pmid stringlengths 8 8 | pmcid stringlengths 8 11 ⌀ | source stringclasses 2
values | rank int64 1 9.78k | sections unknown | tokens int64 3 46.7k |
|---|---|---|---|---|---|
31057689 | PMC6497423 | pmc | 3,825 | {
"abstract": "Surface\nX-ray diffraction has been employed to quantitatively determine\nthe geometric structure of an X-ray-induced superhydrophilic rutile-TiO 2 (110)(1 × 1) surface. A scatterer, assumed to be oxygen,\nis found at a distance of 1.90 ± 0.02 Å above the five-fold-coordinated\nsurface Ti atom, indicating surface hydroxylation. Two more oxygen\natoms, situated further from the substrate, are also included to achieve\nthe optimal agreement between experimental and simulated diffraction\ndata. It is concluded that these latter scatterers are from water\nmolecules, surface-localized through hydrogen bonding. Comparing this\ninterfacial structure with previous studies suggests that the superhydophilicity\nof titania is most likely to be a result of the depletion of surface\ncarbon contamination coupled to extensive surface hydroxylation.",
"conclusion": "Conclusions To summarize, SXRD data have been acquired from\nan X-ray-induced\nsuperhydrophilic rutile-TiO 2 (110)(1 × 1) surface.\nIt is concluded that the five-fold-coordinated surface Ti atom is\nhydroxylated, as indicated by the presence of an atom, assumed to\nbe O, at a distance of 1.90 ± 0.02 Å. There is also evidence\nof hydrogen-bonded H 2 O molecules, which are located somewhat\nfurther from the substrate surface. The examination of the current\nstructure, in tandem with previous work, 14 , 20 suggests that the X-ray-induced superhydrophilicity of titania is\nlikely to be a result of both the depletion of surface carbon and\nincreased surface hydroxylation.",
"introduction": "Introduction Ever since Wang et\nal.’s discovery that UV irradiation of\ntitania results in a superhydrophilic surface, 1 there has been a great deal of effort to both exploit and understand\nthis novel phenomenon. Significant success has been achieved in the\nformer of these two goals, with applications including self-cleaning\nwindows and antifogging mirrors. 2 − 5 In contrast, uncertainty still remains as to the\norigin of the superhydrophilicity. Currently, there are a number of\npotential explanations to be found in the literature, 1 , 3 , 6 − 10 but none are supported by compelling experimental\nevidence. For example, it is proposed that the superhydrophilicity\nis simply a result of the removal of surface carbon contamination. 6 Other researchers suggest that modification of\nthe surface structure/chemistry of the titania substrate (e.g., surface\nhydroxylation) underpins this macroscopic property. 9 Longer range structural changes are also purported to be\nimportant, including the formation of nanoscale hydrophobic and hydrophilic\ndomains. 1 , 11 , 12 Here we directly\naddress this topic, employing surface X-ray diffraction (SXRD) to\nquantitatively determine the structure of a model titania surface,\nrutile-TiO 2 (110), that exhibits superhydrophilicity induced\nthrough X-ray exposure. Previously, Shirasawa et al. (SEL) have\nundertaken SXRD measurements\nfrom rutile-TiO 2 (110) to identify changes in surface structure\nassociated with the UV-induced hydrophobic-to-hydrophilic transition. 13 They report that the application of a wet chemical\npreparation (WCP) recipe to the substrate resulted in a hydrophobic\n(1 × 1) surface termination, which became hydrophilic upon UV\nirradiation. It is suggested that this transition is associated with\nthe presence of surface hydroxyls (OH), as surface five-fold-coordinated\ntitanium atoms (Ti 5c ) and bridging oxygens (O b ) become hydroxylated following the exposure to UV light. Figure 1 illustrates the\nchanges in interface geometry concluded in ref ( 13 ). Figure 1 Ball-and-stick models\nshowing the UV-induced (hydrophobic to hydrophilic)\nchanges in interface geometry for rutile-TiO 2 (110), as\nconcluded by SEL from SXRD data. 13 Red\nspheres are Ti atoms and darker (lighter) blue spheres are substrate\n(adsorbate) oxygen atoms. H atoms (pink spheres) are depicted, although\nthey were not explicitly included in SEL’s structure determination.\nPossible hydrogen bonds are indicated by dashed lines. In this Article, we revisit the structure of the\nsuperhydrophilic\nrutile-TiO 2 (110)(1 × 1) surface. A WCP method\nis again employed for sample preparation, but with X-rays being used\nto induce superhydrophilicity. Similar to ref ( 13 ), the surface is found\nto be extensively hydroxylated, including OH bound to Ti 5c . The diffraction data acquired in this study, however, resemble\nmuch more closely those acquired from the hydrophobic termination\nin ref ( 13 ). As argued\nin detail below, this somewhat unexpected finding suggests that the\nanalysis and interpretation of SEL 13 require\nrevision.",
"discussion": "Results and Discussion The application of our WCP recipe to the rutile-TiO 2 (110) sample resulted in a deionized water contact angle of ∼80°.\nThis value is consistent with that reported in ref ( 14 ) for a (1 × 1) surface\ntermination subsequent to immersion in aqua regia but not exposed\nto UV-ozone treatment. Following exposure to I07’s photon beam,\nthe contact angle was found to fall to essentially 0°; that is,\na superhydrophilic transition was induced by X-ray exposure. All diffraction\nmeasurements were undertaken with the rutile-TiO 2 (110)(1\n× 1) surface in this state; a contact-angle measurement at the\nend of data collection indicated that a value of 0° was maintained\nthroughout this period. Data from the (−1, 0, 0.9) reference\nreflection also revealed no substantive surface degradration. Figure 2 shows four\nof the experimental CTRs acquired in the current study (black markers\nwith error bars), together with equivalent data collected by SEL 13 from rutile-TiO 2 (110) following UV\nexposure (blue markers with error bars). A priori, as both data sets\nwere recorded from superhydrophilic surfaces, it was expected that\nthey would be very similar. However, there are significant differences.\nFor example, the local maxima in our data at (0, 1, ∼3) and\n(1, 0, ∼2), are not replicated in those from SEL. In contrast,\nour CTR profiles are much more comparable to those reported by SEL\nfor their pre-UV exposure (hydrophobic) surface. These data are also\nshown in Figure 2 as\nred markers with error bars. We note that on an adsorbate-free rutile-TiO 2 (110)(1 × 1) surface, prepared in ultrahigh vacuum (UHV),\nthe aforementioned local maxima are associated with significant displacements\nof surface atoms away from their bulk positions; 19 that is, they are not a direct signature of surface hydrophilicity.\nAn absence of such features may be a result of either a more bulk-like\ntermination or surface roughening. Figure 2 Comparison of the (0, 1, l ), (1,0, l ), (1, 1, l ), and (2,\n0, l ) experimental\nCTRs acquired from X-ray-induced superhydrophilic rutile-TiO 2 (110) in the current study with data from SEL. 13 Current study: black markers with error bars; pre-UV exposure\nfrom SEL: 13 red markers with error bars;\npost-UV exposure from SEL: 13 blue markers\nwith error bars. Profiles are systematically offset for clarity. Considering the qualitative comparison\noutlined above, it was expected\nthat fitting of our experimental SXRD data would result in the hydrophobic\nstructure determined by SEL, 13 where molecular\nH 2 O is bound atop Ti 5c (see Figure 1 ). Figure 3 shows the best fit (blue line) achieved\nusing SEL’s hydrophobic structure as a starting point and simply\nallowing the displacement of both atomic positions and nonstructural\nparameter values. As indicated by χ 2 = 2.60, as well\nas visual inspection, the experiment–theory agreement is far\nfrom perfect, suggesting that the correct structural solution had\nnot been found. On this basis, we explored other potential surface\nterminations, including those consistent with the presence of surface\nhydroxyls. The resulting overall best fit to the experimental CTRs\nis shown in Figure 3 (red line). To achieve this fit, 78 parameters were optimized, that\nis, 51 atomic coordinates, 21 Debye–Waller (DW) factors, a\nscale factor, surface roughness (β), three fractional occupancies,\nand surface fraction. The corresponding χ 2 is 1.05;\nthat is, there is an excellent level of agreement between the experimental\nand simulated data. Figure 3 Comparison of experimental CTR data (black markers with\nerror bars),\nacquired from X-ray-induced superhydrophilic rutile-TiO 2 (110), and theoretical best-fit simulations. Solid blue line indicates\nthe best fit achieved following relaxation of the hydrophobic structure\nreported by SEL. 13 Solid red line indicates\nthe overall best fit, with β = 0.24 and a surface fraction of\n0.96. The optimum geometry of the first\nfew atomic layers emerging from\nthe best fit to the experimental CTR profiles is depicted in Figure 4 . Selected corresponding\ninteratomic distances are listed in Table 1 . A ball-and-stick model showing all atoms\ndisplaced during fitting is shown in Figure S1 , along with a complete list of the optimized coordinates, DW factors,\nand fractional occupancies in Table S1 .\nNeglecting the details of atomic relaxation, the surface mimics the\nstoichiometry and geometry of bulk-terminated rutile-TiO 2 (110)(1 × 1) but is decorated with oxygen species. Focusing\non Ti 5c (labeled Ti(2)), an adsorbed oxygen atom (labeled\nO(1′)) is located atop at a distance of 1.90 ± 0.02 Å,\nwhich is consistent with the presence of a bound terminal hydroxyl\n(OH t ). 20 , 21 The experimental distance from\nref ( 20 ) is 1.95 ±\n0.03 Å, with a moderately longer distance of 2.07 Å being\nobtained from molecular dynamics calculations. 20 Two additional nonsubstrate oxygen atoms (labeled O(2′)\nand O(3′)) are at somewhat greater distances from the topmost\nsubstrate atoms. O(2′) is 2.68 ± 0.03 Å above the\nbridging oxygen atom (labeled O(1)), with O(3′)’s nearest\nneighbor being O(1′) at a distance of 2.65 ± 0.05 Å.\nThese interatomic separations suggest that oxygen atoms O(2′)\nand O(3′) arise from water molecules, which are localized through\nhydrogen bonding. 22 For illustrative purposes,\nwe have included H atoms in Figure 4 but stress that these species were not explicitly\nincluded during the generation of simulated SXRD data due to\ntheir negligible X-ray scattering. Figure 4 Ball-and-stick models of the X-ray-induced\nsuperhydrophilic rutile-TiO 2 (110) surface structure determined\nfrom SXRD data. Perspective\n(top) and plane (bottom) views are shown. Red spheres are Ti atoms,\nand darker (lighter) blue spheres are substrate (adsorbate) oxygen\natoms. H atoms (pink spheres) are depicted, although they were not\nexplicitly included in the structure determination. Possible hydrogen\nbonds are indicated by dashed lines. The numerical labeling of the\natoms is employed in Table 1 and Table S1 for identification\npurposes. Symmetry-paired atoms are denoted by *. Table 1 Selected Interatomic Distances Derived\nfrom Atomic Coordinates ( Table S1 ) of Optimized\nSuperhydrophilic TiO 2 (110)(1 × 1) Structure atoms interatomic\ndistance (Å) O(3′)–O(2′) 2.70 ± 0.06 O(3′)–O(1′) 2.65 ± 0.05 O(2′)–O(1) 2.68 ± 0.03 O(1′)–Ti(2) 1.90 ± 0.02 O(1)–Ti(1) 1.83 ± 0.02 O(2)–Ti(1) 1.98 ± 0.02 O(2)–Ti(2) 1.95 ± 0.02 O(3)–Ti(1) 1.94 ± 0.01 O(4)–Ti(2) 1.94 ± 0.01 Given the optimized structure displayed in Figure 4 , it is interesting to compare this result\nwith other pertinent studies. Focusing initially on SEL’s work, 13 the present diffraction data are very similar\nto those acquired from their hydrophobic surface, as demonstrated\nin Figure 2 . Because\nour surface is superhydrophilic, as a result of X-ray exposure, this\nagreement presents a conundrum. One plausible explanation, arising\nfrom discussion with SEL, 13 is that the\n∼1 mm 2 footprint of the X-ray beam employed for\ntheir SXRD measurements induced superhydrophilicity only in this region.\nHence the water contact-angle measurement, where the droplet employed\ncovered a much larger surface area, did not reveal this local X-ray-induced\nsuperhydrophilicity. It should be noted that on I07 almost the entire\nsample surface would have been exposed to the X-ray beam during alignment\nand measurement. On the basis that SEL’s pre-UV-exposure\nsurface is superhydrophilic\nin the area probed by the X-ray beam, then one further issue requires\nresolution. Specifically, despite the similar experimental data, the\ndiscrepancy between our optimized structure and SEL’s needs\nto be understood, for example, the variation in the Ti(2)–O(1′)\ndistance (1.90 ± 0.02 versus 2.09 ± 0.03 Å 13 ). To this end, our experimental data set (eight\nCTRs) was reduced to match that of SEL (six CTRs), and fitting was\nundertaken. Under these conditions, we were able to effectively model\nthe data with SEL’s hydrophobic structure. On this basis, it\nis evident that fewer experimental CTRs leads to a local χ 2 minimum, resulting in a significantly different surface structure. One other matter emerging from the preceding discussion is the\norigin of the UV-induced change in CTR profiles observed by SEL. 13 Assuming that their pre-UV data is acquired\nfrom a superhydrophilic area of the rutile-TiO 2 (110) surface,\nthen the observed changes cannot be accounted for by a hydrophobic–hydrophilic\ntransition. This deduction implies that UV-irradiation leads to additional\ninterfacial modification; that is, a unique surface structure is formed\nupon exposure to UV light. Currently, this suggestion is essentially\nconjecture, but is worthy of further investigation. Having reconciled\nthe results of this study with those of SEL, 13 a comparison of the geometry of the current\nsuperhydrophilic termination with those reported for interfaces formed\nby the exposure of UHV-prepared TiO 2 (110)(1 × 1) to\nliquid water (H 2 O(l)) is worthwhile. 20 Figure 5 compares the current optimum structure ( TiO 2 (110):Super ) to that elucidated with SXRD\nfollowing dipping of TiO 2 (110)(1 × 1) into H 2 O(l) and measuring ex situ in UHV ( TiO 2 (110):Dip-H 2 O(l) ) as\nwell as that determined for TiO 2 (110) submerged in H 2 O(l) ( TiO 2 (110):Sub-H 2 O(l) ). These three structures are similar\nbut not identical. For example, both TiO 2 (110):Super and TiO 2 (110):Sub-H 2 O(l) exhibit\noxygen atoms consistent with hydrogen-bonded H 2 O molecules,\nalthough their configuration differs; such scatterers are not evident\nin the TiO 2 (110):Dip-H 2 O(l) data due to the acquisition in\nUHV. Turning to Ti 5c , for each structure displayed in Figure 5 , the distance to\nthe atop oxygen atom is consistent with hydroxylation (Ti 5c –OH t ). However, TiO 2 (110):Super displays a slightly shorter Ti 5c –OH t distance (1.90 ± 0.02 Å) than either TiO 2 (110):Dip-H 2 O(l) or TiO 2 (110):Sub-H 2 O(l) (1.95 ± 0.03 Å). This\nvariation may be a result of the former substrate being essentially\nfully oxidized, whereas the latter two were somewhat reduced as a\nresult of substrate preparation in UHV. Figure 5 Ball-and-stick models\nof the optimum interfacial structures determined\nfrom SXRD for X-ray-induced superhydrophilic rutile-TiO 2 (110) (current study), rutile-TiO 2 (110) subsequent to\ndipping in H 2 O(l), 20 and rutile-TiO 2 (110) submerged in H 2 O(l). 20 Side (top) and plane (bottom) views are shown. Red spheres are Ti\natoms, and darker (lighter) blue spheres are substrate (adsorbate)\noxygen atoms. Selected interatomic distances are annotated. Fractional\noccupancies of adsorbate oxygen atoms are indicated by the values\ninscribed on the lighter blue spheres. Regarding the origin of the X-ray-induced superhydrophilicity\nof\ntitania, the current SXRD data rule out the coexistence of hydrophilic\nand hydrophobic domains, as analysis indicates that almost the entire\nsurface adopts the same geometry; that is, surface fraction is 0.96.