ]. Introduction of diacylglycerol acyltransferase 2 gene from Brassica napus to Chlamydomonas reinhardtii has also resulted in enhanced lipid production [ 9 ]. Together these studies indicate that understanding the regulation of microalgal lipid metabolism is absolutely essential for developing engineered microalgae with enhanced lipid production capabilities. [ 10 ]. While algal sequence data from genome assembly projects is rapidly increasing, the generated annotation for predicted sequences are usually limited and includes only user-defined function prediction with no detailed pathway, structure or genome-context information [ 11 ]. This limits our understanding of the overall lipid biosynthetic pathway in microalgae [ 12 ]. On contrary, the genes and enzymes involved in plant lipid biosynthetic pathway have been characterized extensively [ 13 ], and a number of biomass-related enzyme databases are also available to promote the development of transgenic biofuel crops [ 14 – 17 ]. Considering the importance of microalgae biofuel, paucity of information on algal lipid biosynthesis and unavailability of dedicated databases on enzymes underpinning the process, the present study was performed to identify a total of 289 enzymes responsible for lipid accumulation in fifteen sequenced microalgal species by using available homologous sequences from the model plant species, Arabidopsis thaliana . Functional annotation of the putative enzymes has also been improved by employing several bioinformatic tools to study metabolic pathways, ontology, subcellular location, secondary and tertiary structure, biophysical properties, cellular processes and protein family information. Furthermore, the emanated data are made publicly accessible through an open-access web-based database, dEMBF (database of Enzymes of Microalgal Biofuel Feedstock, http://bbprof.immt.res.in/embf ). dEMBF is the first integrative platform that provides a complete list of enzymes putatively involved in lipid biosynthesis in microalgae. This database will certainly provide a roadmap for experimental as well as computational studies leading to identification of orthologous lipid synthesis enzymes in newly sequenced algal species and facilitate further R&D research aimed at attaining a sustainable and cost-effective biofuel production from microalgae.",
"discussion": "Discussion After a thorough examination of the fifteen algal genomes, a total of 289 enzymes with putative roles in lipid synthesis were identified ( S1 and S2 Tables). Sequence-structure information of these enzymes, together with the 27 well characterized homologous enzymes from Arabidopsis used as reference dataset in this study, are provided in the database. While previous studies have identified some key enzymes associated with lipid metabolic pathway in few algal species [ 41 – 49 ], the genomes of C . variabilis , M . pusilla , Micromonas sp ., B . prasinos , T . pseudonana , P . tricornutum , E . siliculosus , A . anophagefferens and E . huxleyi have been mined for the first time in this study to collate the entire repertoire of enzymes responsible for lipid accumulation in microalgae. In addition to genome mining, we have assigned pathways, gene ontology terms and cluster of orthologous ( S3 Table ), subcellular location, secondary and tertiary structure, biophysical properties, cellular processes and protein family terms to each of the enzymes. Consequently, we have improved the existing functional annotation of all 289 enzymes including 86 previously uncharacterized sequences for which a putative function in lipid biosynthesis has been determined ( Fig 9 ). We observed that the analyzed algal genomes exhibited an overall comparable enzymatic makeup and each encode the major enzymes for lipid synthesis similar to Arabidopsis ( S2 Table and S1 Fig ). However, we found that four algal species viz., C . variabilis . C . reinhardtii , V . carteri and C . merolae contain both homomeric and heteromeric ACCase enzyme, while the rest contain only the homomeric form of ACCase. This is in agreement to a previous published report, stating that the green (Chlorophyta) and red (Rhodophyta) algae with the exception of the green algal class Prasinophyceae ( O . lucimarinus , O . tauri , M . pusilla , Micromonas sp . and B . prasinos ) contain both homomeric and heteromeric ACCase while other algal species belonging to Heterokontophyta and Haptophyta lack heteromeric ACCase [ 42 ]. Furthermore, we found that the acyltransferases (60% of the total number of enzymes) is the most abundant enzyme class ( Fig 10 ). The increased number of enzymes belonging to this class is probably significant considering that the three acyltransferases including GPAT, LPAT and DGAT catalyzes sequentially to acylate glycerol backbone, to ultimately produce TAG. These enzymes play a vital role in determining the acyl composition of glycerolipids and the final content of TAG [ 12 ]. In particular, relatively more number of DGAT (80 in number) followed by ACCase (39 in number) enzyme was observed in all algal genomes. The fact that ACCase catalyzes the initial rate limiting step of fatty acid biosynthesis by converting acetyl CoA to malonyl CoA while the DGAT enzyme drives the final step of TAG synthesis acylating diacylglycerol to TAG [ 4 , 10 ], clearly reflects the high lipid accumulation capability of microalgae for biofuel production. 10.1371/journal.pone.0146158.g009 Fig 9 Chart showing the annotations of 289 enzymes present in dEMBF that are putatively involved in lipid biosynthesis in fifteen algal species. The dark grey sector indicates the total number of enzymes with functional annotations available from JGI database, but was further evaluated in this study for confirmation or assignment of any missing functional features. The light grey sector indicates the total number of previously uncharacterized enzymes from JGI, for which putative functions were predicted based on UniProt annotations. The medium grey sector indicates the number of enzymes for which no annotations were available in JGI as well as in UniProt. A putative function for each of these enzymes was predicted using various bioinformatics tools. The values inside the chart refer to the total number of enzymes while the values outside the chart indicate the distribution of the enzymes per organism. The names of the fifteen microalgae are indicated in the right panel with different color codes for each species. 10.1371/journal.pone.0146158.g010 Fig 10 Distribution of sequences according to enzyme classification in the dEMBF database. The main chart shows the overall percentage of enzymes belonging to acyltransferase, oxidoreductase, ligase, lyase and hydrolase while the insert charts shows the total number of each enzymes (values indicated) belonging to a particular enzymes class. To our knowledge, dEMBF is the first comprehensive database on enzymes responsible for lipid accumulation in fifteen diverse algal species whose genome sequences are available. This work could be useful towards better understanding of fatty acid and TAG biosynthetic pathways in microalgae, besides facilitating the development of genetically engineered algal strains for a sustainable and economical viable biofuel production."
} | 2,703 |
25100933 | PMC4106401 | pmc | 8,594 | {
"abstract": "Efficient Analog-Digital Converters (ADC) are one of the mainstays of mixed-signal integrated circuit design. Besides the conventional ADCs used in mainstream ICs, there have been various attempts in the past to utilize neuromorphic networks to accomplish an efficient crossing between analog and digital domains, i.e., to build neurally inspired ADCs. Generally, these have suffered from the same problems as conventional ADCs, that is they require high-precision, handcrafted analog circuits and are thus not technology portable. In this paper, we present an ADC based on the Neural Engineering Framework (NEF). It carries out a large fraction of the overall ADC process in the digital domain, i.e., it is easily portable across technologies. The analog-digital conversion takes full advantage of the high degree of parallelism inherent in neuromorphic networks, making for a very scalable ADC. In addition, it has a number of features not commonly found in conventional ADCs, such as a runtime reconfigurability of the ADC sampling rate, resolution and transfer characteristic.",
"introduction": "1. Introduction Circuits for analog-digital-conversion (ADC) are at the heart of every integrated circuit (IC) that deals with sensory or other analog input signals. Their performance and characteristics have a large repercussion on the signal processing carried out in the later (usual digital) stages of the IC, as distortions of the signal introduced in the ADC cannot usually be recovered. In general, ADCs because of their analog nature are handcrafted to achieve optimum characteristics for a given application. They usually require a wide range of custom analog circuit components, such as amplifiers, voltage/charge/current converters, integrators, addition/subtraction circuits, threshold switches, etc (van de Plassche, 2003 ). However, this handcrafted, analog nature of ADCs is not in keeping with todays mostly digital Systems-on-Chip (SoC). SoCs due to their digital nature can be rapidly prototyped and transferred across technology nodes, something not possible with a handcrafted analog circuit. In addition, state-of-the-art deep-submicron technology nodes have become increasingly worse in their analog performance. ADCs have started to partially follow this trend, offering architectures such as Delta-Sigma-Modulators (DSM) that only need low-performance analog components and move a large part of their functionality into the digital domain (Marijan and Ignjatovic, 2010 ; Mayr et al., 2010b ). However, to really comply with the demands placed on modern ADCs, inspiration may be taken from a completely different domain, that of neural information processing and neuromorphic design. Neural networks rely for their overall function on multiple replication of a single, simple base element, the neuron. Thus, scaling and technology transfer of a neuromorphic ADC would be simplified. A neural network represents data across a population, thus inherently smoothing out variations and noise and making the signal representation more robust. Neurons take analog data as input, transferring it immediately into a pseudo-digital, timing based pulse representation. Thus, all subsequent processing would be digital directly after this first stage. Neural networks can replicate non-linear transfer functions of one or several input variables (Lovelace et al., 2010 ). Thus, sensor fusion and analog preprocessing could be achieved, which in conventional ADCs requires separate analog blocks (Chen et al., 2013 ). This paper proposes using the Neural Engineering Framework (NEF) (Eliasmith and Anderson, 2004 ) as a method to build an ADC that incorporates most of the above advantages of neural networks. In the NEF, a signal is encoded across a neuron population by a set of encoder weights and the transfer functions of the neurons. A set of decoder weights can be computed that extracts the signal itself or a transformation of it from the postsynaptic current (PSC) traces of the neurons. By building the encoder step and the neurons in analog circuitry while having the decoding and signal reconstruction done in the digital domain, a straightforward conversion from analog to digital can be established. Specifically, we show in this paper the usage of NEF as a linear, single input ADC comparable to conventional ADCs. The theoretical and simulative analysis is supported by an example design in a 180 nm CMOS technology, proving feasibility of the approach. The remainder of the paper is structured as follows: section 2.1 introduces the NEF framework. In section 2.2, its general application to analog-digital-conversion is given. Section 2.4 details the analog and digital circuit design. Results are given in section 3.1 for an ADC based on idealized neurons in a neural network simulator. Results for the actual hardware implementation of neurons, encoder and decoder network are given in section 3.2. Section 4 discusses the significance of the results.",
"discussion": "4. Discussion 4.1. NEF in a general neuromorphic VLSI context NEF has recently attracted significant interest from the neuromorphic community, with e.g., an implementation on Neurogrid (Choudhary et al., 2012 ). It exhibits several features of interest to engineers. Using it, one can engineer a neural system with a target reliable behavior based on unreliable elements. The target behavior can range from building blocks familiar to an engineer, such as control systems or filters (Dethier et al., 2013 ), up to abstract cognitive functions (Eliasmith, 2007 ). This paper has highlighted another useful aspect: NEF makes it easy to cross timing domains from asynchronous to synchronous and from analog to digital value representation. Traditionally, this has been one of the major bottlenecks when interfacing neuromorphic systems to more conventional processing units. The other main challenge of neuromorphic engineering, i.e., achieving biological real time operation (Giulioni et al., 2012 ), could also be alleviated by NEF. By not representing the system variables directly as spikes, but rather abstracting the single pulses to a time-varying system state vector or scalar variable (Equation 5), the underlying neurons can be dictated by CMOS constraints (i.e., can be operated faster), while the state vector changes could be slower, i.e., able to interact with the outside world in biological real time. By adding this layer of abstraction on top of the neuromorphic network, the CMOS speed advantage can be utilized for e.g., a higher fidelty computation and/or representation of the system state variables, as shown in this paper. This layer of abstraction can also be used to transmit computational variables between neuromorphic units in a more CMOS-friendly fashion. Traditionally, states of neural networks are communicated by the single underlying spikes, requiring large bandwidths in FPGA-based spike routers (Hartmann et al., 2010 ) or even dedicated IC solutions (Scholze et al., 2011 ). By abstracting the single pulses to a time-varying digital state, bandwidth can be reduced significantly. 4.2. Other neuromorphic ADCs There are a number of groups that have built ADCs based on neural networks. Table 10 gives an overview of the salient features of these ADCs. Table 10 Comparison of various neuromorphic ADC concepts . References Description Most similar conventional architecture Parallel/ serial Required analog precision / required design effort Config . Power Sensor fusion possible This work digital decoding of analog input from neuron population signals Flash and feedforward oversampling Parallel Very low / low, repetitive neuron circuit Rate and resolution x10 more power than best reported conventional Yes Tapson and van Schaik, 2012 Parallel noise shaping network with lateral inhibition DSM Parallel Low / low, repetitive neuron circuit rate and resolution No data No data Chande and Poonacha, 1995 Binary threshold neurons in a weighted MSB to LSB decoder network Successive approximation Serial Equal to ADC resolution / low, repetitive neuron circuit resolution No data, but likely comparable to median conventional Yes Yang and Sarpeshkar, 2006 Time-domain pipeline architecture, with neurons handling time domain processing Pipeline Serial Equal to ADC resolution / high, numerous handcrafted components No on par with best reported conventional, subthreshold operation No Some of those use time-invariant threshold neurons in architectures derived from conventional flash or pipeline ADCs (Chande and Poonacha, 1995 ). Neuromorphic principles have also been used to convert conventional architectures into the time domain. For example, Yang and Sarpeshkar ( 2006 ) show a pipeline ADC composed of Integrate-and-Fire (IAF) neurons that transfers the AD conversion into the time domain. While the use of subthreshold operation in Yang and Sarpeshkar ( 2006 ) makes for a very power efficient pipeline design, the entire design is targeted at a single application, without the wide configuration ability of the NEF ADC. For example, a higher resolution can only be achieved in the design of Yang and Sarpeshkar ( 2006 ) by increasing the complexity and power draw of the comparator. Also, a higher sample rate is only achievable through a non-subthreshold-operation of the neurons, loosing the energy advantage. In both Chande and Poonacha ( 1995 ) and Yang and Sarpeshkar ( 2006 ) the performance of the design is ultimately limited by the precision of its handcrafted building blocks. Thus, no significant advantage is gained compared to conventional ADCs. In particular, both the above ADCs do not use the high parallelism of neural networks to increase robustness and/or conversion speed or precision. In contrast, another family of devices uses the noise shaping effect that a group of neurons achieves when recurrently inhibitory connected (Watson et al., 2004 ; Tapson and van Schaik, 2012 ). Here, the signal is represented robustly across a neuron population, i.e., the overall network activity is modulated by the signal (Mayr et al., 2009 ). The distribution across a neuron population even allows representation of signals above the intrinsic frequency of single neurons (Spiridon and Gerstner, 1999 ). One main drawback is that some of these architectures are unstable. There is also no fully established method to extract the digital output signal from such a network (Mayr and Schüffny, 2005 ). 4.3. NEF as an analog-digital converter The NEF ADC shares some characteristics with different conventional ADCs. For instance, time-domain ADCs also integrate the input signal to arrive at analog to time conversion that can then be digitized (Yang and Sarpeshkar, 2006 ). ADCs that oversample the input signal, such as the DSM mentioned in the introduction, also digitize an input signal with high frequency and low initial resolution. Similar to the NEF ADC, they build up resolution by removing high-frequency components with a filter. Also similar to a DSM, for most applications the NEF ADC does not require an analog Nyquist filter due to the low pass filter characteristic of the neurons and the PSC filter. The NEF ADC also shares some characteristics with flash ADCs, as both use a large parallelism of elements to arrive at a coarse fast quantization. Similar to the NEF ADC, some flash ADCs also rely on statistical deviation of elements for their quatization curve Weaver et al., 2011 ). The comparison across Tables 7 – 9 shows that in terms of absolute figures of sample rate and bit resolution achieved, the NEF ADC is competitive. However, it underperforms quite severly with regard to area and power, see the FOM comparison. The major part of the area of the NEF ADC is spent on the digital building blocks, letting it benefit significantly from technology scaling. Conventional ADCs do not shrink well due to their usually significantly larger portion of analog circuitry. Thus, the area comparison would look decidedly different in e.g., a 28 nm technology, where the digital blocks would only occupy approx. 0.080 mm 2 . Also, a large fraction of the digital area is spent on the conservative choice of the decoder weight resolution, the large width of the decaying accumulator and the reconfiguration options. Thus, a more dedicated, less configurable design would realize additional area savings. The analog neurons can also be shrunk with the technology node, as this increases their speed and amplifies their mismatch, both desirable properties for the NEF ADC. Pushing the power consumption of the NEF ADC into a competitive range is harder than for the area. However, as the design of the NEF ADC is intended as a proof-of-principle, no effort has been spent on power optimization. Especially the neuron power draw is quite excessive, with its multiple current paths from VDD to ground. More that 80% of its power draw is not spent on charging the membrane or for switching, but in the offset and gain error stages. Due to downscaling, future neurons in smaller technologies may offer the same variation with significantly less involved circuits, i.e., less power budget. The digital circuitry has also not been optimized for low power draw. Since the NEF is robust to small timing variations in its pulses, the initial digital building blocks such as the decoder weight readout and adder tree could be run asynchronously, only synchronizing directly before the decay register. This would save significant power in the clock tree. For overall clocking, energy-efficient variable clock generators (Eisenreich et al., 2009 ) could be used to adjust the operating frequency of the system, making a system possible that offers the same resolution at different sample rates, similar to (Yip and Chandrakasan, 2011 ). Also, the multiple configuration options and corresponding bit widths at all stages add to the power draw. Here, gating techniques that shut off parts of the circuitry not needed for a given configuration have to be explored. In terms of absolute performance figures, Table 9 shows that the NEF ADC may be especially competitive when it comes to achieving very high resolution digitization, as resolution can be achieved cheaply by digitally averaging over a longer time span. This aspect will be preferentially evaluated once the hardware is available. However, while a one-to-one comparison of the NEF ADC with conventional ADC is informative, it was not the single design target. The main advantages of the NEF used as an ADC are the following:\n In the NEF ADC, the signal is represented in a robust way across a neuron population (see Table 6 ). Since the network is purely feed-forward, stability is not an issue. NEF makes little demand on the specific transfer characteristics of the analog neurons, and the encoder network uses binary weights. Accordingly, no high-fidelity, complex analog circuits are required anywhere in the system. The handcrafted analog circuits usually needed for an ADC are reduced to two simple neuron circuits, that are multiply instantiated. A large part of the processing is carried out in digital, making technology scaling very attractive and enabling design transfer across technologies with minimum effort. The possibility of adjusting the transfer characteristic, resolution and sample rate at runtime make for a very flexible system. In addition, the NEF framework incorporates a simple method to input several signals into this network and do computation with them for e.g., sensor fusion. In addition, NEF represents a theoretically well-explored paradigm, coming complete with a mathematically rigorous method for high-fidelity extraction of the original signal (Eliasmith and Anderson, 2004 ). Scaling and signal representation behavior necessary to achieve a given target ADC characteristic has been partially established in Choudhary et al. ( 2012 ) and treated in depth in this manuscript. 4.4. Limits of the NEF ADC resolution The INL plots of section 3.2 illustrate how insufficient decoder weight resolution, insufficient neuron number or tuning curve variation (represented by setting decoder weights zero) or insufficient tuning curve characterization (represented by perturbed decoder weights) can negatively influence the static INL. Especially for the case of perturbed decoder weights, the ENOB does not provide sufficient characterization of the ADC characteristic, as it stays virtually constant. The INL plots on the other hand provide a clear indication that static INL dominates dynamic INL (i.e., the INL caused by incomplete filtering as seen in the waveform-based INL in Figure 8 ). As can be seen from Table 2 , increasing the number of neurons increases resolution only sublinear, while power draw increases linearily. Thus, an ideal NEF ADC should be operated at the border between the dynamic INL and the static INL (also Figure 8 ). In other words, tuning curve variation, decoder weight resolution and especially neuron number should just be sufficient for the target INL, with τ PSC chosen such that the remaining pulse noise is on the same order as the static INL. The above is valid if the NEF ADC is built for a single conversion characteristic. In contrast, when using the NEF ADC over a wide range of possible τ PSC , there are two different options. Either the number of neurons is chosen very large so that even for the high resolution at large τ PSC , a sufficiently linear overall transfer characteristic can be constructed from the neuron tuning curves. However, this implies that at small τ PSC , the number of neurons is far in excess of those needed and the NEF ADC is dominated by pulse noise. The second option would be to choose the number of neurons only sufficient for linearity at small τ PSC , i.e., at low resolutions. At high resolutions (large τ PSC ), the static INL would intentionally dominate. To still achieve linearity, the digital output codes of the low pass filter would be passed through a look-up-table containing the inverse of the static INL curve. 4.5. Outlook In the current version, the NEF ADC still has a number of drawbacks. It is very susceptible to temperature and VDD variation. The transfer characteristic must thus ideally be measured for all these operating conditions and stored, or a constant on-line characterization has to be carried out. Built-in self-tests (BIST) such as Flores et al. ( 2004 ) look promising, as they would allow enhancing the NEF ADC with a constant self-monitoring at very little reduction of usable sample rate. Especially digital-heavy versions of BIST could be incorporated with little detriment in design time, as most of the functionality would be synthesizable. The area overhead would also be minimal if the NEF ADC is used as part of a larger digital system where existing compute resources could be reused for BIST (Flores et al., 2004 ). A second, more experimental approach might be to adjust the decoder weights online via neuromorphic means, such as synaptic plasticity. NEF has been shown to be amenable to supervised biologically plausible plasticity rules which have as supervisory input the overall transfer characteristic (Bekolay et al., 2013 ). This plasticity could act either in the analog domain as adjustable factor in the single neuron processing chains, or it could act directly on the digital decoder weights. A candidate plasticity rule that can be configured for a wide range of behavior, i.e., for different compensation or decoder characteristics, has recently been demonstrated (Mayr and Partzsch, 2010 ) and implemented efficiently in analog CMOS hardware (Mayr et al., 2010a ). Digital plasticity rules have been shown e.g., on the Spinnaker system (Jin et al., 2010 ). The main point for future work, however, will be to take advantage of the computational capability inherent in NEF. In this paper, NEF has been reduced to a linear representation of a single variable. We will explore various non-linear ADC characteristics and joint conversion of multiple inputs, offering complex sensor fusion and feature extraction (König et al., 2002 ; Mayr and Schüffny, 2007 ). Beyond the usage as ADC, the NEF could pave the way toward a future mixed-signal, mixed neuromorphic/conventional system on chip. The NEF could take various elements (regular CMOS, memristors (Jo et al., 2010 ; Ou et al., 2013 ), other nanoscale elements) and engineer a system with a set of target computations based on these elements. As demonstrated, such a framework can easily cross the barrier between asynchronous and synchronous systems as well as between analog and digital domains, doing the signal reconstruction either digitally as demonstrated here or via compact, configurable analog PSC circuits (Noack et al., 2010 ). Signal reconstruction could be via a decoder learned in memristors (Mayr et al., 2012 ). Thus, one could employ each type of system/device where it is most beneficial and arrive at an amalgan of the state of the art in the neuromorphic discipline, the digital/analog CMOS discipline and in nanodevice systems. 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."
} | 5,325 |
22609369 | null | s2 | 8,597 | {
"abstract": "Host-microbe symbioses involving bacterial endosymbionts comprise some of the most intimate and long-lasting interactions on the planet. While restricted gene flow might be expected due to their intracellular lifestyle, many endosymbionts, especially those that switch hosts, are rampant with mobile DNA and bacteriophages. One endosymbiont, Wolbachia pipientis, infects a vast number of arthropod and nematode species and often has a significant portion of its genome dedicated to prophage sequences of a virus called WO. This phage has challenged fundamental theories of bacteriophage and endosymbiont evolution, namely the phage Modular Theory and bacterial genome stability in obligate intracellular species. WO has also opened up exciting windows into the tripartite interactions between viruses, bacteria, and eukaryotes."