\nIt should be emphasized that the present study cannot be used to definitively\nrule out the existence of such domains on UV-exposed titania. Furthermore,\nbecause the diffraction data from TiO 2 (110):Dip-H 2 O(l) were\nacquired in UHV from a surface not irradiated with either UV or X-rays\nduring dipping, the mere presence of OH t cannot be directly\nrelated to photoinduced superhydrophilicity. Given this result, one\ncould suggest that the simple removal of surface carbon most likely\nunderpins this property. 6 It is, however,\nnotable that the fractional occupancy of OH t (O(1′))\nfor TiO 2 (110):Super is\napproximately double that for either TiO 2 (110):Dip-H 2 O(l) or TiO 2 (110):Sub-H 2 O(l) , that is, 1.00 compared to 0.45 and 0.50, respectively,\nas indicated in Figure 5 . Hence, increased surface hydroxylation may play a role in TiO 2 superhydrophilicity, coupled to the loss of surface carbon.\nWe remark that in ref ( 20 ) ab initio modeling suggests that the presence of OH t is\na result of (near) surface defects driving surface H 2 O\ndissociation. Because the substrate in the current study is not expected\nto possess any significant concentration of defects, the hydroxyl\nspecies must arise from elsewhere. Almost certainly, it is photon-induced\n(or photoexcited electron) chemistry that produces these OH t adsorbates, which may be the reason that a higher coverage is achieved;\nwe note that this increase in surface hydroxylation is not simply\nrelated to carbon removal, as the surfaces in ref ( 20 ) are reported to be relatively\ncarbon-free (≤0.1 monolayer). Finally, we would like\nto comment on a recent elegant study suggesting\nthat air- or aqueous-solution-exposed rutile-TiO 2 (110)\nis commonly decorated by carboxylate species. 23 On the basis that SXRD is not a spectroscopic probe, there is always\nthe potential for misidentification of adsorbates, especially those\nexhibiting similar X-ray scattering characteristics (e.g., C and O).\nFor the current study, however, we argue that this is not the case.\nSupporting evidence is two-fold. First, Auger spectra acquired from\na superhydrophilic rutile-TiO 2 (110) surface, prepared following\nour WCP recipe, show no discernible carbon signal. 14 Second, SXRD data were acquired from a superhydrophilic\nsurface, which is inconsistent with the presence of adsorbed carboxylates. 6 , 23"
} | 4,495 |
37915545 | null | s2 | 3,826 | {
"abstract": "This paper presents a theoretical framework for probably approximately correct (PAC) multi-agent reinforcement learning (MARL) algorithms for Markov games. Using the idea of delayed Q-learning, the paper extends the well-known Nash Q-learning algorithm to build a new PAC MARL algorithm for general-sum Markov games. In addition to guiding the design of a provably PACMARL algorithm, the framework enables checking whether an arbitrary MARL algorithm is PAC. Comparative numerical results demonstrate the algorithm's performance and robustness."
} | 136 |
33879791 | PMC8058080 | pmc | 3,828 | {
"abstract": "Acaryochloris marina is one of the cyanobacterial species that can use far-red light to drive photochemical reactions for oxygenic photosynthesis. Here, we report the structure of A. marina photosystem I (PSI) reaction center, determined by cryo-electron microscopy at 2.58 Å resolution. The structure reveals an arrangement of electron carriers and light-harvesting pigments distinct from other type I reaction centers. The paired chlorophyll, or special pair (also referred to as P740 in this case), is a dimer of chlorophyll d and its epimer chlorophyll d ′. The primary electron acceptor is pheophytin a , a metal-less chlorin. We show the architecture of this PSI reaction center is composed of 11 subunits and we identify key components that help explain how the low energy yield from far-red light is efficiently utilized for driving oxygenic photosynthesis.",
"introduction": "Introduction Photosynthesis is driven by two types of photoreaction systems, namely type I and type II reaction centers. Photosystems I (ref. 1 ) and II (ref. 2 ; PSI and PSII, belonging to type I and type II reaction centers, respectively) work sequentially in oxygenic photosynthesis in plants and cyanobacteria 3 . They typically use chlorophyll (Chl) a (Supplementary Fig. 1 ) as the major pigment that absorbs visible light (670–700 nm in the red-light region; Supplementary Fig. 2 ). Anoxygenic photosynthetic bacteria use bacteriochlorophylls (BChl) that absorb 800–900-nm light in addition to blue/ultraviolet light and have either a type I or type II reaction center. Acaryochloris marina ( A. marina ) was found as a cyanobacterial species that uses Chl d (Supplementary Fig. 1 ), which absorbs 700–750-nm light in vivo 4 (Supplementary Fig. 2 ). Later, several cyanobacteria were found that induce a limited amount of Chl f for capturing far-red light under the far-red light condition 5 – 7 (Supplementary Fig. 2 , and see below). A. marina was isolated from colonial ascidians, which harbor mainly Chl a -type cyanobacteria, resulting in an environment with low visible light and high far-red light. Analogous to the cortex effect within a lichen body 8 , high-energy blue light absorbed by the Chls likely has diminished intensity within the symbiotic host ascidians. A. marina has exploited this environment with the niche-filling introduction of Chl d that absorbs far-red light 9 , 10 . The discovery of A. marina prompted intensive research into the mechanism of this low-energy-driven system. Although the organism also contains Chl a at ~5% relative abundance, spectroscopic and pigment analyses revealed that it is Chl d that plays a central role in the photoreaction 4 , 9 , 11 . With the acquisition of Chl d , A. marina harnesses far-red light, the energy of which is lower than that of visible light that is utilized in most oxygenic photosynthetic reactions. Chl d has a peak wavelength of absorption at ~697 nm in methanol (Supplementary Fig. 3 ; so-called Qy band; 700–750 nm in vivo, Supplementary Fig. 2 ), which is longer than the 665.2 nm of Chl a (670–700 nm in vivo, Supplementary Figs. 2 and 3 ) by ~30 nm. This means that the photon energy absorbed is ~80 mV (10%) lower. How PSI and PSII of A. marina drive the similar photochemical reactions that occur in the Chl a -dependent systems of other oxygenic photosynthetic organisms has been a long-standing puzzle. PSI generates reducing power for NADPH production by accepting electrons originating from PSII 3 . PSI of A. marina has a similar subunit composition to that of the PSIs of other oxygenic photosynthetic organisms 12 , 13 (Supplementary Table 1 ). However, the peak wavelength of the light-induced redox difference absorption spectrum of the paired Chls, the so-called special pair, in A. marina PSI is longer (740 nm) than that of Chl a (700 nm) in the PSI of plants and typical cyanobacteria 14 (Fig. 1a ). The special pair P740 have been assumed to be a heterodimer of Chl d and d ′ (refs. 15 , 16 ) as in P700, which is composed of a heterodimer of Chl a /Chl a ′ in all other PSIs (Fig. 1a ) 1 , 17 . Cofactors of the PSI electron transfer chain in A. marina (Fig. 1b ) and the reduction potential of P740 are also similar to those of P700 in other organisms 12 , 13 , 18 . Fig. 1 Comparison of electron transfer chains in PSI. a Cyanobacterial and higher plant PSI. b \n A. marina PSI identified in this work. In cyanobacterial and higher plant PSI, electrons released from the paired Chls, so-called special pair, of P700, a heterodimer of Chl a and a ′, are transferred to ferredoxin (Fd) to reduce NADP + via Acc (Chl a ), A 0 (Chl a ), A 1 (PhyQ), F X (iron–sulfur center), and F A /F B (iron–sulfur center). The components are arranged in a pseudo-C 2 axis on a heterodimer reaction center protein complex (PsaA/PsaB) and the two routes (A- and B-branch located on PsaA and PsaB, respectively) are thought to be equivalent. Acc accessory chlorophyll, A 0 primary electron acceptor, A 1 secondary electron acceptor, F X , F A , and F B iron–sulfur centers, PhyQ phylloquinone. In A. marina PSI, the electron transfer chain is the same, but some cofactors differ from those in the PSIs of other organisms; the Chls of special pair P740 are Chl d / d ′, while Acc could be Chl d (see text for detail) and A 0 is pheophytin (Pheo) a . The far-red light-absorbing Chl, Chl f , is also utilized in some cyanobacteria to capture far-red light (Supplementary Fig. 2 ). How Chl f , as well as Chl d , has succeeded in expanding harvesting the low-energy far-red light region is of great interest 5 – 7 . Chl f is only induced under far-red light conditions, and even then makes up less than ~10% of total Chls in PSI (seven Chl f and 83 Chl a in Halomicronema hongdechloris PSI 19 ), and the locations of these Chl f molecules is still debated 20 , 21 . Recent cryo-EM analyses show that Chl f is not part of the charge-separating pigments, including and following the special pair 19 , 22 , 23 . Thus, Chl f in these PSIs is likely entirely responsible for light harvesting. High-resolution 24 and ultrafast 25 spectroscopic studies support this. Still, PSII in some cyanobacteria may use Chl f (and/or Chl d ) for electron transfer when grown under far-red light conditions 21 , 26 . In contrast to Chl f -carrying PSIs 19 , 20 , 22 , 23 , 25 , 27 , Chl d in A. marina PSI is always induced even under natural white light and does, in fact, take the place of the special pair Chls in the photochemical reaction chain. The Chl d -driven photoreaction occurs by low-energy far-red light corresponding to the Qy band of Chl d . Chl d also absorbs high-energy blue light, corresponding to the Soret band, but the resulting excited state relaxes to a lower excited state corresponding to the Qy level for the photochemical reaction. In this sense, the Chl d -driven PSI-system is unique among photochemical reaction centers, utilizing as it does far-red light level energy. To date, several structures of type I reaction centers have been determined from higher plants 28 , 29 , green algae 30 , 31 , red algae 32 , diatoms 33 , cyanobacteria 1 , 19 , 22 , 23 , 34 , 35 , and an anoxygenic bacterium 36 . All structures of type I reaction centers including Chl f -carrying cyanobacteria 19 , 22 , 23 have Chl a in the electron transfer chain with the one exception, namely, the one from the anoxygenic bacterium that has BChl g ′ and 8 1 -OH-Chl a (ref. 36 ). The PSI of A. marina with Chl d in its electron transfer chain for utilizing low-energy light directly for the photochemistry is thus unique. Here, we show the structure of the PSI trimer isolated from A. marina revealed by cryo-electron microscopy (cryo-EM) at 2.58 Å resolution. The special pair P740 is a dimer of Chl d and its epimer Chl d ′, and the primary electron acceptor is pheophytin (Pheo) a not Chl a nor Chl d , which is embedded in membrane protein complex composed of 11 subunits. The structure helps gain insight into the mechanism for the utilization of far-red light (Fig. 1b ).",
"discussion": "Results and discussion Overall structure, protein subunits, and cofactors of A. marina PSI The PSI trimer complex was isolated from A. marina . Detailed methods of sample preparation, data processing, and structural refinement by cryo-EM, and biochemical data, are presented in Supplementary Information (Supplementary Figs. 4 – 12 and Supplementary Tables 1 – 4 ). The model of the PSI trimer was refined to give a correlation coefficient and Q -score, which are indices for validating the correctness of a model, of 0.80 and 0.65, respectively (Supplementary Table 4 ). These index values can be considered reasonable at 2.58 Å resolution (Supplementary Fig. 11 ). The PSI trimer has dimensions of 100 Å depth, 200 Å length, and 200 Å width, including the surrounding detergent micelle (Fig. 2 ). The overall structure resembles those of PSI trimers reported for other cyanobacteria 1 , 22 , 23 , 35 . The root-mean-square deviations using secondary structure matching between the monomer model (chains A–M) and Protein Data Bank (PDB) structures 1JB0 of Thermosynechococcus elongatus 1 , 5OY0 of Synechocystis sp. PCC 6803 35 , and 6KMX of H. hongdechloris 19 are 1.09 Å, 1.05 Å, and 1.11 Å, respectively. Fig. 2 Overall structure of the photosystem I (PSI) trimer from Acaryochloris marina . a Cryo-electron microscopy (cryo-EM) density map of PSI trimer viewed from the stromal side perpendicular to the membrane plane. Monomer-1, multicolored; monomer-2, orange; monomer-3, gray; detergent micelle, cloudy gray. b Cryo-EM density map of PSI trimer viewed from the side of the membrane plane. c Structural model of PSI monomer with view in the same direction as in a . PsaA, green; PsaB, cyan; PsaC, orange; PsaD, yellow; PsaE, red; PsaF, gray; PsaJ, light blue; PsaK, magenta; PsaL, dark green; PsaM, teal blue; Psa27, blue. d Structural model of PSI monomer with view in the same direction as in b . The A. marina PSI monomer contains 11 subunits (PsaA, PsaB, PsaC, PsaD, PsaE, PsaF, PsaJ, PsaK, PsaL, PsaM, and Psa27), all of which could be assigned to the cryo-EM density map (Fig. 2 , Supplementary Figs. 6 and 13 , and Supplementary Table 1 ). PsaX (a peripheral subunit in other cyanobacteria such as T. elongatus ), PsaG, and PsaH (subunits in higher plants) were missing in the density map, consistent with the absence of their genes from the A. marina genome 10 . The name Psa27 was given to a subunit protein in A. marina that has low sequence identity (29.4%) with PsaI of T. elongatus (Supplementary Table 5 ) 12 . Here, we found that Psa27 is in the same location as PsaI, a transmembrane alpha helix, in T. elongatus . Psa27 also contributes to the structural stabilization of the PSI trimer just as PsaI does in T. elongatus . Thus, we concluded that Psa27 and PsaI are in effect the same subunit. The loop regions of PsaB (residues 290−320 and 465−510) were not identified due to disorder because of the absence of PsaX, suggesting that PsaX stabilizes these regions in other cyanobacteria. Most of subunits PsaE, PsaF, and PsaK were modeled using polyalanine because of poor regional map quality (Supplementary Figs. 13 and 14 , and Supplementary Table 1 ), possibly suggesting that some part of those subunits dissociated during sample preparation. The cofactors assigned (Supplementary Fig. 15 ) are 70 Chl d (Supplementary Data 1 ), 1 Chl d ′ (an epimer of Chl d ), 12 α-carotenes (α-Car) but no β-Car (Supplementary Table 6 ) 10 , 14 , 2 Pheo (a derivative of Chl; however, unlike Chl, no Mg 2+ ion is coordinated by the tetrapyrrole ring), 2 phylloquinones (PhyQs), 3 iron–sulfur clusters, 2 phosphatidylglycerols (PGs), 1 monogalactosyl diacylglycerol (MGDG), and 84 water molecules. Two molecules of Pheo were assigned as Pheo a on the basis of pigment analysis (Supplementary Table 6 ). In addition, a small amount of Chl a (approximately one Chl a per Chl d ′) was detected by pigment analysis (Supplementary Table 6 ). In this study, we assigned the Chls of A. marina PSI as Chl d . This is because the amount of Chl a in the A. marina PSI is minimal, and it is not possible to distinguish between Chl d and Chl a at 2.58 Å resolution, as is described later. In the A. marina PSI, a large number of Chl d and α-Car harvest light energy and finally transfer the energy to P740. The arrangements of Chl d and α-Car in A. marina resemble those of Chl a and β-Car, respectively, in T. elongatus , as shown in Supplementary Fig. 16 . The pigments are numbered according to the nomenclature of Chl a in T. elongatus 37 (Supplementary Data 1 ). However, there are differences in some amino acid residues surrounding the cofactors, slight gaps in the arrangement of the cofactors, and absence of some Chls in the A. marina PSI compared with those of T. elongatus (Supplementary Table 5 and Supplementary Figs. 16 – 18 ). For example, Phe49/J in A. marina , the counterpart of His39/J that is an axial ligand of Chl a 88 in T. elongatus PSI, explains the absence of the corresponding Chl d (capitalized letters following the slash (/) indicate the subunit names, such as PsaA or PsaB; Supplementary Fig. 16d ). Owing to the absence of PsaX in A. marina , the Chl d corresponding to Chl a 95 (refs. 1 , 38 ) in T. elongatus whose axial ligand is Asn23/X is also missing. The absence of Chl a 94 is seen in some other reported PSI structures, such as Synechocystis sp. PCC 6803 (ref. 35 ). The arrangements of pigments adjacent to Psa27 (PsaI) and PsaL, whose sequence identity with other cyanobacterial PsaLs is low, are specific to A. marina (Supplementary Fig. 16a , black dashed line region, and Supplementary Figs. 17 and 18 ). Chl d 38, 52, 53, and the ring (α or ε) of Car4007 near Chl d 53 are set within the surrounding structure differently from those in T. elongatus (Supplementary Fig. 17a ). These differences when compared with the structure of T. elongatus help explain the specific features of light harvesting in A. marina . In addition, the C3-formyl groups of some Chl d in the A. marina PSI form hydrogen bonds with their surrounding amino acid residues (Supplementary Data 1 and Supplementary Fig. 18 ). These characteristic structural features will be important in future theoretical studies of the light-harvesting mechanism in A. marina . Although the amount of Chl d per Chl d ′ determined by pigment analysis (67.0 ± 0.66, n = 5 of independently prepared PSI) and assigned by structural analysis (70 including one or two Chl a ) is lower in this study than those in previous studies (145 ± 8, ref. 14 and 97.0 ± 11.0, ref. 12 ), the semi-stoichiometric amount of Phe a at 1.92 ± 0.022 (0.3 ± 0.2 per reaction center in a previous study 12 ) is consistent with our structural analysis. Notwithstanding the fact that the local resolution of some parts of the outer region is somewhat lower, the PSI was stable enough to keep its integrity for 5 days (Supplementary Fig. 5 ). Chls that were unable to be assigned in A. marina PSI compared with those in T. elongatus PSI are shown by marking N/D in the column of the ID number in Supplementary Data 1 . Such Chls can be recognized by the chlorins in transparent gray in Supplementary Fig. 16 . Chls corresponding to Chl-88 and Chl-94 in T. elongatus appear to be absent in A. marina PSI. The numbers of Chls in the PSI of Chl f -carrying Fischerella thermalis (89, ref. 22 ) and H. hongdechloris (90, ref. 19 ) are also lower than those of T. elongatus PSI (96, ref. 1 ). The type I reaction center of Heliobacterium modesticaldum carries a much smaller amount of Chl species; 60 molecules per reaction center 36 . Due to the somewhat lower resolutions of the outer regions of the A. marina PSI, we mainly focus on the structure and function of the central part of A. marina PSI at high local resolution, that is, P740 and the electron transfer components. Electron transfer components The configuration of the electron transfer chain in A. marina PSI is similar to that of the PSI from T. elongatus (Fig. 1 ), although cofactor compositions are different. The important assigned cofactors involved in electron transfer (Fig. 1b ) are four Chls, two Pheos, two PhyQs (A 1 ), and three iron–sulfur clusters (F X , F A , and F B ; Fig. 3a ). The Chls and PhyQs are arranged in two branches, the A-branch and the B-branch, which are related by a pseudo-C 2 axis as in other type I reaction centers, and are stabilized by amino acid residues of subunits PsaA and PsaB. The Chls of special pair P740 are Chl d ′ (P A ) and Chl d (P B ), which are coordinated by residues His678/A and His657/B, respectively (Fig. 3b, c ). The distance between the ring planes (the π–π interaction distance) of Chl d ′ and Chl d is 3.5 Å (Fig. 3d ; 3.6 Å in T. elongatus ). Two Tyr residues, Tyr601/A and Tyr733/A, are positioned within hydrogen bonding distance of P A as observed in PSI from T. elongatus . Fig. 3 Cofactors involved in the electron transfer reaction in A. marina PSI and their surrounding amino acid residues. a Arrangement of cofactors involved in the electron transfer reaction. Values on both sides represent the center–center distances (in Å) between the cofactors. b Arrangement of P740 and its surrounding structure. View from the side of the membrane plane. c As in b , but view from the pseudo-C 2 axis. d Arrangement of the amino acid residues influencing E m (P A ) and E m (P B ). Superposition of the structural model of PSI from A. marina with that from T. elongatus (PDB code 1JB0). Transparent green and blue, PSI from A. marina ; transparent gray, PSI from T. elongatus . Chlorophyll (Chl) d , green; Chl d ′, blue; water molecule, red; pheophytin (Pheo) a , orange, phylloquinone (PhyQ), yellow. In A. marina PSI, we identified three water molecules (W1 − W3) around P A . They form hydrogen bonds with surrounding amino acid residues (Tyr601/A, Ser605/A, Asn608/A, Ser741/A, and Try745/A) and one Chl d (Chl d 32; Fig. 3b, c ). For the T. elongatus P A , only one water molecule forms hydrogen bonds with surrounding amino acid residues (Tyr603/A, Ser607/A, Ile610/A, Thr743/A, and Phe747/A) and Chl a ′ (P A ). There are no water molecules around P B in A. marina , but some water molecules surround P B without forming hydrogen bonds. The arrangement of these water molecules differs from that of T. elongatus PSI, and this difference around P B may come from the difference in two amino acid residues: Val594/B in A. marina vs. Thr597/B in T. elongatus , and Asn598/B in A. marina vs. His601/B in T. elongatus . The hydrogen bonding pattern around P A probably contributes to the charge distribution ratio (P A ˙ + /P B ˙ + ) 1 , 39 , and therefore, is likely different in the two organisms. The midpoint potential values, E m (P A ) and E m (P B ), are influenced by the protein environment, in particular by the presence of charged residues. The E m difference, E m (P A ) − E m (P B ) (=Δ E m ), is an important factor in determining the P A ˙ + /P B ˙ + ratio 40 . The orientations of the formyl group in P A (Chl d ′) and P B (Chl d ) were identified by considering the distribution of the cryo-EM density (Fig. 4a, b ). No amino acid residues and pigments capable of forming hydrogen bonds with these formyl groups were found in the vicinity of P A and P B . This suggests that these formyl groups form hydrogen bonds with the C5 H atom in Chl d ′ and Chl d , respectively. Similarly, there were no amino acid residues and pigments capable of forming a hydrogen bond around the formyl group of Acc (Acc A and Acc B ; Fig. 4c, d ). However, the orientations of the formyl group of Acc (Acc A and Acc B ) were altered by the hydrophobic environment caused by Trp (Trp585/B and Trp599/A) when compared with P740. Fig. 4 Cryo-EM density maps around P740, Acc and A 0 . Each map around P740 ( a , b ), Acc ( c , d ), and A 0 ( e , f ) is shown in a mesh representation at 3.0 sigma contour level for a , b , e , and f , and 5.0 sigma for c and d for ease of recognition of the water molecule coordinating the Mg 2+ of chlorin. Colors are the same as in Fig. 3 . Previous studies have suggested that the primary electron acceptor A 0 in A. marina PSI (Fig. 1a ) is Chl a (refs. 13 , 41 ), in line with other species 1 , 42 . However, we found that there was no Mg 2+ -derived density at the center of the tetrapyrrole rings of A 0A and A 0B , but rather a hole in the cryo-EM density map (Figs. 4e, f and 5a ). This indicates, when combined with the result of pigment analysis (Supplementary Table 6 and Supplementary Fig. 19 ), that A 0 is actually Pheo a , a derivative of Chl a . Furthermore, the position of Leu665/B (Supplementary Fig. 20b ) 43 pointing to A 0B , supports Pheo a as A 0B in A. marina , because Leu cannot serve as a ligand to Mg 2+ in Chl. Thus, surprisingly, A 0A and A 0B are assigned as Pheo a in A. marina PSI (Figs. 1b , 3a , and 5a–d ). The present study reveals that the axial amino acid residues at the A 0A and A 0B sites are Met686/A and Leu665/B, respectively. In comparison, the A 0A and A 0B sites in T. elongatus PSI are both occupied by Chl a molecules whose central Mg 2+ ions are both coordinated by Met residues 1 , 43 . At present, it is unclear how the A. marina PSI selectively binds Pheo a at the A 0 sites instead of other pigments (Chl a or Chl d ). However, the switch from Chl a to Pheo a in the A 0A and A 0B sites of A. marina PSI is probably not entirely dependent on the difference in its central ligand. Fig. 5 Arrangement of A 0 , Acc, and the surrounding protein environment. The structural model of PSI from A. marina is superposed with that from T. elongatus (PDB code 1JB0). a Cryo-EM density map of P A , P B , A 0A , and A 0B . Each map is shown in a mesh representation at 1.0 sigma contour level. b Arrangement of A 0 , Acc, and the surrounding protein environment. c As in b , but view from the pseudo-C 2 axis. d Geometry of P A /P B and Acc (Acc A and Acc B ). Chl d , green; Pheo, orange; water molecule, red. Colors are the same as in Fig. 3 . The amino acid residue nearest to A 0A in A. marina is Met (Met686/A; Supplementary Fig. 20a ) 43 and the identity of close amino acid residues may modify the reduction potentials of the two A 0 sites and their absorption peak wavelength(s). Two mechanisms have been proposed for electron transfer in PSI reaction centers, one using both A- and B-branches 44 , 45 , and the other using the A-branch preferentially 46 , 47 . Only the A-branch may be active in A. marina PSI 43 . Two previous studies have reported different mechanisms for the delocalization of the charge distribution in P740 (refs. 48 , 49 ). Future theoretical studies using the A. marina PSI structure may throw more light on the details of the electron transfer mechanism in relation to the P A ˙ + /P B ˙ + ratio. Why is Pheo a the primary acceptor, A 0 , in A. marina PSI? According to the midpoint potential value ( E m ), it seems reasonable for the Chl d -driven PSI to use Pheo a as A 0 , since the energy gap is sufficient for the primary electron transfer step as is estimated below. The E m of P740 vs. the standard hydrogen electrode (SHE) is 439 mV (refs. 12 , 13 , 18 , 50 ), which is comparable with that of P700 (470 mV, ref. 51 ). While the special pair P740, which are Chl d / d ′, use low-energy far-red light of 740 nm (1.68 eV), it generates reducing power almost equivalent to that of the P700 (Chl a / a ′, 1.77 eV) in plants and most cyanobacteria 12 . The produced reducing power of the excited state of P740 (P740*) is weaker by 0.09 eV than that of P700* (i.e., 1.77–1.68 eV). This could result in a slower electron transfer rate to A 0 and an increase in reverse reaction without a change in E m of A 0 , due to the smaller driving force. It is reported that the rates of electron transfer from P740* to A 0 , and to PhyQ are actually comparable to those from P700* to A 0 and to PhyQ 41 . Then, E m of A 0 has to change for a proper forward reaction. Because most of the amino acid residues around A 0 in A. marina PSI are similar to those in T. elongatus , it is unlikely that the protein structure around A 0 influences the E m value. Therefore, we looked at the E m value of the cofactor molecule itself, and obtained E m (vs. SHE) values of purified Chl a , Chl d , and Pheo a in acetonitrile of −1100, −910, and −750 mV, respectively (Supplementary Fig. 21 ) 51 , 52 . The E m of Pheo a is the highest. Accordingly, Pheo a as A 0 should achieve the same electron transfer efficiency as the Chl a -type PSI. Then, reinvestigation of a possible effect on rate of the following A 0 to PhyQ (A1) step may be warranted as the charge recombination kinetics between P740 + and A 1 − has been reported to be comparable to those between P700 + and A 1 − of Chl a -type PSI 50 . A. marina contains a limited but distinct amount of Chl a (refs. 12 , 14 , 43 ), and we found a small amount in the PSI (1–2 Chl a per PSI monomer; Supplementary Table 6 ). It was once assumed that A 0 is Chl a (ref. 41 ), but we now know this to be incorrect. One possible place for Chl a is as the accessory Chls, Acc A , and Acc B (Figs. 1 and 5 ). Unfortunately, the functional group containing C3 (ref. 1 ) of Acc Chl could not be precisely defined from the cryo-EM density map. While quantum mechanical/molecular mechanical calculations show that the formyl group of Chl d adopts two orientations (Supplementary Fig. 1c, d ) 53 , the oxygen atom of the formyl group on the Chl d molecule in a vacuum is more stable when oriented toward the C5 H atom, suggesting the conformer in Supplementary Fig. 1c . In contrast, the vinyl group of Chl a can adopt either orientation. Therefore, Acc could not be conclusively identified as Chl d or Chl a at the present resolution, and we assigned Accs as Chl d in this study. The Mg 2+ in Acc A and Acc B are coordinated by water molecules forming hydrogen bonds with Asn588/B and Asn602/A, respectively (Fig. 5b, c ). In T. elongatus PSI, the methyl ester groups of the Chl a in Acc A and Acc B affect the charge and spin distributions on P700 (refs. 40 , 54 – 56 ). In A. marina PSI, these distances were estimated to be 6.2 Å (Acc A –P A ) and 6.6 Å (Acc B –P B ), respectively, in the present structure. The amino acid residues surrounding PhyQ (A 1A and A 1B ) are switched compared with those in T. elongatus PSI—Met720/A and Leu665/B in A. marina vs . Leu722/A and Met668/B in T. elongatus (Supplementary Fig. 22a, b ). The arrangement of the phytol chain of A 1B is also different from that in other cyanobacteria (Supplementary Fig. 22b ). These structural differences may suggest that the protein environment within A. marina PSI containing Chl d is modified to support forward electron transfer by suppressing the reverse electron transfer from PhyQ to A 0 . However, cofactors F X , F A , and F B and their surrounding structures in A. marina PSI are nearly identical to those in other cyanobacteria (Fig. 3a and Supplementary Fig. 22 ). This indicates that the reduction potentials are likely at the same level, although the three Fe/S clusters, Fx, F A, and F B , transfer electrons from PhyQ (A 1 ) to ferredoxin and the environment does not need to be conserved as long as the overall energy trajectory is downhill. A. marina PSI is a trimer as the other cyanobacterial PSIs, but the amino acid sequence identity of each protein subunit is low, and, not unexpectedly, the divergence is most pronounced around the pigments specifically observed in A. marina PSI, for example Chl d , α-Car, and Pheo a -A 0 (Supplementary Table 5 ). Again, the alterations must reflect optimization to drive efficient photochemistry utilizing low-energy light. In conclusion, the structure and identification of electron carriers and key light-harvesting pigments provide a basis for understanding how low-energy far-red light is utilized by A. marina PSI. However, the full picture must wait for the structure of A. marina PSII, the system responsible for oxygen evolution through water-splitting, to be elucidated 3 ."