} | 206 |
39404445 | PMC11580455 | pmc | 8,598 | {
"abstract": "ABSTRACT Type IVa pili (T4aP) are widespread and enable bacteria to translocate across surfaces. T4aP engage in cycles of extension, surface adhesion, and retraction, thereby pulling cells forward. Accordingly, the number and localization of T4aP are critical to efficient translocation. Here, we address how T4aP formation is regulated in Myxococcus xanthus , which translocates with a well-defined leading and lagging cell pole using T4aP at the leading pole. This localization is orchestrated by the small GTPase MglA and its downstream effector SgmX that both localize at the leading pole and recruit the PilB extension ATPase to the T4aP machinery at this pole. Here, we identify the previously uncharacterized protein SopA and show that it interacts directly with SgmX, localizes at the leading pole, stimulates polar localization of PilB, and is important for T4aP formation. We corroborate that MglA also recruits FrzS to the leading pole, and FrzS stimulates SgmX recruitment. In addition, FrzS and SgmX separately recruit SopA. Precise quantification of T4aP-formation and T4aP-dependent motility in various mutants supports a model whereby the main pathway for stimulating T4aP formation is the MglA/SgmX pathway. FrzS stimulates this pathway by recruiting SgmX and SopA. SopA stimulates the MglA/SgmX pathway by stimulating the function of SgmX, likely by promoting the SgmX-dependent recruitment of PilB to the T4aP machinery. The architecture of the MglA/SgmX/FrzS/SopA protein interaction network for orchestrating T4aP formation allows for combinatorial regulation of T4aP levels at the leading cell pole resulting in discrete levels of T4aP-dependent motility. IMPORTANCE Type IVa pili (T4aP) are widespread bacterial cell surface structures with important functions in translocation across surfaces, surface adhesion, biofilm formation, and virulence. T4aP-dependent translocation crucially depends on the number of pili. To address how the number of T4aP is regulated, we focused on M. xanthus , which assembles T4aP at the leading cell pole and is a model organism for T4aP biology. Our results support a model whereby the four proteins MglA, SgmX, FrzS, and the newly identified SopA protein establish a highly intricate interaction network for orchestrating T4aP formation at the leading cell pole. This network allows for combinatorial regulation of the number of T4aP resulting in discrete levels of T4aP-dependent motility.",
"introduction": "INTRODUCTION Bacterial motility is important for the colonization of environmental niches, interactions with host cells, virulence, biofilm formation, and fitness by directing cells toward nutrients and away from toxins and predators ( 1 ). For translocation on solid surfaces, bacteria most commonly use type IVa pili (T4aP), long thin filaments that are also important for adhesion to host cells and abiotic surfaces, biofilm formation, virulence, predation, protein secretion, DNA uptake, and surface sensing ( 2 ). T4aP undergo cycles of extension, surface adhesion, and retraction ( 3 – 5 ). During these cycles, retractions generate a force of up to 150 pN that is sufficient to pull a cell forward ( 3 , 5 , 6 ). Efficient T4aP-dependent translocation depends on the number and cellular localization of T4aP ( 7 , 8 ). The T4aP extension/adhesion/retraction cycles are powered by the highly conserved T4aP machine (T4aPM) ( 2 ). In Gram-negative bacteria, this nanomachine is composed of 15 highly conserved proteins and spans from the outer membrane (OM) across the periplasm and inner membrane (IM) to the cytoplasm ( 9 – 11 ) (Fig. S1A). The hexameric PilB and PilT ATPases ( 12 – 15 ) associate with the cytoplasmic base of the core T4aPM in a mutually exclusive fashion to power T4aP extension and retraction, respectively ( 10 ). With the exception of PilT, all T4aPM proteins are important for T4aP extension, while PilT is only important for retraction ( 2 ). The T4aP is composed of thousands of copies of the major pilin subunit and contains a tip complex composed of minor pilins and the PilY1 adhesin ( 11 , 16 – 18 ). During extensions, major pilins are extracted from the IM and inserted at the T4aP base ( 4 , 19 , 20 ); during retractions, this process is inverted, with major pilin subunits being removed from the T4aP base and reinserted into the IM ( 4 , 21 ). While the highly conserved T4aPM constitutes the basis for the extension/adhesion/retraction cycles, much less conserved regulatory proteins determine where and how many T4aP are formed ( 7 , 18 , 22 – 29 ). However, their mechanism of action is poorly understood. Here, we address the regulation of T4aP formation in Myxococcus xanthus , a predatory soil bacterium with a social lifestyle and a model organism for understanding T4aPM function and regulation. The rod-shaped M. xanthus cells move across surfaces in the direction of their long axis using two motility systems, one for gliding and one for T4aP-dependent motility ( 30 , 31 ). Motility is important for the social behaviors of M. xanthus including predation and formation of swarming colonies in the presence and spore-filled fruiting bodies in the absence of nutrients ( 30 – 32 ). The T4aPM core is present at both cell poles ( 11 , 33 – 37 ). However, T4aP only assemble at one pole at a time ( 38 , 39 ). This localization enables M. xanthus cells to move unidirectionally with a piliated leading and a non-piliated lagging cell pole ( 7 , 39 ) and is essential for efficient translocation across surfaces ( 7 ). Consistent with the unipolar T4aP formation, the PilB extension ATPase almost exclusively localizes to the leading cell pole, while the PilT retraction ATPase localizes in a more bipolar asymmetric pattern and with the large cluster at the lagging cell pole ( 34 ) (Fig. S1B). In response to signaling by the Frz chemosensory system, M. xanthus cells reverse their direction of translocation ( 40 ) and after a reversal, T4aP assemble at the new leading pole ( 39 ); in parallel, PilB and PilT switch polarity ( 34 ) (Fig. S1B). The activity of the T4aPM in M. xanthus is regulated by the polarity module ( 41 – 43 ). The output of this module is generated by the small Ras-like GTPase MglA, which is a nucleotide-dependent molecular switch that is inactive in the GDP-bound and active in the GTP-bound state ( 44 , 45 ). In its GTP-bound state, MglA localizes to and defines the leading cell pole ( 44 , 45 ) (Fig. S1C). At this pole, MglA interacts with effectors to stimulate the T4aPM resulting in T4aP formation ( 7 , 46 ) and is essential for T4aP-dependent motility ( 47 , 48 ). The remaining five proteins regulate the nucleotide-bound state and localization of MglA by acting as a guanine nucleotide exchange factor in case of the RomR/RomX complex ( 49 ) or as a GTPase-activating protein in case of the MglB/RomY complex ( 44 , 45 , 50 ). MglA and the RomR/RomX and MglB/RomY complexes together with the MglC protein interact to bring about their asymmetric polar localization ( 43 , 51 ) (Fig. S1C). During the Frz-induced reversals, these six proteins switch polarity, thereby enabling the activation of the T4aPM at the new leading cell pole ( 43 – 45 , 49 , 50 , 52 – 54 ) (Fig. S1C). At the leading pole, MglA directly interacts with and recruits SgmX, a protein containing 14 tetratricopeptide repeats (TPR) ( 7 , 46 ), and has also been suggested to interact directly with FrzS ( 55 ), which is also important for T4aP-dependent motility ( 56 , 57 ). FrzS also interacts directly with SgmX and stimulates the recruitment of SgmX to the leading pole ( 58 ). SgmX, in turn, brings about PilB localization at the leading pole by an unknown mechanism and is essential for T4aP formation and, consequently, also for T4aP-dependent motility ( 7 ). Based on these observations, it has been suggested that SgmX stimulates T4aP formation by enabling PilB interaction with the base of the T4aPM ( 7 ). Here, to increase our understanding of how T4aP formation is regulated in M. xanthus , we searched for putative SgmX interaction partners. We identify the previously uncharacterized protein MXAN_0371 (reannotated to MXAN_RS01825 in the NCBI Reference Sequence NC_008095.1; henceforth S timulation of p ili formation p rotein A , SopA) and demonstrate that SopA interacts directly with SgmX, localizes at the leading pole, stimulates polar PilB localization, and is important but not essential for T4aP-formation and T4aP-dependent motility. We confirm that MglA is important but not essential for FrzS polar localization, and FrzS interacts directly with SgmX, thereby stimulating the polar recruitment of SgmX. In doing so, FrzS indirectly stimulates PilB polar localization, T4aP-formation, and T4aP-dependent motility. Additionally, SgmX and FrzS can separately recruit SopA to the leading pole. Altogether, our data support a model whereby MglA, SgmX, FrzS, and SopA interact to establish a protein interaction network that allows for combinatorial regulation of T4aP formation at the leading cell pole resulting in discrete levels of T4aP-dependent motility.",
"discussion": "DISCUSSION In this study, we addressed how T4aP formation is regulated in the rod-shaped cells of M. xanthus . Altogether, the detailed quantification of protein localization and T4aP formation supports a model in which the four proteins MglA, SgmX, FrzS, and SopA establish a highly interconnected protein interaction network to regulate T4aP formation ( Fig. 6 ). In this network, the small GTPase MglA is recruited to the leading pole via the RomR/RomX complex of the polarity module. MglA and its downstream effector protein SgmX are required and sufficient for the unipolar formation of T4aP and jointly bring about a low level of T4aP formation. By contrast, FrzS and SopA are dispensable for T4aP formation, and our data suggest that these two proteins function to stimulate the MglA/SgmX pathway for T4aP formation. In agreement with previous observations, FrzS is recruited to the leading pole by MglA-dependent and MglA-independent mechanisms. At this pole, FrzS stimulates SgmX polar localization and, thus, T4aP formation. In the case of SopA, it is separately recruited to the leading pole by SgmX and FrzS, where it stimulates the MglA/SgmX pathway for T4aP formation. Because SgmX and SopA are essential for the polar localization of the PilB extension ATPase while FrzS is important, we propose that the output of this pathway is to stimulate PilB interaction with the cytoplasmic base of the core T4aPM ( Fig. 6 ), thereby licensing T4aP formation. Because SopA does not affect the polar localization of MglA, SgmX, and FrzS, we suggest that SopA stimulates the function of SgmX in PilB polar recruitment ( Fig. 6 ). Fig 6 Model of protein interaction network for combinatorial regulation of T4aP-formation and T4aP-dependent motility in M. xanthus . The box shown by stippled lines indicates interactions that stimulate polar recruitment of proteins; gray circle surrounding MglA-GTP indicates the polar recruitment of MglA-GTP by the RomR/RomX complex of the polarity module. The figure presents a schematic network model of protein interactions involving GTP-bound MglA, FrzS, SgmX, SopA, and PilB, which further leads to interactions with the cytoplasmic base of the core T4aPM, T4aP formation, and T4aP-dependent motility. The detailed quantification of T4aP formation in different mutants supports a model in which the MglA/SgmX/FrzS/SopA interaction network allows for combinatorial regulation of the level of T4aP formation. In this model, the MglA/SgmX/FrzS/SopA interaction network can distinguish different input states that generate output states characterized by discrete levels of T4aP formation. Specifically, (i) in the absence of MglA and SgmX, no T4aP are formed ( 7 ), (ii) in the presence of only MglA and SgmX, a low level of T4aP is formed, (iii) in the presence of only MglA, SgmX, and SopA, the level is increased, (iv) in the presence of only MglA, SgmX, and FrzS, an even higher level of T4aP is assembled, and finally, (v) in the presence of all four proteins, the WT level of T4aP formation is accomplished. Thus, this pathway allows the regulation of the number of T4aP by integrating the input from MglA, FrzS, and SopA on the central protein SgmX. Under the conditions of the assay for T4aP-dependent motility, the defects in T4aP formation correlated with the level of T4aP-dependent motility in the different mutants except for the Δ sopA Δ frzS mutant. This mutant had a ~ 10-fold reduced amount of PilA in the sheared fraction compared to WT and did not display T4aP-dependent motility, suggesting that the number of T4aP in this mutant is too low to enable the pulling of cells across the surface used in the assay for T4aP-dependent motility. SgmX with its 14 TPRs contains three functional regions ( 7 , 46 , 58 ). The eight N-terminal TPRs mediate the activation of T4aP-dependent motility, the three middle TPRs engage in the interaction to FrzS, and the three C-terminal TPRs in the interaction to MglA ( 46 , 58 ). FrzS is a pseudo-response regulator with an N-terminal receiver domain, which lacks critical residues for phosphorylation, and a large C-terminal coiled-coil domain ( 56 , 57 ). The pseudo-receiver domain of FrzS interacts with SgmX ( 58 ), while the C-terminal coiled coil is sufficient for polar localization of FrzS ( 71 ). Previously, MglA was suggested to interact directly with FrzS ( 55 ); however, it is not known how MglA might interact with FrzS. Using a BACTH assay, we observed that SopA interacts directly with SgmX; however, we did not detect an interaction between SopA and FrzS. Based on the dissection of SgmX by Bautista et al. and Mercier et al. ( 46 , 58 ), we suggest that the eight N-terminal TPRs of SgmX are involved in the polar recruitment of PilB to the T4aPM. PilB interacts directly with PilM and PilC at the cytoplasmic base of the T4aPM [Fig. S1A; ( 15 , 72 , 73 )]. However, direct interactions between SgmX and PilB and/or PilM have not been detected ( 7 ). Interestingly, despite PilB not being polarly localized in the absence of SopA, the Δ sopA mutant still makes T4aP, suggesting that the formation of a visible polar PilB cluster may not fully reflect the interaction of PilB with the cytoplasmic base of the core T4aPM. Therefore, important goals for the future will be to determine how SgmX stimulates the interaction of PilB with the cytoplasmic base of the T4aPM and how SopA might further stimulate this interaction. In other bacteria, the regulation of T4aP formation also centers on the PilB extension ATPase. Specifically, in Vibrio cholerae and Clostridium perfringens , the second messenger c-di-GMP binds directly to the MshE and PilB2 ATPase, respectively, to stimulate T4aP formation ( 74 – 76 ). In Xanthomonas axonopodis pv. citri, c-di-GMP binds to the effector protein FimX, which then interacts with PilZ that, in turn, interacts with PilB, likely to stimulate T4aP formation ( 23 , 77 , 78 ). Similarly, in Pseudomonas aeruginosa , the c-di-GMP binding effector proteins FimX stimulate T4aP formation by interacting directly with PilB ( 25 ). The genetic and cell biological analyses demonstrate that the MglA/SgmX/FrzS/SopA network for T4aP formation is able to distinguish different input states with the formation of discrete levels of T4aP. However, the pathway is based on complete loss of function of MglA, SgmX, FrzS, and SopA. Therefore, in the future, it will be interesting to investigate under which physiological conditions these four proteins have altered accumulation and/or localization. In this context, we note that biosynthetic mutants unable to synthesize the secreted polysaccharide exopolysaccharide (EPS) have reduced but not abolished T4aP formation ( 79 ). This defect is caused by reduced T4aP extension and not increased retraction ( 79 ), but it is not known what causes this extension defect. MglA, SgmX, FrzS, and SopA accumulate at WT levels in an Δ epsZ mutant ( 80 ) that lacks the phosphoglycosyl transferase EpsZ that initiates EPS biosynthesis ( 79 ). In the future, it will be of interest to determine the localization of MglA, SgmX, FrzS, and SopA in EPS biosynthetic mutants."
} | 4,073 |
30214133 | PMC6129946 | pmc | 8,600 | {
"abstract": "Many animals display vigilance behaviors in order to detect predators. We found that the guards of a social bee coordinate their vigilance, which increases nest defense. Guards hover and distribute themselves evenly around the nest entrance. This increases the group’s field of view and ability to detect predators. We discuss how and why this organized pattern might arise.",
"introduction": "INTRODUCTION Vigilance against predators is one potential benefit of group living, as it can increase predator detection and individual survival ( Pulliam 1973 ; Krebs and Davies 1993 ; Cresswell 1994 ; Beauchamp 2017 ). An increase in group size also leads to a reduction in the time that individual group members spend being vigilant ( Bertram 1980 ; Elgar and Caterall 1981 ; Lima 1995 ). This group size effect is commonly explained by either the many-eyes hypothesis, where the proportion of time at least one individual is scanning increases ( Bertram 1980 ; Lima 1995 ; Fairbanks and Donson 2007 ), or the dilution effect whereby each individual is at lower risk of being targeted by a predator ( Hamilton 1971 ; Dehn 1990 ; Roberts 1996 ). By spending less time vigilant, individuals can dedicate more time to foraging or other activities that enhance fitness ( Elgar and Caterall 1981 ). For example, Lima (1995) found that dark-eyed juncos, Junco hyemalis , consumed food items over 50% faster as group size increased from 1 to 6. The collective vigilance of a group would be increased if group members also coordinated their vigilance efforts, such as by looking in different directions. Alternatively, coordinated vigilance may be organized so that some individuals focus on vigilance allowing others to focus on foraging. Although models predict benefits of coordination to collective vigilance ( Bednekoff and Lima 1998 ; Ferriere et al. 1999 ), such behavior is rarely observed in nature ( Ward 1985 ; Pays et al. 2007 ). This may be because individuals are selfish ( Hamilton 1971 ) or that the need to monitor the vigilance status of neighbors is itself costly and provides only marginal benefits over noncoordinated vigilance ( Ward 1985 ; Rodríguez-Gironés and Vasquez 2002 ). Where coordinated vigilance has been observed, it usually involves a sentinel system of only 1 or 2 vigilant individuals (meerkats, Clutton-Brock et al. 1999 ; cranes, Ge et al. 2011 ; rabbitfish, Brandl and Bellwood 2015 ). However, how vigilant individuals position themselves relative to each other and how this affects collective vigilance have received less attention than the effect of group size. Vigilance in social insects differs from most vertebrate examples in that, rather than fleeing from predators, vigilance may improve the defense of a fixed location, the nest. The nest contains reproductive individuals, offspring (brood), and food stores, such that its defense provides large fitness benefits. Early detection of predators is important for social insects because the first predators to arrive are often scouts of other social-insect colonies that can recruit nestmates for a mass attack ( Blum et al. 1970 ; Michener 1974 ; Ono et al. 1995 ). Detecting and disabling these scouts is, therefore, critical for colony survival. The second important distinction is that social insects often possess dedicated defenders (guards), which sometimes have morphological specializations and are not constrained by the need to forage or reproduce. Rather, time and effort are traded-off at the colony level through division of labor, with workers allocated among different tasks. The stingless bee Tetragonisca angustula (Apidae: Meliponini) presents an excellent opportunity to study the group-level coordination of vigilance. In addition to guards that stand at the nest entrance, which is normal in social insects, T. angustula colonies also have guards that hover near the entrance ( Grüter et al. 2011 ). To date, hovering guards are only known in T. angustula and the closely related T. fiebrigi (Grüter C, personal communication). Hovering guards are normally positioned to the left and right of the entrance and face inwards to form a corridor through which most bees entering the nest must pass ( Figure 1A , Wittman 1985 ). Guards inspect incomers, intercept nonnestmates and wrestle them to the ground ( Wittman et al. 1985 ). In agreement with studies of vigilance in vertebrates, larger groups of hovering guards are better at detecting intruders ( van Zweden et al. 2011 ). Furthermore, T. angustula guards are morphologically specialized, being the first described and most prominent example of a soldier caste within the eusocial bees ( Grüter et al. 2012 ; Grüter et al. 2017 ). The main natural enemy of T. angustula is the obligate robber bee Lestrimelitta limao ( Figure 1B ), which probably drove the evolution of the soldier caste ( Grüter et al. 2017 ) and whose local density influences colony investment in defense ( Segers et al. 2016 ). Figure 1 (A) Hovering guards of the stingless bee Tetragonisca angustula at a nest entrance in São Paulo State, Brazil. Hovering guards are positioned one on each side of the entrance tube to form a corridor through which incoming bees must pass. Standing guards can be seen in and around the entrance. (B) T. angustula guard (left) fighting with Lestrimelitta limao robber bee (right). We investigated how T. angustula hovering guards are positioned relative to each other and how this affects vigilance and predator detection. Guards typically hover on either side of the entrance tube, looking inwards and pointing left or right ( Figure 1A ). As a result, a guard facing left of the entrance will have a more limited view of the right of the entrance and vice versa . To enhance their collective vigilance, we predict that guard groups should have individuals facing both directions. Our first aim was to establish whether hovering guards were positioned more evenly, left and right of the entrance, than expected if they positioned themselves randomly. We then compared vigilance of even versus skewed left–right distributions of hovering guards. Finally, we investigated the effect of guard distribution on the ability to detect predators using a bioassay to simulate an attack by L. limao robber bees.",
"discussion": "DISCUSSION Our results show that multiple hovering guards of T. angustula coordinate themselves in a way that improves the collective vigilance of the group. Hovering guards were distributed more evenly, left versus right, than would be expected if each individual was positioned at random. This effect was significant in each of 3 situations: 2, 3, and 4 guards, providing strong evidence for colony-level adaptive organization. The effect was weaker when 3 guards were present. However, this was likely because the expected frequency of guards in the most even ratio (2:1) was 75%, meaning that the maximum possible effect size in the direction of evenness was only 25% ( Figure 2B ) versus 50% (50% expected) when 2 guards were present ( Figure 2A ). The coordination of hovering guards into an even ratio increased the collective vigilance of the group, but did not have an effect at the individual level ( Figure 3B ,D). Meanwhile, an increase in group size caused a decrease in individual vigilance but an increase in collective vigilance, consistent with the group-size effect ( Figure 3A ,C). The decrease in individual rotation may be beneficial, if rotation somehow reduces the quality of vision of the guard and, presumably, the guard saves a small amount of energy. The individual response to group size may be adaptive, resulting from an awareness that other hovering guards are present. Alternatively, the increased level of rotation in small groups may be because every guard has to inspect incoming bees, whereas in large groups, some inspect while some remain in position and so rotate less. The collective response to group size was greatest between 1 and 2 guards and was enhanced by coordination, because the second guard was typically on the opposite side to the first, which generally doubled the total field of view. Hovering guards seldom face outwards from the nest entrance, which would seemingly limit the group’s collective view of the environment. However, the compound eyes of T. angustula extend to the side of the head (see Grüter et al. 2012 ), allowing the bee to see outwards even when its body is perpendicular to the nest entrance. Coupled with the generally poor visual acuity of the insect compound eye ( Mallock 1894 ; Kirschfield 1976 ; Snyder 1977 ; Land 1997 ), this suggests that the addition of guards facing outwards would not greatly increase predator detection. The positioning of hovering guards to face a flight corridor has the additional function of increasing the ability of guards to intercept intruders flying towards the entrance ( Wittman 1985 ). Guards facing in the direction of attack were better able to detect intruders, as shown in our 2 complementary bioassays. Lone guards were 3 times as likely to detect a dummy robber bee when it approached from the front, rather than the rear. When there were 2 hovering guards, the guard facing the intruder was twice as likely to initiate an attack as the guard facing away. This second result is all the more powerful because the guard facing the model predator was always the further from it of the two. The diffusive nature of larger guard groups may lead to the breakdown of this rule, because an intruder will have to bypass several guards facing away before it encounters a guard facing towards it. The direct defensive benefits of coordinated over noncoordinated vigilance is a topic for further study. In particular, it would be valuable to investigate whether coordinated vigilance in T. angustula increases the ability of a colony to defend against the robber bee L. limao , which is probably the most important enemy of T. angustula ( Segers et al. 2016 ; Grüter et al. 2017 ). Furthermore, is coordinated vigilance more efficient than noncoordinated vigilance? For example, do 2 hovering guards in an even left–right ratio may have greater collective vigilance than 3 guards that all hover on the same side of the entrance, meaning that fewer guards are needed? Coordinated vigilance in T. angustula is presumably adaptive in the context of the behavior and strategy employed by L. limao robber bees, especially scouts, when approaching a T. angustula nest entrance. If robber bees approach from the side then the coordination of vigilance is clearly of value, as shown by our bioassays. However, if robber bees approach from the front then we would not expect coordination to be more effective than if guards were positioned at random. If robber bees do not employ any positional strategy and instead attack from a random direction, then the coordination of vigilance will be of use at least some of the time, and there is presumably little additional cost of coordinated versus uncoordinated vigilance. Unfortunately, to witness the beginning of a raid, where robber bee scouts first find the host nest, is extremely rare ( von Zuben 2012 ; Grüter C, personal communication), and we have not ourselves witnessed the initial stages of an attack. It would therefore be of great value to observe the initial stages of a raid and to study the response of hovering guards. The benefits of coordinated vigilance relative to the more established role of group size remain unknown. Although we studied groups of 1–4 hovering guards, the number may exceed 15 ( van Zweden et al. 2011 ). We predict that as group size increases, the importance of coordination relative to group size will diminish for 2 reasons: first, because coordination will become more difficult, analogous to the costs of monitoring other group members proposed by Ward (1985) ; second, with many guards even a random configuration would likely cover all directions. Furthermore, as guard number increases, we expect collective vigilance ( Figure 3C ) to plateau as it approaches the limit of 360°. However, higher guard number could still increase collective vigilance through the occupation of a greater area ( van Zweden et al. 2011 ). There would also be defensive benefits unrelated to vigilance, in particular, the ability to fight, harass, or confuse predators should continue to increase with group size ( Shields 1984 ; Landeau and Terborgh 1986 ; Shackleton et al. 2015 ). Indeed, there are also several guards that stand at the nest entrance, ready to attack any threats once they are detected. How do hovering guards achieve an even left–right distribution? We hypothesize that the pattern is self-organized, which is a common mechanism in insect societies, including nest defense ( Bonabeau et al. 1997 ; Millor et al. 1999 ; Boomsma and Franks 2006 ). There is also evidence that self-organization works in conjunction with group size to produce greater collective vigilance in fish shoals ( Ward et al. 2011 ). The pattern in T. angustula could arise through individual guards reacting to their own local environment and experience, with the application of 2 simple rules: first, if a guard detects another guard on the same side of the entrance as itself, then its propensity to switch sides increases; second, if after switching a guard detects another guard on the same side as itself, it remains for some time before moving, in order to prevent continuous switching. Alternatively, a guard may react to the absence of guards on its side or the guard state on the opposite side to itself. There is some evidence that bees can count, at least up to 4 ( Chittka and Geiger 1995 ; Dacke and Srinivasan 2008 ), which might also be used in distributing hovering guards into an even ratio. In contrast to individuals in an ungulate herd, bird flock, or fish shoal, social-insect guards should always be vigilant. Because worker fitness is tied more closely to the colony than personal safety, guard groups should be free from the limitations of the selfish herd ( Hamilton 1971 ), which may prevent the emergence of collective vigilance arising from the cooperation of unrelated individuals. In selfish herding, unrelated individuals should strive for the safe positions with little regard for the interests of their neighbors ( Hamilton 1971 ). The study of organized patterns in animals with high intra-group relatedness (e.g. Santema and Clutton-Brock 2013 ) may reveal new rules governing vigilance and the benefits of group living in general."