} | 7,179 |
30177768 | PMC6120894 | pmc | 3,830 | {
"abstract": "While biofilms are known to cause problems in many areas of human health and the industry, biofilms are important in a number of engineering applications including wastewater management, bioremediation, and bioproduction of valuable chemicals. However, excessive biofilm growth remains a key challenge in the use of biofilms in these applications. As certain amount of biofilm growth is required for efficient use of biofilms, the ability to control and maintain biofilms at desired thickness is vital. To this end, we developed synthetic gene circuits to control E . coli MG1655 biofilm formation by using CRISPRi/dCas9 to regulate a gene (wcaF) involved in the synthesis of colanic acid (CA), a key polysaccharide in E . coli biofilm extracellular polymeric substance (EPS). We showed that the biofilm formation was inhibited when wcaF was repressed and the biofilms could be maintained at a different thickness over a period of time. We also demonstrated that it is also possible to control the biofilm thickness spatially by inhibiting wcaF gene using a genetic light switch. The results demonstrate that the approach has great potential as a new means to control and maintain biofilm thickness in biofilm related applications.",
"introduction": "Introduction Biofilms are widely found in nature and they are commonly formed by group of microorganisms sticking onto surfaces and forming slimy extracellular matrix. Biofilms are known to cause problems in many areas of human health 1 and the industry 2 , 3 , including food, marine, and environment. However, by exploiting the biofilms’ unique characteristics of being tolerant and persistent in harsh environments, biofilms have also been found to be beneficial in a number of applications including wastewater management 4 , 5 , bioremediation 6 , continuous bioproduction of valuable chemicals 7 – 9 and production of biomaterials 10 – 12 . Biofilms are known to have high tolerance against low pH, toxicities and antimicrobial agents which makes it useful for processing wastes and bioproduction of chemicals that are toxic to the host cells (e.g., succinic acid 8 and ethanol 13 ). Biofilms are also able to achieve high cell density and cells in biofilm state usually have higher productivity compared to planktonic cells 14 . Together with the property that the biomass can be retained during the changing of medium, biofilms are attractive as a means for continuous bioproduction 15 , 16 . A common and important challenge in utilising biofilm in these applications is the excessive growth of biofilms and associated extracellular polymeric substances (EPS), causing clogging, biofouling and loss in productivity 7 , 15 , 17 , 18 . Excessive growth of biofilms and associated EPS would also greatly limit the diffusion of substrates and nutrients to the cells 19 , 20 . Consequently, the cells in the inner layers of biofilm would become less active. For the use of biofilms in wastewater treatment, the thickness of biofilm has shown to influence pollutant removal efficiency 16 , 21 , 22 . Pollutant removal efficiency of biofilm increases within certain thickness due to increased biomass, but starts to decrease as the biofilm becomes thicker because of the diffusion limitation 21 , 22 . In bioproduction, excess growth in biofilm limits the feed availability, thus reducing the bioreactor efficiency and productivity 15 , 18 . Hence, to enable more efficient use of biofilm, there is a need to be able to effectively control the growth of biofilm and maintain the thickness of the biofilm. Due to the problems caused by biofilms, most of the previous studies have focused on either eradicating biofilms or preventing biofilm formation. These biofilm prevention/removal methods include: (i) mechanical methods (e.g., mechanical scraping; bubbling and vibration); (ii) chemical methods (e.g., the use of surfactant such as δ-hemolysin to remove biofilm 23 and the use of nitric oxide which can lead to biofilm dispersal 24 ); and (iii) biological methods. A number of biological methods which target different aspects of biofilm formation have been reported 25 . These methods include using enzymes, such as proteinase and DNase to degrade the extracellular proteins and DNA respectively 25 ; expressing proteins/peptides that can target second messengers (e.g. cyclic di-guanosine monophosphate (c-di-GMP)) that regulate biofilm development 26 – 28 , and peptide that reduces stress responsive guanosine pentaphosphate (ppGpp) concentration 29 . Other biological methods also include quorum quenching that interrupts autoinducer-mediated quorum sensing 20 , 30 and controlling the production of adhesive proteins such as csgA which could inhibit initial attachment of the bacteria 12 . These methods have been demonstrated to be able to either remove biofilm or prevent biofilm formation. Among these methods, targeting c-di-GMP and quorum quenching are most well studied. Although targeting the signaling molecules such as c-di-GMP and quorum sensing autoinducers can be effective in reducing biofilm formation, it might affect the cell’s performance in the bioprocess as c-di-GMP and autoinducers are also involved in many other cellular activity pathways 31 . Methods that inhibit cell attachment could prevent biofilm formation but they would not be suitable for the applications that require certain amount of biofilm 32 . Thus, the use of these methods is limited to achieve long term biofilm control in continuous bioprocess. Taken together, there still lack methods that could maintain the biofilm at a certain desired thickness. In this paper, we aimed to engineer a strain of E . coli (MG1655) in which its biofilm thickness can be controlled and maintained. This is achieved by developing synthetic gene circuits coupled with CRISPRi/dCas9 to regulate a gene (wcaF) within the E . coli MG1655 genome involved in exopolysaccharide colanic acid synthesis 33 . Previous studies have shown that CA affects the formation of biofilm thickness and it does not appear to be involved in the initial attachment 34 , 35 . Domka et al . showed that genes encoded for CA synthesis were up-regulated in mature biofilms 36 , while Danese et al . showed that E . coli K-12 with mutated wcaF gene could only form biofilm at one or two cell layers in depth 37 . Here, we hypothesised that by directly controlling the expression of wcaF gene which is involved in the synthesis of CA we would be able to achieve a more direct control of the biofilm thickness. To this end, synthetic gene circuits controlled by chemicals or light were developed to regulate the expression of wcaF using CRISPRi/dCas9. We designed gRNA that targets the wcaF gene within the E . coli MG1655 chromosome. Potentially, other genes involved in colanic acid synthesis could also be targeted 33 . Our results showed that it is possible to control and maintain the E . coli MG1655 biofilm thickness over time by regulating the expression of wcaF gene. To the best of our knowledge, this is the first study on inhibiting wcaF gene within the CA synthesis using CRISPRi/dCas9 for the control of biofilm growth. Our approach differs from commonly studied strategy which targets global regulators such as cyclic-di-GMP and quorum sensing molecules. The presented approach has great potential as a new means to control and maintain established E . coli 1655 biofilm thickness, which would be very useful in biofilm related applications (e.g., controlling biofilm used in wastewater treatment at certain thickness to prevent clogging and potentially increase pollutant removal efficiency).",
"discussion": "Discussion In this paper, we have engineered E . coli MG1655 in which its wcaF gene within the genome can be regulated using aTc or light. wcaF was hypothesised to encode acetyltransferase which is involved in the synthesis of colanic acid 33 . We showed that regulating the wcaF gene affects biofilm formation. Previous studies have showed that colanic acid is essential for biofilm maturation and is related to biofilm thickness 34 . Moreover, E . coli MG1655 with mutated colanic acid cluster gene wcaF was not able to build three-dimensional biofilm structure 37 . It was hypothesised that wcaF gene is involved in the acetylation of colanic acid synthesis 33 . Besides, acetyl groups are able to increase the adhesive and cohesive properties of biofilm 40 . Unlike commonly used biofilm control methods such as using dispersin and quorum quenching which target global regulator that might have effect on many aspects of the cells, targeting wcaF gene would mainly affect colanic acid synthesis. Here, we showed that regulating wcaF gene that involved in colanic acid synthesis could potentially be an effective method to control biofilm formation, specifically maintaining established biofilm thickness. We first investigated the use of CRISPRi to interfere the wcaF gene in the E . coli genome. wcaF gene has been suggested to be involved in colanic acid synthesis by a study involving mutations of the gene 37 . It was unclear whether directly regulating wcaF gene using CRISPRi could be used to control biofilm formation. Here, we showed that the formation of biofilm was effectively inhibited when the wcaF gene was targeted by the designed gRNA. In addition, cell growth was minimally affected when wcaF gene was repressed using the synthetic gene circuits. Consequently, this result suggests that it could be possible to maintain the biofilm at different established thickness using our approach. By inducing gRNA at different time point, we found that the thickness of biofilms with wcaF gene repressed after 6, 8 or 24 hours, but all cultured for a total of 30 hours remain comparable with the wild type biofilms that were grown for 6, 9 and 24 hours respectively. Our study has been performed using petri dishes. It will be interesting to further study the method in a more continuous manner with the use of microfluidics 41 . Nonetheless, as a proof of concept, these results demonstrate that repressing wcaF gene could be a promising method to maintain established biofilm of different thickness. To the best of our knowledge, this is the first study to demonstrate the maintenance of biofilm thickness through the inhibition of wcaF gene using CRISPRi. This could be particularly useful for applications related to wastewater treatment and bioproduction in which the thickness of biofilm has shown to influence pollutant removal efficiency in wastewater treatment and bioproduction 15 , 16 , 18 , 21 , 22 . We were interested to test whether we could spatially or locally control biofilm formation by regulating wcaF gene, as it would provide an added capability to control biofilm thickness at different location of a surface. Hence, we studied the expression of gRNA wcaF under a blue light repressible promoter. The results showed that blue light could be used to regulate wcaF gene expression and, consequently, control biofilm formation locally. Interestingly, using light to control the expression of gRNA that regulates wcaF gene could create biofilm 2D patterning. On the same substrate surface, only the biofilm grew in the blue light shined area could form three-dimensional structure. This offers a potential means for localised control of biofilm thickness using light which could be useful in biomaterial patterning 10 – 12 . However, as repressing wcaF gene would not disperse biofilm, additional control such as introduction of dispersin into the gene circuit would be required for applications in which dispersal is necessary. The EPS forms a protective layer that protects the cells within the biofilm against extreme/harsh conditions (e.g., low pH, high toxicity and antibiotics). To investigate whether the tolerance of the engineered biofilm would be significantly affected with the inhibition of wcaF gene expression, we designed experiments to test its tolerance against antibiotic, erythromycin, which would cause bacterial cell death by binding to bacterial ribosome and inhibiting protein synthesis. The tolerance of the biofilm was tested by studying the ratio of live cells against the total cells which include both live and dead cells 42 . The results show that the live cells ratio of the engineered biofilm (in which wcaF gene expression had been repressed after 6 hours) was similar with the wild type biofilm (grown for 6 hours) after treating with erythromycin. This finding suggests that inhibiting wcaF gene did not significantly reduce biofilm tolerance as compared to the biofilm grown for the same period of time. This implies that the proposed method could maintain the property needed for applications involving harsh conditions such as the production of toxic chemical and wastewater management. In this paper, we targeted wcaF gene. Other than wcaF gene, other genes involved in CA synthesis pathway, such as wzx gene which is hypothesised to participate in CA polymerization and export, gene manB and manC which are required for synthesis of nucleotide sugar precursors of CA 33 , could be potential targets to inhibit CA synthesis. Targeting CA to regulate biofilm formation could potentially be applied to other strains that also have CA, such as Salmonella enterica serovar Typhimurium LT2 43 . For mixed species, if this genetic modification can be used to control biofilm of each species, the mixed species biofilms could be possibly controlled by tuning the CRISPRi. In summary, the results presented in this paper show that repressing wcaF involved in colanic acid synthesis using CRISPRi is potentially an effective means to control and maintain the thickness of biofilm, particularly for applications in wastewater treatment and bioproduction."
} | 3,437 |
37738348 | PMC10516496 | pmc | 3,831 | {
"abstract": "A neuromuscular junction (NMJ) is a particularized synapse that activates muscle fibers for macro-motions, requiring more energy than computation. Emulating the NMJ is thus challenging owing to the need for both synaptic plasticity and high driving power to trigger motions. Here, we present an artificial NMJ using CuInP 2 S 6 (CIPS) as a gate dielectric integrated with an AlGaN/GaN-based high-electron mobility transistor (HEMT). The ferroelectricity of the CIPS is coupled with the two-dimensional electron gas channel in the HEMT, providing a wide programmable current range of 6 picoampere per millimeter to 5 milliampere per millimeter. The large output current window of the CIPS/GaN ferroelectric HEMT (FeHEMT) allows for amplifier-less actuation, emulating the biological NMJ functions of actuation and synaptic plasticity. We also demonstrate the emulation of biological oculomotor dynamics, including in situ object tracking and enhanced stimulus responses, using the fabricated artificial NMJ. We believe that the CIPS/GaN FeHEMT offers a promising pathway for bioinspired robotics and neuromorphic vision.",
"introduction": "INTRODUCTION The somatosensory system in biological organisms is responsible for detecting and responding to various external stimuli, including vision, sound, odor, pressure, and temperature ( 1 – 3 ). To achieve these reactions, external stimuli detected by the afferent nerve (sensory neurons) are first transferred to the central nervous system (CNS) ( 4 ). The CNS then generates an action potential for the efferent nerve (motor nerve) that actuates the target muscle through the neuromuscular junction (NMJ). The NMJ is a unique and essential synaptic connection between the efferent nerve and muscle fibers that triggers motion via the transmission of action potentials through it ( 5 ). Consequently, the stimulated muscle fibers contract and relax, becoming capable of triggering macro-motions. Macro-motions generally require substantially higher energy than that required for computation. Therefore, it has been challenging to emulate NMJs to fulfill both synaptic plasticity and the capability to drive large amounts of energy ( 6 – 9 ). To address these challenges, here we demonstrate synaptic transistors by heterogeneously integrating a CuInP 2 S 6 (CIPS) ferroelectric membrane as a gate dielectric material with an AlGaN/GaN high-electron mobility transistor (HEMT). The programmable transconductance of ferroelectric transistors allows for artificial synaptic plasticity, which is attributed to the polarization of the ferroelectric gate dielectric layer. Among many ferroelectric materials, including Zr-doped Hf 1− x Zr x O ( 10 ), PbZr 0.52 Ti 0.48 O 3 ( 11 ), BaTiO 3 ( 12 ), α-In 2 Se 3 ( 13 ), and BiFeO 3 ( 14 ), CIPS is a unique two-dimensional (2D) van der Waals (vdW) material with a relatively wide bandgap and out-of-plane ferroelectricity that offers high integrability, flexibility, and responsivity to electrical signals ( 15 – 20 ). The out-of-plane ferroelectricity in the CIPS is attributed to the ionic dynamics of the Cu and In cations, which are vertically displaced in the sulfur framework ( 21 ). We use a CIPS/GaN ferroelectric HEMT (FeHEMT) to achieve high-power driving capability. GaN-based HEMTs are widely used in radio frequency (RF) and power electronics applications because of their high output current and fast switching capability by using an AlGaN/GaN heterostructure, which forms a 2D electron gas (2DEG) transport channel ( 22 ). Thus, it offers a high-electron saturation velocity, a high breakdown electrical field, and a high-electron mobility suitable for high-power and high-frequency applications ( 23 , 24 ). The fabricated CIPS/GaN FeHEMT exhibited hysteresis in the current-voltage ( I - V ) characteristics that emulate biological short/long-term plasticity (STP/LTP) ( 25 ). The feasibility of the CIPS/GaN FeHEMT as an artificial synapse was experimentally verified by characterizing its programmability, retention time, and endurance. In addition, we demonstrate enhanced reflexes by accelerating the spike timing of the ferroelectric switching in CIPS/GaN FeHEMT, analogous to the enhanced response to the stimulus in biological systems. The enhanced reflexes were achieved by connecting a CIPS/GaN FeHEMT with a complementary metal-oxide semiconductor (CMOS)–based integrate fire unit (IFU) that accumulates the input spikes and fires the output spike when the accumulated spike exceeds a threshold. Last, we used a CIPS/GaN FeHEMT for direct synaptic power transmission to the microelectromechanical system (MEMS)–based actuators without amplifier circuits, mimicking the oculomotor NMJ that triggers muscular motion of the eyeball for in situ object tracking. We believe that the proposed device can potentially play a key role in the implementation of artificial NMJ for robotic and artificial muscle systems.",
"discussion": "DISCUSSION We have demonstrated that an artificial NMJ can generate a programmable high current by using the tunable ferroelectric coupling between the CIPS gate dielectric and 2DEG channel at the AlGaN/GaN interface. Figure S10 shows the simulated CIPS thickness-dependent hysteresis loops, revealing that the largest hysteresis with an enhanced ∆ V th is achieved with a CIPS membrane of 100 nm, while we used a 60-nm-thick CIPS membrane in this study. We believe that the programmability of CIPS/GaN FeHEMTs can be further improved by precisely controlling the CIPS thickness. Analogous to conventional GaN HEMTs, the proposed CIPS/GaN FeHEMT also potentially attains nanosecond operation by etching the AlGaN barrier layer to minimize the surface potential and by using the thinner CIPS membrane to alleviate the voltage drops at CIPS ( 22 , 37 ). The direct actuation of the mechanical platform using the artificial NMJ provides a wide range of neuromorphic sensing-to-action applications, including time-of-flight ranging ( 38 – 43 ), in-sensor/near-sensor computing ( 44 – 48 ), and human-computer interaction ( 49 ). In this study, we achieved a normalized output current of 200 mA/mm with the CIPS/GaN FeHEMT, which is notably greater than that of recently reported synaptic transistors (for more details, see table S1) ( 50 – 59 ). Therefore, the CIPS/GaN FeHEMT is potentially deployable as an artificial NMJ in robotic systems to operate mechanical actuators that require a milliampere-scale driving current for macro-motion. Moreover, the 2DEG transport channel of HEMT offers a gigahertz range frequency response that can be coupled with integrated ferroelectricity for reconfigurable RF applications ( 13 , 29 ). We also applied a CIPS ferroelectric membrane integrated with a GaN FeHEMT for artificial oculomotor dynamics. The high output current achieved by the AlGaN/GaN 2DEG enabled amplifier-less actuation for the adduction and abduction motions. The polarization of the CIPS tuned the output current of the CIPS/GaN FeHEMT using the enhancement pulse at its gate node. This nonvolatile artificial synaptic device was connected to the CMOS-based efferent system that integrates and fires the spike to actuate the mechanical platform, such as a mechanical object tracker. The temporal dynamics of the CIPS/GaN FeHEMT were modulated by the enhancement process, analogous to the biological stimulus response. We believe that artificial NMJs with high-power and high-frequency electronic components have the potential to realize functional bioinspired elements for artificial muscles and smart robotic applications."
} | 1,885 |
36012853 | PMC9409915 | pmc | 3,834 | {
"abstract": "To explore the effect of arbuscular mycorrhizal fungi (AMF) on the environmental migration of cadmium (Cd), a sand column-maize system containing 20 mg·L −1 Cd solution was used to investigate the AMF effect on maize growth, Cd uptake by maize, Cd adsorption by sand and Cd leaching loss. The results showed that AMF significantly increased the content of EE-GRSP and T-GRSP by 34.9% and 37.2%, respectively; the secretion of malonic acid, oxalic acid and succinic acid increased by 154.2%, 54.0% and 11.0%, respectively; the secretion of acetic acid and citric acid increased by 95.5% and 59.9%, respectively; and the length, surface area, volume, tip number and cross number of maize roots decreased by 10%, 15%, 17%, 20% and 36.4%, respectively. AMF significantly increased Cd adsorption by sand by 6.2%, Cd uptake by maize by 68.1%, and Cd leaching loss by 84.6%. In the sand column-maize system, 92.3% of the total Cd was adsorbed by sand, 5.9% was taken up by maize and 1.8% was lost due to leaching. Moreover, Cd adsorption by sand was significantly positively correlated with the GRSP content and oxalic acid secretion, and Cd uptake by roots was significantly negatively correlated with Cd leaching loss. Overall, AMF reduced the loss of Cd in the leaching solution by promoting the release of oxalic acid and GRSP, increasing the adsorption of Cd in the sand and fixing the Cd in the plant to the roots.",
"conclusion": "5. Conclusions This study showed that AMF inoculation plays a significant role in reducing the leaching of Cd from contaminated sand. The subtle changes of Cd adsorption by sand were closely related to environmental Cd migration (Cd uptake by plants and solution Cd loss). However, further studies are necessary to determine the influence of AMF on the leaching of cadmium from contaminated soil under different environmental conditions.",
"introduction": "1. Introduction Due to industrial and agricultural production activities such as metal mining and smelting, sewage irrigation and sludge application, the pollution of heavy metal cadmium (Cd) in farmland soil has become increasingly serious and one of the major environmental problems in the world [ 1 ]. The majority of the Cd that enters the soil is absorbed by soil particles and eventually generates a Cd pool in the environment [ 2 ]. However, the activity and migration ability of Cd is strong [ 3 ]. Soil Cd easily migrates with water flow generated by precipitation and irrigation [ 4 ] and is taken up by plant roots [ 5 ]. The environmental migration of Cd derived from soil Cd pools, such as leaching and loss with water flow and uptake by plants, destroys the permeability of the cell membrane, makes the binding enzymes and other substances on the membrane dysfunctional, accelerates the extravasation of the substances needed for plant growth, causes the physiological and biochemical processes in the body to be disordered, changes the hormone levels and nucleic acid metabolism levels in plants, and accelerates plant death [ 6 , 7 ]. Therefore, the problem of Cd leaching and loss in contaminated soil has attracted widespread attention. Arbuscular mycorrhizal fungi (AMF) are widely distributed soil fungi that can form symbiotic relationships with approximately 71% of angiosperms (including most crops) [ 8 , 9 ]. According to Janouková and Pavlková [ 10 ], AMF spread out and expand in the soil to produce dense hyphal networks, the hyphae wrap around the soil particles, and there are binding sites on the hyphae that can adsorb and fix Cd, reducing the loss of Cd. Hyphae secrete a protein called glomalin-related soil protein (GRSP), which can alter the surface characteristics of soil particles and increase Cd adsorption [ 11 ]. AMF can also boost the nutrition and growth of the host plant, promote root growth and the secretion of low molecular organic acids, further promote the chelation and adsorption of Cd, reduce the toxicity of Cd and play an important role in soil Cd fixation [ 12 , 13 , 14 , 15 ]. AMF significantly affects the Cd distribution in plants; some studies found that Cd was mainly fixed in AMF-inoculated roots, and Cd transport to shoots was reduced, resulting in less Cd uptake in shoots [ 16 , 17 , 18 ]. In addition, AMF reduces soil Cd leaching loss under rainfall conditions [ 19 ]. Therefore, AMF significantly affects the environmental migration of Cd in soil-plant systems. Because most of the previous studies used polluted soil as the matrix, the soil composition was complex, which made the quantitative analysis of Cd migration in the soil-plant system difficult to control. Amir et al. [ 20 ] found that although AMF could promote Cd adsorption by soil, this effect was subtle. The influence of the minor changes in Cd adsorption in soil induced by AMF on environmental Cd migration, such as plant Cd uptake and Cd leaching loss, is still unclear. Replacing the soil matrix with quartz sand with simple composition can effectively reduce the interference of complex soil composition and is often used in quantitative experimental studies to control a single variable [ 21 ]. Due to the large biomass, developed root and certain Cd tolerance, maize is often used in the study of AMF on crop growth and Cd absorption under Cd stress [ 22 ]. Therefore, in the present study, the experiment using sand culture with Cd treatment was carried out to study the effect of AMF inoculation on maize growth, Cd content and Cd migration in the sand column-maize system. We expected that AMF would have a minor impact on sand Cd adsorption, but would have a significant impact on plant Cd absorption and Cd leaching loss, as well as modify the Cd distribution in the sand column-maize system.",
"discussion": "4. Discussion 4.1. Effect of Arbuscular Mycorrhizal Fungi on Cadmium Adsorption by Sand The soil Cd pool is a Cd repository, and its changes will affect the reactivity, mobility, and bioavailability of trace elements, thereby altering biodiversity, plant metabolism, and physiological processes [ 33 ]. The GRSP produced by the decomposition of AMF hyphae can enhance soil stability and has a strong binding ability to Cd, thus promoting the adsorption capacity of soil for Cd [ 34 , 35 , 36 ]. In addition, inoculation with AMF can promote the release of low molecular weight organic acids from roots, reduce soil pH, change the form of Cd, promote the chelation and adsorption of Cd, and affect the fixation of Cd in soil [ 37 ]. This is consistent with the conclusion found in this experiment that inoculation with AMF and Cd stress significantly increased the contents of malonic acid, oxalic acid, succinic acid and malic acid. This experiment also found that EE-GRSP has a very significant positive correlation with oxalic acid, succinic acid, and malonic acid, and oxalic acid has a very significant positive correlation with Cd adsorption by sand, indicating that AMF can promote Cd adsorption by sand by promoting root exudates. This is because organic acid molecules can chemically react with free Cd ions to generate stable compounds that exist in soil solutions and promote soil desorption of Cd [ 38 ]. However, this experiment also found that T-GRSP and EE-GRSP are significantly negatively correlated with acetic acid and citric acid, and Cd adsorption by sand is significantly negatively correlated with acetic acid and citric acid. The reason for this phenomenon is the biological characteristics of the strain itself or the inconsistency of the affinity between the strain and the host plant, and additionally, the differences in the soil environment and the physiological and biochemical characteristics of the plant itself can also cause this phenomenon [ 39 ]. 