} | 3,641 |
39555692 | PMC11982788 | pmc | 8,602 | {
"abstract": "Summary \n Cross‐kingdom RNA interference (ckRNAi) is a mechanism of interspecies communication where small RNAs (sRNAs) are transported from one organism to another; these sRNAs silence target genes in trans by loading into host AGO proteins. In this work, we investigated the occurrence of ckRNAi in Arbuscular Mycorrhizal Symbiosis (AMS). We used an in silico prediction analysis to identify a sRNA ( Rir2216 ) from the AM fungus Rhizophagus irregularis and its putative plant gene target, the Medicago truncatula MtWRKY69 transcription factor. Heterologous co‐expression assays in Nicotiana benthamiana , 5′ RACE reactions and AGO1‐immunoprecipitation assays from mycorrhizal roots were used to characterize the Rir2216–MtWRKY69 interaction. We further analyzed MtWRKY69 expression profile and the contribution of constitutive and conditional MtWRKY69 expression to AMS. We show that Rir2216 is loaded into an AGO1 silencing complex from the host plant M. truncatula , leading to cleavage of a host target transcript encoding for the MtWRKY69 transcription factor. MtWRKY69 is specifically downregulated in arbusculated cells in mycorrhizal roots and increased levels of MtWRKY69 expression led to a reduced AM colonization level. Our results indicate that MtWRKY69 silencing, mediated by a fungal sRNA, is relevant for AMS; we thus present the first experimental evidence of fungus to plant ckRNAi in AMS.",
"conclusion": "Conclusions In summary, our work describes fungus to plant ckRNAi in the AMS for the first time. Starting with in silico target predictions, we identified a plausible ckRNAi interaction between the fungal sRNA Rir2216 and the plant transcript MtWRKY69 . We propose that R. irregularis exports Rir2216 to cortical cells which establish intimate contact with the fungus, and, by hijacking the AGO1‐equipped plant RNA silencing machinery, Rir2216 downregulates the MtWRKY69 encoding gene at the posttranscriptional level. As many WRKY transcription factors are involved in the response to pathogens (Jiang et al ., 2017 ; Chen et al ., 2019 ), we speculate that the Rir2216 ‐mediated silencing of MtWRKY69 , particularly in cortical cells that house arbuscules, could contribute to local suppression of the plant immune response, which would favor successful colonization. This comports with the previously described mechanism of action of the SP7 protein effector in R. irregularis (Kloppholz et al ., 2011 ). Further investigations on the function and regulation of MtWRKY69 in more detail are needed to elucidate its specific regulatory role in AMS. In addition, it so far remains unknown how R. irregularis exports Rir2216 , and possibly other sRNAs, into plant cells; indeed, whether ckRNAi is mediated by extracellular vesicles, ribonucleotide binding proteins or passive diffusion remains to be deciphered (Zand Karimi et al ., 2022 ; He et al ., 2023 ). Finally, whether ckRNAi in the AMS is a bidirectional phenomenon, as has been previously reported in other interactions (Weiberg et al ., 2013 ; Wang et al ., 2016 ; Zhang et al ., 2016 ; Cai et al ., 2018a ; He et al ., 2023 ), is still unknown. Taken together, our findings describe a new layer of plant–fungus communication in the AMS and are a stimulus for further research into the molecular mechanisms underlying one of the most important symbioses on the planet.",
"introduction": "Introduction RNA interference (RNAi) is a biological process, almost universally present in eukaryotes, which, based on the recognition of target nucleic acids by small RNAs (sRNAs), leads to regulation of gene expression at the transcriptional and/or posttranscriptional level. Small RNAs and RNAi were shown to play a role in different interspecies, and even inter‐kingdom, communication as sRNAs can move from one organism to a distantly related one, leading to the silencing of target genes through the exploitation of RNAi (Cai et al ., 2018a ). This process, known as cross‐kingdom RNAi (ckRNAi), has been described in many pathogenic interactions involving animal and plant systems (Weiberg et al ., 2013 ; Buck et al ., 2014 ; Zhang et al ., 2016 ; Wang et al ., 2017b ; Shahid et al ., 2018 ; Cai et al ., 2018b ; Dunker et al ., 2020 ; He et al ., 2021 ) and in a few plant mutualistic associations (Ren et al ., 2019 ; Wong‐Bajracharya et al ., 2022 ). In silico work has hinted at the possibility of cross‐kingdom RNAi occurring in the Arbuscular Mycorrhizal Symbiosis (AMS) (Silvestri et al ., 2019 ), one of the most widespread and ancient symbiotic associations on the planet (Genre et al ., 2020 ). Most land plants, including many crops, engage in this symbiosis with soil fungi from the subphylum Glomeromycotina, which provides key benefits to host plants in both natural and agricultural systems (Genre et al ., 2020 ). The cellular and metabolic reprogramming of plant cells upon colonization by AM fungi implies a complex network of transcriptional regulation and molecular signaling. Small RNAs are emerging as essential elements within this gene regulatory network (Ledford et al ., 2023 ; Zeng et al ., 2023 ). There is also indirect evidence that translocation of sRNA occurs in AMS and may be involved in ckRNAi (Qiao et al ., 2023 ). In particular, host‐ and virus‐induced gene silencing techniques have been shown to be successful tools for downregulating fungal genes in mycorrhizal roots, pointing to a movement of functional sRNAs from the plant to the AM fungus (Helber et al ., 2011 ; Kikuchi et al ., 2016 ). Moreover, the observation of extensive membrane rearrangements and the formation of extracellular vesicles at the plant‐AM fungal interface (Ivanov et al ., 2019 ; Roth et al ., 2019 ) bolsters curiosity since in pathogenic interactions, extracellular vesicles represent a pathway of cross‐kingdom communication and sRNA transfer in ckRNAi (Buck et al ., 2014 ; Cai et al ., 2018a , 2018b ; He et al ., 2023 ). In a previous study, we investigated the potential for fungal‐to‐plant sRNA transfer and showed that the model AM fungus Rhizophagus irregularis possesses RNAi machinery and produces functional sRNAs, with some predicted to potentially target mRNAs from the host plant Medicago truncatula (Silvestri et al ., 2019 ). In this work, we set up multiple assays to validate the results of an in silico sRNA‐target mRNA prediction analysis and offer experimental evidence that a fungal sRNA guides the silencing of a plant gene through ckRNAi, favoring the establishment of AMS.",
"discussion": "Results and Discussion The fungal sRNA \n Rir2216 is predicted to target the \n WRKY69 \n gene of the host plant M. truncatula \n Using a sequence complementarity approach based on sRNAs from the AM fungus R. irregularis , we predicted targets in M. truncatula transcriptome. As input, we selected fungal sRNA‐generating loci known to be upregulated in intraradical relative to extra‐radical mycelium (Silvestri et al ., 2019 ). We then singled out the most abundant fungal sRNAs from each locus (589 sRNAs in total). Based on target expectation scores, sRNAs from an intergenic locus named ‘cluster_832’ were among the highest ranked (Table S1 ). After filtering for sRNAs with a length of 21 nucleotides – the length of previously described ckRNAi sRNAs (Cai et al ., 2018a ) – cluster_832 became especially pronounced (Table S1 ). We named the most abundant sRNA of cluster_832 Rir2216 . We identified Rir2216 in other publicly available RNA‐Seq data from R. irregularis (Table S4 ). Rir2216 presented isoforms ranging from 21 to 24 nucleotides in length as it was observed in other sRNAs involved in cross‐kingdom RNAi (Wang et al ., 2017a ). As revealed by two independent prediction tools, among the highest scoring targets of Rir2216 was a gene encoding for the M. truncatula WRKY transcription factor 69 ( MtWRKY69 ; Medtr2g083870, Fig. 1a ; Table S2 ). WRKY transcription factors are known to regulate several aspects of plant biology, including responses to biotic factors (Jiang et al ., 2017 ; Chen et al ., 2019 ), making MtWRKY69 a particularly interesting target gene. The theoretical hybridization energy between Rir2216‐MtWRKY69 is −36.37 kcal mol −1 (88.30%) (Fig. 1a ), which is within the range of hybridization energies found in endogenous miRNA : target duplexes in plants (Alves‐Junior et al ., 2009 ). Fig. 1 \n In silico analysis of the interaction between Rir2216 and MtWRKY69 . (a) (Upper panel) Target prediction results of potential targets of Rir2216 in the Medicago truncatula transcriptome with their associated description, expectation, and alignment. (Lower panel) Alignment of the target site within the MtWRKY69 sequence with that of Rir2216 and the associated hybridization energy of the sRNA‐mRNA pair. For the alignment, the following symbols are used: colon (:) denotes standard Watson‐Crick base pairs; dot (.) denotes G‐U wobble pairs. (b) Phylogenetic relationship of MtWRKY69 orthologues. Protein sequences were aligned with mafft v.7.511 (Katoh & Standley, 2013 ), and their phylogenetic relationship was inferred with IQ‐T ree 2.0.7 (Minh et al ., 2020 ) with the tree rooted on the outlier sequence of Arabidopsis thaliana WRKY1. Sequences, whose corresponding CDSs contain the Rir2216 binding site, are indicated in green. (c) Nucleotide diversity (pi) across the CDS alignment of MtWRKY69 orthologous belonging to Fabaceae family, calculated with a 21‐mer sliding‐windows method (i.e. Pi for all the possible 21‐mer across the alignment in steps of 1 nt) excluding sites having alignment gaps. The most conserved 21‐mers in the alignment overlaps with the binding site for Rir2216 . Notably, the predicted binding site for Rir2216 has a higher level of conservation than the sequence encoding the WRKY domain. We hypothesized that, if Rir2216 ‐dependent regulation of MtWRKY69 was functional and important for AMS establishment, the binding site of Rir2216 should be conserved in WRKY69 orthologs from other species able to engage in AMS with R. irregularis . To test this, the protein sequences of MtWRKY69 orthologues (98 sequences in total) were retrieved from EnsemblPlant and used to build a phylogenetic tree (Fig. 1b ). The MtWRKY69 orthologs clustered into groups based on species phylogeny, including Poaceae, Brassicaceae, and Fabaceae members. Examination of their coding sequences led to the identification of those carrying potential binding sites of Rir2216 by repeating the target analysis on the orthologues using Rir2216 as a guide. All but one of the species that clustered to Fabaceae (an AMS‐forming family) presented the predicted binding site for Rir2216 within the CDS of their orthologous MtWRKY69. The species that clustered to other families did not contain a predicted binding site for Rir2216 . Among the Fabaceae family orthologs, the nucleotide diversity (Pi index) was calculated using a window‐based method (Pi for all possible 21‐mers; from position 1 to the end of the alignment). The predicted binding sequence of Rir2216 was found to overlap with the most conserved 21‐mer site in the entire alignment, even more conserved than 21‐mers corresponding to the WRKY domain itself (Figs 1c , S1 ). Such a degree of conservation supports the biological regulatory relevance of the specific sequence motif. We can speculate that the Rir2216 – WRKY interaction evolved specifically in the Fabaceae family to control mycorrhizal formation in this group of plants. However, as mentioned before, we identified Rir2216 isoforms in publicly available RNA‐Seq data from R. irregularis ‐colonized roots of other plant species ( Nicotiana attenuata , Solanum lycopersicum ; Table S4 ); we can hypothesize that in these hosts that belong to the Solanaceae family, Rir2216 may have other target mRNAs. Such instance has been described for plant miRNAs, where for example miR396 can target GRF, bHLH and/or MADs box transcription factors in different species, being those different targets related to defense responses (Silvestri et al ., 2024 and references therein). A wider knowledge on the population of sRNA in mycorrhizal roots of different plant–fungus combinations would be instrumental to clarify this issue. \n \n MtWRKY69 \n is a bona fide cross‐kingdom RNAi target of Rir2216 \n If Rir2216 suppresses the expression of MtWRKY69 by ckRNAi, a lower abundance of its transcripts in mycorrhizal roots should be observed. Quantitative reverse transcription polymerase chain reaction assays on RNA extracted from whole roots revealed that MtWRKY69 expression did not change in mycorrhizal samples compared to control roots (Fig. S2 ). As mycorrhizal roots are a heterogeneous environment consisting of different plant cell types and fungal structures, gene expression profiles associated to specific cell types can be masked by a dilution effect when RNA is analyzed at the whole root level. Laser Microdissection (LMD) technology was therefore exploited to isolate arbuscule‐containing cortical cells (Fig. 2a ), which are considered the critical functional structures of the AMS in which fungal and plant cells achieve the most intimate interaction and where the nutrient exchange is thought to occur (Genre et al ., 2020 ). The quality and identity of the LMD samples were confirmed by the transcript abundance of MtPT4 , a phosphate transporter encoding gene specifically expressed in arbusculated cells (Javot et al ., 2007 ) (Fig. 2a ). MtWRKY69 was found to be downregulated in arbuscule‐containing cells compared to cortical cells from nonmycorrhizal roots (Fig. 2a ). As AGO1 is the nuclease central to posttranscriptional gene silencing and previously described as involved in ckRNAi (Dunker et al ., 2020 ) we also monitored the expression of MtAgo1 in these LMD samples. A significant upregulation of MtAgo1 in arbusculated cortical cells relative to those from control roots was observed (Fig. 2a ), indicating that this component of the plant RNAi machinery is activated in this specific cell type. Fig. 2 \n Rir2216 is a bona fide ckRNAi sRNA targeting MtWRKY69. (a) On the left: transversal section of a mycorrhizal root under the laser microdissector before (upper panel) and after (down) the cut. Arbuscule‐containing cells are indicated by a red line. Bar, 50 mm. On the right: Normalized expression values of MtWRKY69 , MtPT4 , and MtAGO1 transcript abundance in cells collected by laser microdissection: CC‐cortical cells from noncolonized roots; Arb‐cortical cells containing arbuscules. Box plots display the median (horizontal line), the quartiles (boxes) and 1.5 interquartile range (whiskers); each dot corresponds to an independent replicate. Statistical analysis was performed using one‐way analysis of variance (Kruskal–Wallis test; *, P < 0.05). (b) Western blot of co‐expression assays using proteins extracted from Nicotiana benthamiana leaves expressing WRKY‐YFP alone, in combination with Rir2216 , or with the plant miR159b. The lower panel corresponds to a Ponceau staining of the gel showing the Rubisco protein. (c) The cleavage site (arrow) and frequencies (indicated by the ratio of the number of clones showing that 5′ end to the total number of sequenced recombinant clones) detected using a 5′ RACE on Medicago truncatula mycorrhizal roots. (d) (Upper panel left) Western blot (WB) of anti‐AGO1 (left lane) and anti‐myc‐tag (central lane) and no‐antibody (right lane) immunoprecipitations with anti‐AGO1 and anti‐myc‐tag antibodies. Full images of blots are shown in Supporting Information Fig. S6 ; (lower panel left) Gel electrophoresis of stem‐loop reverse transcription polymerase chain reaction assays on Rir2216 , Rir434 , and Rir196 using RNA extracted from the immunoprecipitated fractions. (Right) Gel electrophoresis of stem‐loop reverse transcription polymerase chain reaction assays on Rir2216 , Rir434 , Rir196 , and Mt‐miR398a‐3p from input RNA from noninoculated (Mock) and mycorrhizal roots (Myc). To assess whether Rir2216 is capable of silencing its predicted target MtWRKY69 in planta , we set up transient co‐expression assays in Nicotiana benthamiana leaves. We inserted Rir2216 into two miRNA backbones: miR159b from M. truncatula (Fig. 2b ) and miR319a from A. thaliana (Fig. S3 ). We separately cloned the YFP (yellow fluorescent protein) coding sequence fused to MtWRKY69 at the C‐terminal end (MtWRKY69‐YFP) under the strong and constitutive 35S promoter from the Cauliflower Mosaic Virus (CaMV35S). When MtWRKY69‐YFP was expressed alone, a robust accumulation of the chimeric protein was detected by Western blot (Figs 2b , S3 ). By contrast, when MtWRKY69‐YFP was co‐expressed with miRNA backbones containing Rir2216 , a visible decrease in protein accumulation was observed. The decrease in protein accumulation was not observed when MtWRKY69‐YFP was co‐expressed with sRNAs of plant ( MtmiR159b , Fig. 2b ) or fungal ( Rir773 ; Fig. S3 ) origin that lacked predicted target sites in the MtWRKY69 sequence. These results demonstrate that Rir2216 is able to silence its target gene, MtWRKY69 , in planta in a sequence‐specific manner. To confirm that Rir2216 can target MtWRKY69 in mycorrhizal roots, we performed a 5′ RLM‐RACE assay to amplify the cleavage fragment of MtWRKY69 based on predicted AGO‐catalyzed Rir2216 ‐guided endonucleolysis. We first successfully validated the RACE library by corroborating the presence of MtHB8 cleaved transcripts within the known miR166 binding site (Fig. S4a ; Boualem et al ., 2008 ). Subsequently, we obtained a unique PCR product showing the expected size when mapping MtWRKY69 cleaved products (Fig. S4b ). Sequencing of the amplified DNA fragment showed that it did indeed belong to MtWRKY69 and that the 5′‐end was located within the predicted binding site of Rir2216 (Fig. 2c ). Nevertheless, target cleavage mapped 4 nucleotides downstream of a canonical plant miRNA‐guided cleavage site that is found between the nucleotides 10–11 from the 5′‐end of the miRNA. Deviations from the canonical cleavage position are often observed when mapping miRNA‐ and cross‐kingdom sRNA‐mediated cleavage (Llave et al ., 2002 ; Jones‐Rhoades & Bartel, 2004 ; Zhao et al ., 2012 ; Tsikou et al ., 2018 ; Ren et al ., 2019 ; Ji et al ., 2021 ). The reason behind the production of noncanonical cleavage fragments around small RNA binding sites is not completely understood. However, it is possible that exonucleases like XRN4 (Souret et al ., 2004 ) may trim the cleavage product after AGO‐catalyzed cleavage. Inefficient AGO‐catalyzed target cleavage may also lead to stalling of RISC at the target site. Thus, site‐specific cleavage independent of catalytic activity of AGO, possibly mediated by stalled ribosomes (Arribas‐Hernández et al ., 2016 ), could occur following AGO binding. To verify the possible association of Rir2216 to components of M. truncatula RNAi machinery, we performed RNA immunoprecipitation (RIP) to pull down sRNAs associated with M. truncatula AGO1 from mycorrhizal roots, and then followed with stem‐loop reverse transcription polymerase chain reaction to examine specific sRNAs. We focused on AGO1 for three reasons: Rir2216 contains the hallmarks of AGO1 binding (5′ U and 21 nt in length) (Mi et al ., 2008 ), in other biological systems AGO1 is the member of the AGO family principally involved in ckRNAi (Wang et al ., 2016 ; Shahid et al ., 2018 ; Cai et al ., 2018a ; Cui et al ., 2019 ), and we had previously demonstrated upregulation of MtAGO1 in arbusculated cells (Fig. 2a ). We first confirmed the cross‐reactivity of a commercially available A. thaliana anti‐AGO1 antibody against M. truncatula AGO1 by performing a Western blot using proteins extracted from shoots and mycorrhizal roots. We were able to detect a corresponding band ( c . 122 kDa) from shoots but not from roots, possibly due to the low concentration of AGO1 in the protein extract (Fig. S5 ). Therefore, we generated composite plants with roots expressing MtAGO1 – tagged with a myc epitope at the N‐terminal – under the CaMV35S promoter. Through this approach, the recombinant AGO1 was successfully immunoprecipitated from mycorrhizal roots of composite plants using the A. thaliana anti‐AGO1 antibody but not the anti‐myc‐tag antibody, as confirmed by Western blot (Fig. 2d , top left). RNA was extracted from the immunoprecipitated fractions and used for stem‐loop reverse transcription polymerase chain reaction assays. We identified Rir2216 in the anti‐AGO1‐immunoprecipitated fraction but neither in the anti‐myc‐tag fraction (in which the AGO1 pull‐down was unsuccessful) nor in the control sample without antibody (Fig. 2d , bottom left). As negative controls, two fungal sRNAs, Rir434 and Rir196 , were analyzed in parallel. Both sRNAs showed higher expression levels compared to Rir2216 in the sRNAseq dataset (Silvestri et al ., 2019 ) and were detected in the input of RNA immunoprecipitation from mycorrhizal roots (Fig. 2d , right). In addition, Rir434 possesses similar structural characteristics to known AGO1‐binding sRNAs (Mi et al ., 2008 ), while Rir196 does not. Notably, no amplification was observed in the immunoprecipitated fraction for either Rir434 or Rir196 . This suggests that there is selectivity in the transfer of fungal sRNAs to plants and/or their association with host AGO1 proteins during R. irregularis root colonization. These findings indicate that Rir2216 is a functional sRNA involved in ckRNAi. Increased levels of \n MtWRKY69 \n expression reduce AM fungal colonization of host roots To establish the biological significance of Rir2216 ‐mediated regulation of its target gene, we overexpressed MtWRKY69 and tested how plants responded to fungal colonization. To that end, we generated composite plants expressing MtWRKY69 or, as a control, the fluorescent protein Scarlet, under the CaMV35S promoter, and inoculated them with R. irregularis . By quantitative reverse transcription polymerase chain reaction, we first confirmed the upregulation of MtWRKY69 and then observed that MtWRKY69 overexpression significantly reduced the level of AM colonization as indicated by transcript abundance of the M. truncatula AM marker gene MtPT4 compared to Scarlet expressing plants (Fig. 3a ). Morphological analyses of mycorrhizal formation confirmed the molecular data: a lower percentage of frequency, intensity and arbuscule abundance was observed in MtWRKY69 ‐overexpressing plants compared to control plants (Fig. 3b ). No alteration in arbuscule morphology was observed (Fig. 3c ). As the overexpression of the gene driven by the CaMV35S promoter could lead to pleiotropic effects, which, in turn, might affect AM symbiosis development, we also analyzed the mycorrhizal phenotype of composite plants expressing MtWRKY69 under the arbuscule‐specific promoter of MtPT4 , along with Scarlet as control. MtPT4 promoter conditionally increased the expression of MtWRKY69 or Scarlet solely upon AMS, resulting again in a reduced mycorrhizal formation level only when MtWRKY69 was induced (Fig. 3d ). Fig. 3 MtWRKY69 transcription factor modulates fungal colonization. (a) Normalized expression value of MtWRKY69 and the AM‐responsive gene MtPT4 in plants expressing Scarlet or MtWRKY69 under the CaMV35S promoter (OE = overexpressing). (b) Frequency of mycorrhizal colonization (F%), intensity of colonization (M%) and arbuscules abundance (A%) in composite plants expressing Scarlet or MtWRKY69 under the CaMV35S promoter at 60 d post inoculation (dpi). (c) Representative images of R. irregularis ‐colonized roots from composite plants overexpressing Scarlet or MtWRKY69 . Arrows indicate arbuscule‐containing cells. cc, central cylinder; c, cortical cells; ep, epidermal cells. Bars correspond to 50 μm. (d) Frequency of mycorrhizal colonization (F%), intensity of colonization (M%) and arbuscules abundance (A%) in composite plants expressing Scarlet or MtWRKY69 under the MtPT4 promoter at 60 dpi. In a, b and d, box plots display the median (horizontal line), the quartiles (boxes) and 1.5 interquartile range (whiskers); each dot corresponds to an independent replicate. Statistical analysis was performed using one‐way analysis of variance (ANOVA; *, P < 0.05; **, P < 0.01). On the whole, these results indicate that regulation of MtWRKY69 expression levels plays a role in controlling the extent of fungal colonization in mycorrhizal roots. Conclusions In summary, our work describes fungus to plant ckRNAi in the AMS for the first time. Starting with in silico target predictions, we identified a plausible ckRNAi interaction between the fungal sRNA Rir2216 and the plant transcript MtWRKY69 . We propose that R. irregularis exports Rir2216 to cortical cells which establish intimate contact with the fungus, and, by hijacking the AGO1‐equipped plant RNA silencing machinery, Rir2216 downregulates the MtWRKY69 encoding gene at the posttranscriptional level. As many WRKY transcription factors are involved in the response to pathogens (Jiang et al ., 2017 ; Chen et al ., 2019 ), we speculate that the Rir2216 ‐mediated silencing of MtWRKY69 , particularly in cortical cells that house arbuscules, could contribute to local suppression of the plant immune response, which would favor successful colonization. This comports with the previously described mechanism of action of the SP7 protein effector in R. irregularis (Kloppholz et al ., 2011 ). Further investigations on the function and regulation of MtWRKY69 in more detail are needed to elucidate its specific regulatory role in AMS. In addition, it so far remains unknown how R. irregularis exports Rir2216 , and possibly other sRNAs, into plant cells; indeed, whether ckRNAi is mediated by extracellular vesicles, ribonucleotide binding proteins or passive diffusion remains to be deciphered (Zand Karimi et al ., 2022 ; He et al ., 2023 ). Finally, whether ckRNAi in the AMS is a bidirectional phenomenon, as has been previously reported in other interactions (Weiberg et al ., 2013 ; Wang et al ., 2016 ; Zhang et al ., 2016 ; Cai et al ., 2018a ; He et al ., 2023 ), is still unknown. Taken together, our findings describe a new layer of plant–fungus communication in the AMS and are a stimulus for further research into the molecular mechanisms underlying one of the most important symbioses on the planet."
} | 6,685 |
30345402 | PMC6190515 | pmc | 8,603 | {
"abstract": "Microbial natural products are a tremendous source of new bioactive chemical entities for drug discovery. Next generation sequencing has revealed an unprecedented genomic potential for production of secondary metabolites by diverse micro-organisms found in the environment and in the microbiota. Genome mining has further led to the discovery of numerous uncharacterized ‘cryptic’ metabolic pathways in the classical producers of natural products such as Actinobacteria and fungi. These biosynthetic gene clusters may code for improved biologically active metabolites, but harnessing the full genetic potential has been hindered by the observation that many of the pathways are ‘silent’ under laboratory conditions. Here we provide an overview of the various biotechnological methodologies, which can be divided to pleiotropic, biosynthetic gene cluster specific, and targeted genome-wide approaches that have been developed for the awakening of microbial secondary metabolic pathways.",
"conclusion": "5 Concluding thoughts and future perspectives The abundance of silent and cryptic pathways in microbial genomes provides an exciting opportunity for the discovery of new chemical entities with high therapeutic potential. The diverse strategies developed for awakening these pathways offer an excellent starting point, but to date no single superior methodology has been developed. This is not unexpected, since the BGCs reside in vastly diverse micro-organisms and in many cases are strictly regulated to be activated only under highly specific conditions. The availability of the numerous methods described above, which all have their strengths and weaknesses, can therefore be considered as a significant advantage and it will remain the responsibility of the researcher to choose the most appropriate one for any given project."
} | 453 |
38904481 | PMC11231971 | pmc | 8,604 | {
"abstract": "Triboelectric nanogenerators (TENGs) have garnered substantial\nattention in breeze wind energy harvesting. However, how to improve\nthe output performance and reduce friction and wear remain challenging.\nTo this end, a blade-type triboelectric-electromagnetic hybrid generator\n(BT-TEHG) with a double frequency up-conversion (DFUC) mechanism is\nproposed. The DFUC mechanism enables the TENG to output a high-frequency\nresponse that is 15.9 to 300 times higher than the excitation frequency\nof 10 to 200 rpm. Coupled with the collisions between tribomaterials,\na higher surface charge density and better generating performance\nare achieved. The magnetization direction and dimensional parameters\nof the BT-TEHG were optimized, and its generating characteristics\nunder varying rotational speeds and electrical boundary conditions\nwere studied. At wind speeds of 2.2 and 10 m/s, the BT-TEHG can generate,\nrespectively, power of 1.30 and 19.01 mW. Further experimentation\ndemonstrates its capacity to charge capacitors, light up light emitting\ndiodes (LEDs), and power wireless temperature and humidity sensors.\nThe demonstrations show that the BT-TEHG has great potential applications\nin self-powered wireless sensor networks (WSNs) for environmental\nmonitoring of intelligent agriculture.",
"conclusion": "4 Conclusions In this study, a novel\nblade-type triboelectric-electromagnetic\nhybrid generator (BT-TEHG) has been proposed for effectively capturing\nthe breeze wind energy. A double frequency up-conversion (DFUC) mechanism\nwas constructed to realize a high-frequency output from the TENG under\nlow-frequency breeze wind excitation and ultimately augment the output\npower. Simulations and experiments were conducted to optimize the\nmagnetization direction in the EMG and the dimensional parameters\nof the TENG unit. The generating characteristics of the BT-TEHG under\nvarying rotational speeds and electrical boundary conditions were\nexplored. Results show that the response frequency of the TENG can\nbe amplified by 15.9 to 300 times when subjected to an excitation\nof rotational speed varying from 10 to 200 rpm. At wind speeds of\n2.2 and 10 m/s, the BT-TEHG can generate, respectively, power of 1.30\nand 19.01 mW when TENG and EMG are connected to the matched resistances.\nThe optimal rotational speed for the TENG is 160 rpm, corresponding\nto a wind speed of 7.84 m/s. It also demonstrates the capabilities\nto quickly charge capacitors and feed 248 LEDs in series and 496 LEDs\nin parallel at a wind speed of approximately 6 m/s, indicating the\nsuperiority of the hybrid energy harvesting mechanism. Through an\nenergy\nmanagement circuit, a self-powered wireless temperature and humidity\nmonitoring system was developed and powered by the BT-TEHG. In summary,\nthis work demonstrates the potential application of the BT-TEHG as\na distributed energy source for self-powered WSNs and provides promising\napplication prospects in intelligent agriculture.",