4.2. Effect of Arbuscular Mycorrhizal Fungi on Plant Growth and Cadmium Uptake The crop root is the part of the plant that comes into direct contact with the soil and is the principal organ for plant uptake, and its development is frequently influenced by Cd toxicity [ 40 ]. AMF inoculation has been demonstrated to enhance the root absorption range, increase the root-to-soil contact area, and promote root growth [ 41 ]. However, another study also found that if there is enough soil fertility in the growth medium to meet the needs of the roots, then the AMF hyphae will have difficulty obtaining nutrients to promote plant growth, thereby inhibiting the growth of the host plants [ 9 ]. Therefore, in our study, inoculation with AMF significantly reduced root volume, root tip number and root crossings number and reduced root and shoot biomass. AMF symbionts produce a huge mycelial network, which requires a large amount of nutrients to maintain their own growth and development and forms a part of the competition relationship with maize plants in nutrient absorption, which does not meet the growth and development needs of maize plants [ 42 ]. In addition, AMF inoculation leads to an increase in Cd content in maize roots, and the uptake of toxic Cd ions could mask the positive effect of mycorrhizae and change maize root characteristics, thereby inhibiting maize growth [ 43 ]. Cd uptake can reflect the absorption capacity of plants to soil Cd and can also reflect the degree of Cd stress in plant growth and development, that is, the tolerance of plants to Cd stress. In Cd-contaminated soil, AMF can play the role of filtration and chelation, adsorb many Cd through a large mycelial network, or “compartmentalize” Cd through structures such as arbuscules and vesicles and fix them in the symbiotic interface to inhibit the transport ability of Cd in plants and further weaken the toxic effects of Cd. At the same time, the inoculation of AMF can enrich the Cd ions at the binding sites of hyphae through the physical adsorption, coordination and microprecipitation of metal ions on the outer surface of cells and the binding effects of hyphae cell walls and intracellular thermostable proteins, polyphosphoric acid and organic acids, thus enhancing the adsorption performance and ability of plants to Cd [ 44 ]. In this experiment, it was found that inoculation with AMF also promoted the Cd uptake of plants but decreased the transfer of Cd to shoots, which also alleviated the damage of Cd to shoots. This is because inoculation with AMF will fix Cd in the vacuoles or cell walls of plant roots, reduce the accumulation of Cd in stems and leaves, change the absorption and distribution of Cd in maize, and hinder the transfer of Cd to stems and leaves [ 45 , 46 ]. However, some studies have found that inoculation with AMF reduced the Cd uptake of plants [ 47 ], which is because GRSP precipitates Cd in the soil through binding transformation, reducing the available state of Cd in soil and reducing the uptake of Cd by plants [ 8 ]. The effects of AMF inoculation on the Cd uptake of plants also had different results depending on the plant species, soil pollution, AMF species and their different combinations. In addition, the change in soil pH caused by root exudates also affects the content of Cd in solution and the content of available Cd in soil, which in turn affects Cd leaching loss and Cd uptake in plants [ 18 ]. 4.3. Effect of Arbuscular Mycorrhizal Fungi on Cadmium Leaching Loss Leaching Cd from the soil is a complicated process. In addition to crop roots, other factors, such as soil physicochemical qualities, soil texture, pollution levels, and others, might influence Cd leaching [ 48 , 49 ]. The matrix in this experiment was pure quartz sand, and exogenous Cd (CdCl 2 ·2.5H 2 O) was introduced to the quartz sand column, with no influence on Cd loss. There are multiple migratory pathways for Cd after it enters the sand column-maize system: sand adsorption, maize absorption, and leaching loss. In this experiment, the content of Cd applied to the sand column-maize system was known. After maize growth and leaching, the Cd content and proportion of each part can be determined to clarify the effect of AMF on the distribution of Cd in the maize system. It was found that the proportion of Cd in the sand column-maize system decreased from high to low, followed by Cd adsorption by sand > Cd uptake in maize > Cd leaching loss, and inoculation with AMF reducing the Cd leaching loss by 85.17%, indicating that AMF played a significant role in reducing Cd leaching loss. In the experiment, it was found that Cd leaching loss was significantly negatively correlated with the uptake of Cd in shoot and was significantly negatively correlated with the uptake of Cd in the roots, indicating that AMF reduced the amount of Cd leaching loss by promoting the uptake of Cd by plants [ 50 ]. Second, AMF acts on soil particles through hyphae to form the skeleton of soil aggregates and further forms microaggregates and macroaggregates, which improves the soil structure and reduces Cd leaching loss. In summary, AMF play a significant role in immobilizing Cd and reducing their migration from leaching from contaminated soils. The subtle changes in the soil Cd pool were closely related to environmental Cd migration (Cd uptake by plants and solution Cd loss). However, this experiment was an indoor sand culture experiment, and the process of crop growth and leaching is affected by many environmental factors. It is necessary to further study the influence of AMF on the leaching of contaminated soil under different environmental conditions and its mechanism. In addition, the forms of Cd in leaching solution also need to be further studied."
} | 3,362 |
19701714 | PMC2767516 | pmc | 3,835 | {
"abstract": "A novel anaerobic, thermophilic, Gram-positive, spore-forming, and sugar-fermenting bacterium (strain TLO) was isolated from a geothermal spring in Ayaş, Turkey. The cells were straight to curved rods, 0.4–0.6 μm in diameter and 3.5–10 μm in length. Spores were terminal and round. The temperature range for growth was 40–80°C, with an optimum at 70°C. The pH optimum was between 6.3 and 6.8. Strain TLO has the capability to ferment a wide variety of mono-, di-, and polysaccharides and proteinaceous substrates, producing mainly lactate, next to acetate, ethanol, alanine, H 2 , and CO 2 . Remarkably, the bacterium was able to grow in an atmosphere of up to 25% of CO as sole electron donor. CO oxidation was coupled to H 2 and CO 2 formation. The G + C content of the genomic DNA was 35.1 mol%. Based on 16S rRNA gene sequence analysis and the DNA–DNA hybridization data, this bacterium is most closely related to Thermoanaerobacter thermohydrosulfuricus and Thermoanaerobacter siderophilus (99% similarity for both). However, strain TLO differs from Thermoanaerobacter thermohydrosulfuricus in important aspects, such as CO-utilization and lipid composition. These differences led us to propose that strain TLO represents a subspecies of Thermoanaerobacter thermohydrosulfuricus , and we therefore name it Thermoanaerobacter thermohydrosulfuricus subsp. carboxydovorans.",
"introduction": "Introduction Diverse thermophilic heterotrophic anaerobes have been isolated from a variety of habitats. Members of the genus Thermoanaerobacter , in the order Thermoanaerobacteriales , are widely distributed in hydrothermal and oil-producing vents, volcanic hot springs, non-volcanic geothermally heated subsurface aquifers, soil, sugar beet, and sugar cane extraction juices (Klaushofer and Parkkinen 1965 ; Wiegel and Ljungdahl 1981 ; Schmid et al. 1986 ; Cayol et al. 1995 ; Cook et al. 1996 ; Kozianowski et al. 1997 ; Larsen et al. 1997 ; Slobodkin et al. 1999 ; Fardeau et al. 2000 ; Kim et al. 2001 ; Onyenwoke et al. 2007 ; Wagner et al. 2008 ). Thermoanaerobacter species are strictly anaerobic, thermophilic, rod-shaped bacteria, growing between 55 and 75°C, and most of them form round to oval terminal spores. Thermoanaerobacter species have been investigated for their sensitivity to different antibiotics but no differences have been found. A wide range of carbohydrates can be utilized by this group of organisms. Although the end products are mainly acetate, lactate, ethanol, H 2 , and CO 2 , the most abundant end product depends on the species and the growth conditions. Generally, thiosulfate can be used as electron acceptor in anaerobic respiration. Here we describe a new anaerobic thermophilic bacterium which belongs to the genus Thermoanaerobacter , and which differs from its closest relatives with respect to CO-utilization, lipid composition and fermentation pattern.",
"discussion": "Discussion Here we describe the characterization of a novel thermophilic, Gram-positive, anaerobic bacterium, which was isolated from a geothermal spring in Ayaş in Turkey. Based on the 16SrRNA sequence, strain TLO is phylogenetically closely related to T. thermohydrosulfuricus and T. siderophilus (Fig. 1 ), although its G + C content of 35.1 mol% is different from that of T. hydrosulfuricus (37.6%) and T. siderophilus (32.0%). Strain TLO shares several phenotypic features with its close relatives, like a broad substrate specificity, optimal pH- and T-range and the facultative use of thiosulfate. However, several obvious differences were found as well (Table 1 ). First, in contrast to the type strain T. thermohydrosulfuricus and also T. siderophilus , strain TLO is able to use CO as electron donor, which is converted to H 2 and CO 2 . This feature is not new among Gram-positive anaerobes, and various recently isolated thermophiles have been shown to grow chemolithoautotrophically through the conversion of CO + H 2 O to H 2 + CO 2 (Svetlichnyi et al. 1994 , 2001 ; Sokolova et al. 2001 , 2002 , 2004 , 2005 , 2007 ; Slepova et al. 2006 ). Utilization of CO has also been demonstrated for several representatives of a subdivision of the Thermoanaerobacter genus (Group 3) (Subbotina et al. 2003 ), which was recently reassigned to the genus Caldanaerobacter (Fardeau et al. 2004 ). For example, Caldanaerobacter subterraneus subsp. tengcongensis and C. subterraneus subsp. pacificius are able to use CO. However, among the more distantly related, true Thermoanaerobacter strains, CO-dehydrogenase activity has not been demonstrated or determined yet, with the exception of strain TLO, as shown here (Table 1 ). It is important to realize that the bacterium either did not grow or did not completely oxidize CO at concentrations higher than 25% (v/v) in the gas phase. Often, growth is only tested in the presence of 100% CO, which may be toxic and then gives rise to false conclusions regarding the ability to grow on CO. Moreover, levels of CO in natural hot environments are probably much lower, and microorganisms growing on CO in such environments are likely to be able to use very low CO concentrations. For example, Carboxydothermus hydrogenoformans consumed CO to below detectable levels of 2 ppm, when the CO 2 concentration was kept low (A.M. Henstra et al., unpublished data). Table 1 Phenotypic characteristics of strain TLO in comparison with phylogenetically closely related species Feature Strain TLO \n T. thermohydrosulfuricus \n \n T. siderophilus \n Source Geothermal spring Extraction juices of beet sugar factory Hydrothermal vent Gram reaction + V + Spores + + + Temperature range ( o C) 40–80 40–78 39–78 Optimum temperature (°C) 68–70 65–68 69–71 pH range 4.5–9 5–9 4.8–8.2 Optimum pH 6.3–6.8 6.9–7.5 6.3–6.5 DNA G + C content (mol%) 35.1 37.6 32 Reduction of arsenate + + a \n + a \n Growth substrates Ribose + + NR Mannitol W W a \n V Starch + + + Pectin + + NR Xylan + + – Carboxymethylcellulose W + – CO (up to 25% v/v) + – a \n – b \n Data for reference species were obtained from Klaushofer and Parkkinen ( 1965 ) and Slobodkin et al. ( 1999 ) \n NR not reported, V variable, W weak \n a Tested in this report; b Data from Gavrilov et al. ( 2003 ) \n Second, significant differences in the lipid composition of strain TLO and the phylogenetically closely related species, T. thermohydrosulfuricus and T. siderophilus , were detected as shown in Table 2 . In T. siderophilus the iso-C17 FA was substantially more abundant than in the TLO strain and T. thermohydrosulfuricus . Interestingly, in all three strains the uncommon membrane-spanning α,ω -13,16-dimethyl-octacosanedioic acid, previously detected in the phylogenetically related T. ethanolicus (Jung et al. 1994 ; Lee et al. 2002 ) was present in substantial amounts, especially in strain TLO. Since increasing ethanol concentrations were found to be associated with high levels of C 30 fatty acids in T. ethanolicus (Burdette et al. 2002 ), the high tolerance to ethanol further verifies also the potential of strain TLO. Table 2 Lipid composition (in % of total quantified lipids) of strain TLO in comparison with phylogenetically closely related species Lipids Strain TLO \n T. thermohydrosulfuricus \n \n T. siderophilus \n FLF BLF FLF BLF FLF BLF Summed concentration (mg/g dry weight) 6.7 2.0 17.4 2.9 8.6 0.6 \n n -C 14 FA 4 1 n.d. n.d. 7 2 Iso-C 15 FA 39 25 81 60 3 40 Anteiso-C 15 FA 2 2 2 2 0 8 \n n -C 16 FA 5 1 2 2 19 5 Iso-C 17 FA 16 5 12 22 66 16 Anteiso-C 17 FA 2 1 1 2 5 4 \n n -C 18 FA n.d. n.d. n.d. n.d. n.d. 1 Iso-C 15 -OH 10 n.d. 1 n.d. <1 3 \n n -C 16 -OH 7 1 n.d. n.d. n.d. 1 Iso-C 17 -OH 9 9 1 n.d. n.d. n.d. Anteiso-C 17 -OH 4 <1 n.d. n.d. n.d. n.d. \n α,ω -13,16-Dimethyl-octacosanedoic acid 2 50 n.d. 12 n.d. 20 \n α,ω -13,16-Dimethyl-triacontanedioic acid n.d. 2 n.d. n.d. n.d. n.d. 30-Hydroxy-13,16-dimethyl-triacontanoic acid n.d. 3 n.d. n.d. n.d. n.d. The strains were grown in glucose-containing bicarbonate-buffered (BM) medium at 65°C for 24 h \n n.d. not detected, FA fatty acid, OH alcohol, FLF free lipid fraction, BLF bound lipid fraction \n Strain TLO showed some additional features that are noteworthy. The use of thiosulfate as terminal electron acceptor is a common property of many thermophilic fermentative bacteria. Both hydrogen sulfide and elemental sulfur have been reported as reduced end product. The formation of elemental sulfur deposits was observed for Thermoanaerobacter italicus (Kozianowski et al. 1997 ) and T. uzonensis (Wagner et al. 2008 ), whereas Thermoanaerobacter sulfurigignens (Lee et al. 2007 ) only produces elemental sulfur, similar to most species of Thermoanaerobacterium (Schink and Zeikus 1983 ; Lee et al. 1993 ). The ability to produce either sulfide or sulfur from thiosulfate has been proposed as a differentiating feature between the genera Thermoanaerobacter and Thermoanaerobacterium (Lee et al. 1993 ). However, for strain TLO we could show that the formation of hydrogen sulfide or sulfur is very much dependent on the type of medium and the pH. At pH 6.7 (BM-medium), sulfide was exclusively formed, whereas at pH 5.8 (PB-medium) substantial sulfur deposition was observed. The reason for this drastic change in thiosulfate reduction is not known, but it indicates that the ability to produce either hydrogen sulfide or sulfur is not as group-specific as previously thought. The pH of the growth medium severely influenced the fermentation pattern. Whereas at basic and neutral conditions significant amounts of acetate, ethanol, and alanine were formed in addition to the major product lactate, at pH values between 4.5 and 5, an almost homolactic fermentation was observed. This latter feature combined with the ability to convert a wide array of substrates, and the growth on non-complex media, make this organism an interesting candidate for industrial lactic acid production. In the absence of thiosulfate, strain TLO also produced substantial amounts of alanine (0.3 mol/mol glucose). Alanine has been reported before as fermentation end product in sugar-fermenting thermophiles, and it is regarded as a sink for reducing equivalents (Kengen and Stams 1994 ; Balk et al. 2002 ). Accordingly, in the presence of the electron acceptor thiosulfate, alanine formation was almost negligible (0.02 mol/mol glucose). In the last two decades, intensive research on anaerobic, thermophilic, carbohydrate-fermenting microorganisms from marine and terrestrial volcanic hot springs has led to the isolation of several new genera and species in the domains Bacteria and Archaea. The major aim for this research stems from the biotechnological potential and the basic evolutionary traits of these microbes. CO-utilizing thermophilic microorganisms are able to grow by converting CO with water to H 2 and CO 2 . This feature makes these microorganisms interesting for cost effective hydrogen production. Hydrogen gas attracts great interest as a potential clean future fuel. Besides its potential as a future energy carrier, H 2 is a potent electron donor in various reductive processes, both in chemical and biotechnological applications. The use of thermophilic microorganisms for these processes could offer some advantages; although to date, few thermophiles are known that grow well on CO. The identification of new isolates that would broaden the product range of synthesis gas fermentations is desirable. Strain TLO can be one of the possible candidates for further research in this area. In conclusion, a new thermophilic anaerobic bacterium is described that differs from its closest phylogenetic relatives, T. thermohydrosulfuricus and T. siderophilus, in the ability to use carbon monoxide, its G + C content, and fatty acids composition. The lipid profile of strain TLO is different from T. thermohydrosulfuricus and T. siderophilus and characterized by the predominance of membrane-spanning lipids. Moreover, in the absence of thiosulfate strain TLO produced almost entirely lactate instead of ethanol or acetate. These differences led us to propose that strain TLO represents a subspecies of Thermoanaerobacter thermohydrosulfuricus, and we therefore name it Thermoanaerobacter thermohydrosulfuricus subsp. carboxydovorans. \n The description of T. thermohydrosulfuricus is as given by Klaushofer and Parkkinen ( 1965 ) with the following modifications. Arsenate is used as electron acceptor. The most abundant fatty acids are iso-C 15:0 and iso-C 17:0 and the membrane-spanning lipids, α,ω -13,16-dimethyl-octacosanedioic acid, was also found. One of the subspecies of T . thermohydrosulfuricus , strain TLO, is able to grow on CO (<25% v/v)."