"introduction": "1 Introduction Wireless sensor networks\n(WSNs), comprising a multitude of processors,\nsensors, and radio nodes, promote the information and intelligent\nevolution of human society. This sophisticated network finds widespread\napplication across diverse domains, including industrial control,\nsmart home systems, consumer electronics, security equipment, logistics\ninfrastructure, intelligent agriculture, environmental perception,\nand health monitoring systems. However, the efficacy of battery-powered\nsensors within WSNs faces challenges due to inherent limitations in\ncommonly employed batteries, such as lithium-ion and fuel cells. These\nlimitations encompass constraints related to energy storage capacity,\nportability, and environmental impact, thereby posing significant\nhurdles to endurance, operational efficiency, and reliability. 1 − 4 Consequently, researchers are engaged in investigating methodologies\nto harness wind energy from the ambient environment with the goal\nof establishing self-powered WSNs. Traditional wind generators, characterized\nby their substantial weight and volume, remote installation, and elevated\nmanufacturing and installation costs, have constrained their widespread\nadoption in self-powered WSNs. Miniaturized wind generators, which\nrely on the principle of electromagnetic induction, represent a prevalent\nalternative. Despite their merits, including a simple structure, a\nhigh electromechanical coupling coefficient, and simple processing,\nthese generators exhibit suboptimal power generation efficiency under\nconditions of low wind speed. 5 , 6 Triboelectric\nnanogenerator (TENG) invented by Wang’s group\ncan adeptly convert mechanical energy into electrical energy through\nthe synergistic mechanisms of contact electrification and electrostatic\ninduction. Distinguished by its cost-effectiveness and broad material\nselectivity, the TENG exhibits notable advantages. Of particular significance\nis its high energy conversion efficiency, especially under low-frequency\nstimuli, which enables it being a promising and viable alternative\nfor powering WSNs by harvesting breeze wind energy. 7 Typical structures of TENG for wind energy harvesting can\nbe divided into two categories based on their generation mechanisms.\nThe first category is the rotation-sliding mode, such as rotation-disk, 8 − 11 rotation-cylinder, 12 , 13 and rotation-blade architectures. 14 − 16 To mitigate the challenges associated with rigid contact, namely,\nfriction and wear, scholars have proposed innovative designs featuring\nsoft contact. These include the revision of dielectric material arrangements 17 − 19 and the incorporation of unconventional tribomaterials such as rabbit\nhair, 20 , 21 Ag fiber cloth, 22 and cotton. 23 The second one is vertical\ncontact-separation (CS) mode. 24 − 26 To augment generation efficiency\nunder low wind speeds, scholars also explored novel structures tailored\nfor breeze wind energy harvesting. 27 − 30 Particularly, the integration\nof wind-induced vibration with TENG facilitates the contact and separation\nbetween tribomaterials and consequently lowers the start-up wind speed. 31 − 35 It is evident that the rigid contact in the rotation-sliding mode\ngives rise to non-negligible frictional resistance and wear. In contrast,\nthe challenge associated with soft contact lies in optimizing the\neffective contact area between tribomaterials, which determines the\nquantity of tribo-charge. The vertical CS mode ensures a substantial\neffective contact area, but the low operating frequency leads to large\ncapacity reactance, high cost of matched load resistance, but poor\ngenerating performance. Consequently, there exists a compelling need\nto develop efficient TENGs for capturing breeze wind energy. In this study, we present a novel blade-type triboelectric-electromagnetic\nhybrid generator (BT-TEHG) constructed from blade-type TENG units\nand a rotating disk electromagnetic generator (EMG). The novelty of\nour work lies in a double frequency up-conversion (DFUC) mechanism\nthrough the systematic arrangement of the TENG units and multiple\nplectrums. This configuration is conducive to heighten output power,\nparticularly at low wind speeds. The collision between two tribomaterials\namplifies the effective contact area, consequently elevating the charge\ndensity on the contact surfaces and mitigating frictional wear. To\nimprove output power, we conducted parameter optimizations of the\nTENG and EMG. The generating properties of the BT-TEHG were experimentally\nmeasured. At wind speeds of 2.2 and 10 m/s, the BT-TEHG can output\naverage power of 1.30 and 19.01 mW, respectively. Furthermore, it\nwas used to charge capacitors, light up light emitting diodes (LEDs),\nand power commercial wireless sensors through an energy management\nmodule. These demonstrations validate its potential utility as a distributed\npower source for WSNs.",
"discussion": "2 Results and Discussion 2.1 Structure Design and Working Principle The BT-TEHG with a DFUC mechanism is illustrated in Figure 1 . It comprises eight blade-type\nTENG units and a rotating disk EMG enclosed within a shell ( Figure 1 a). This enclosure\nprotects against environmental disturbances, such as humidity and\ndust, ensuring operational stability and compactness. Figure 1 b provides a detailed schematic\nof the blade-type TENG unit, employing poly(tetrafluoroethylene) (PTFE)\nand copper as the tribomaterials. The bottom blade features a copper\nfoil affixed to an acrylic substrate, acting as a stationary contact\nelectrode. Correspondingly, a copper foil positioned between a PTFE\nfilm and an acrylic substrate is named as a rotatable back electrode,\nconstituting the top blade with a poly(ethylene terephthalate) (PET)\nsheet together. Figure 1 c showcases a photograph of two blades of a TENG unit. The prototype\nof the proposed BT-TEHG is presented in Figure 1 d. In this device, the breeze wind energy\nis collected by an acrylic wind scoop and converted into rotational\nenergy of shaft, which in turn drives the plectrums to rotate. The\nrotating plectrums pluck the top blades and then release them, resulting\nin periodic contact and separation between top and bottom blades,\nthus generating an alternative current. Figure 1 (a) Explosion diagram\nof the blade-type triboelectric-electromagnetic\nhybrid generator (BT-TEHG). (b) Schematic of the blade-type triboelectric\nnanogenerator (TENG) unit. (c) Photograph of two blades of a TENG\nunit. (d) Prototype of the BT-TEHG. (e) Comparison of surface charge\ndensity between BT-TEHG and other works. Through the interplay between plectrums and TENG\nunits, a double\nfrequency up-conversion (DFUC) mechanism is introduced to achieve\nhigh-frequency responses under breeze wind excitation. In this context,\nfour plectrums are employed to modulate the plucking frequency, achieving\na quadrupled frequency relative to the rotating shaft, denoted as\nthe first frequency up-conversion (FUC). Upon the release of the separated\ntop blade, it collides with the bottom one at the natural frequency\nof the PET sheet, which surpasses the trigger frequency, termed the\nsecond FUC. This design yields numerous advantages: it facilitates\nan augmentation in the CS frequency between two triboelectric layers,\nthereby mitigating the capacitive reactance and matched resistance\nof the TENG, ultimately resulting in elevated output power. 36 The CS mode concurrently attenuates the friction\nresistance and wear between triboelectric layers, thereby prolonging\nthe operational lifespan of the system. Furthermore, the collision\npromotes an expansion of the effective contact area between two triboelectric\nlayers, engendering a higher surface charge density and enhancing\noverall output performance. 37 As shown\nin Figure 1 e, since\nthe TENG works on the principle of contact electrification and electrostatic\ninduction, the surface charge density on the triboelectric layer is\nan important indicator to measure the output performance of the generator.\nBenefiting from the novel design, a TENG unit generates a higher charge\ndensity of 71.57 μC/m 2 than other works in recent\nyears. 14 , 19 , 28 , 38 − 40 In addition, previous studies\nhave demonstrated that the TENG generates higher energy conversion\nefficiency than EMG under breeze wind stimuli. In contrast, the EMG\ndominates power generation at high-speed wind condition. 41 Accordingly, the BT-TEHG takes advantage of\nthe complementary nature of TENG and EMG under different wind speeds,\nachieving synergistically efficient power generation within a wide\nrange of wind speeds. The electricity generation process of\nthe BT-TEHG can be divided\ninto the TENG and EMG parts. Illustrated in Figure 2 a, the operating principle of TENG is based\non triboelectric and electrostatic induction effects. In the initial\nstate ( Figure 2 a(I)),\nelectrons undergo transfer from the copper foil of the contact electrode\nto the PTFE film due to contact electrification, instigating net negative\ncharges at the PTFE film surface and an equivalent number of positive\ncharges on the copper foil surface. 42 Notably,\nan absence of an electric potential difference between back and contact\nelectrodes characterizes this stage. Upon the plucking of the top\nblade by the plectrum, the PTFE film undergoes separation from the\ncontact electrode ( Figure 2 a(II)). This separation prompts a potential difference between\nthe two electrodes, forming an instantaneous current as electrons\nmigrate from the back electrode to the contact electrode. At the maximum\nseparation angle, all positive charges aggregate on the back electrode\n( Figure 2 a(III)). Subsequently,\nas the plectrum continues its rotation, the top blade is released.\nDuring the reduction of the separation distance, electrons transition\nfrom the contact electrode back to the back electrode, producing a\nreverse instantaneous current ( Figure 2 a(IV)). Upon recontact between the PTFE film and the\ncontact electrode, all induced charges are neutralized ( Figure 2 a(I)). Due to the continuous\nrotation of plectrums, a periodic alternating current is engendered\nfrom the TENG units in this cyclical process. To acquire a more intuitive\ncomprehension of the potential distribution between two triboelectric\nlayers, a finite element simulation is conducted utilizing COMSOL\nMultiphysics 6.0. The outcomes depicted in Figure 2 b align with the stages in Figure 2 a. Obviously, a discernible\ncorrelation exists between the separation angle and the electric potential\ndifference. Figure 2 Working principle of the BT-TEHG: (a) Working principle of the\nTENG. (b) Surface potential distribution of a TENG unit during contact-separation\noperation. (c) Working principle of the electromagnetic generator\n(EMG). Figure 2 c describes\nthe operational principle of the EMG based on electromagnetic induction.\nIn the initial state ( Figure 2 c(I)), the magnet within the rotor disk is aligned with the\ninduction coil, maintaining a constant magnetic flux within the coil,\nwhich outputs no current. Upon the rotation of the rotor disk, the\nmagnet’s position relative to the induction coil varies, inducing\na flux change. Consequently, a current is generated in the coil, according\nto Lenz’s law ( Figure 2 c(II)). As the rotor disk persists in its rotation, the magnetic\nflux traverses the coil from the opposing direction, leading to an\ninduced current in the reverse direction ( Figure 2 c(III)). This cyclic process realizes a complete\ncycle of alternating current within the EMG. 2.2 Modeling In order to describe theoretically\nthe electrical characteristics of the TENG part, a nonparallel plate\ncapacitor model filled with air and PTFE film is established based\non the specific configuration, as depicted in Figure 3 a. The thickness and dielectric constant\nof PTFE are d 1 and ε 1 , respectively. The separation angle, amount of transferred\ncharge, and potential difference between two electrodes are defined\nas θ , Q , and V , respectively. The surface tribo-charge density of the PTFE film\nis σ . Since the contact area ( S ) between two blades is much larger than their separation distance\nin the experiments, an approximate analytical V – Q – θ relationship can be derived\nby neglecting the edge effect. Under ideal conditions, the nonparallel\nplate capacitor can be divided into multiple parallel plate microcapacitors\nconnected in parallel. Known that the capacitance of a parallel plate\ncapacitor with air as dielectric is C = ε 0 S / d , where d is the separation distance between two plates. Accordingly,\nthe capacitance of the blade-type TENG can be calculated as 1 where W and\n( R 2 – R 1 ) are the width and length of the contact area. Figure 3 (a) Nonparallel plate\ncapacitor model for TENG. (b) Dimensional\nparameters of the top blade. According to Gauss’s theorem, the electric\nfield strength\nof a microcapacitor in each medium at the position of r is expressed as follows 2 3 where Q 0 is the\namount of triboelectric charge. The voltage between the two electrodes\nis V ( r ) = E PTFE d 1 + E air rθ . Since there is no charge transfer\n( Q = 0) in the open-circuit (OC) state, the OC voltage\nis given by 4 Since the total tribo-charge is unchanged,\nwhich is expressed as 5 Substituting eq 4 into eq 5 , the OC voltage of TENG is obtained as 6 According to the electrical\npotential superposition principle, the V – Q – θ relationship can be given\nby 7 Under short-circuit (SC) condition,\nthe transferred charges and SC current can be derived as 8 9 Since the TENG can be simplified\nto be a serial connection of an ideal voltage source and a capacitor,\nits average impedance is approximately X g = 1/( ωC avg ), where ω is the angular frequency of the signal source and C avg is the average inherent capacitance. When it is externally\nconnected with a load resistance R L , the\npower delivered to the load is calculated as 43 10 Therefore, the optimal power can be derived\nby letting 11 Obviously, the matched resistance for the\noptimal power depends on the average impedance X g . Furthermore, all approaches that can lower the average impedance X g are conducive to reducing the matched resistance\nand increasing the output power of the TENG, including expanding contact\narea, amplifying response frequency, and parallel connection of multiple\nTENGs. 36 These equations theoretically\ndemonstrate the advantages and benefits of the proposed BT-TEHG. The detailed schematic diagram of the top blade is shown in Figure 3 b. Herein, the PET\nsheet acts as a cantilever beam with length ( L p ), width ( W p ), and height ( H p ), while the acrylic substrate possesses dimensions\nof length ( L a ), width ( W a ), and height ( H a ). As mentioned\nabove, the top blade will impact the bottom one during the DFUC mechanism,\nwhich in turn generates a contact force. It is worth noting that the\ncontact force introduces surface deformation and changes the microscopic\ncontact area and interfacial dangling bonds of the tribomaterials,\nultimately improving the surface charge density and the amount of\ntransferred charge. Therefore, the contact force is a key factor to\naffect the triboelectric behavior and electrical output, whose maximum\nis described as 44 12 where v and M are impact velocity and equivalent mass of the top blade, respectively,\nand K p is the equivalent stiffness of\nthe cantilever beam, which can be expressed as 45 13 where E p and I p are the Young’s modulus and moment\nof inertia of the PET sheet, respectively. It predicts that the maximum\ncontact force is proportional to v , K p , and M . To analyze the dynamic\nresponses of the top blade, it is simplified\nto be a lumped-parameter model, as illustrated in Figure S1 . At the initial stage ( Figure S1a ), the top blade is plucked away from the static position\nand moves along with the plectrum until their overlap length is reduced\nto zero, in which case the plectrum rotates at a constant angular\nvelocity ω . Within the contact time t , the moving speed is ż = L 1 ω cos( ωt ), 0 ≤ t ≤ t f , where L 1 is the\nrotational radius of plectrum and t f is\nthe separation time. When they get separated, the blade undergoes\nfree vibration ( Figure S1b ) and the governing\nequation can be expressed as 14 where F e , C p , and z are the electrostatic\nforce, mechanical damping coefficient, and the relative displacement\nof the top blade tip relative to the initial static position, respectively. When the top blade collides with the bottom one or plectrum, its\nmoving speed changes at the moment of impact, which can be described\nby ż + = − e i ż – ( i = 1, 2), where e 1 and e 2 are the coefficients of restitution\nbetween two blades and between the top blade and plectrum, respectively,\nand ż + and ż – denote the moving speeds of the top blade just\nafter and before impact, respectively. At the moment that the top\nblade is just in contact with plectrum, a contact force F = Mz̈ + C p ż + K p z + F e ( z > 0) is\nexerted\non the blade. When the value is zero, it indicates the state of separation. With respect to the EMG part, the OC voltage ( V OC-EMG ) and SC current ( I SC-EMG ) can be expressed as 15 16 where N is the number of\nturns of the induction coil, Φ is the magnetic flux, and R coil is the internal resistance of the coil.\nHence, the performance of EMG is determined by the rate of change\nof magnetic flux, N , and R coil . 2.3 Performance To investigate the influences\nof dimensions on output characteristics and determine optimal parameters\nof the top blade, 7 groups of TENGs consisting of different sizes\nof the PET sheet and acrylic substrate are produced. The detailed\ndimension parameters are given in Table S1 . It should be noted that only the length ( L p ) and width ( W p ) of the cantilever\nbeam and the height ( H a ) of the acrylic\nsubstrate are variable, so as to simplify the optimization procedure.\nThe shaft of the BT-TEHG is connected to a DC motor, offering tunable\nrotational speeds. The motor test platform is depicted in Figure S2 . Under a constant rotational speed\nof 100 rpm, the output SC current ( I SC ) and transferred charge ( Q SC ) were measured\nand are compared in Figure 4 . For Groups 1, 2, and 3 with constant W p and H a , the root-mean-square\n(RMS) current ( I rms ) and transferred charge\nquantity exhibit a gradual increase with the augmentation of L p . This behavior is attributed to the elongation\nof the cantilever beam, leading to an expanded overlap length for\nthe plectrum and acrylic substrate, thereby increasing the separation\ndistance between the top and bottom blades. Consequently, more elastic\npotential energy is accumulated in the cantilever beam and ultimately\nconverted into more electrical energy. Nonetheless, an overextended\ncantilever beam will pose an obstacle to the rotation of plectrum,\nculminating in the failure of the DFUC mechanism. When maintaining\nconstant values for L p and W p , an augmented height H a of\nthe acrylic substrate results in amplified M and\nenhanced contact force between the two blades, as described in eq 12 . This significantly\nexpands the effective contact area between two triboelectric layers,\nintroducing an ascending surface charge density and improved generating\nperformance, as evidenced in the outcomes of Groups 3, 4, and 5. Although\nthere is no W p in the expressions of I SC and Q SC , the\ncomparative analysis among Groups 3, 6, and 7 reveals that a broader\ncantilever beam introduces greater stiffness and more elastic potential\nenergy at the same separation distance, consequently achieving more\nelectrical output. Experimental findings demonstrate that at dimensions L p = 5 mm, W p = 50\nmm, and H a = 3 mm, the maximum I rms attains 4.93 μA, accompanied by a\ntransferred charge quantity of 98.02 nC. Hence, the size parameters\nof Group 3 are designated as the benchmark for TENG. Figure 4 Output performance of\n7 groups of TENGs with different parameters\nat a rotational speed of 100 rpm. (a) Short-circuit (SC) current I SC and its root-mean-square (RMS) value I rms . (b) SC transferred charge Q SC and transferred charge quantity. In order to assess the impact of the DFUC mechanism\non the dynamic\nresponse and output performance, the electrical characteristics of\na TENG unit are measured and compared at different rotational speeds. Figure 5 a presents the interaction\nbetween the rotational plectrum and top blade, realizing four pluckings\nwithin one cycle of rotating shaft. This process is named as the first\nFUC. Prior to the release of the top blade, elastic potential energy\nis stored in the cantilever beam, which is then converted into kinetic\nenergy upon release. Figure 5 b depicts the displacement responses ( z )\nof the top blade when it is plucked twice at different rotational\nspeeds. For the speeds of 10 and 100 rpm, two distinct CS processes\ncan be observed after each plucking, revealing two primary inelastic\ncollisions. Despite part of kinetic energy dissipated in the forms\nof acoustic, thermal, potential energy, etc. during the first collision,\nthe residual is still sufficient to overcome the effect of electrostatic\nadsorption force and cause the top blade to rebound. More importantly,\nthe vibration frequency after being released is much higher than the\nplucking frequency, achieving high-frequency response under low-frequency\nexcitation, denoted as the second FUC. However, when the speed reaches\n200 rpm, the speed is so fast that there is no time for a secondary\ncollision. Correspondingly, the OC voltage ( V OC ), SC current ( I SC ), and transferred\ncharge ( Q SC ) at a rotational speed of\n10 rpm are shown in Figure 5 c. During the contact state, no charge transfer occurs under\nSC conditions, and there is an absence of potential difference between\nthe two electrodes. Once the plucking happens, the separation of two\ntriboelectric layers ensues, promoting a substantial charge transfer\nand forming an instantaneous current. This process exhibits a rapid\nsaturation trend. Notably, the output characteristics manifest a high\nsensitivity to the initial separation distance. The transferred charge\nquantity is 161.03 nC, yielding a charge density of 71.57 μC/m 2 . The peak–peak values of V OC and I SC are 1.36 kV and 69.00 μA,\nrespectively. Although the trend of the V OC waveform resembles that of I SC , it deviates\nfrom the trend outlined in eq 6 . This difference may be caused by the smaller input impedance\nof the oscilloscope than the internal impedance of TENG, resulting\nin charge flowing through the instrument. Therefore, the measured V OC represents the voltage across divider resistor\nin the oscilloscope. 46 After the top blade\nis released, the FUC phenomenon also appears in the electrical signal.\nTaking the waveform of I SC as an example,\nthe time interval between two peaks A and B is 0.02 s. That is to\nsay, the vibration frequency of the top blade is approximately 50\nHz, a frequency 300 times greater than the rotational frequency of\nthe shaft (0.17 Hz). As a result, the heightened CS frequency diminishes\nthe impedance of TENG, thereby amplifying the output power, which\nis a notable advantage conferred by the DFUC mechanism. 36 At last, the top blade is again in contact with\nthe contact electrode as the kinetic energy disappears. Figure 5 (a) Interaction\nbetween the rotational plectrum and the top blade\nrealizes the first FUC. (b) Displacement responses of the top blade\nwhen plucked twice at different rotational speeds. (c) Electrical\ncharacteristics of a TENG unit at a rotational speed of 10 rpm. (d)\nComparison of electrical characteristics at different rotational speeds. Figure 5 d provides\nthe electrical characteristics at varying rotational speeds. Obviously,\nthe RMS value I rms exhibits a gradual\nincrement alongside a concurrent reduction in the transferred charge\nquantity with escalating rotational speed. In contrast, the RMS voltage V rms experiences an initial increase, followed\nby a subsequent decline. Such divergent trends can be attributed to\nthe following reasons. As shown in Figure 5 c, it takes about 0.761 s for charge transfer\nto reach the saturation state. With the increase in the rotational\nspeed, the time devoted to charge transfer within each plucking period\ngradually decreases, resulting in a gradual reduction in the transferred\ncharge quantity. However, I rms represents\nthe amount of transferred charge in unit time. Since the induced charge\nis highly sensitive to the initial separation distance, the transfer\ntime for 80% charges is only 0.077 s, close to the plucking cycle\n(0.075 s) at 200 rpm. Therefore, I rms shows\nan upward trend with the enhancement of the plucking frequency or\nrotational speed. As expressed in eq 7 , the voltage is related to the amounts of tribo-charge Q 0 and transferred charge Q .\nWithin the rotational speed range of 10 to 100 rpm, charge transfer\nplays the predominant role. However, at the speed of 200 rpm, there\nis only one impact caused by one plucking and its contribution to\ntriboelectricity is weakened. The inadequate contact between two triboelectric\nlayers leads to a pronounced reduction in tribo-charge and an ultimately\ndeclining trend in V rms . Most importantly,\nat this rotational speed, the vibrational frequency of the top blade\npersists at approximately 15.9 times the rotational frequency of the\nshaft, thereby affirming the validity and feasibility of the DFUC\nmechanism. To assess the effect of magnetization direction of\nadjacent magnets\non the output performance of the EMG, we investigated the V rms across the induction coil under two conditions:\nthe same magnetization direction (N–N) and different magnetization\ndirections (N–S) ( Figure S3 ). The\npower generation characteristics at varying rotational speeds were\nsimulated and measured, as presented in Figure 6 a,b. The investigation disclosed a progressive\nincrease in V rms with the rotational speed,\nregardless of the magnetization direction. However, the values under\nthe same magnetization direction consistently surpassed that under\ndifferent magnetization directions. Consequently, for subsequent experiments,\nmagnets were arranged in the same magnetization direction. Furthermore,\nthe output power across load resistances connected to the induction\ncoil was measured at different rotational speeds, as illustrated in Figure 6 c. The findings indicated\na gradual rise in matched load resistance from 540 to 620 Ω\nas the rotational speed ranged from 40 to 200 rpm. This value was\ncomparable to the internal resistance of the coils in series (578\nΩ). The observed change was attributed to the variation of the\ninductive reactance, which is proportionate to the rotational speed. Figure 6 (a) Simulated V rms from the EMG at\ndifferent rotational speeds. (b) Measured V rms from the EMG at different rotational speeds. (c) Output power of\nEMG with different load resistances. (d–f) V OC and V rms from different\nTENG units in parallel connection at different rotational speeds.\n(g–i) I SC and I rms from different TENG units in parallel connection at\ndifferent rotational speeds. Figure 6 d–i\npresent the rectified V OC and I SC for 2, 4, and 8 TENG units in parallel connection\nat different rotational speeds. All units are distributed evenly around\nthe circumference. Consistent trends are observed in both the voltage\nand current. As the rotational speed increases from 40 to 200 rpm, I rms rises monotonically and levels off at 200\nrpm. In contrast, V rms initially ascends\nand subsequently descends. This phenomenon may be attributed to the\nescalating CS frequency with an increase in rotational speed, resulting\nin a general rise in I rms . However, when\nthe plucking frequency of the plectrum is high enough, there is insufficient\ntime for adequate contact. This leads to a reduction in the effective\ncontact area and tribo-charges, causing a downward trend in V rms and saturation in I rms . It also reveals that more units in parallel connection\ncorrespond to increased V rms and I rms at the same rotational speed, indicating\nenhanced generating performance. Nonetheless, it is imperative to\nconsider the limitations of the shell’s capacity and the operational\nspace of the TENG. Within these constraints, this design can accommodate\nup to 8 TENG units. The output characteristics of the standalone\nTENG and the BT-TEHG\nare critical to evaluate their generating performance. First, we investigated\nthe output power from 2, 4, and 8 TENG units connected to varying\nload resistances under diverse rotational speeds ( Figure 7 a–c). Irrespective of\nthe specific quantity of TENG units, the output power exhibits a rising\ntrend followed by a decline with increasing load resistance, reaching\nits maximum when connected to the matched resistance. Concurrently,\nas the rotational speed increases, the maximum output power initially\nascends and subsequently descends, but the corresponding matched resistance\ngradually decreases. The fluctuations observed in output power are\ncaused by changes in tribo-charges and transferred charge quantities\nat distinct rotational speeds. The reduction of matched resistance\nis induced by the decrease of capacitive reactance that is inversely\nproportional to the rotational speed, as described in eq 11 . The maximum output powers yielded\nby 2, 4, and 8 TENG units amount to 1.10, 2.40, and 3.66 mW, respectively.\nThe associated matched resistances stand at 100, 20, and 16 MΩ.\nThroughout the experiments, the optimal rotational speed consistently\nremained at 160 rpm. By comparing the matched resistances of the three\ngroups, it can be found that the greater the number of TENG units\nconnected in parallel, the smaller the matched resistance. This is\nbecause the TENG can be considered as a capacitor, where parallel\nconnections serve to mitigate capacitive reactance, subsequently leading\nto a decrease in matched resistance. In short, augmenting the rotational\nspeed and implementing parallel connections are effective strategies\nfor reducing matched resistance. When connected to the matched resistances,\nthe output power of 2, 4, and 8 standalone TENG units is compared\nwith that of the standalone EMG, as shown in Figure S4 . Figure 7 (a–c) Output power of 2, 4, and 8 TENG units with different\nload resistances at distinct rotational speeds. (d) Output power of\nthe BT-TEHG, TENG part, and EMG part and the corresponding rotational\nspeeds at different wind speeds. Figure 7 d shows\nthe output power of the BT-TEHG connected to the matched resistances\nand the corresponding rotational speed in the actual wind field. The\ncorresponding output power from TENG and EMG parts are also presented.\nObviously, output powers steadily increase with the increase of wind\nspeed but with different slopes. As the wind speed rises, the contribution\nof TENG to the total output power gradually decreases, and EMG gradually\ntakes the upper hand. At a wind speed of 2.2 m/s (about 33 rpm), the\nBT-TEHG can steadily generate power of 1.30 mW and the value reaches\nto 19.01 mW at 10 m/s (about 210 rpm), indicating better output performance\ncompared to standalone TENG and EMG. Correspondingly, the maximum\nenergy conversion efficiency of TENG part is 23.01% at a wind speed\nof 2.2 m/s, while that of the EMG part is 3.80% at 10 m/s. The detailed\nvalues are listed in Table S2 , and the\ncalculation is shown in Note S1 . The results\nindicate that the efficiency of the plucked TENG is superior to that\nof previous TENGs using solid–solid CS mode, 47 − 50 demonstrating the advantage of\nDFUC mechanism in improving energy conversion efficiency. The relationship\nbetween wind speed and rotational speed is linearly fitted as Y = 22.25 X – 14.45, where X is the wind speed and Y is the rotational\nspeed. Accordingly, the wind speed corresponding to the optimal rotational\nspeed of the TENG (160 rpm) is about 7.84 m/s. Moreover, the relationship\nbetween output power and rotational speed is close to the data measured\nin the motor test platform, demonstrating the robustness of the generating\nperformance of the BT-TEHG."
} | 8,798 |
38485920 | PMC10940629 | pmc | 8,605 | {
"abstract": "Contractile rings are formed from cytoskeletal filaments during cell division. Ring formation is induced by specific crosslinkers, while contraction is typically associated with motor protein activity. Here, we engineer DNA nanotubes and peptide-functionalized starPEG constructs as synthetic crosslinkers to mimic this process. The crosslinker induces bundling of ten to hundred DNA nanotubes into closed micron-scale rings in a one-pot self-assembly process yielding several thousand rings per microliter. Molecular dynamics simulations reproduce the detailed architectural properties of the DNA rings observed in electron microscopy. Theory and simulations predict DNA ring contraction – without motor proteins – providing mechanistic insights into the parameter space relevant for efficient nanotube sliding. In agreement between simulation and experiment, we obtain ring contraction to less than half of the initial ring diameter. DNA-based contractile rings hold promise for an artificial division machinery or contractile muscle-like materials.",
"introduction": "Introduction Cell division is a hallmark of life. After duplication and separation of the genetic information, the cellular compartment has to be divided to give rise to two daughter cells. Nature’s solution for compartment division is the formation of contractile rings made from cytoskeletal filaments 1 . In eukaryotic cells, micron-scale actomyosin rings assemble at the cell’s equatorial plane at the end of mitosis and meiosis 2 . The ring contraction can be powered by two distinct mechanisms, namely molecular motor activity 3 and diffusable actin crosslinkers 4 . Whereas the first is an active, energy-consuming process, the latter mechanism mediates passive entropy driven contraction. Recently, it has been shown that passive filament crosslinkers can generate filament sliding and contractile forces, that are sufficient to antagonize motor protein action in microtubule 5 and actin networks 4 . Bottom-up synthetic biology has pursued the long-term goal to reconstitute a minimal division machinery inside lipid vesicles to establish fundamental physical principles in isolation of the complex environment of a cell and to eventually engineer a self-replicating cellular system from scratch. Towards this goal, actomyosin rings have been formed in vitro 6 and their contraction has been demonstrated in confinement 7 . This led to membrane deformations in lipid vesicles 8 . Nevertheless, the reconstitution of a division machinery that can complete the division of lipid vesicles remains an open challenge in bottom-up synthetic biology 9 , 10 . It is important to critically ask which physical features are required for a minimal division system and how they contribute to the contraction process. Towards this end, it would be ground breaking to establish an entirely synthetic division machinery, which does not rely on nature’s building blocks. A fully engineered contractile ring could yield mechanistic insights into the biophysics of the process towards an alternative set of molecular components for the division of synthetic cells. There are ongoing efforts in the field of DNA nanotechnology to recreate functional mimics of cytoskeletal elements from DNA. Of particular interest are DNA nanotubes 11 , which have been equipped with different features that mimic functions of a cytoskeleton, such as reversible assembly 12 , 13 , directional growth 14 , signaling 15 , and transport with engineered molecular motors 16 or enzyme activity 17 . However, a key functionality remains unachieved, namely the formation of contractile DNA rings. Closed DNA rings have been assembled on the nanoscale 18 , 19 . They have been used to template liposomes 20 or gold nanoparticles 21 , to engineer liposome fusion and lipid transfer 22 or mechanically interlocked molecules 23 and they have been used as large-diameter membrane pores 24 . However, these nanoscale DNA rings are one to two orders of magnitude too small to span the circumference of synthetic cellular compartments and beyond that, no mechanism for their contraction has been proposed or experimentally realized. Therefore, our aim is the self-assembly of DNA-based contractile rings on the micron scale. We reason that the assembly and contraction can take inspiration from the mechanisms at play for nature’s cytoskeletons. While ring assembly clearly requires crosslinkers, contraction could either be achieved with suitable molecular motors or only by passive crosslinkers, as it has been discovered recently 4 , 25 . Assuming that the passive crosslinker approach is more straight forward to adapt to DNA nanotubes, we require a synthetic crosslinker for DNA nanotubes. Multivalent positively charged peptides have been shown to crosslink microtubules 26 . Here, we revert to the DNA nanotubes as an entirely synthetic system that is well established in the bottom-up synthetic biology community as an alternative route to the reconstitution of proteins. The reconstruction of a protein-like machinery from a different material is not only exciting in itself, but it may also provide a shortcut towards a truly self-replicating system since DNA replication requires fewer components than the replication of proteins. In the long term, one could thus envision a synthetic cell that operates outside of the central dogma of molecular biology.",
"discussion": "Results and discussion We show that we can bundle DNA nanotubes and achieve the one-pot self-assembly of closed micron-scale DNA rings upon addition of such multivalent positively charged peptides. We control the DNA bundle thickness as well as the ring diameter. With theory and molecular dynamics simulations we gain mechanistic insights into the formation of DNA nanotube rings and the architecture of its contraction mechanism. We translate the simulation parameters of interest into physical properties of our system and realize the predicted conditions experimentally. Thereby, we achieve the contraction of the DNA rings to less than 45 percent of their initial diameter. We relate this to the theory and adapt it to the particularities of the physical system so that we can reduce the entangled relationships of the experiment to quantitative parameters. Synthetic peptides as crosslinkers for DNA nanotubes We first assemble DNA nanotubes from the well-established double-crossover DNA tile design, whereby each tile consists of five DNA oligomers (Fig. 1 a, Supplementary Table 1) 11 . Due to their sticky-end overhangs and their intrinsic curvature, these tiles assemble into hollow nanotubes consisting of 8 to 20 DNA duplexes (4 to 10 tiles) 11 , resulting in an experimentally determined diameter of 11.8 ± 2.1 nm. To form bundles and contractile rings from these DNA nanotubes, we need a synthetic crosslinker that satisfies the following two requirements imposed by the nature of the DNA nanotubes: First, a crosslinker that binds to DNA nanotubes by electrostatic interactions has to be positively charged because of the negatively charged backbone of the DNA. Second, it needs to act as a multivalent crosslinker that can connect multiple DNA nanotubes. Thus, we can make use of a multivalent positively charged peptide construct, which has been shown to electrostatically crosslink microtubules 26 . Fig. 1 StarPEG-(KA7) 4 bundles DNA nanotubes. a Schematic illustration of DNA nanotubes formed from double-crossover DNA tiles 11 . b Schematic illustration of tetravalent starPEG-(xA7) 4 composed of four branches of 7 lysine- or aspartate-alanine repeats. c Schematic illustration of DNA nanotubes in the absence and presence of different synthetic peptide constructs. d Confocal images of DNA nanotubes (30 nM DNA tiles, labeled with Atto633, λ e x = 640 nm) without any peptide; with 2 μM positively charged monovalent KA7-peptide; with 500 nM negatively charged tetravalent starPEG-(DA7) 4 composed of four branches of 7 aspartate-alanine repeats and with 500 nM positively charged tetravalent starPEG-(KA7) 4 composed of four branches of 7 lysine-alanine repeats (from left to right). Scale bar: 10 μm. The construct consists of a four-arm 10 kDa starPEG backbone which is coupled to seven lysine-alanine amino acid repeats on each of the four arms (Fig. 1 b). We will refer to it as starPEG-(KA7) 4 . The end-to-end distance R 0 of the polymer (i.e. a long chain with no significant hindrances to backbone rotation, as is the case for PEG) can be derived from treating the polymer as a self-avoiding freely jointed chain with R 0 = b ⋅ N (3/5) with b , the Kuhn monomer (1.1 nm for PEG), and N , the number of Kuhn monomers in the chain (18.25 for 2.5 kDa PEG arm) 27 . According to this, each of the four arms has an apparent length of about 6.3 nm. By applying the formula R m i n = 0.066 M 1/3 for M in Da 28 , one can estimate the minimal radius for alanine (89 Da) being 0.29 nm, and for lysine (146 Da) being 0.35 nm (PubChem release 2021.10.14). Thus, the peptide chain measures 4.5 nm. In a fully extended conformation, (for PEG treated as a freely jointed chain and the peptide as a linear chain) the starPEG construct would have a max. length of approximately 21.6 nm – long enough to compliantly connect two DNA nanotubes with a diameter of ~12 nm each. Hence, the construct exhibits four flexible arms with positively charged amino acid repeats that can bind to and crosslink the negatively charged DNA nanotubes by electrostatic interactions. As a control, we additionally synthesize monovalent KA7 peptides, which are not expected to crosslink DNA nanotubes, as well as a construct that features seven negatively charged aspartate-alanine repeats on a tetravalent starPEG backbone (starPEG-(DA7) 4 ) which is not expected to bind to the negatively charged DNA backbone. All constructs are further labeled with 5-TAMRA ( λ e x = 561 nm) to allow for their visualization with confocal microscopy. To test whether starPEG-(KA7) 4 indeed can bundle DNA nanotubes, we mix fluorescently-labeled DNA nanotubes with starPEG-(KA7) 4 or the respective control constructs and image them by confocal microscopy (Fig. 1 c, d). In the presence of monovalent KA7 peptides, single DNA nanotubes remain homogeneously distributed across the observation chamber similar to DNA nanotubes in absence of any peptide. Weak DNA nanotube bundling can be observed in the presence of the negatively charged starPEG-(DA7) 4 (Fig. 1 d), likely due to the positively charged magnesium-ions in solution. Notably, in the presence of starPEG-(KA7) 4 , the DNA nanotubes form DNA nanotube bundles with much higher fluorescence intensity extending to a length of several tens of micrometer (Fig. 1 d, right image). The extent of bundling correlates with the concentration of starPEG-(KA7) 4 relative to the concentration of DNA tiles (Supplementary Fig. 1) , forming a hybrid material with engineerable properties. The bundles curve into closed rings. To tune ring size and accomplish contraction, we first need to understand the action of the peptide crosslinker on DNA nanotubes. In the following sections, bundle formation is discussed in particular. The properties of ring formation and ring structures are then considered. Characterization of DNA nanotube crosslinking with synthetic multivalent peptides In particular, we need to establish whether the DNA nanotube bundling is caused by the depletion effect only, whereby starPEG-(KA7) 4 acts as a molecular crowder, or whether the peptide can actually crosslink DNA nanotubes. We therefore immobilize DNA nanotubes on the surface of the observation chamber and preload them with starPEG-(KA7) 4 . We then wash out excess starPEG-(KA7) 4 from the solution when adding a second type of DNA nanotubes, which we labeled with a different fluorophore in order to distinguish them from the immobilized DNA nanotubes (immobilized DNA nanotubes A: green; free DNA nanotubes B: yellow, Fig. 2 a). After another washing step, only starPEG-(KA7) 4 that is bound to the DNA nanotube A is present and responsible for the colocalization of the two DNA nanotube types (Fig. 2 b). Thereby we conclude that the binding must be induced by crosslinking and not by the depletion effect only. Fig. 2 DNA nanotube colocalization and bundling by starPEG-(KA7) 4 . a Schematic illustration of the DNA nanotube colocalization assay with a biotinylated DNA nanotube A (green) bound to biotinylated bovine serum albumin-coated glass slides upon addition of 0.2 wt% neutravidin. StarPEG-(KA7) 4 crosslinks DNA nanotube B (yellow) to the immobilized nanotube A. b Composite confocal image of the immobilized DNA nanotube A (green, 6-FAM, λ e x = 488 nm) and DNA nanotube B (yellow, Atto633-labeled, λ e x = 640 nm) illustrating their colocalization. Scale bar: 10 μm. c Confocal images of DNA nanotubes (30 nM DNA tiles, yellow, Atto633-labeled, λ e x = 640 nm) and 200 nM starPEG-(KA7) 4 (cyan, TAMRA-labeled, λ e x = 561 nm). Scale bar: 10 μm. d Colocalization intensity of starPEG-(KA7) 4 (TAMRA-labeled, λ e x = 561 nm) with DNA nanotubes (30 nM DNA tiles) (Atto633-labeled, λ e x = 640 nm, Mean ± SD, n = 10 DNA nanotube bundles analyzed per condition, exponential fit plotted as red line, \\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}$$y={y}_{0}+(P-{y}_{0})(1-\\exp (-kx)),\\,{y}_{0}=0.85,\\,P=53.49,\\,k=0.018$$\\end{document} y = y 0 + ( P − y 0 ) ( 1 − exp ( − k x ) ) , y 0 = 0.85 , P = 53.49 , k = 0.018 ). e TEM micrographs of DNA nanotubes (30 nM DNA tiles) in the absence of starPEG-(KA7) 4 (left) and in presence of 200 nM starPEG-(KA7) 4 (right). Scale bar: 200 nm. f DNA nanotube bundle thickness over starPEG-(KA7) 4 concentration (Mean ± SD, n = 100 DNA nanotube bundles analyzed per condition, 30 nM DNA tiles, exponential fit plotted as red line, \\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}$$y={y}_{0}+(P-{y}_{0})(1-\\exp (-kx)),\\,{y}_{0}=11.79,P=138.80,\\,k=0.017$$\\end{document} y = y 0 + ( P − y 0 ) ( 1 − exp ( − k x ) ) , y 0 = 11.79 , P = 138.80 , k = 0.017 ). Source data are provided as a Source Data file. To assess how much starPEG-(KA7) 4 can bind to DNA nanotubes, we label the synthetic starPEG peptides with fluorescent 5-TAMRA and analyze its colocalization with the DNA nanotubes (Fig. 2 c), i.e. the fluorescence of starPEG-(KA7) 4 colocalizing with the DNA nanotubes (see ‘Colocalization Assay’ in the “Methods” section and Supplementary Fig. 2) . The colocalization intensity saturates at around 200 nM starPEG-(KA7) 4 for DNA nanotubes assembled from 30 nM DNA tiles (Fig. 2 d and Supplementary Fig. 3) . The concentration at which binding saturates agrees with previous results for the microtubule binding of starPEG-(KA7) 4 26 . Additionally, we analyze the DNA bundle thickness with negative stain transmission electron microscopy (TEM) for varying starPEG-(KA7) 4 concentrations. Electron micrographs reveal the transition from single DNA nanotubes (single (11.8 ± 2.1) nm or weakly bundled (15.7 ± 5.3) nm) to DNA bundles consisting of tens of DNA nanotubes in the presence of starPEG-(KA7) 4 (Fig. 2 e and Supplementary Fig. 4) . Concomitantly, the bundle thickness increases by one order of magnitude from single DNA nanotubes with an apparent cross section of (15 ± 5) nm to bundles with a cross section of (145 ± 67) nm for 0 and 500 nM starPEG-(KA7) 4 , respectively. In accordance with the colocalization intensity, the bundle thickness also does not increase further than at 200 nm starPEG-(KA7) 4 indicating a maximal occupancy of starPEG-(KA7) 4 on the DNA nanotubes (Fig. 2 f). We expect both the association of starPEG-(KA7) 4 and the association of DNA nanotubes to bundles, to follow association binding kinetic models. Therefore, it can be fitted with an equation of the form: \\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}$$y={y}_{0}+(P-{y}_{0})(1-\\exp (-kx))$$\\end{document} y = y 0 + ( P − y 0 ) ( 1 − exp ( − k x ) ) with x being the starPEG-(KA7) 4 concentration. StarPEG-(KA7) 4 binding to DNA nanotubes saturates within minutes (Supplementary Fig. 5) . Self-assembly of micron-scale DNA rings To study the ability of starPEG-(KA7) 4 to promote the self-assembly of DNA nanotubes into higher-order structures, we observe DNA nanotubes without immobilization in an unconstrained 3D environment (Supplementary Movie 1) . We find that starPEG-(KA7) 4 promotes the efficient formation of DNA nanotube rings with several micrometers in diameter. Figure 3 a provides a confocal overview image that depicts the high abundance of ring-like DNA structures. The sample is taken directly from the storage solution without additional purification steps. We reproducibly obtain 5300 ± 2500 closed DNA rings per μL (mean ± SD, Supplementary Fig. 6) . StarPEG-(KA7) 4 thus mimics the behavior of septins in actin filament networks 29 . The formation of closed rings and their microscopic structure are verified by STED (Fig. 3 b and Supplementary Fig. 7) and transmission electron microscopy (Supplementary Fig. 7) , which also reveals that the DNA rings typically consist of tens to hundred of DNA nanotubes. Since the persistence length to mean length ratio for DNA nanotubes (Poisson distributed lengths) is around 1 and the persistence length is defined as the length over which correlations in the direction of the tangent are lost, we can assume that an encountering of DNA nanotube ends is possible on the experimentally relevant time scale of several seconds up to a few minutes. Once the ends have met, some DNA nanotubes may undergo end-to-end joining, others overlap to maximize the crosslinking. STED reveals free DNA nanotube ends at the edges of the ring suggesting that the observed ring formation cannot be a result of end-to-end joining only, but is rather mediated by crosslinkers along the DNA nanotubes that induce further growth of the bundle thickness by recruiting more single or prebundled DNA nanotubes (Fig. 3 b and Supplementary Fig. 7) . Fig. 3 StarPEG-(KA7) 4 induces self-assembly of multiple DNA nanotubes into closed micron-scale DNA rings. a Confocal overview image of self-assembled DNA nanotube rings formed from 50 nM DNA tiles (yellow, Atto633-labeled, λ e x = 640 nm) in presence of 500 nM starPEG-(KA7) 4 . Rings are highlighted with white arrows. Scale bar: 20 μm. b Representative confocal (top) and STED (bottom) images of an individual DNA ring formed from 50 nM DNA tiles (yellow, Atto633-labeled, λ e x = 640 nm) in presence of 500 nM starPEG-(KA7) 4 . Scale bar: 2 μm. c Representative confocal images of individual DNA rings formed at starPEG-(KA7) 4 to DNA tile (yellow, Atto633-labeled, λ e x = 640 nm) ratios from 0.1 to 30. The DNA tile concentration is constant at 50 nM. Scale bar: 2 μm. d Histogram of the DNA ring diameters for starPEG-(KA7) 4 to DNA nanotube ratios from 0.1 to 30 ( n = (7, 15, 15, 15, 15) DNA nanotube rings per condition). Source data are provided as a Source Data file. We analyze the ring diameter for a range of starPEG-(KA7) 4 to DNA tile ratios. For an excess of DNA tiles, the amount and the diameter of rings is significantly reduced compared to equimolar ratios (Fig. 3 c, d). The ring diameter is, however, largely independent of the magnesium-ion concentration and the absolute concentrations of starPEG-(KA7) 4 or DNA tiles (Supplementary Fig. 8) . The number of DNA rings in a given sample volume (Supplementary Figs. 5 and 6) as well as their diameter remains constant over time (Supplementary Fig. 9) . Taken together, we obtain the first self-assembled and free-standing micron-scale DNA nanotube bundle rings to the best of our knowledge. This ring formation mechanism seems to mimic naturally occurring ring formation by cytoskeletal filament crosslinking 4 , 29 . To gain a deeper understanding of DNA nanotube ring formation and to derive strategies for potential ring contraction, we next develop a theoretical framework and subsequently combine it with coarse-grained molecular dynamics (MD) simulations. Theory and predictions for DNA ring formation and contraction Ring formation and contraction of bundles of semiflexible filaments such as DNA nanotubes can be described by a balance of an adhesion energy gain, which is reduced by a surface energy term and a bending energy term 30 , 31 . The adhesion energy gain can be, for example, crosslinker-mediated, electrostatic or from depletion attraction, whereas the surface energy term results from DNA nanotubes exposed to solvent and lacking adhesion. Crosslinkers generate an adhesive energy between two DNA nanotubes by presenting one adhesive end to each DNA nanotube 31 . Thereby they accumulate in the overlap region between DNA nanotubes and can be viewed as a one-dimensional gas of particles confined to the overlap region. We show that, generally, this dense and mobile gas of adhesive crosslinkers gives rise to an additional effective free energy of adhesion inside a nanotube bundle, which is of entropic nature. In case of actin filaments it has been confirmed that the entropy of the crosslinker gas tends to maximize overlaps between the filaments 4 . With our generalization, the same should hold true for DNA nanotubes. We consider a torus of diameter D consisting of bundled DNA nanotubes, which are uniformly bent (Fig. 4 a). The bundle’s circular cross section contains N DNA nanotubes resulting in a total length L tot = π N D . The bundle has a bending rigidity κ b ( N ), which is related to the bending rigidity κ of individual DNA nanotubes via κ b ( N ) = κ N α , with α = 1 for decoupled sliding DNA nanotubes and α = 2 if crosslinking resists shear 32 . We assume uniform DNA nanotube bending rigidities throughout the entire torus. We assume an adhesion energy g per length for the DNA nanotubes in the interior of the bundle from crosslinker-mediated attraction or depletion attraction. For roughly circular cross sections, there should be N s = a s N 1/2 out of N DNA nanotubes at the bundle surface with a geometric factor a s of order unity. This results in an effective DNA nanotube length L i = L tot − π a s N 1/2 D within the interior of the DNA nanotube bundle that is fully accessible to the attraction of strength g . Fig. 4 Theoretical description and simulation of DNA nanotube rings. a Schematic illustration of the theory model representing the associated parameters. The DNA nanotubes are drawn as a continuous line and colored to distinguish surface (red) and interior (yellow) of the ring. The diameter of the bundle cross section is assumed to be negligible compared to the ring diameter D . Zoom: A discretized bead-spring representation of the DNA nanotubes is used in the MD simulations with parameters as indicated. b Snapshots of an isotropic initialization (left) and a DNA nanotube ring (right) taken from MD simulations. For clarity, the DNA nanotubes widths are increased. The cubic boxes show the simulation volume. c Coarse-grained MD simulation of the DNA ring formation from a solution of DNA nanotubes represented as bead-spring polymers. Individual nanotubes involved in ring formation are colored for clarity. A reduced persistence length is employed to facilitate the ring closure. d Kinetically trapped structure in incomplete ring formation after simulated annealing (right) (starting at temperature T 1 (left) and annealing to a high temperature T 2 = 8 T 1 ). e Transmission electron microscopy image of a kinetically trapped DNA ring as observed in experiments. Scale bar: 500 nm. Contraction of a toroidal bundle of DNA nanotubes can then be described by the free energy 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}$$F={E}_{{{{{{{{\\rm{bend}}}}}}}}}+{E}_{{{{{{{{\\rm{ad}}}}}}}}}+{F}_{c}$$\\end{document} F = E bend + E ad + F c 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}$$=2\\pi {\\kappa }_{b}(N){D}^{-1}-g{L}_{i}+{F}_{c}({L}_{i}),$$\\end{document} = 2 π κ b ( N ) D − 1 − g L i + F c ( L i ) , which is the sum of bundle bending energy E bend , adhesion energy in the interior E ad and the entropic free energy F c of the crosslinker gas in the interior of the bundle, which will arise if crosslinkers are mobile. We approximate F c by the entropic free energy of a Tonks gas of N c non-overlapping particles of crosslinker size b c in a one-dimensional volume of length L i with a line density of 1/ b c of possible binding sites (see Eq. ( 4 ) in Supplementary Note 1) . In this sense b c can be viewed as the size of the “footprint” of the crosslinker on the DNA nanotube. Minimizing the total free energy F with respect to the diameter D at fixed L tot , i.e., using N ( D ) = L tot / π D , gives the equilibrium diameter of the toroidal DNA nanotube bundle (see Supplementary Note 1) . The bending energy will favor large D , while adhesion energy and crosslinker entropy favor ring contraction via increasing the overlapping interior length L i of the bundle. Neglecting prefactors of order unity we find for the equilibrium diameter 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}$${\\left(\\frac{D}{{L}_{{{{{{{{\\rm{tot}}}}}}}}}}\\right)}^{(3/2)+\\alpha } \\sim \\, \\frac{{L}_{p}}{{L}_{{{{{{{{\\rm{tot}}}}}}}}}}\\frac{1}{{N}_{c}}{\\left(\\frac{g{d}_{c}}{{k}_{B}T}+\\frac{1}{1-{b}_{c}/{d}_{c}}\\right)}^{-1},$$\\end{document} D L tot ( 3 / 2 ) + α ~ L p L tot 1 N c g d c k B T + 1 1 − b c / d c − 1 , where d c = L i / N c ≈ L tot / N c is the average distance between crosslinkers (while b c is their minimal possible distance) and L p = κ / k B T is the persistence length of individual DNA nanotubes (see Eq. (9) in Supplementary Note 1) . This result predicts several experimentally testable scenarios under which ring contraction could occur: (i) Rings contract for decreasing persistence length or bending rigidity, \\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}$$D\\propto {L}_{p}^{1/(3/2+\\alpha )}$$\\end{document} D ∝ L p 1 / ( 3 / 2 + α ) . If the persistence length L p = κ / k B T is decreased by increasing the temperature, additional temperature-dependencies in Eq. ( 3 ) become relevant and are discussed under points (iii) and (iv). (ii) Rings contract if the total DNA nanotube length is decreased, for example, by depolymerization, \\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}$$D\\propto {L}_{{{{{{{{\\rm{tot}}}}}}}}}^{(1/2+\\alpha )/(3/2+\\alpha )}$$\\end{document} D ∝ L tot ( 1 / 2 + α ) / ( 3 / 2 + α ) . (iii) If crosslinker entropy can be neglected, the ring diameter will contract with increasing g according to D ∝ ( κ / g ) 1/(3/2+ α ) . This means that if g is an entropic depletion attraction with g ∝ T ring contraction occurs for increasing temperature or decreasing persistence length of individual DNA nanotubes, \\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}$$D\\propto {T}^{-1/(3/2+\\alpha )}\\propto {L}_{p}^{1/(3/2+\\alpha )}$$\\end{document} D ∝ T − 1 / ( 3 / 2 + α ) ∝ L p 1 / ( 3 / 2 + α ) (iii.a). If g is a crosslinker-mediated attraction, it is largely temperature-independent and D is independent of temperature although the persistence length will decrease (iii.b). (iv) For dominant crosslinker entropy, rings will contract with increasing temperature or decreasing persistence length, \\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}$$D\\propto {T}^{-1/(3/2+\\alpha )}\\propto {L}_{p}^{1/(3/2+\\alpha )}$$\\end{document} D ∝ T − 1 / ( 3 / 2 + α ) ∝ L p 1 / ( 3 / 2 + α ) . The crosslinker entropy is dominant for g d c < k B T /(1 − b c / d c ), where g d c is the average adhesion energy per crosslinker; in particular, it becomes dominant for a dense crosslinker gas with b c / d c ≤ 1 if the average adhesion per crosslinker is of the order of several k B T . An additional entropic depletion attraction with g ∝ T will further contract rings under these conditions (iv.a). In summary, the key parameters for ring constriction are L p and g . These can be controlled in simulations as well as experiments with DNA nanotube rings, making the theoretical predictions testable. Simulation of the formation and contraction of DNA nanotube bundles and rings To test the theoretical predictions for ring contraction, we first set up a simulation framework that reproduces the formation of DNA nanotube rings based on coarse-grained MD simulations using LAMMPS 33 with a bead-spring representation of the DNA nanotubes and attraction modeled by a Lennard-Jones potential of strength ε and particle size parameter σ (see Fig. 4 a, b and ‘Molecular Dynamics Simulation’ in the “Methods” section). MD simulations reproduce ring formation from individual DNA nanotubes in solution as shown in Fig. 4 c. The lengths of the simulated DNA nanotubes are Poisson distributed, as experiments from previous publications revealed 13 , 17 and set to a mean value of 580 σ ≈ 6.96 μm. Simulations suggest that ring assembly proceeds via nucleation of an initial ring containing few DNA nanotubes and further growth by incorporation of single DNA nanotubes or DNA nanotube bundles. Sliding of bead-spring polymers in simulations and, thus, equilibration of DNA nanotube rings is impeded by a “lock-in” of beads by attraction to two neighboring beads on a neighboring polymer generating energy barriers for sliding. To facilitate equilibration, we employ an annealing protocol in the MD simulations, where we increase temperature to T 2 ~ 2 − 8 T 1 for short time intervals for a simulation at temperature T 1 . Only for very small ring diameters (small κ / ε ), this annealing procedure is not sufficient to reach equilibrium in available simulation times (three light yellow crosses in Fig. 5 c, see Supplementary Note 2 for details). Interestingly, if the annealing temperature T 2 is chosen too high, simulated annealing can give rise to kinetically trapped partially unbundled ring structures, which are strikingly similar to experimentally obtained incomplete ring structures observed in TEM (Fig. 4 d, e). The simulations can thus reproduce architectural details of the experimentally observed DNA rings, which provides additional validation for our approach (we note that experimental and simulation protocols giving rise to trapped structures can not be compared). Fig. 5 Coarse-grained MD simulations of DNA nanotube rings and their contraction. a , b Bundle contraction during equilibration (ring diameter as a function of MD simulation time) for different single DNA nanotube rigidities κ ( a , for ε / k B T = 0.3) and different potential strengths ε / k B T ( b , for κ / k B T σ = 600). Smaller equilibrium diameters result from decreasing κ or increasing ε . Simulation snapshots for ε / k B T = 0.8. Scale bar: 60 σ . c Double-logarithmic plot of equilibrium ring diameter (Mean ± SD, n = 545 measurements, error bars are smaller than symbols) as a function of κ / ε σ for increasing bending rigidity κ (red circles, for ε / k B T = 0.3; yellow crosses for ε / k B T = 1.0) or decreasing potential strength ε (blue circles, for κ / k B T σ = 600) in comparison to the theory ( 3 ) with D ∝ ( κ / ε ) 2/5 corresponding to α = 1, i.e., sliding decoupled nanotubes (solid line) and D ∝ ( κ / ε ) 2/7 corresponding to α = 2, i.e., shear-resisting coupling between nanotubes (dashed line). Three simulations at smallest κ / ε could not be fully equilibrated (three light yellow crosses). Source data are provided as a Source Data file. In order to quantify ring diameters of bundles in simulations we measure the 3x3 gyration tensor \\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}$${S}_{mn}={N}_{{{{{{{{\\rm{tot}}}}}}}}}^{-1}\\mathop{\\sum }\\nolimits_{i=1}^{{N}_{{{{{{{{\\rm{tot}}}}}}}}}}{r}_{m}^{(i)}{r}_{n}^{(i)}$$\\end{document} S m n = N tot − 1 ∑ i = 1 N tot r m ( i ) r n ( i ) from all bead positions \\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}$${\\overrightarrow{r}}^{(i)}$$\\end{document} r → ( i ) and obtain a ring diameter by comparing its eigenvalues with corresponding eigenvalues for a homogeneous torus (Supplementary Note 2) . We reproduce the theoretically predicted ring contraction in MD simulations. The simulations show contraction both upon decreasing bending stiffness κ or persistence length L p (according to theoretical predictions (i), (iii.a) and (iv), Fig. 5 a) and upon increasing ε or adhesion strength (according to theoretical predictions (iii) and (iv.a), Fig. 5 b). Starting from a pre-assembled ring of diameter D ~ 250 σ (corresponding to 3 μm and similar to the rings observed with confocal microscopy) we equilibrate rings over ~ 10 9 MD simulation steps, where the ring diameter approaches its equilibrium value. Typically, rings contract during equilibration to diameters in the range of D ~ 80 − 250 σ corresponding to a reduction of the diameter down to D ~ 1 μm for contracted rings (Supplementary Movies 2 and 3) . We clearly observe smaller equilibrium ring diameters for decreasing κ or increasing ε (Fig. 5 a–c) and find a dependence D ∝ ( κ / ε ) 2/5 characteristic for sliding adhesive DNA nanotubes with α = 1 in agreement with the theory result (Eq. ( 3 )) (blue circles and yellow crosses in Fig. 5 c). The theory assumes strong interactions deep in the bundled phase. For weak interactions close to the bundling threshold, the bare adhesion energy g should be replaced by the bundling free energy f = g − T s per length, which is reduced by entropic contributions s from the DNA nanotube shape fluctuations 31 . This gives rise to deviations from the proposed scaling (Eq. ( 3 )) for small but realistic values of ε (red circles in Fig. 5 c). Investigations on a translation of the low viscosities applied in the MD simulation to the more realistic viscosity of water predict an equilibration time of ~ 20s (see Supplementary Note 2 for details). These results indicate that ring contraction happens too fast to be observed in our experimental setting because it takes at least half a minute before the microscopy experiment can be started due to mixing of DNA nanotubes with crosslinkers and crowders and filling the solution into an observation chamber. In addition, we can only analyze DNA rings immobilized at the surface which prevents imaging during ring contraction. Experimental realization of DNA ring contraction Finally, we set out to realize DNA ring contraction experimentally and use simulations and theory to rationalize the results quantitatively. In experiments, a decrease in persistence length (according to theoretical prediction (i)) can be achieved by increasing the temperature, as temperature is known to decrease the persistence length of DNA 34 . At the same time, a temperature increase leads to nanotube depolymerization, which should cause further contraction (according to theoretical prediction (ii)). We thus form the DNA rings in the presence of starPEG-(KA7) 4 . Indeed, increasing the temperature from room temperature to 40 °C leads to a reduction of the mean ring diameter from (3.0 ± 0.7) μm to (1.9 ± 0.7) μm (Fig. 6 a, b). Note that the ring diameter does not change over time at constant temperature (Supplementary Fig. 9) . Fig. 6 Rings formed from DNA nanotubes contract upon addition of a molecular crowder or heating. a Confocal images of uncontracted (left) and contracted (right) DNA nanotube rings formed from 50 nM DNA tiles, 500 nM starPEG-(KA7) 4 in 1x PBS and 10 mM MgCl 2 without and with 25 wt% 500 kDa dextran. Scale bar: 2 μm. b DNA nanotube ring diameter after 15 min heating to 35 and 40 °C, respectively (Mean ± SD, n = (31, 30, 30) DNA nantube rings analyzed per condition). Mann-Whitney test with p-values from left to right: ****≤0.0001, *0.0106. c DNA nanotube ring diameter for different molecular weights of 25 wt% dextran (2'500, 35'000, 70'000, 250'000, 500'000 g/mol) (Mean ± SD, n = (32, 32, 31, 32, 31, 30) DNA nanotube rings analyzed per condition). Mann–Whitney test with p-values from left to right: ****≤0.0001, *0.0221, *0.0434, ns 0.5787. d Experimental DNA nanotube ring diameter reduction by depletion attraction as a function of molecular weight of dextran at 25 wt% (black, identical to data points in c , constant total number of dextran monomers). Depletion theory (turquoise, see text and Supplementary Note 3) assumes a penetrable layer of crosslinkers (thickness p ) around DNA nanotubes (thickness d 0 ). Parameters fitted to experimental ring diameters (black, see c (Mean ± SD, n = (32, 32, 31, 32, 31, 30) DNA nanotube rings analyzed per condition)): DNA nanotube diameter d 0 = 18.6 ± 2.4 nm, penetration depth p = 11.0 ± 3.3 nm, attraction strength due to crosslinkers ε cross / k B T = 0.31 ± 0.14. Source data are provided as a Source Data file. To relate the experimentally observed ring contraction to the MD simulations, we have to consider the following dependencies: On the one hand, an increase from room temperature to 40 °C decreases the persistence length of double-stranded DNA by ~10% 34 . On the other hand, we have to consider depolymerization of the free ends of the DNA nanotubes. The critical melting temperature for DNA nanotubes can be calculated as a function of the enthalpy of disassembly, the entropy, the number of sticky end bonds and base pairs per bond and the concentration of free tiles 35 , 36 . Our DNA nanotube design has a maximum melting temperature of 37.2 °C (at a maximum free tile concentration of 5 nM, compared to 42.0 °C at 50 nM free tiles). Experiments have shown that in absence of free tiles, the depolymerization rate measures around 0.3 layers per second at 40 °C 35 . Hence, we assume that the free DNA nanotube ends, which we saw with STED microscopy (Fig. 3 b) are depolymerizing at 40 °C, which makes sense if we compare the extent of the ring contraction in the MD simulation to the experiments: In the MD simulation, the bending stiffness κ was decreased from 1000 to 200 k B T σ for a fixed attractive strength ε resulting in smaller ring diameters (see Fig. 5 a and red circles in Fig. 5 c). By assigning the temperature-dependent contraction to a purely bending stiffness-dependent contraction we can calculate corresponding effective bending stiffnesses from the simulation data. Measured ring diameters decreasing from 3.0 μm to 1.9 μm (Fig. 6 b) for increasing temperatures (23, 35, 40) °C correspond to decreasing effective bending rigidities κ = (2010, 890, 500) k B T σ according to the simulation data for a fixed attractive strength of 0.3 k B T (red circles in Fig. 5 c). This predicted four-fold decrease in effective bending stiffness is not sufficient to explain the experimentally observed ring contraction. It confirms that both, a decrease of the persistence length and depolymerization reduce the effective ring parameter in experiments, validating theoretical predictions (i) and (ii). Indeed, depolymerization is also relevant for the contraction of actin rings 4 , 37 . To test theoretical predictions (iii) and (iv), we need to increase g with an additional depletion force. We can achieve this experimentally by adding dextran as a molecular crowding agent, which allows us to control the depletion attraction with the concentration and the molecular weight of the crowding agent. We thus add 25 wt% of dextran to the DNA ring-containing solution. We find that the additional molecular crowding induces DNA ring shrinkage to less than 45% of the initial diameter from a mean diameter of (3.3 ± 0.7) μm to (1.4 ± 0.4) μm at constant temperature (Fig. 6 c), confirming theoretical prediction (iii). Importantly, the rings’ shape, quantified by their circularity, remains unaltered and close to 1 (circularity of 1 corresponds to a perfect circle, Supplementary Fig. 10) . We observe similar results with methylcellulose, confirming that the contraction is induced by crowding and not due to the chemical nature of the agent (Supplementary Fig. 11) . The high uncertainty of ring diameters might mainly result from highly heterogeneous bundles that consist of DNA nanotubes of different lengths (Poisson-distributed). Overall, we observe that a higher molecular weight of dextran gives rise to smaller ring sizes, i.e., an effectively increased attractive interaction (theoretical prediction (iv.a)). This ring contraction can be understood quantitatively by calculating the additional depletion interaction which arises from exclusion of dextran from an overlap volume between DNA nanotubes. Details of the calculation can be found in Supplementary Note 4 . The molecular crowding effect of a macromolecule is, on the one hand, dependent on its radius of gyration (and thus on its molecular weight), which sets the thickness δ of the depletion layer (the depletion length) and, on the other hand, on its concentration. It is important to state that the number concentration of dextran molecules in the experimental sample decreases with higher molecular weights if the mass concentration is kept constant (and thus the number of monomers per sample). For hard rods of diameter d , the combined effects of an increasing radius of gyration R g ∝ M 1/2 of the molecular crowder (increasing the depletion layer thickness), and a decreasing number concentration c / M (at fixed mass concentration c ) give rise to a depletion attraction that decreases with molecular weight ∝ M −1/2 for small radii of gyration R g ≪ d before it saturates for large radii of gyration R g ≫ d corresponding to large molecular weights M . The model of a hard rod is, however, not completely adequate in the presence of additional crosslinkers, which “decorate” the DNA nanotubes and act as a penetrable layer of thickness p around the DNA nanotubes of bare diameter d 0 . The glycocalix around red blood cells constitutes a similar penetrable layer that has been shown to give rise to increased aggregation of red blood cells by dextran of higher molecular weight 38 . Following Neu et al. 38 , we treat the crosslinker-decorated DNA nanotubes as penetrable rods of total diameter d = d 0 + 2 p with a reduced depletion length δ − p . The depletion layer thickness δ is reduced by a penetration depth p , which is the thickness of the penetrable layer: a part of the crowding agent of size p can be “burried” within the penetrable layer, which reduces the size of the depletion layer accordingly. As a result, smaller crowding agents with less molecular weight become less effective. For p as small as a few nanometers, this gives rise to a depletion attraction ε dep ( M ) that increases with molecular weight M for R g ≪ d before it saturates for high molecular weight ( R g ≫ d ). Using the ring diameter result \\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}$$D \\, \\approx \\, 100{d}_{0}{(\\varepsilon /{k}_{B}T)}^{-2/5}$$\\end{document} D ≈ 100 d 0 ( ε / k B T ) − 2 / 5 from our MD simulations with an attractive strength ε = ε cross + ε dep ( M ) that is the sum of a crosslinker contribution and the depletion contribution, we can fit the experimental data for the ring diameter as a function of dextran molecular weight quantitatively (see Fig. 6 d, blue line) with fit parameters d 0 = 18.6 ± 2.6 nm and p = 11.0 ± 3.3 nm for the bare DNA nanotube diameter and the penetration depth, respectively. A DNA nanotube diameter of 11.8 ± 2.1 nm has been measured by TEM (unbundled DNA nanotubes in the absence of starPEG-(KA7) 4 , Supplementary Fig. 4) . The size of the starPEG-(KA7) 4 crosslinkers can be estimated to be below 22 nm (see above and earlier publications 26 ): These values are compatible with the fit results in view of additional complications, such as tilted orientations of attached crosslinkers and the possibility that the KA7 peptides align due to their electrostatic interactions in parallel with the DNA nanotube, which reduce the value of p . The fit from the MD simulation reproduces the experimental data qualitatively as well as quantitatively. With temperature increase and the addition of molecular crowders, we were able to obtain contraction of the DNA nanotube rings to less than half of their initial diameter. With this, we validate the four theoretical predictions, which we used as guidance to derive the parameters relevant for ring contraction. Our DNA rings thus contract significantly more than rings formed from actin filaments when they are contracted by passive crosslinkers 4 , likely due to the lower persistence length of the DNA nanotubes compared to actin. In conclusion, we engineered synthetic micron-scale DNA rings, self-assembled from a bundle of tens of DNA nanotubes and crosslinked via electrostatic interactions with custom-designed starPEG-(KA7) 4 peptides. Based on theoretical considerations, we derived conditions for DNA ring contraction, which we validate with experiments and coarse-grained MD simulations. Since our micron-sized DNA rings consist of well-established DNA nanotubes, further investigations, e.g. involving mutations like stiffness control 39 , 40 are facilitated compared to protein-based materials like actin filaments. In the future, these micron-sized DNA rings could be used as tracks for molecular assembly and transport or embedded into adaptive materials. In addition, synthetic DNA rings are already equipped with features responding to temperature change similar to their biological counterparts (polymerization of actin filaments 41 and DNA nanotubes 35 , 36 ). By contrast, DNA nanotubes can easily be equipped with other types of molecular functionalization, reprogrammed, and repurposed for diverse systems. An entirely synthetic division machinery for liposomes, based on DNA nanotechnology and peptide design, is a highly attractive albeit far-reaching goal. It has to be acknowledged that besides complete contraction several challenges have to be overcome to induce vesicle division due to membrane fission. The rings have to be positioned in the equatorial plane of the liposome, which will likely be achievable by self-assembly if the persistence length of the DNA nanotubes is sufficiently high 13 . Secondly, the rings have to be linked to the membrane which could be achievable e.g. with cholesterol-tags 13 or transmembrane entities 15 . The temperature-induced contraction would in principle be compatible with liposome-encapsulation and the contraction force could potentially be sufficiently high to induce vesicle deformation 26 . Our approach could be complemented by engineered molecular motors that walk on DNA nanotubes 16 . For ring disassembly, which will be necessary to complete the division of the compartment, it is plausible to use mechanisms that have already been described for DNA nanotubes 12 , 13 , 17 . Each of these steps, however, warrants detailed investigation and presents a fruitful challenge for future research. The symbiosis of DNA nanotechnology and peptide engineering may lead to advanced and highly functional molecular hardware for bottom-up synthetic biology and hybrid materials with a wide range of applicability."