} | 3,196 |
28989864 | PMC5628509 | pmc | 3,836 | {
"abstract": "The low cost of natural gas has driven significant interest in using C 1 carbon sources (e.g. methane, methanol, CO, syngas) as feedstocks for producing liquid transportation fuels and commodity chemicals. Given the large contribution of sugar and lignocellulosic feedstocks to biorefinery operating costs, natural gas and other C 1 sources may provide an economic advantage. To assess the relative costs of these feedstocks, we performed flux balance analysis on genome-scale metabolic models to calculate the maximum theoretical yields of chemical products from methane, methanol, acetate, and glucose. Yield calculations were performed for every metabolite (as a proxy for desired products) in the genome-scale metabolic models of three organisms: Escherichia coli (bacterium), Saccharomyces cerevisiae (yeast), and Synechococcus sp. PCC 7002 (cyanobacterium). The calculated theoretical yields and current feedstock prices provided inputs to create comparative feedstock cost surfaces. Our analysis shows that, at current market prices, methane feedstock costs are consistently lower than glucose when used as a carbon and energy source for microbial chemical production. Conversely, methanol is costlier than glucose under almost all price scenarios. Acetate feedstock costs could be less than glucose given efficient acetate production from low-cost syngas using nascent biological gas to liquids (BIO-GTL) technologies. Our analysis suggests that research should focus on overcoming the technical challenges of methane assimilation and/or yield of acetate via BIO-GTL to take advantage of low-cost natural gas rather than using methanol as a feedstock.",
"introduction": "1 Introduction Abundant, low cost C 1 compounds such as methane, methanol, and carbon monoxide have garnered attention as potentially inexpensive sources of carbon and energy in biocatalytic processes for producing commodity chemicals ( Conrado and Gonzalez, 2014 , Haynes and Gonzalez, 2014 , Whitaker et al., 2015 ). Over the last decade, natural gas supply has reached all-time highs with costs consistently lower than petroleum. Despite these economic advantages, large volumes of natural gas are flared at wellheads daily to reduce greenhouse gas emissions or directly leaked, both intentionally and unintentionally, to the environment during production ( Howarth et al., 2011 , Salmon and Logan, 2013 ). Nighttime satellite images of these areas show light intensities equivalent to major US cities and illustrate the enormous potential that is wasted. Alternative uses, such as pipelining to refineries, catalytic conversion to syngas or methanol, or combustion for electricity and heat have not been deployed due to costs, wide geographic distribution, and/or poor proximity to end-users. Given that feedstocks are the major operating cost of producing biomanufactured chemicals ( Klein-Marcuschamer et al., 2011 ), the choice of carbon source can have a significant impact on profitability. Given the potential process advantages, the economic potential of C 1 feedstocks for biomanufacturing of commodity chemicals warrants evaluation. Cost is not the only criteria when considering methane, as gas-phase feedstocks suffer from a few bioprocess drawbacks. First, uptake of gas-phase feedstocks can be mass-transfer limited and significantly slower than uptake of traditional aqueous feedstocks, such as sugars and organic acids, depending on bioreactor conditions ( Conrado and Gonzalez, 2014 ). Second, methane utilization by methylotrophs requires an electron acceptor – most frequently oxygen – which raises safety concerns over potentially explosive mixtures of feedstock gases ( Whitaker et al., 2015 ). Third, the conversion of methane to methanol, catalyzed by methane monooxgynease, is a slow step that limits overall productivity. For these reasons, we were curious if alternative derivatives of C 1 compounds, methanol and acetate, would have economic advantages over glucose. Methanol, produced by steam reformation to syngas and catalytic conversion to methanol, is the first intermediate in methane assimilation. Acetate can be produced from natural gas by combining catalytic water-gas shift with anaerobic fermentation of the resulting syngas by acetogens. These so-called biological gas-to-liquid (Bio-GTL) processes have been recently demonstrated for producing lipids and biodiesel ( Hu et al., 2016 ). This low-cost process makes acetate an interesting potential feedstock to replace glucose in biocatalytic processes. Feeding aqueous methanol or acetate would circumvent many of the technical hurdles in a bioreactor while potentially leveraging the low-cost and abundant supply of C 1 feedstocks. The last factor in selecting a feedstock is the relative amount needed to generate a given amount of product. Sugars are the dominant feedstock for bioconversions because metabolism efficiently extracts energy and electrons for use in synthesizing chemical products. Assimilation of C1 compounds is not as energy efficient given higher energy and reducing power costs. There are several pathways for assimilating methane and methanol; each differs in energetic yield, connections to central metabolism, and kinetics. Methane and methanol assimilation both occur through the assimilation of formaldehyde. The first step in methane catabolism is oxidation to methanol by a methane monooxygenase ( Hwang et al., 2014 ). Methanol is further oxidized to formaldehyde by an alcohol oxidase or methanol dehydrogenase ( Müller et al., 2015 , Whitaker et al., 2015 ). Formaldehyde is then assimilated through one of three pathways: the ribulose monophosphate (RuMP) pathway, the dihydroxyacetone (DHA) pathway, or the serine pathway ( Fig. 1 ) ( Yurimoto et al., 2005 ). The DHA — found only in fungi — and RuMP — common in gamma-proteobacteria and only found in bacteria and archea — pathways both use a five carbon sugar as a substrate to assimilate formaldehyde and produce a six carbon sugar ( Hwang et al., 2014 , Müller et al., 2015 , Whitaker et al., 2015 , Yurimoto et al., 2005 ). Every three turnovers of these cycles produces a single dihydroxyacetone phosphate. In contrast, the serine pathway — common in alpha-proteobacteria — assimilates formaldehyde through reaction with glycine to create serine ( Hwang et al., 2014 , Yurimoto et al., 2005 ). Every two turnovers of this cycle assimilates two formaldehydes and one carbon dioxide to produce a 2-phosphoglycerate (the base cycle can be augmented with other reactions to produce acetyl-CoA and TCA cycle intermediates with additional cycles and assimilation of CO 2 ). To obtain energy, formaldehyde is oxidized to carbon dioxide to generate reducing equivalents that can be converted to ATP via the electron transport chain and ATP synthase ( Yurimoto et al., 2005 ). Overall, the RuMP pathway is considered the most efficient pathway in terms of energetic yield and is the preferred C 1 assimilation pathway in studies of C 1 catabolism as a feedstock ( Müller et al., 2015 ). Beyond the formaldehyde assimilating pathways, there is a small class of C 1 catabolizing, non-photosynthetic bacteria that use the Calvin-Benson-Bassham (CBB) cycle to assimilate carbon dioxide by oxidation of C 1 carbon sources to produce the energy needed to run the CBB ( Hwang et al., 2014 ). This does not appear to be an efficient pathway and is not biologically common, so we did not include it in our analysis. In addition, Bogorad et al., 2014 , Bogorad et al., 2013 created a synthetic methanol condensation cycle (MCC) for the efficient assimilation of methanol by using a combination of the RuMP cycle and synthetic non-oxidative glycolysis in Escherichia coli . We tested this pathway for improved yield as part of a set of analysis looking at the impact of including carbon efficient assimilation pathways and show little difference on predicted product yields in E. coli . Fig. 1 Formaldehyde assimilation pathways. The ribulose mono-phosphate (RuMP) pathway is a bacterial formaldehyde assimilation pathway that uses ribulose-5-phosphate as a substrate for formaldehyde assimilation. The dihydroxyacetone (DHA) pathway is a fungal formaldehyde assimilation pathway that uses xylulose-5-phosphate as a substrate for formaldehyde assimilation. The serine pathway is a bacterial pathway that uses glycine to assimilate formaldehyde. The serine pathway also assimilates one carbon dioxide for every two formaldehydes assimilated. All three pathways produce glycolytic intermediates. The abbreviations are defined as follows: CH4 – methane, MeOH – methanol, CHO – formaldehyde, CO2 – carbon dioxide, H6P – hexulose 6-phosphate, F6P – fructose 6-phosphate, FBP – fructose 1,6 bisphosphate, DHAP – dihydroxyacetone phosphate, G3P – glyceraldehyde 3-phosphate, E4P – eyrthrose 4-phosphate, S7P – septulose 7-phosphate, R5P – ribose 5-phosphate, Xu5P – xylulose 5-phosphate, Ru5P – ribulose 5-phosphate, DHA – dihydroxyacetone, Ser – serine, HPyr – hydroxypyruvate, Glyc – glycerate, 2PG – 2-phosphoglycerate, 3PG –3-phosphoglycerate, PEP – phosphoenolpyruvate, HCO3 – bicarbonate, OAc – oxaloacetate, MAL – malate, MALC – malyl-CoA, AcC – acetyl-CoA, GLX – glyoxylate, GLY – glycine. Fig. 1 Acetate is typically assimilated into central metabolism as acetyl-CoA. Acetate is a common by-product during rapid growth in bacteria and re-assimilates during stationary phase at the cost of one ATP. For this reason, most organisms, including those studied here, have pathways for assimilating acetate. In E. coli , and many other microorganisms, acetate assimilation proceeds through acetyl-phosphate (catalyzed by ackA/pta). In other organisms, such as the cyanobacterium Synechococcus sp. strain PCC 7002 studied here, acetate assimilation proceeds through a transient acyl-AMP intermediate as part of an acyl-CoA ligase mechanism ( Begemann et al., 2013 ). Other pathways for acetate assimilation involve CoA transfer from other metabolites (e.g. propionyl-CoA, succinyl-CoA). To assess the tradeoffs between feedstock cost and product yield, we have performed an economic analysis using maximum theoretical yields calculated by flux balance analysis of genome-scale metabolic models. We performed the analysis on models of three organisms, the bacterium E. coli , the yeast Saccharomyces cerevisiae , and the cyanobacterium Synechococcus sp. strain PCC 7002, augmented with all three common methane assimilation pathways (serine, RUMP, DHA). For each model, we calculated the theoretical yield for every metabolite in each model from each feedstock – glucose, methane, methanol, and acetate as well as xylose and glycerol for E. coli only. We use the calculated theoretical yields and current feedstock prices to create surfaces of feedstock costs. Our analysis shows that despite the lower stoichiometric yield from methane, methane feedstock costs are consistently lower than glucose when used as carbon and energy sources to make products. Conversely, methanol is equal to or costlier than glucose. Acetate is much more difficult to give an accurate estimate of comparable feedstock cost with glucose given current technological change and price variation. Our analysis suggests that methane is the most promising C 1 feedstock, assuming it is possible to overcome any technical hurdles associated with its use as a feedstock.",
"discussion": "3.4 Discussion Our analysis, based on feedstock price and theoretical conversion yields, suggests that methane is the most attractive feedstock for microbial cell factories. This assumes that technical barriers to using methane can be overcome. For instance, biological assimilation kinetics of methane tend to be relatively slow and are a rate-limiting step in methane conversion to value-added chemicals. The main cause of the slow kinetics is the long turnover time of the methane monooxygenase active site ( Hwang et al., 2014 ). The reactivation of the active site is a slow process that results in an overall low methane assimilation rate. With protein engineering, it could be possible to overcome this barrier. There are also difficulties in scaling-up microbial production using a methane feedstock. Methane assimilation pathways require both methane and oxygen supply for effective function. This combination of gases can be explosive, which introduces a safety concern when using industrial-size reactors. Another difficulty is the slow mass transfer of methane into aqueous solutions ( Conrado and Gonzalez, 2014 ). If these barriers are overcome, methane has the potential to be a more profitable feedstock than glucose. A caveat to this analysis is that it only considers feedstock costs and does not consider differences in other costs associated with these feedstocks, such as operating costs, capital costs, conversion inefficiencies (i.e. operating at a fraction of theoretical yield), separations, wastes, etc. A true measure of profitability will require a more detailed technoeconomic analysis of any potential process. For instance, if a technological advance in methane-to-methanol catalytic conversion occurs, the price of methanol could drop considerably and make methanol more competitive with glucose. Methanol provides other technical advantages over methane such as faster mass transfer, faster uptake rate, simplified gassing, and reduced safety concerns that our feedstock analysis does not consider. However, considering current prices, it is clear that methane is the preferred feedstock. If we are unable to engineer microbes for efficient methane assimilation for chemical production (i.e. achieve yields approaching theoretical limits), the more attractive strategy would be to convert methane to acetate (via syngas) instead of converting methane to methanol. The price of acetate is a large uncertainty in our analysis, but new BIO-GTL technologies that leverage less-expensive, nitrogen-containing syngas could make the lower end of our estimate range a reality. Assuming a low-cost bio-GTL process works at industrial scale, then acetate would compete with glucose as a feedstock. Our analysis shows a consistent trend when comparing theoretical yields on different feedstocks, suggesting that a general relationship is derivable. This trend is interesting because it suggests that product yield varies with carbon source identity independent of the specific product (i.e. products have similar relative yields on different carbon sources). We postulate that this linear relationship is obtained because of the structure of metabolism in which catabolic and anabolic pathways are linked by common central metabolites. In other words, all of the substrates we simulated connect to central metabolism through the catabolic pathways described in the introduction, in the process generating NAD(P)H and ATP to varying extents depending on the substrate. Conversely, all products (except those metabolites found in the catabolic pathways) are produced from the same anabolic pathways independent of carbon source. Therefore the difference in a product's yield on different carbon sources can be traced to the difference in NAD(P)H/ATP produced via the catabolic pathways used to convert carbon sources into precursors for that product. Thermodynamics could provide a possible explanation for the linear relationship between product yields on different feedstocks. A thermodynamic estimate of the maximum theoretical product yield from a specific substrate can be calculated using the carbon number and degree of reduction of both the substrate and product ( Doran, 2012 ). This is calculated using Eq. (3) , where f max is the number of moles of product that can be made from 1 mol of substrate, w and j are the number of carbon atoms in the substrate and product and γ s and γ p are the degree of reduction of substrate and product, respectively ( Doran, 2012 ). (3) f max = w γ s j γ p For a given product, j and γ p will remain constant when comparing different substrates, while w and γ s will be unique to the substrate used. This allows for a comparison between maximum theoretical yields of different substrates on a thermodynamic basis. The numerator of Eq. (3) is 8 for methane, 6 for methanol, 8 for acetate, and 24 for glucose. This implies yields from glucose will be 3 times higher than from methane or acetate and 4 times higher than yields from methanol. When normalized for carbon content, methane generates a theoretical yield twice that of glucose. In our analysis, when yields were normalized to the carbon number in the feedstock, the largest difference between feedstocks was significantly less than 2 ( Fig. 2 ). Therefore, the degree of reduction is not completely responsible for remaining difference in yields on various feedstocks. One reason for the discrepancy is the requirement of additional electrons and energy in the specific biological pathways (e.g. MMO) used to conduct the transformations. For this reason, analysis using genome-scale metabolic models is an improvement over analysis using just stoichiometry."