} | 12,530 |
39249534 | PMC11604758 | pmc | 8,608 | {
"abstract": "Arbuscular mycorrhizal (AM) fungi not only play a crucial role in acquiring nutrients for plants but also serve as a habitat for soil microbes. Recent studies observed that AM fungal hyphae are colonized by specific bacterial communities. However, so far it has not been explored whether fungal hyphae and mycorrhizal networks also harbor specific communities of protists, a key group of microbes in the soil microbiome. Here, we characterized protist communities in soil in a compartment with plant roots and on hyphae collected from hyphal compartments without plant roots. We detected specific protist communities on fungal hyphae. Fourteen protistan amplicon sequences variants (ASVs) were significantly associated with fungal hyphae, half of which belonged to the Cercozoa group. This research, for the first-time detected specific protist ASVs directly associated with abundant AM fungus hyphae, highlighting the complexity of the hyphal food web. Supplementary Information The online version contains supplementary material available at 10.1007/s00572-024-01167-3.",
"introduction": "Introduction Arbuscular mycorrhizal (AM) fungi colonize plant roots and form extensive hyphal networks in the surrounding soil, facilitating the uptake of water and nutrients by the plant (Smith and Read 2010 ; Martin and van der Heijden 2024 ). In return, the plants translocate up to 5–20% of their photosynthetic production through the AM fungal hyphae, enriching the surrounding soil with carbon (Jakobsen and Rosendahl 1990 ; Wang et al. 2022 ; Hawkins et al. 2023 ). The interface between AM fungi hyphae and soil is not solely a conduit for chemical exchange but also provides a habitat for a diverse range of soil microbes, including bacteria, archaea and fungi (Zhang et al. 2018 , 2024 ; Emmett et al. 2021 ; Nuccio et al. 2022 ). Moreover, recent studies have provided evidence that specific ‘mycorrhiza helper bacteria’ colonize AM fungal hyphae, thereby affecting both the development of mycorrhizal associations and nutrient cycling in the soil (Zhang et al. 2024 ). Nevertheless, the interactions of AM hyphae with other key groups of soil organisms including protists are still poorly understood. Protists are abundant in soils, with densities ranging from 10 4 to 10 8 per gram (Adl and Coleman 2005 ). Protists can have various lifestyles, including phototrophy, heterotrophy, mutualism, and parasitism. Phototrophic protists (known as algae) contribute significant amounts of organic carbon to soil (Schmidt et al. 2016 ), while heterotrophic protists (known as protozoa) consume bacteria and release nitrogen into the soil because of their higher C: N than their prey (Sherr et al. 1983 ). Furthermore, the nitrogen released by heterotrophic protists into the soil is generally in a form that is readily accessible to plants and leads to enhanced plant growth (Bonkowski 2004 ; Gao et al. 2019a , b ). De Gruyter et al., (De Gruyter et al. 2022 ) identified differences in the protist community between pots with and without AM fungi. Moreover, a recent study demonstrated that protists can enhance the utilization of organic nitrogen by AM fungi (Rozmoš et al. 2021 ). However, it is poorly understood whether protists are associated with AM hyphae in soil, and it is unknown whether AM hyphae harbor specific communities of protists. Here, we conducted a study using compartmentalized microcosms in a greenhouse study to investigate if AM fungi host specific protist communities. We collected root and soil samples from a plant-colonized compartment and extraradical hyphal samples from another compartment from which plant roots were excluded. We then analyzed the hyphal and soil protist communities using 18 S rRNA gene amplicon sequencing. Our findings provide direct evidence of specialized protist taxa accompanying mycorrhizal fungal hyphae.",
"discussion": "Results and discussion The protist communities differed between soil and hyphal samples (Fig. 1 b). Sample type (hyphal or soil sample) accounted for a significant 30.9% of observed variation between treatments (PERMANOVA, R 2 = 0.309, F = 9.883, p < 0.001). Further investigation was undertaken to identify the specific protist phyla contributing to this difference. Ochrophyta, often represented by phototrophic algae and belonging to Stramenopiles, were significantly enriched in hyphal samples (RA: 12.5%) compared to soil samples (RA, 3.7%; Fig. S3 ). Pseudofungi, Apicomplexa, Conosa, Mesomycetozoa, and Chrompodellids exhibited higher abundance in soil samples than in hyphal samples (Fig. S3 ). It is unresolved what are the specific mechanisms explaining why particular protists are more or less abundant on fungal hyphae. This is linked to the ecological niche of individual protists, their most-important food source (e.g., specific carbon sources or bacteria), the presence or absence of other microbes (competitors, facilitators etc.) and their abiotic niche (including soil texture and water availability). Subsequently, we investigated the differences of protist ASVs between hyphal and soil samples using Indicspecies . A total of 14 ASVs were significantly more abundant on fungal hyphae compared to soil (Fig. 1 c, Table S1 ). Of these 14 hyphal ASVs, 7 were classified as Cercozoa, the protists phylum that generally is most common in soils and mostly bacterivorous (Dumack et al. 2022 ). Cercozoa were previously also shown to be enriched in rhizospheres (Sapp et al. 2018 ). Several studies suggest that Cercozoa species have a specific preference for predation on bacteria (Glücksman et al. 2010 ; Amacker et al. 2022 ). The bacterial community assembled on fungal hyphae, compared with that of the soil (Zhang et al. 2024 ), may consequently harbor specific bacterivorous cercozoans that feed on bacteria particularly associated with mycorrhizal fungi. However, it remains uncertain whether these Cercozoa exclusively feed on bacteria attached to hyphae or interact directly or indirectly with fungal hyphae. Other ASVs that are significantly more abundant on hyphae comprised 2 Ochrophyta ASVs, 2 Chlorophyta ASVs, 2 Choanoflagellida ASVs, and 1 Lobosa ASV (Fig. 1 c, Table S1 ). Only 6 of the 14 hyphal ASVs were identified at the genus level, including Brachysira , Massisteria , Vermamoeba , Choanoflagellida , and Bracteacoccus . The remaining eight ASVs were classified with an unclear genus designation (Table S1 ). The two hyphae enriched Ochrophyta ASVs (ASV_fd9f, ASV_3fab) were identified among all ASVs. Intriguingly, when comparing ASVs only in Ochrophyta between hyphal and soil samples using the Wilcox test, these two ASVs also emerged as the most differentially abundant hyphal ASVs (Fig. S4). ASV_fd9f and AS_3fab exhibit RA of 1.1% and 0.5% in hyphal samples, respectively. Conversely, the RA of these ASVs in soil samples is nearly negligible, approaching zero. This suggests that these two ASVs are the predominant taxa that differentiated Ochrophyta between hyphal and soil samples. To test the preference of these two ASVs for AM fungal hyphae, further studies need to isolate them and experimentally test their function. Overall, we observed that 7 protist groups together comprised 99.3% of the protist relative abundance (RA) in hyphal samples. These top-seven most-abundant groups were Rhizaria (RA, 41.3%), Alveolata (RA, 21.3%), Stramenopiles (RA, 15.6%), Archaeplastida (RA, 11%), Amoebozoa (RA, 6.6%), Hacrobia (RA, 2.5%), and Opisthokonta (RA, 1.3%; Fig. 2 ). The abundances of these broad taxonomic groups did not differ significantly between soil and hyphal samples (Fig. S5). From those 7 taxonomic groups, 80 ASVs were detected in both soil and hyphal samples (Fig. S6a). These shared ASVs were abundant in our system with a RA of 64.7% in the hyphal samples and 31.9% in the soil samples (Fig. S6b). This implies that, the majority of the hyphal protist communities are derived from the soil samples. \n Fig. 2 Hyphal and soil samples share common protist groups. Relative abundance of protist groups in soil and hyphal samples. Colors represent different protist groups. The protist groups with relative abundance below 1% were aggregated and categorized as low abundance \n Using internally transcribed spacer (ITS) amplicon sequencing, we found that the hyphal compartment was colonized by both AM fungi and other non-AM fungi. The relative abundance of Glomeromycota fungal phylum (AM fungi) was 51% while the remaining sequences primarily consisted of Chytridiomycota , Ascomycota , and Basidiomycota . Furthermore, within the Glomeromycota phylum, two AM fungal species, Rhizophagus irregularis (RA: 36%) and Septoglomus viscosum (RA: 14%), were most abundant. Detailed data can be found in Zhang et al. 2024 . Further studies, thus need to test whether the observed hyphae-specialized protist communities are specific for AM fungi or fungi in general. Note that the soil substrate from the hyphal compartment (COMP5) contained 20% more sand and nearly 80% less soil than the substrate in the plant compartment (COMP3). This was done to facilitate hyphal extraction and to reduce the amount of organic material attached to the fungal hyphae. The varying ratios of soil and sand contribute to differences in substrate texture and nutrient availability. Consequently, disparities in protist communities between hyphae and soil could potentially be because of differences in sand/soil content or differences in the ability of protists to disperse and move through the soil (Zhao et al. 2019 ). Thus, further studies should verify the observations made here. Moreover, prior to plant growth and mycorrhiza development, the soil substrate in COMP5 was autoclaved to diminish non-AMF fungi, thereby enhancing the likelihood of sampling mycorrhizal fungi originating from COMP3. This may also contribute to the subsequent differences in protist communities between hyphae and soil. This aspect warrants future experiments to validate the conclusions drawn in this study. In addition to collecting hyphae, roots, and soil from different compartments, future studies should also collect hyphae from the compartments with roots to assess whether microbial communities associated with hyphae differ between ‘hyphae-only’ compartments and those associated with root compartments. Alternatively, future studies could collect microbial communities from soil samples taken from the hyphal compartment to avoid any potential bias arising from differences in soil composition. By extracting protist DNA attached to and surrounding fungal hyphae, our work shows, for the first time, that protists directly colonize fungal hyphae. Furthermore, our work highlights that protist communities developed on fungal hyphae differ from the original protist community in field soil. Further work now needs to test whether protists play a role in the functioning of the plant-AM fungi symbiosis and whether protists may use fungal networks and mycelia as hyphal highways to spread through the soil. These findings highlight the intricate nature of the food web associated with AM hyphae and elucidate a significant connection between AM fungi and their associated microbes."
} | 2,800 |
37112070 | PMC10140994 | pmc | 8,611 | {
"abstract": "Lignin and cellulose derivatives have vast potential to be applied in polymer materials. The preparation of cellulose and lignin derivatives through esterification modification is an important method to endow cellulose and lignin with good reactivity, processability and functionality. In this study, ethyl cellulose and lignin are modified via esterification to prepare olefin−functionalized ethyl cellulose and lignin, which are further used to prepare cellulose and lignin cross−linker polymers via thiol–ene click chemistry. The results show that the olefin group concentration in olefin−functionalized ethyl cellulose and lignin reached 2.8096 mmol/g and 3.7000 mmol/g. The tensile stress at break of the cellulose cross−linked polymers reached 23.59 MPa. The gradual enhancement in mechanical properties is positively correlated with the olefin group concentration. The existence of ester groups in the cross−linked polymers and degradation products makes them more thermally stable. In addition, the microstructure and pyrolysis gas composition are also investigated in this paper. This research is of vast significance to the chemical modification and practical application of lignin and cellulose.",
"conclusion": "4. Conclusions In this study, ethyl cellulose and lignin were modified via esterification to prepare olefin−functionalized ethyl cellulose and lignin, which were further used to prepare cellulose and lignin cross−link polymers via thiol–ene click chemistry. The results show that the olefin group concentration in the olefin−functionalized ethyl cellulose and lignin reached 2.8096 mmol/g and 3.7000 mmol/g, respectively. The tensile stress at break of the cellulose cross−linked polymers reached 23.59 MPa. The gradual enhancement in the mechanical properties was positively correlated with the olefin group concentration. Compared with EC, the tensile strength of the cross−linked polymers decreased significantly, but the tensile strain increased sharply. The flexible aliphatic hydrocarbon chain from undecylenic acid contributed to plasticization by functioning as an internal plasticizer. The changes in mechanical properties make the cross−linked polymers easier to process, expanding their application range. The EC–lignin polymers had an uneven surface structure and there were many large cracks and micropores due to the rigid benzene ring structure of lignin. The infrared characteristic absorption peaks at 3684, 2979, 2306 and 1747, 1391, 1057 cm −1 indicated that H 2 O, aliphatic hydrocarbon segments, CO 2 and degradation products containing ester groups were released. The existence of ester groups in the cross−linked polymers and degradation products makes them more thermally stable. The obtained ethyl cellulose and lignin cross−link polymers are of great significance for practical applications and contribute towards the high−value−added utilization of lignin and cellulose.",
"introduction": "1. Introduction With the increasing depletion of fossil resources and the increasing environmental problems of “white pollution” and “microplastics” caused by the extensive use of non−degradable petroleum−based polymers, the use of bio−based raw materials, instead of petroleum−based compound raw materials, to prepare polymer materials has attracted widespread attention [ 1 ]. Cellulose has become an ideal raw material for bio−based polymer materials due to its rich sources, low price, excellent biodegradability, easy modification and many other advantages. Cellulose and its derivatives have been widely used in the fiber, paper, film, plastics, coatings and other industrial fields [ 2 , 3 , 4 ]. However, due to the existence of a large number of hydrogen bonds within and between the molecules of natural cellulose, the complexity of cellulose aggregation structure and high crystallinity, cellulose is insoluble in water and general organic solvents, in addition to not being able of being melted or processed as traditional plastics, which severely limits the application of cellulose materials. The chemical modification of cellulose includes oxidation, esterification, etherification and other grafting methods [ 5 , 6 , 7 ]. At present, the esterification derivatives of cellulose include cellulose acetate, cellulose propionate, cellulose butyrate and various cellulose−mixed esters, which are widely used in plastics, coatings, separation membranes, cigarette filters and other daily necessities [ 8 , 9 , 10 ]. Lignin is a rich natural resource. Compared with other biomass products, lignin has a relatively complex structure and contains a variety of functional groups [ 11 ]. It has broad research prospects for the development of appropriate methods for separating and extracting lignin, and then prepare functional composites. At present, lignin has been applied to the preparation of high−value materials, such as porous carbon materials, adsorption materials, capacitor electrode materials, graphene materials, surfactants and hydrogels, and has broad application prospects in the energy, medical, construction, agriculture and other fields [ 12 , 13 , 14 , 15 , 16 ]. The structural unit of lignin is similar to that of phenol. Preparing lignin−based phenolic resin by partially replacing phenol with lignin is the most feasible method [ 16 , 17 , 18 , 19 , 20 ]. Lignin and its derivatives can be used to synthesize bipolar plate materials for phenolic resin fuel cells, phenolic resin catalysts, phenolic resin foams, phenolic resin adhesives and other phenolic resin materials [ 21 , 22 , 23 ]. Since the 20th century, researchers have been exploring the preparation of lignin−based phenolic resin by replacing phenol with structurally modified lignin derivatives. After the 1990s, the preparation technology of lignin−based phenolic resin developed rapidly because many kinds of structurally modified lignin derivatives with flame retardancy were produced. However, due to the shortcomings of high pollution, high energy consumption, and a complex preparation process, lignin−based phenolic resin cannot be produced on a large scale. As a simple C–S bonding reaction, “thiol–ene” reaction was discovered over 100 years ago [ 24 ]. This reaction has very attractive advantages: First, the C–S bonding reaction can be conducted under a variety of conditions. Secondly, various olefins can be used as suitable substrates, including activated and inactive polysubstituted olefins. Third, almost all mercaptans can be used for reactions, including highly functional substances. Finally, this reaction is very rapid, and the reaction conditions are mild, which can be conducted in the air environment [ 25 , 26 , 27 ]. The “thiol–ene “click reaction can be initiated by free radicals, initiated by ultraviolet light and free radicals using natural light, red light and redox system. The click reaction is fast and can be performed under normal temperature and pressure. If it is initiated by ultraviolet light, the reaction yield can reach more than 90% in a few seconds. As a means of constructing new materials, the “thiol–ene” click reaction has many advantages: olefin compounds are very rich, with high selectivity, and can synthesize a large number of compounds with various structures [ 28 , 29 , 30 , 31 ]. Recently, Jawerth reported that the ethanol−soluble fraction of Lignoboost Kraft lignin was selectively allylated using allyl chloride by means of a mild and industrially scalable procedure. The obtained modified lignin was then subsequently cross−linked to prepare thermosetting resin via thermally induced thiol–ene chemistry [ 29 ]. Cao et al. prepared thermosetting lignin−based polyurethane coatings with superior corrosion resistance and a high content of lignin by the polymerization of lignin−based polyol. Firstly, the phenolic hydroxyls of enzymatic hydrolysis lignin were, firstly, selectively converted to primary aliphatic hydroxyls by an allylation reaction. Subsequently, the thermal radical initiated thiol–ene click reaction was applied to efficiently prepare the lignin−based polyol [ 30 ]. Zeng et al. developed durable a superhydrophobic and oleophobic coating based on perfluorodecanethiol fluorosilicone polyurethane (PFDT−FSPU) and thiol−modified cellulose substrate. The cross−linked network structure was formed by the radical polymerization of double bonds in PFDT−FSPU when the polyurethane was irradiated with ultraviolet light, and it was anchored on the surface of the cotton fibers by click reaction between the thiol−modified cellulose substrate and PFDT−FSPU. The coated fabric showed excellent durability and can still maintain superhydrophobicity and oleophobicity even after 600 cycles of abrasion or 30 times of washing cycles or 168 h of accelerated aging test [ 31 ]. The chemical modification of ethyl cellulose and lignin to prepare olefin−functionalized lignin and ethyl cellulose can effectively increase reaction activity and enrich their application performance. The chemical modification of ethyl cellulose and lignin to construct new functional materials has become a major strategy to increase their added value. In this study, ethyl cellulose (EC) and lignin are modified via esterification to prepare olefin−functionalized ethyl cellulose and lignin, which are further used to prepare cellulose and lignin cross−linker polymers via thiol–ene click chemistry. The chemical structure of olefin−functionalized EC and lignin, as well as ethyl cellulose and lignin cross−linked polymers are characterized. In addition, the thermal stability, microstructure, mechanical property and pyrolysis gas composition are also investigated.",
"discussion": "3. Results and Discussion The esterification products of lignin and EC with undecenoic acid were characterized with FT−IR and 1 H NMR, As seen in Figure 1 , there are two strong proton signals at 5.0 ppm and 5.82 ppm in the 1 H NMR of undecenoic acid, which are attributed to the protons of olefin [ 33 , 34 ]. There are no proton signals at 5.0 ppm and 5.82 ppm in the 1 H NMR of EC. After esterification, proton signals at 5.0 ppm and 5.82 ppm gradually increased in the 1 H NMR of EC−1−0.5, EC−1−1.0 and EC−1−1.3. The results indicate that OFE products were obtained. As seen in Figure 2 , the strong peak at 1725 cm −1 is attributed to the infrared absorption peak of the carbonyl group in the FT−IR of undecenoic acid [ 35 , 36 , 37 ]. There is no peak at 1725 cm −1 in FT−IR of EC. After esterification, infrared absorption peaks of the carbonyl group gradually increased in the FT−IR of EC−1−0.5, EC−1−1.0 and EC−1−1.3, which indicates that the carbonyl groups were formed after esterification. The peak at 1675 cm −1 in the FT−IR of EC−1−0.5, EC−1−1.0 and EC−1−1.3 is attributed to the olefin groups [ 35 , 36 , 37 ]. The formation of carbonyl groups and the appearance of olefin groups suggests that OFE products were obtained. The 1 H NMR of OFL showed a similar characteristic absorption peak compared with that of OFE. As seen in Figure 3 , there are no proton signals at 5.0 ppm and 5.82 ppm in the 1 H NMR of lignin. After esterification, proton signals gradually increased at 5.0 ppm and 5.82 ppm in the 1 H NMR of lignin−1−0.5, lignin−1−1.0 and lignin−1−1.3. The results indicate that OFL products were obtained. Similarly, the FT−IR of OFL showed a similar characteristic absorption peak compared that of with OFE. As seen from Figure 4 , the strong peak at 1725 cm −1 is attributed to the infrared absorption peak of the carbonyl group in the FT−IR of undecenoic acid [ 33 , 34 ]. There is no peak at 1725 cm −1 in the FT−IR of lignin. After esterification, infrared absorption peaks of the carbonyl group gradually increased in the FT−IR of lignin−1−0.5, lignin−1−1.0 and lignin−1−1.3, which indicates that the carbonyl groups were formed after esterification. The peak at 1675 cm −1 in the FT−IR of lignin−1−0.5, lignin−1−1.0 and lignin−1−1.3 is attributed to olefin groups [ 35 , 36 , 37 ]. The formation of carbonyl groups and the appearance of olefin groups suggests that OFL products were obtained. The olefin group concentration in OFL and OFE was determined with the internal standard method using 2,3,5,6−Tetrachloro−3−nitrobenzene as the interior label. The 1 H NMR of OFL and OFE with the interior label was detected and the results are shown in Figure 5 and Figure 6 . The strong signal at 7.75 ppm in Figure 5 and Figure 6 is attributed to the protons of 2,3,5,6−Tetrachloro−3−nitrobenzene [ 38 , 39 ]. The olefin group concentration in OFL and OFE was calculated according to the procedure presented in a recent study [ 32 ]. The results are shown in Table 2 , revealing that the olefin group concentration in OFL increased from 1.8060 to 2.8096 mmol/g when the n(−COOH):n(−OH) increased from 0.5 to 1.3. The olefin group concentration in OFE increased from 2.9200 to 3.7000 mmol/g when the n(−COOH):n(−OH) increased from 0.5 to 1.3. The chemical structure of EC cross−linked polymers and lignin cross−linked polymers were investigated with FT−IR and 1 H NMR. As seen from Figure 7 , the FT−IR of EC−1−1−2SH and EC−1−1−4SH was compared with EC−1−1. The peak at 1675 cm −1 in the FT−IR of EC−1−1 is attributed to the olefin groups [ 35 , 36 , 37 ]. After the thiol–ene click reaction, there is no peak at 1675 cm −1 in the FT−IR of EC−1−1−2SH and EC−1−1−24H. The chemical structure of the dissolved part for EC cross−linked polymers was investigated with 1 H NMR. As seen from Figure 8 , the 1 H NMR of EC cross−linked polymer was compared with that of EC−1−1. The two strong proton signals at 5.0 ppm and 5.82 ppm in the 1 H NMR of EC−1−1 are attributed to the protons of olefin [ 33 , 34 ]. When the thiol–ene click reaction finished, there were no protons at 5.0 ppm and 5.82 ppm in the 1 H NMR of EC−1−1−2SH and EC−1−1−4SH, which indicates that there was a thiol–ene click reaction between OFL and the cross−linker (3,6−Dioxa−1,8−octanedithiol, pentaerythritol tetra(3−mercaptopropionate)), and that the EC cross−linked polymers were obtained. Figure 9 shows the FT−IR of ligin−1−1−2SH and lignin−1−1−4SH compared with that of lignin−1−1. The peak at 1675 cm −1 in the FT−IR of ligin−1−1 is attributed to olefin groups [ 35 , 36 , 37 ]. When the thiol–ene click reaction finished, the peak at 1675 cm −1 in the FT−IR of ligin−1−1−2SH and ligin−1−1−4SH decreased obviously, which shows that there is still a small amount of the olefin group in the lignin cross−linked polymer. Figure 10 shows the 1 H NMR of the chemical structure of the dissolved part for the lignin cross−linked polymers. The 1 H NMR of the lignin cross−linked polymer was compared with that of lignin−1−1. The two strong proton signals at 5.0 ppm and 5.82 ppm in the 1 H NMR of ligin−1−1 are attributed to the protons of olefin [ 34 , 34 ]. After the thiol–ene click reaction, the proton signals at 5.0 ppm and 5.82 ppm in the 1 H NMR of lignin−1−1−2SH and lignin−1−1−4SH appeared weak and almost disappeared, which indicates that the thiol–ene click reaction between OFL and cross−linker (3,6−Dioxa−1,8−octanedithiol, pentaerythritol tetra(3−mercaptopropionate)) is incomplete, because the precise proportion of olefin and sulfhydryl is difficult to control. Figure 11 shows the FT−IR of the EC–lignin cross−linked polymer compared with that of lignin−1−1, EC−1−1 and undecenoic acid. When the thiol–ene click reaction occurred, the peak at 1675 cm −1 in the FT−IR of EC−ligin−2SH and EC−ligin−4SH was weaker than those of lignin−1−1 and EC−1−1, which shows that there is still a small amount of the olefin group in the EC–lignin cross−linked polymer. Figure 12 shows the 1 H NMR of the chemical structure of the dissolved part for the EC–ligin cross−linked polymers, and the 1 H NMR of the lignin cross−linked polymers compared with that of lignin−1−1 and EC−1−1. The proton signals at 5.0 ppm and 5.82 ppm in the 1 H NMR of EC−ligin−2SH and EC−ligin−4SH were low, which was caused by the thiol–ene click reaction. The tensile stress and strain at break of the EC and cross−linked polymers were investigated and the results are shown in Figure 13 , Figure 14 , Figure 15 and Figure 16 . As seen in Figure 13 , the tensile stress and strain at break of EC were 146.8 MPa and 1.32%, respectively. For the cross−linked polymers, the tensile stress and strain at break increased when the olefin group concentration increased in OFE. When the olefin group concentration increased from 2.9200 to 3.7000 mmol/g, the tensile stress at break increased from 16.41 MPa to 23.59 MPa, as seen in Table 3 , while the tensile strain at break firstly increased from 17.25% to 19.41% and then decreased to 18.12%. When the pentaerythritol tetra(3−mercaptopropionate) was used as cross−linker, as seen in Figure 15 , the tensile stress at break increased from 6.86 MPa to 13.94 MPa, and the tensile strain increased from 14.85% to 22.38. EC−lignin−4SH showed excellent tensile properties compared to EC−lignin−2SH. The tensile stress at break for EC−lignin−4SH was 15.19 MPa, which is higher than that of EC−lignin−2SH at 13.72 MPa, and the tensile strain at break increased from 15.12% to 17.25%. The gradual enhancement in mechanical properties is positively correlated with the olefin group concentration. The mechanical properties of the lignin cross−linked polymers (lignin−1−1−2SH and lignin−1−1−4SH) were not tested because the polymers were in powder form. Compared with EC, the tensile strength of the cross−linked polymers decreased significantly, but the tensile strain increased sharply, which is caused that the flexible aliphatic hydrocarbon chain from undecylenic acid as the branched chains of EC and lignin increased the distance among the main chains, weakened the interaction force among the main chains and reduced the hydrogen bond interaction in the matrix of EC and lignin. The flexible aliphatic hydrocarbon chain from undecylenic acid contributed to plasticization by functioning as an internal plasticizer. The C gel of the cross−linked polymers were detected and the results are shown in Table 3 . The results show that C gel is positively correlated to the tensile stress for the same type of polymers. For the EC−n−2SH cross−linked polymers, when the tensile stress increased from 16.41 MPa to 23.59 MPa, the C gel increased from 94.9% to 98.0%. The other cross−linked polymers showed the same relationship between the gel content and the tensile strength. The microstructure of the cross−linked polymers were investigated using a Leica DM750M optical microscope. As seen in Figure 17 , the EC cross−linked polymers showed a uniform surface structure without large cracks or micropores, while the EC−lignin polymers showed an uneven surface structure with many large cracks and micropores, which may be due to solvent evaporation. Figure 18 shows the thermal stability of the EC and cross−linked polymers. Only one thermal degradation stage occurred at 360–370 °C for EC and all cross−linked polymers. It has been reported that, when cellulose−based polymers are degraded at high temperatures, their polymerization degree is reduced, the chemical composition also changes and the carbonyl groups increase [ 40 , 41 ]. When cellulose−based polymers are fully degraded, carbon monoxide, carbon dioxide, ethylene, water and carbon are produced [ 42 , 43 ]. Table 4 shows the thermal degradation temperature (T d ), the peak value of the thermal degradation temperature (T P ) and the char residue of the cross−linked polymers. The thermal stability of the cross−linked polymers (EC−1−1.0−2SH, EC−1−1.0−4SH and EC−1−1.3−4SH) was higher than that of EC. The T d , T P and char residue of the cross−linked polymers increased compared with those of EC. This is because OFE and the cross−linkers, such as 3,6−Dioxa−1,8−octanedithiol and pentaerythritol tetra(3−mercaptopropionate), contain thermostable ester groups, which makes the cross−linked polymers difficult degrade. In order to further investigate the composition of the thermal degradation products, TGA−FTIR was carried out. Figure 19 and Figure 20 show the 3D and 2D FT−IR, respectively, of EC (a), EC−1−1.0−2SH (b), EC−1−1.0−4SH (c) and EC−1−1.3−4SH (d). The infrared characteristic absorption peak of the gas phase of the thermal degradation products can be clearly observed in Figure 20 . The data were collected at the fastest decomposition temperature of 340 °C. The infrared characteristic absorption peaks at 3684, 2979, 2306 and 1747, 1391, 1057 cm −1 were attributed to H 2 O, aliphatic hydrocarbon segments, CO 2 and degradation products containing ester groups, respectively [ 44 , 45 , 46 ]. The existence of ester groups in the cross−linked polymers and degradation products make them more thermally stable."
} | 5,195 |
31245773 | PMC6589526 | pmc | 8,612 | {
"abstract": "Abstract Legumes develop root nodules that harbor endosymbiotic bacteria, rhizobia. These rhizobia convert nitrogen to ammonia by biological nitrogen fixation. A thorough understanding of the biological nitrogen fixation in legumes and its regulation is key to develop sustainable agriculture. It is well known that plant hormones affect nodule formation; however, most studies are limited to model legumes due to their suitability for in vitro, plate‐based assays. Specifically, it is almost impossible to measure the effects of exogenous hormones or other additives during nodule development in crop legumes such as soybean as they have huge root system in soil. To circumvent this issue, the present research develops suitable media and growth conditions for efficient nodule development under in vitro, soil‐free conditions in an important legume crop, soybean. Moreover, we also evaluate the effects of all major phytohormones on soybean nodule development under identical growing conditions. Phytohormones such as abscisic acid (ABA) and jasmonic acid (JA) had an overall inhibitory effect and those such as gibberellic acid (GA) or brassinosteroids (BRs) had an overall positive effect on nodule formation. This versatile, inexpensive, scalable, and simple protocol provides several advantages over previously established methods. It is extremely time‐ and resource‐efficient, does not require special training or equipment, and produces highly reproducible results. The approach is expandable to other large legumes as well as for other exogenous additives.",
"conclusion": "5 CONCLUSION Nodule development in legumes directly affects nitrogen fixation efficiency during plant growth. Here, we present a method for determining the effects of ABA, auxin, BAP, GA, BR, SA, and JA on soybean nodulation that is rapid, accurate, technically simple, and requires minimal resources. This method provides several advantages over other methods as these approaches do not require time‐consuming additional steps such as changing solvents and maintaining hormonal concentrations day‐by‐day nor use of large containers, which require large quantities of hormones, space, and tedious handling. Continued manipulations often increases potential technical errors. The method can also be applied to roots coming out of seeds, if generating the transgenic hairy roots is not a requirement. Moreover, the standardization of hormone concentrations and the description of resultant phenotypes will support further targeted studies, and in combination with additional genetic and genomic tools being developed in multiple labs, will greatly increase its use for the study of the effects of exogenous factors affecting nodulation in soybeans as well as in other larger legumes. Similarly, for the study of gene silencing, overexpression or gene editing on nodule development, Agrobacterium cells ( K599 ) expressing the appropriate gene constructs can be used in this method by mixing it with a nitrogen‐free nutrient solution, which will result in the generation of transgenic soybean hairy roots.",
"introduction": "1 INTRODUCTION Nitrogen is an essential element for plant growth, development and productivity. Improving the nitrogen availability to plants results in significant increases in crop yields. Although present in huge quantities in the atmosphere (78% of earth's atmosphere), this nitrogen is not available to plants, unless fixed by biological nitrogen fixation (BNF). BNF happens by the activity of specialized groups of bacteria called rhizobia, which exists as symbionts with the roots of leguminous plants in specialized structures called root nodules. Root nodule formation is a sophisticated process that requires strict synchronization of bacterial infection and growth as well as plant organogenesis and nodule development. The successful interactions between the host plant and the soil bacteria of Rhizobium spp. begin with the secretion of flavonoids from plant roots. In response, the rhizobia produce lipochito‐oligosaccharides, known as nodulation factors or nod factors (NFs). The secreted NF from symbiotically compatible rhizobia directly bind with and activate the nod factor receptors (NFRs) of plants, which are LysM (Lysine motif)‐containing receptor like kinases (Limpens et al., 2003 ; Madsen et al., 2003 ; Radutoiu et al., 2003 ). NFR activation induces root hair deformation, curling, and consequently entrapment of bacteria in those root hairs. The entrapped bacteria form infection threads, which enters in the root hair cells and elongates from the root hair tips to the inner cells to initiate early infection. Additionally, active NFRs stimulate downstream signaling pathways through nuclear Ca 2+ oscillations and Ca 2+ spiking to begin nodule organogenesis from the cortical cells (Gleason et al., 2006 ; Tirichine et al., 2006 ). All these signaling and organogenesis events are considerably affected by the hormonal balance in plants (Ryu, Cho, Choi, & Hwang, 2012 ). Phytohormones both positively and negatively regulate nodulation and nitrogen fixation in legumes. The positive effects of plant hormones auxins and cytokinins in nodule development have been established for a long time. Auxins are a prerequisite during the development and differentiation of nodule primordia and the formation of the vasculature within the nodules (Kohlen, Ng, Deinum, & Mathesius, 2018 ; Takanashi, Sugiyama, & Yazaki, 2011 ; Thimann, 1936 ). Similarly, cytokinins are responsible for the cortical cell division, differentiation, and nodule organogenesis (Frugier, Kosuta, Murray, Crespi, & Szczyglowski, 2008 ; Gonzalez‐Rizzo, Crespi, & Frugier, 2006 ; Reid et al., 2017 ). In addition to auxins and cytokinins, gibberellins (gibberellic acid, GA) are also involved during regulation of nodulation likely via their cross talk with cytokinin signaling pathways (Maekawa et al., 2009 ). Conversely, stress‐related hormones such as jasmonic acid (JA), salicylic acid (SA), and abscisic acid (ABA) typically reduce nodulation by disrupting NF‐induced Ca 2+ spiking and downstream signaling pathways (Martinez‐Abarca et al., 1998 ; Nakagawa & Kawaguchi, 2006 ; Phillips, 1971 ). Nodulation is an energy‐demanding process, therefore to control the number of nodules, legumes have evolved a systemic auto‐regulation of nodulation (AON) as well as local hormonal inhibition of nodulation, which are considered the negative feedback systems. The molecular mechanism of AON has been actively investigated using different supernodulation mutants, such as hyper nodulation and aberrant root 1 (har1), super numeric nodules 1 (sunn) , and nodule autoregulation receptor kinase (nark) in Lotus japonicus, Medicago truncatula , and Glycine max , respectively (Ferguson et al., 2010 ; Krusell et al., 2002 ; Nishimura et al., 2002 ; Oka‐Kira & Kawaguchi, 2006 ; Oka‐Kira et al., 2005 ; Searle et al., 2003 ). Numerous studies suggest that auxin, JA, and brassinosteroids (BRs) modulate AON signaling pathways (Kinkema & Gresshoff, 2008 ; Nakagawa & Kawaguchi, 2006 ; Oka‐Kira et al., 2005 ; Terakado, Yoneyama, & Fujihara, 2006 ) whereas ABA, JA, ethylene, and SA appear to act as during local inhibitory regulation of nodulation (Biswas, Chan, & Gresshoff, 2009 ; Ding et al., 2008 ; Oldroyd, Engstrom, & Long, 2001 ; Penmetsa & Cook, 1997 ; Sun et al., 2006 ). Coordinated action of the hormone levels and signaling controls nodule organogenesis and mature nodule development. Previous studies on the hormonal control of nodulation are based on physiological approaches using a variety of leguminous species and exogenous application of phytohormones to study their effect on nodule formation. For example, exogenous application of cytokinins and auxins to pea root cortical explants induced cell proliferation required for root nodule formation (Libbenga, van Iren, Bogers, & Schraag‐Lamers, 1973 ). On the other hand, exogenous ABA reduced the number of root nodules by inhibiting the cortical cell divisions during nodule organogenesis (Phillips, 1971 ). GA, an important growth regulator, also modulates root nodule formation in legumes by exogenous application (Maekawa et al., 2009 ). SA, a key molecule in plant disease resistance, was shown to inhibit the indeterminate nodules of Vicia sativa , but not the determinate nodules of Lotus japonicus after exogenous application (van Spronsen et al., 2003 ). Although each hormone had a characteristic physiological effect, it was evident that different hormones may also follow additive, synergistic, or antagonistic interactions to regulate nodule formation. The availability of excellent mutant populations in plants such as L. japonicus and M. truncatula provided genetic evidence for the effect of hormones on nodule formation. These plants serve as useful models due to their modest genome sizes, short seed‐to‐seed generation time, high plant transformation efficiency, and the formation of a restricted number of root nodules. These traits are also useful for performing highly controlled in vitro assays with multiple exogenous additives, which has led to several key discoveries (Bensmihen, 2015 ; Maekawa et al., 2009 ; Nakagawa & Kawaguchi, 2006 ; Stacey, McAlvin, Kim, Olivares, & Soto, 2006 ; Sun et al., 2006 ). Conversely, it is difficult to perform similar assays with exogenous additives in crop legumes such as soybean, due to their large stature, long life cycle, the formation of a huge number of root nodules, and the requirement of soil for nodule formation. To circumvent these problems, split‐root system was used in soybean to study the effects of various exogenous variable during rhizobia‐legume symbioses (Chaillou, Rideout, Raper, & Morot‐Gaudry, 1994 ; Gil‐Quintana et al., 2013 ; Lin, Gresshoff, & Ferguson, 2012 ; Singleton & Bohlool, 1984 ). Most recently, a new split‐root system was developed for continuous monitoring of soybean roots throughout the whole experiment after rhizobial infection (Hidalgo, Ruiz‐Sainz, & Vinardell, 2018 ). Although these protocols are useful, it is still technically challenging to do some of these experiments and is difficult to apply to a large plant population. The goals of this study were to identify suitable media and growth conditions for efficient nodule development under in vitro conditions in soybean G. max , which is an important crop but not amenable to the standard plate‐based assays used to study nodule formation in Medicago or Lotus sp. In addition, the effect of each of the major phytohormones was analyzed on nodule development under a standard set of conditions. The results presented in the following sections describe a set of optimum growth and treatment conditions for soil‐free soybean nodulation and effects of phytohormones on it, which will be useful for the community at large.",
"discussion": "4 RESULTS AND DISCUSSION 4.1 Optimization of soil‐free nodule development in soybean Symbiotic nitrogen fixation in nodules plays a key role in the maintenance of soybean seed production. While in plants such as Medicago , their small stature allows for growth under in vitro conditions on sterile media plates, the large size of soybean plants precludes such a possibility. The study of soybean nodule development in soil by using different additives is relatively hard, inaccurate, and expensive. To overcome these issues, soil‐free nodule production under in vitro condition is a suitable choice. We optimized a method where the nodules were allowed to develop in germination papers, after initial infection in soil (Figure 1 ). Briefly, trifoliate leaves with a stalk were cut from 2‐week‐old plants grown on soil and inserted in the middle of sterilized rock wool cubes (Govindarajulu et al., 2009 ; Libault et al., 2009 ). The rock wool in cubes causes abrasion, which promotes the development of hairy roots, which are well developed after 2 weeks of growth. After this time plants with hairy roots were transferred to small pots (85 mm × 85 mm) containing soilrite (vermiculite:perlite:sand in a 3:1:1 ratio) and allowed to grow for 1 week. For nodule development, plants in pots were infected with compatible rhizobium strains. Infection of plants in soilrite was critical, because infecting plants in germination papers by direct application of rhizobia would result in uncontrolled bacterial infection on the paper itself. Plants were transferred 72 h post‐infection between the two sheets of pre‐wet germination paper and were arranged in rolls. One plant was placed per roll and each 250 ml beaker can accommodate six rolls. The plants were allowed to grow in nitrogen‐free nutrient solution and nodule number was counted after 4 weeks of further growth. For nodule development in soilrite, the plants can continue to grow in pots after infection, but will require one ½ gallon pot per plant. Achieving uniform root growth and nodule formation in soybean under in vitro conditions has been a challenge as it shows enormous inconsistencies in nodule numbers. In our experiments, the time of bacterial infection and the duration of subsequent growth of plants in soilrite were critical for efficient nodule development at later stages. To optimize the conditions for consistent, reproducible results, we transferred plants from soilrite to germination paper rolls at different time points after rhizobium infection and counted the nodules after 4 weeks of growth. Plant relocation time after rhizobium infection had a huge effect on nodule formation (Figure 2 ). No nodules were formed if we transferred the plants from 0 to 24 h after rhizobium infection, whereas ~2, 8, and 12 nodules on average were formed per plant if the plants were transferred 48, 72, and 96 h after rhizobium infection, respectively. Based on these results, all through our experiments, we have transferred the plants from soilrite to germination paper rolls 72 h post‐infection as it generates a reasonable number of nodules needed for any comparative analysis. Figure 2 Optimization of nodule development time after infection. Rhizobia‐infected soybean hairy roots were transferred to pre‐wet germination paper rolls at different time points. Nodule number was counted 4 weeks after infection. All experiments were repeated two times independently and data were averaged. Each replicate consisted of 24 plants. Asterisks denote significant difference, * p < 0.5, Student's t test We are cognizant of the fact that the need to grow roots in soilrite for up to 72 h post‐infection excludes the possibility of assaying the effects of various additives during the early stages of infection such as on root hair curling, or infection thread formation. The method described in this research is therefore suitable for the study of additives at the later stages of nodule development only. However, as detailed in the next sections, it does result in reproducible and consistent effects of exogenous additives on nodule developments and is extremely resource‐ and cost‐effective. 4.2 Optimization of hormone treatment time For this study, we focused on assaying the effects of various plant hormones as exogenous additives, as these have a significant influence not only on nodule initiation but also on organogenesis. Similar assays can be optimized for other additives such as salt, mannitol (for osmotic stress) etc. as needed. The hormone treatments were started when the plants were transferred from soilrite, and germination paper rolls containing plants were transferred to either control or treatment media. An obvious effect on nodule development was observed for each of the plant hormones. 4.3 Effect of ABA on root nodule development Abscisic acid regulates multiple aspects of rhizobia‐legume symbiosis (Bano & Harper, 2002 ; Bano, Harper, Auge, & Neuman, 2002 ; Cho & Harper, 1993 ; Phillips, 1971 ). Because the addition of exogenous ABA causes a marked increase in the endogenous ABA levels in plants, the application of ABA to media in which plants are growing is a suitable system to study its effect on nodulation. In L. japonicus , the number of root nodules is reduced at higher ABA concentrations, and increased in the presence of abamine (which causes lower ABA concentrations) suggesting its inhibitory role during nodule development (Suzuki et al., 2004 ). In soybean, exogenous ABA decreased nodule number in both the wild‐type and a supernodulation mutant (Bano & Harper, 2002 ; Bano et al., 2002 ; Cho & Harper, 1993 ). We evaluated the effect of ABA on soybean nodulation in a concentration‐dependent manner using ABA concentrations ranging from 2 to 25 μM; concentrations lower than 2 μM were ineffective under our growth conditions. An equimolar amount of ethanol was used as a control in all assays. Nodule formation showed extreme sensitivity to exogenously added ABA. On average, seven nodules were formed per plants after 4 weeks of growth under control conditions, which was reduced by 15%–30% in presence of 2–15 μM ABA (Figure 3 ). It was noticeable that plants treated with ABA had fewer large (>2 mm in diameter) and medium (0.5–2 mm in diameter) compared to the control plants but a comparable number of small nodules (<0.5 mm in diameter) were produced (Table S1). Concentrations higher than 15 μM severely affected nodule formation, with the nodule numbers reduced by more than 60% per plant in the presence of 20–25 μM ABA. At these higher concentrations, all types of nodules (large, medium, and small) were significantly affected. These data confirm the inhibitory effects of exogenous ABA on soybean nodule formation, and by extension of other abiotic stresses, which cause an increase in ABA concentrations in planta. Figure 3 Effect of exogenous ABA on nodule formation. Rhizobia‐infected soybean hairy roots were treated with different concentrations of ABA when transferred to the nitrogen‐free media. ABA concentration was maintained throughout the experiment (4 weeks). Nodule number was counted 4 weeks after infection. All experiments were repeated two times independently and data were averaged. Each replicate consisted of 24 plants. Asterisks denote significant difference, * p < 0.5, Student's t test 4.4 Effect of auxin (IAA) and cytokinin (BAP) on nodule development Several studies have demonstrated that auxin and cytokinin intricately control root nodule formation. IAA (indole‐3‐acetic acid), a native auxin in plants, derived from the phenylpropanoid biosynthetic pathway is a key member of the auxin family. The root nodules have a higher IAA content than uninfected root tissues, suggesting a role for IAA in nodule development (Thimann, 1936 ). The higher IAA content of infected roots promotes root cells to undergo cell division, elongation, differentiation, and vascular bundle formation to develop nodules (Boivin, Fonouni‐Farde, & Frugier, 2016 ; Kuppusamy et al., 2009 ; Nagata & Suzuki, 2014 ; Suzaki et al., 2012 ). Moreover, rhizobia alter the root auxin balance, which is a prerequisite for nodule formation (Boot, van Brussel, Tak, Spaink, & Kijne, 1999 ; Hirsch, Bhuvaneswari, Torrey, & Bisseling, 1989 ; Mathesius et al., 1998 ; Pacios‐Bras et al., 2003 ; Wasson, Pellerone, & Mathesius, 2006 ). Besides auxin content, rhizobia‐legume symbiosis is also regulated by shoot‐to‐root auxin transport, which is an active transport process involving auxin efflux protein complexes (van Noorden, Ross, Reid, Rolfe, & Mathesius, 2006 ). Overall, it is well established that the changes in auxin accumulation and transport, both are essential for lateral root development, nodule primordium activation, and nodule organogenesis (Mathesius et al., 1998 ; Suzaki et al., 2012 ; Takanashi et al., 2011 ). To test the effect of exogenous auxins on soybean nodule formation, plants were treated with different IAA concentrations ranging from 10 nM to 100 μM. IAA at 10 nM caused the most obvious phenotypic differences in the nodule number, although all treatment conditions resulted in higher nodule number per plant. Approximately four times more nodules were formed per plant at 10 nM IAA compared to control plants with a huge increase in the number of small‐ and medium‐size nodules (Figure 4 a). Two times more nodules were formed compared to control conditions in the presence of 1 μM IAA, whereas a modest increase in nodule numbers was observed in the presence of 100 μM IAA in the media. These results indicate that even at a very high concentration, auxin still has a limited but positive effect on nodule formation. Furthermore, 10 nM exogenous auxin is an optimal concentration for increased nodule numbers in soybean. Figure 4 Effect of exogenous auxin (IAA) and cytokinin (BAP) on nodule formation in soybean. Rhizobia‐infected soybean hairy roots were treated with different concentrations of (a) IAA and (b) BAP, when transferred to the nitrogen‐free media. The hormone concentrations were maintained throughout the experiment (4 weeks). Nodule number was counted 4 weeks after infection. All experiments were repeated two times independently and data were averaged. Each replicate consisted of 24 plants. Asterisks denote significant difference, * p < 0.5, Student's t test Extensive research on L. japonicus and M. truncatula have demonstrated that cytokinins (CKs) are key players in the regulation of rhizobium infection and nodule development (Frugier et al., 2008 ; Gonzalez‐Rizzo et al., 2006 ; Tirichine et al., 2007 ). The activation of NF signaling pathway rapidly induces CK accumulation and response in hairy roots (Buhian & Bensmihen, 2018 ; Gamas, Brault, Jardinaud, & Frugier, 2017 ; Murray et al., 2007 ). The exogenous application of CKs also promotes cortical cell divisions and the expression of early nodulation markers in different legumes (Bauer, Ratet, Crespi, Schultze, & Kondorosi, 1996 ; Jimenez‐Zurdo, Frugier, Crespi, & Kondorosi, 2000 ; Mathesius, Weinman, Rolfe, & Djordjevic, 2000 ; Murray et al., 2007 ; Tirichine et al., 2007 ). To study the effect of exogenous applications of cytokinins on soybean nodulation, we included different concentrations of BAP (6‐Benzylaminopurine) (50 nM, 100 nM, 250 nM, 1 μM, and 10 μM) in the media when the plants were transferred from soilrite to germination paper rolls. We detected a clear, concentration‐dependent effect of BAP on nodulation. Approximately 2.5 times more nodules were formed in the presence of 50 nM BAP, whereas a modest increase was observed in the presence of 100 M BAP, compared to the control conditions (Figure 4 b). Interestingly, BAP concentrations higher than 250 nM were inhibitory and lead to the development of fewer nodules compared to control media grown plants. For example, ~80% less nodule developed in the presence of 10 μM BAP. These data suggest that nodule formation in soybean strongly regulated by precisely controlled CKs level. 4.5 Effect of GA and BR on root nodule development Gibberellic acid is one of the vital growth regulators in higher plants. Several studies have highlighted the involvement of GA in the regulation of the rhizobium‐legume symbiotic interaction (Ferguson, Ross, & Reid, 2005 ; Lievens et al., 2005 ; Maekawa et al., 2009 ; McAdam, Reid, & Foo, 2018 ). Interestingly, both positive and negative effects of GA have been reported in previous studies. For examples, one study reported the inhibitory effects of an exogenous application of potassium gibberellate on nodule formation in Phaseolus vulgaris (Thurber, Douglas, & Galston, 1958 ). On the other hand, nodules aborted in pea mutant ( na‐1 ) which was deficient in GA 3 , but were re‐established by the application of exogenous GA 3 (Ferguson et al., 2005 ) implying requirement of GA for nodule formation. This discrepancy in the results could be either the concentration of GA used or the specific plant species used in the assays. To assess the effect of GA 3 on soybean nodule formation, plants post‐infection were treated with different GA 3 concentrations (10 nM to 1 μM) and compared with control media with no added GA 3. Plants treated with 10 nM and 100 nM GA 3 showed considerably increased nodule formation with ~40% and ~20% more nodules, respectively. Interestingly, most of the nodules in these treatments were small (Table S1). However, similar to what was observed for cytokinins, at higher concentrations (1 μM) of GA 3 , the nodule formation was severely affected and ~70% fewer nodules were formed on treated roots compared with control media (Figure 5 a). These data suggest that the reported discrepancies in the previous publications could be due to the different GA concentrations used in different assays. These observations suggest that the endogenous GA concentration in plants is tightly regulated to achieve effective nodulation. Figure 5 Effect of gibberellic acid (GA 3 ) and brassinosteroids (brassinolide, BR) on nodule formation in soybean. Rhizobia‐infected soybean hairy roots were treated with different concentrations of (a) GA 3 and (b) Brassinolide, when transferred to the nitrogen‐free media. The hormone concentrations were maintained throughout the experiment (4 weeks). Nodule number was counted 4 weeks after infection. All experiments were repeated two times independently and data were averaged. Each replicate consisted of 24 plants. Asterisks denote significant difference, * p < 0.5, Student's t test Brassinosteroids play pivotal roles during many aspects of plant growth and development (Clouse & Sasse, 1998 ; Wei & Li, 2016 ). BRs affect both cell proliferation and cell elongation to control shoot and root lengths and hypocotyl growth in plants (Gonzalez‐Garcia et al., 2011 ; Tong et al., 2014 ). Although the effects of BRs are well documented on root and shoot growth, their effects on nodule development in different legumes are not as well described. In one example, BRs affected nodule development by reducing lateral root numbers in pea as is evident from the BR synthesis mutants lk and lkb and the BR response mutant lka (Ferguson et al., 2005 ). To test the effect of BRs on soybean nodule development, plants were treated with media containing different brassinolide (active BR) concentrations ranging from 10 nM to 1 μM. In comparison to plant grown on control media, a concentration‐dependent increase in nodule numbers was observed in plants grown on BR containing media. Highest nodule numbers were seen in response to 100 nM brassinolide, where ∼85% more nodules were present compared to the control media roots (Figure 5 b). At lower (10 nM) and higher (1 μM) concentrations of exogenous BL, ∼50% more nodules were formed, suggesting the BRs are positive regulators of nodule formation. 4.6 Effect of SA and JA on root nodule development Salicylic acid strongly affects nodule formation at early stages of nodulation (Sato et al., 2002 ). The exogenous application of SA resulted in both reduced and delayed nodule formation on Medicago roots inoculated with wild‐type S. meliloti . Moreover, exogenous SA that inhibited nodulation also strongly reduced the growth of the bacterial symbiont (Martinez‐Abarca et al., 1998 ). The effect of SA was additionally documented in different legumes. For example, the inhibition of nodule formation after exogenous SA treatment was observed in plants like vetch ( Vicia sativa ), pea ( Pisum sativum ), and white clover ( Trifolium repens ) (van Spronsen et al., 2003 ). This study shows that the exogenous SA inhibits indeterminate but not determinate nodulation (van Spronsen et al., 2003 ), which is inconsistent with two other previous reports (Lian, Zhou, Miransari, & Smith, 2000 ; Sato et al., 2002 ), where they showed that exogenous application of SA to soybean, which forms determinate nodules, reduced nodulation. Another report demonstrated that SA enhances the efficiency of nitrogen fixation and assimilation in Cicer arietinum (Hayat, Hayat, Alyemeni, & Ahmad, 2014 ). To examine the role of exogenous SA in soybean nodulation in our system, we used different concentrations of SA ranging from 10 μM to 1 mM in the exogenous media. All SA‐treated soybean plants exhibited a significantly higher nodule number, ranging from 57% to 76% compared to the control media (Figure 6 a). The differences were more prominent in case of higher concentration of SA treatment (100 μM to 1 mM). Our data thus suggest that exogenous SA levels positively regulate nodulation in soybean at these concentrations. Figure 6 Effect of salicylic acid (SA) and jasmonic acid (JA) on nodule formation in soybean. Rhizobia‐infected soybean hairy roots were treated with different concentrations of (a) SA and (b) JA, when transferred to the nitrogen‐free media. The hormone concentrations were maintained throughout the experiment (4 weeks). Nodule number was counted 4 weeks after infection. All experiments were repeated two times independently and data were averaged. Each replicate consisted of 24 plants. Asterisks denote significant difference, * p < 0.5, Student's t test Jasmonic acid negatively regulates plants’ response to the rhizobial bacterial signal, NF (Sun et al., 2006 ). Primarily, JA inhibits nodule formation by suppressing calcium spiking and the frequency of calcium oscillations to modulate the NF‐induced gene expression (Sun et al., 2006 ). In addition, even in the shoots, the AON pathway is modulated by JA (Kinkema & Gresshoff, 2008 ). However, a recent study by suppression of allene oxide cyclase in M. truncatula , an enzyme involved in committed step in JA biosynthesis, suggested that jasmonates are not involved in the development and function of root nodules (Zdyb et al., 2011 ). To examine if exogenous JA has any effect on soybean nodulation, we tested its effects at different concentrations (10 μM, 100 μM, and 1 mM). Surprisingly, at low concentration (10 μM), exogenous JA showed a positive effect and approximately twice as many nodules were formed compared to the control media. However, the nodule number was significantly inhibited (up to 95%) after increasing the JA concentrations from 10 to 100 μM. No nodules were formed in the presence of 1 mM JA (Figure 6 b). These data suggest that regulation of nodule formation by JA is complex and is highly dependent on the exogenous concentrations. As hormones such as JA and SA not only modulate the signaling pathways in plants but are also a core part of plant–microbe interaction (Durner, Shah, & Klessig, 1997 ; Koornneef & Pieterse, 2008 ; Niki, Mitsuhara, Seo, Ohtsubo, & Ohashi, 1998 ), it is expected that different concentrations may have altered effects on growth and development versus survival."
} | 7,684 |
28004884 | PMC5270724 | pmc | 8,614 | {
"abstract": "Summary Growth of biodiesel industries resulted in increased coproduction of crude glycerol which is therefore becoming a waste product instead of a valuable ‘coproduct’. Glycerol can be used for the production of valuable chemicals, e.g. biofuels, to reduce glycerol waste disposal. In this study, a novel bacterial strain is described which converts glycerol mainly to ethanol and hydrogen with very little amounts of acetate, formate and 1,2‐propanediol as coproducts. The bacterium offers certain advantages over previously studied glycerol‐fermenting microorganisms. Anaerobium acetethylicum during growth with glycerol produces very little side products and grows in the presence of maximum glycerol concentrations up to 1500 mM and in the complete absence of complex organic supplements such as yeast extract or tryptone. The highest observed growth rate of 0.116 h −1 is similar to that of other glycerol degraders, and the maximum concentration of ethanol that can be tolerated was found to be about 60 mM (2.8 g l −1 ) and further growth was likely inhibited due to ethanol toxicity. Proteome analysis as well as enzyme assays performed in cell‐free extracts demonstrated that glycerol is degraded via glyceraldehyde‐3‐phosphate, which is further metabolized through the lower part of glycolysis leading to formation of mainly ethanol and hydrogen. In conclusion, fermentation of glycerol to ethanol and hydrogen by this bacterium represents a remarkable option to add value to the biodiesel industries by utilization of surplus glycerol.",
"conclusion": "Conclusion As A. acetethylicum strain GluBS11 T naturally has a high tolerance towards elevated glycerol concentrations, it could be a potentially useful agent for treating glycerol‐rich wastewaters coming from the biodiesel industries. Our study shows that the maximum initial glycerol concentration that did not inhibit growth and metabolic activity of the cells was 1500 mM of pure glycerol. However, 100 mM of initial glycerol concentration was optimal for efficient conversion of glycerol to ethanol (79% conversion efficiency) and hydrogen as compared to other tested initial concentration. Solvent toxicity tests of A. acetethylicum for glycerol and ethanol showed that ethanol is the key solvent that strongly inhibits growth and fermentation activity. Although the strain could produce about 61 mM of ethanol during growth with 100 mM of glycerol, addition of 50 mM of initial ethanol completely inhibited growth (Fig. S1). The inability to accumulate high concentrations of ethanol (more than 60 mM) during growth is possibly a drawback for large‐scale applications. Moreover, pure glycerol was used as a substrate in this study and glycerol derived from biodiesel production might contain compounds that inhibit growth. Yet, due to its high tolerance for glycerol and its fermentation pattern to mainly ethanol and hydrogen, the strain has a high potential for future industrial application in biodiesel industries to convert crude glycerol to value‐added biofuel. Future research should therefore focus on increasing the strain's ability to tolerate ethanol concentrations higher than 60 mM. This could be accomplished, e.g. by genetic engineering of the strain and introducing metabolic pathways for synthesis of oleic acid, which is believed to protect the cell membranes of yeast from the toxic effect of ethanol (You et al ., 2003 ). This would, however, require the development of a genetic system for the strain. It is known that ethanol tolerance in yeast can also be increased by addition of Tween‐80 and oleic acid to growth media, which did not increase ethanol tolerance in A. acetethylicum , however, (data not presented) and probably led to inhibition of growth (Andreasen and Stier, 1954 ). Besides ethanol and hydrogen as main fermentation products, the strain produces very little amount of undesired fermentation products such as acetate and formate; therefore, future efforts in metabolically engineering the strain could aim at deleting the enzymes leading to acetate production, i.e., acetate kinase, and formate production, i.e., formate dehydrogenase, for the production of bioethanol.",
"introduction": "Introduction Fossil fuels are the main source of energy being used worldwide and cover about 80% of the global energy demand (Sarma et al ., 2012 ). Fossil fuels are limited, non‐renewable and associated with many problems such as global warming, ecosystem imbalance and health hazards (da Silva et al ., 2009 ). Therefore, there is a huge demand for alternative energy sources that are renewable, eco‐friendly and sustainable to replace the conventional fossil fuels. Moreover, Campbell and Laherrere ( 1998 ) predicted that petroleum reserves will be completely depleted by 2050 (Nwachukwu et al ., 2012 ). This concern has highlighted the future need for the use of biofuels such as ethanol, biodiesel, butanol, hydrogen or electricity produced from renewable plant biomass as one of the promising alternatives over fossil fuels (Elmekawy et al ., 2013 ; Speers et al ., 2014 ). Therefore, in recent years there has been a significant increase in the production and use of biofuels worldwide, such as biodiesel and bioethanol. In the last decade, the European Union (EU) was the principal biodiesel producer which contributed about 82% of global biodiesel production (Demirbas and Balat, 2006 ). According to the European Biodiesel Board (EBB, 2006 ), the estimated production of biodiesel in 2005 was about 3.2 million tons with a production capacity of 6 million tons (da Silva et al ., 2009 ), which has now increased to about 10.4 million tons in 2013 with a production capacity of 23 million tons. Germany is currently the largest producer and consumer of biodiesel in the EU, producing more than 2.5 million tons in 2013 (EBB, 2013; http://www.ebb-eu.org/stats.php ). The top five global producers of biodiesel are Argentina, Brazil, France, Germany and the United States of America (Sarma et al ., 2012 ). Glycerol (1,2,3‐propanetriol) is a simple trivalent alcohol that results from the natural degradation of the glyceride component of plant cell wall phospholipids or reserve lipids of plant seeds (Roger et al ., 1992 ; Nwachukwu et al ., 2013 ). It is produced in major amounts during transesterification of vegetable oils and animal fats (Solomon et al ., 1995 ; Barbirato et al ., 1997a , b , 1998 ; Colin et al ., 2001 ) and has wide applications in different industries such as food and drinks, toothpaste, cosmetics, toiletries, plastics, tobacco, pulp and paper, paint, leather and textile, pharmaceuticals and automotive (Choi, 2008 ; Nicol et al ., 2012 ; Rossi et al ., 2012 ). The economic value of industrial glycerol has decreased due to the surplus crude glycerol generated during biodiesel production, and it cannot be utilized directly in any industrial applications due to the presence of impurities. Furthermore, it cannot be directly released into the environment without treatment as the cost of such treatment is not economical (Nwachukwu et al ., 2013 ). Recently, fermentative conversion of crude glycerol into valuable products such as, e.g., bioethanol has gained interest for the development of biodiesel‐producing industries, and also for replacing conventional carbohydrate sugars used in industrial microbial fermentation processes to convert it into a broad range of value‐added organic products such as bioethanol (Dharmadi et al ., 2006 ; Rossi et al ., 2012 ). Bioethanol is considered as an alternative to fossil fuels, as it is a renewable, bio‐based resource, and provides the potential to reduce particulate emissions (Hansen et al ., 2005 ). Several microorganisms produce ethanol as a natural fermentation end‐product, sometimes even in a homo‐ethanologenic type of fermentation (Otero et al ., 2007 ). Bioethanol is one of the fermentation products that can be generated from glycerol via anaerobic fermentation, which is more economical than the use of corn or lignocellulosic biomass for bioethanol production (Choi, 2008 ). Moreover, the cost of ethanol produced from glycerol is about 40% lower than when it is produced from corn (Yazdani and Gonzalez, 2007 ). Fermentation of glycerol most often leads to 1,3‐propanediol as reduced end‐product (Homann et al ., 1990 ). Escherichia coli was shown to ferment glycerol anaerobically to ethanol, hydrogen and formate, thus providing a bioagent to produce value‐added biofuel from glycerol (Dharmadi et al ., 2006 ; Trchounian and Trchounian, 2015 ). Other microorganisms are able to perform similar fermentations of glycerol, especially several members of the genus Clostridium (Biebl, 2001 ). Also mixtures of microorganisms, e.g. buffalo slurry, were used to optimize hydrogen production from glycerol (Marone et al ., 2015 ). The main problems with glycerol‐fermenting bacteria are the accumulation of undesired by‐products such as 2,3‐butanediol or butyric acid, and the low tolerance of these strains towards solvents, i.e. glycerol and ethanol. The latter two dissolve cellular membranes at higher concentrations and are therefore lethal for any kind of microorganism. However, yeasts can tolerate ethanol concentrations up to about 120 g l −1 (15% v/v; Lam et al ., 2014 ), which is similar to some bacteria, e.g. Zymomonas sp. (Swings and De Ley, 1977 ). Recently, an anaerobic bacterium representing the new genus Anaerobium within the order Clostridiales was enriched and isolated from sludge samples obtained from a biogas reactor at Odendorf, Germany. Anaerobium acetethylicum strain GluBS11 T was originally described for gluconate fermentation, but it grows also with glycerol under strictly anoxic conditions (Patil et al ., 2015 ). Unlike many other members of the order Clostridiales , fermentation of glycerol by A. acetethylicum mainly produces ethanol and hydrogen and does not coproduce undesired by‐products such as butyrate, 1,3‐propanediol or 2,3‐butanediol under any growth condition (Patil et al ., 2015 ). In this study, we describe the optimum conditions for glycerol fermentation to ethanol and hydrogen by A. acetethylicum using pure glycerol at different concentrations and elucidate the biochemical reactions involved in anaerobic glycerol fermentation based on proteomics and in vitro enzyme assays. Based on our findings, we propose a glycerol fermentation pathway that mainly leads to ethanol and hydrogen and does not involve the formation of 1, 3‐propanediol or 2,3‐butanediol. Application of A. acetethylicum as a potential future candidate for bioethanol and biohydrogen production from glycerol is discussed in the context of the proposed pathway.",
"discussion": "Discussion Glycerol was studied as a substrate for biofuel production mainly because of its abundance, low price and its highly reduced state that makes it prone to generate reduced products like ethanol, hydrogen and also other industrially relevant compounds (Dharmadi et al ., 2006 ; Clomburg and Gonzalez, 2013 ). In comparison with other glycerol‐fermenting strains, A. acetethylicum has higher or at least similar glycerol tolerance, but a low ethanol tolerance. Clostridium pasteurianum converts 691 mM (63.6 g l −1 ) of glycerol to mixed fermentation products including butanol, 1,3‐propanediol, ethanol, butyrate, acetate and lactate, when grown with 1250 mM (114.6 g l −1 ) of initial glycerol concentration (Biebl, 2001 ). Glycerol fermentation by E. coli at an initial glycerol concentration of 108 mM (10 g l −1 ) yields mainly ethanol, hydrogen and formate, similar to A. acetethylicum , but requires complex growth supplements such as yeast extract, tryptone or corn steep liquor (Dharmadi et al ., 2006 ; Murarka et al ., 2008 ). Similarly, Paenibacillus macerans, a glycerol‐fermenting bacterium, produces ethanol and 1,2‐propanediol but depends as well on tryptone as supplement in the growth medium ((Table 4 ); Gupta et al ., 2009 ). In contrast to these reports, Anaerobium acetethylicum did not require additional organic supplements for fermentation of glycerol, except for the defined seven vitamins (Pfennig, 1978 ) present in the medium which include biotin. Biotin could replace yeast extract in cultures of C. pasteurianum , but the overall fermentation time was three times longer than with yeast extract (Biebl, 2001 ). When grown in defined mineral medium containing the seven vitamins and glycerol as substrate, A. acetethylicum had growth rates of 0.101–0.116 h −1 which are about 2–3 times higher than those reported for E. coli (0.04 h −1 ) when grown with glycerol in the presence of tryptone (Murarka et al ., 2008 ), but about four times lower than those reported for P. macerans ((0.4 h −1 ); Gupta et al ., 2009 ). However, growth of A. acetethylicum was not exponential any more after 44 h, indicating that growth is inhibited at this time point. Exponential growth in defined medium is basically possible when grown with gluconate at a growth rate of 0.693 h −1 ; therefore, the medium itself should allow exponential growth as well for cells grown with glycerol (Patil et al ., 2015 ). Table 4 Comparison of ethanol production and growth rates between A. acetethylicum and other anaerobic glycerol‐fermenting bacterial strains Organisms Max. glycerol tolerated (M; (g l −1 )) Organic supplements required Fermentation products Max. growth rate observed (h −1 ) References \n Anaerobium acetethylicum \n 1.5 (138) 7‐vitamins E, H, a,f, pg, CO 2 \n 0.116 This study \n Escherichia coli MG1655 0.108 (10) CTS E, (H, F) a ,s, CO 2 \n 0.04 Dharmadi et al . ( 2006 ) \n Escherichia coli SS1 0.375 (34.5) T, YE E, (H, F), a s, CO 2 \n No data available Adnan et al . ( 2014 ) \n Paenibacillius macerans \n 0.108 (10) T E, H, pg, f 0.4 Gupta et al . ( 2009 ) \n Clostridium pasteurianum \n 1.25 (115) Biotin E, 1,3‐Pd, ButOH, B, A 0.37 Biebl ( 2001 ) Fermentation products H = hydrogen, E = ethanol, F/f = formate, S/s = succinate, Pg/pg = propylene glycol, ButOH = butanol, 1,3‐Pd = 1,3‐propanediol, B = butyrate, A/a = acetate (capital letters – major products and small letter – minor products); complex supplements: CTS = corn steep liquor, T = tryptone, YE = yeast extract. \n a. Production of hydrogen or formate is pH dependent. John Wiley & Sons, Ltd In addition, the maximum glycerol concentration tolerable by A. acetethylicum was 1500 mM (138 g l −1 ), which is 13.8 times higher than the glycerol concentrations tested for E. coli (Dharmadi et al ., 2006 ; Murarka et al ., 2008 ). In a study aimed at optimizing glycerol utilization by E. coli , the optimal glycerol concentration was 375 mM (34.5 g l −1 ; Adnan et al ., 2014 ). However, even though A. acetethylicum can grow at comparably high initial glycerol concentrations, maximally tolerable ethanol concentrations reached during glycerol fermentation were in the range of 60–70 mM, which is similar to the maximal concentrations observed for E. coli (Dharmadi et al ., 2006 ; Murarka et al ., 2008 ), but much lower than the observed maximum ethanol concentration of 342 mM (15.72 g l −1 ) for a growth‐optimized E. coli strain (Adnan et al ., 2014 ). However, it is unclear whether the ethanol concentrations accumulating in cultures of the latter two bacteria do not increase further due to lysis of the cells by ethanol or due to thermodynamic inhibition. This is, however, unlikely as the overall free reaction enthalpy of glycerol conversion to ethanol and hydrogen is negative enough to allow complete conversion of substrate into product (Eq. (1) ).\n (1) C 3 H 8 O 3 → C 2 H 6 O + H 2 + CO 2 Δ G 0 ′ = − 87.6 kJ mol − 1 . \n The free reaction enthalpy required to generate one ATP from phosphorylation of ADP to ATP is about −60 to −70 kJ mol −1 (Schink, 1997 ). Therefore, equation (1) should allow the production of at least 1 mole of ATP per mole of glycerol. When considering the reaction of glycerol fermentation carried out by cultures of A. acetethylicum , which also produced small amounts of side products, the reaction becomes even more favourable allowing an overall ATP yield of 1–2 ATP per mole of glycerol (Eq. (2) ).\n (2) C 3 H 8 O 3 → 0.82 C 2 H 6 O + 1 H 2 + 1.1 CO 2 + 0.11 C 2 H 3 O 2 − + 0.11 H + + 0.03 C 3 H 8 O 2 Δ G 0 ′ = − 145.2 k J mol − 1 when Eq. (1) reaches its equilibrium (ΔG’ = 0), including 60 kJ mol −1 for formation of 1 ATP, the equilibrium constant has a value of about 10 5 , meaning that the reaction equilibrium is far on the side of the reaction products. Thus, thermodynamic inhibition can be ruled out as a possible reason for the incomplete fermentation of glycerol to ethanol and hydrogen, and inhibition of growth by fermentation metabolites is very likely, i.e., via solvent toxicity. A. acetethylicum could ferment glycerol ranging from 10 to 1500 mM initial concentrations, but ethanol production did not exceed 63 mM (Fig. 3 ). Therefore, increased concentrations of glycerol in the growth medium did not increase ethanol production beyond this latter concentration and growth was most likely inhibited by the ethanol toxicity. Ethanol is known as a growth‐inhibiting agent for bacteria as it acts as hydrophobic stressor, especially at concentrations higher than 25% (w/v; 5400 mM), and therefore destabilizes biological membranes by weakening hydrophobic interactions (reviewed in Cray et al ., 2015 ; Ingram, 1990 ). This effect on hydrophobic interactions also causes a reduction of water activity, which was shown to induce water stress in fungi (Hallsworth et al ., 1998 ). This could explain the higher ethanol concentration observed after 166 h of incubation in late stationary to decline phase (62 mM) compared with the ethanol concentration present after 91 h in early stationary phase (38 mM). Most likely, cells were partially lysed by ethanol after they reached stationary phase while still being metabolically active and continuing to ferment glycerol to ethanol and hydrogen. This lytic effect of ethanol might also be reflected by the fact that growth yields decreased with increasing substrate and ethanol concentrations. Consequently, a certain proportion of assimilated substrate was possibly underestimated, which might explain the incomplete electron recoveries at glycerol concentrations higher than 500 mM. Growth could also be inhibited through acidification of the medium. At a substrate concentration of 100 mM of glycerol, the pH dropped from initial 7.2 to 6.3 at the end of growth and earlier investigations revealed a pH range of 6.5–8.5 of strain GluBS11T for growth with gluconate (Patil et al ., 2015 ). Therefore, glycerol fermentation could possibly be improved by increasing the buffer strength. In this study, the fermentation of glycerol to ethanol, CO 2 and hydrogen by A. acetethylicum was biochemically characterized (Fig. 4 ). Glycerol fermentation by A. acetethylicum also produced small amounts of 1,2‐propanediol (Table 1 ). Clomburg and Gonzalez ( 2013 ) reported that 1,2‐propanediol is derived from dihydroxyacetone phosphate. The enzyme methylglyoxal synthase which dephosphorylates dihydroxyacetone phosphate to methylglyoxal was identified in the proteome of A. acetethylicum (Table 3 ). Methylglyoxal (2‐oxopropanal) could theoretically be reduced to 1,2‐propanediol via acetol (hydroxyacetone) or lactaldehyde (Clomburg and Gonzalez, 2011 ). Both pathways involve glycerol dehydratase and aldehyde oxidoreductase, of which one at least the gene for glycerol dehydratase could be identified in the genome of A. acetethylicum (glycerol dehydratase large subunit Ga0116910_100557). 2‐oxopropanal could therefore possibly be reduced to 1,2‐propanediol by one or both of the aforementioned pathways (Fig. 4 ). 2‐oxopropanal is also known as a highly toxic metabolite in bacterial cells and could therefore be inhibitory for growth of A. acetethylicum (Booth et al ., 2003 ; Clomburg and Gonzalez, 2011 ). Despite the fact that HPLC chromatograms of culture supernatants of A. acetethylicum occasionally showed small peaks at the same retention time as 2‐oxopropanal, accumulation of this metabolite could not be reliably verified and 2‐oxopropanal did possibly not exceed concentrations higher than 1 mM (data not shown). Interestingly, batch fermentation experiments with glycerol revealed that hydrogen accumulated at high concentrations, while almost no formate was produced (Table 1 ). This finding was supported by the approximately 2000‐fold lower activity of formate dehydrogenase compared with hydrogenase (Table 3 ). In contrast to this, cultures grown with gluconate or glucose as substrates produced higher concentrations of formate than hydrogen at a ratio of hydrogen to formate of about 1:2 (Patil et al ., 2015 ). Therefore, formate is a more prominent metabolite when cultures are grown with sugars, which might be due to the fact that twice as much CO 2 is released per mole of hexose oxidized compared with glycerol. Even though formate could be detected as a metabolite and formate:benzyl viologen oxidoreductase activity was detected in cell‐free extracts, we were unable to find the corresponding formate dehydrogenase genes in the genome sequence. Possibly, the observed activity is a side reaction of formate C‐acetyltransferase (pyruvate:formate lyase) which was identified in the proteome. When A. acetethylicum was grown with a larger headspace‐to‐culture volume ratio, the maximum concentrations of ethanol and hydrogen were slightly higher compared with cultures with a low headspace‐to‐culture volume ratio, indicating that hydrogen is also inhibitory for glycerol degradation. Similar observations were also made for E. coli grown in a fermenter sparged with argon gas, which vastly increased the amount of glycerol degraded (Murarka et al ., 2008 ). Although experiments with cultures of A. acetethylicum permanently sparged with argon or nitrogen have not been done yet, it can be assumed that this might stimulate fermentation of glycerol as well. Even though A. acetethylicum has certain advantages over other glycerol‐fermenting organisms, the fact that side products are formed in batch cultures especially at higher glycerol concentrations might be disadvantageous for large‐scale bioethanol production. Among acetate, formate and 1,2‐propanediol frequently observed at small concentrations, a further product was released in cultures of A. acetethylicum that could not be identified by HPLC. It was previously reported that during fermentation of glycerol by Clostridium pasteurianum , 1,3‐propanediol is produced via 3‐hydroxypropionaldehyde (Dabrock et al ., 1992 ). However, in this study we could detect neither 1,3‐propanediol nor 3‐hydroxypropionaldehyde as metabolites, and the activities of the respective enzymes were absent in in vitro assays, although we recently reported that occasionally very small amounts of 1,3‐propanediol could be detected by HPLC in cultures of A. acetethylicum (Patil et al ., 2015 ). Likewise, succinate, lactate, 1‐butanol, 1‐propanol, 2‐propanol, butyrate, propionate and 1,3‐butanediol were ruled out as possible side products and neither one of the corresponding metabolic pathways is present in the genome of A. acetethylicum with a complete set of genes (Patil et al ., 2016 ). Interestingly, the percentage of ethanol produced per glycerol slightly increased with increasing substrate concentration, with an optimal initial glycerol concentration of 100 mM (9.2 g l −1 ) at which the strain showed maximal efficiency of glycerol‐to‐ethanol conversion (79% glycerol conversion to ethanol)."
} | 5,940 |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.