} | 4,270 |
34940014 | PMC8698420 | pmc | 3,838 | {
"abstract": "Desertification is a growing risk for humanity. Studies show that water access will be the leading cause of massive migration in the future. For this reason, significant research efforts are devoted to identifying new sources of water. Among this work, one of the more interesting strategies takes advantage of atmospheric non-liquid water using water harvesting. Various strategies exist to harvest water, but many suffer from low yield. In this work, we take inspiration from a Mexican plant ( Echeveria pulvinate ) to prepare a material suitable for future water harvesting applications. Observation of E. pulvinate reveals that parahydrophobic properties are favorable for water harvesting. To mimic these properties, we leveraged a combination of 3D printing and post-functionalization to control surface wettability and obtain parahydrophobic properties. The prepared surfaces were investigated using IR and SEM. The surface roughness and wettability were also investigated to completely describe the elaborated surfaces and strongly hydrophobic surfaces with parahydrophobic properties are reported. This new approach offers a powerful platform to develop parahydrophobic features with desired three-dimensional shape.",
"conclusion": "4. Conclusions In this work, we report for the first time the combination of 3D printing and post-functionalization for parahydrophobic surface elaboration. Surface functionalization of 3D-printed substrates was explored following a two-step process and with a broad range of carboxylic acids. The functionalized surfaces were characterized using different techniques. While infrared spectroscopy did not allow us to confirm the reaction of the carboxylic acid with the oxidized surface, the modification between the Cu surface and the Cu(OH) 2 surface was confirmed. Roughness and size measurements were conducted to confirm that the functionalization of the surface preserved the printed shape. Finally, the surface wettability was investigated. Our study reveals the significant impact of functionalization on surface properties. In particular, surfaces functionalized with linear carboxylic acids revealed a strong hydrophobic character with apparent contact angles near 140°. All the prepared surfaces exhibited parahydrophobic features (e.g., hydrophobicity and strong water adhesion), thus confirming the potential of this approach for water harvesting technology development. This strategy has also been applied for more complex printed structures to improve water harvesting potential. Future and ongoing work will evaluate the harvesting capacity of the elaborated surfaces.",
"introduction": "1. Introduction Water is a sine qua none condition for human development. Water is a key parameter critical to life, agriculture, and even hygiene. A lack of liquid water is one of the most significant problems facing a large portion of the global population. Due to climatic evolution, it has been estimated that in the near future (2025), a significant fraction of the global population will be in dramatic distress with regard to local water supply [ 1 ]. Such a scenario can lead to a catastrophic humanitarian situation, likely resulting in massive population migrations and related conflicts. To avoid such an undesirable outcome, it is necessary to find a sustainable solution to supply water in these at-risk regions. For this reason, a significant body of research aimed at solving this problem and identifying solutions has been pursued. As one example, desalinization techniques that remove salt from seawater to produce drinkable water have been described [ 2 , 3 , 4 , 5 , 6 ]. This type of strategy is very efficient and can also produce a large quantity of water. This strategy is employed in Saudi Arabia, mainly for irrigation [ 7 , 8 ]. Despite the success of desalinization, this strategy presents a major drawback with regards to sustainability. Mainly, if salt is removed to yield water, what are we doing with the leftover salt? Desalinization produces significant quantities of salt, generating hyper-saline regions and salt deserts that sterilize the surrounding area [ 9 , 10 ]. Another solution used to supply water in desert regions is through water harvesting [ 11 , 12 , 13 ]. In many remote locations, if liquid water is lacking, atmospheric water can be abundant and harvested to yield liquid water. Fog harvesting is employed in Chile for irrigation, as one example, and it is clear that this strategy is compatible with sustainable development [ 14 , 15 ]. Water harvesting is also generally cheap, since fog is often harvested on a network of simple nets. The major drawback to fog harvesting is the poor associated efficiency; only a small portion of the available water is collected. A sustainable solution to increase harvesting efficiency while maintaining low cost is a topic of interest for many researchers. One approach to identify a solution is to observe living species effective at water harvesting to understand how this strategy has been optimized in nature. A bioinspired water harvesting device can be an interesting solution to provide liquid water naturally and at a low cost, without significant energy consumption. It is well reported that in arid regions, many species (animal and vegetal) developed techniques to harvest atmospheric water. As one example, the Namib desert beetle ( Stenocara gracilipes ) traps water on their legs and body to supply water [ 16 , 17 , 18 ]. Similarly, cacti can collect water with their pines and thus hydrate their growth [ 19 , 20 ]. In previous work, we also reported the possibility for Mexican plants such as Echeveria pulvinata to harvest water from the air [ 21 ]. The capability for E. pulvinata or cacti to harvest water is due to different parameters. First, plants effective at harvesting water increase their contact surface via macrostructures, such as pines or hairs for cacti and E. pulvinata , respectively. Secondly, these macrostructures have unique wettability that facilitates the collection of water from the air. More specifically, the surface of pines or hairs has a high affinity with water, thus trapping water droplets so they remain stuck to the surface. However, the surface of these macrostructures is also patterned with domains that are highly hydrophobic, this ultimately aids in water droplet collection as efficient transport of droplets can occur in these hydrophobic domains. With these observations, it is obvious that to create similar functionalities and features it is critical to control both surface macrostructure and wettability. In the present work, we leverage 3D printing and post-functionalization to prepare surfaces inspired by E. pulvinata . The desired surface macrostructure is elaborated using a 3D-printing strategy. This choice is motivated by the fact that in the past decade 3D printing has been reported and explored as a powerful tool since it allows one to directly control and adapt the macrostructure of a surface to meet the needs of an application [ 22 , 23 , 24 , 25 , 26 ]. Furthermore, 3D printing allows us to easily prepare a variety of surfaces, with or without macroscopic features so the influence of these structures can be determined explicitly. Surface wettability, which is also a critical variable, is ultimately more difficult to control. The wettability of a surface depends on two key variables: surface energy and morphology, and many strategies manipulating these two variables to tailor surface wettability are presented in the research literature [ 27 , 28 , 29 , 30 , 31 , 32 , 33 ]. In the present work, we employ a chemical strategy to manipulate wettability, specifically a copper post-functionalization strategy with carboxylic acid. Prior works have demonstrated that this approach is effective for the preparation of highly hydrophobic surfaces on copper foils and plates [ 34 , 35 , 36 , 37 ]. In this study, we will use this approach to functionalize 3D-printed surfaces comprised of materials loaded with copper. With this combination of 3D printing and post-functionalization, we will be able to manipulate all critical parameters of the surfaces, mainly macrostructure and wettability ( Figure 1 ). Here, the successful surface modification is confirmed using infrared spectroscopy (IR) and post-functionalized surfaces were characterized for size and roughness to determine the degree of microstructure modification due to the functionalization. Additionally, the surface wettability was also investigated via contact angle measurements."
} | 2,134 |
31506311 | PMC6737243 | pmc | 3,840 | {
"abstract": "A new concept in biofilm science is introduced: “adhesion force sensitivity of genes,” defining the degree up to which expression of different genes in adhering bacteria is controlled by the environmental adhesion forces they experience. Analysis of gene expression as a function of height in a biofilm showed that the information about the substratum surface to which initially adhering bacteria adhere is passed up to a biofilm height of 20 to 30 μm above a substratum surface, highlighting the importance and limitations of cell-to-cell communication in a biofilm. Bacteria in a biofilm mode of growth, as opposed to planktonic growth, are responsible for the great majority of human infections, predicted to become the number one cause of death in 2050. The concept of adhesion force sensitivity of genes provides better understanding of bacterial adaptation in biofilms, direly needed for the design of improved therapeutic measures that evade the recalcitrance of biofilm bacteria to antimicrobials.",
"introduction": "INTRODUCTION Biofilms are surface-adhering and surface-adapted communities of microorganisms ( 1 ), in which adhesion to a substratum surface is the initial step. Two surfaces, including the surface of bacteria adhering on a substratum surface, can be attracted to each other by a combination of Lifshitz-van der Waals, electrostatic double-layer, and acid-base forces ( 2 ). The sum total of these forces is generally called the “adhesion force.” The environmental adhesion forces by which a bacterium adheres to a surface are orders of magnitude larger than the gravitational forces bacteria experience and give rise to nanoscopic deformation of the cell wall ( 3 , 4 ). Cell wall deformation in its turn causes changes in lipid membrane surface tension that provides a stimulus for the environmentally triggered expression of a great number of genes in adhering bacteria ( 5 ) to facilitate their surface adaptation. This leads to new, so-called “emergent” properties of adhering bacteria in their biofilm mode growth ( 6 ). Emergent properties reflect bacterial surface adaptation and arise only after bacteria have adhered to a surface. According to their definition ( 6 ), emergent properties of bacteria in biofilm mode growth are alien to their planktonic counterparts and cannot even be predicted on the basis of the properties of planktonic bacteria. The most prominent, landmark emergent property of adhering bacteria is the production of an extracellular polymeric matrix in which biofilm bacteria protect themselves against host defenses ( 7 ) and antimicrobial agents ( 8 , 9 ) and through which they enforce their bond with a substratum surface ( 10 ). Adhesion-force-induced surface adaptation in adhering bacteria has been observed in Staphylococcus aureus biofilms for the icaA gene, regulating production of extracellular polymeric substances (EPS). However, adhesion-force-induced surface adaptation was not observed for the cidA gene, which is associated with cell lysis and extracellular DNA (eDNA) release ( 11 ). Also, nisin clearance in staphylococci through the two-component NsaRS intramembrane-located sensor NsaS and NsaAB efflux pump ( 12 ) was enhanced when staphylococci adhered more strongly to a substratum surface ( 13 ). Hitherto, adhesion force sensing and associated cell wall deformation have appeared as an appealing concept to explain what environmental stimulus externally triggers the development of emergent properties of bacteria in biofilm mode growth. Yet, there still are many questions to be addressed, most urgently concerning the range over which adhesion force sensing operates in a biofilm. Typically, biofilms are much thicker than the range of the adhesion forces extending from a substratum surface. Adhesion forces can yield an attraction that can be sensed up to maximally 0.5 μm into a biofilm ( 2 , 3 ). The exact magnitude and range of an adhesion force depend on the hydrophobicity and charge properties of the bacterial cell and substratum surfaces. Compared with the thickness of a biofilm, the range over which adhesion forces operate is relatively short. This suggests that quorum sensing plays a role in spreading the “news” that initial colonizers in a biofilm have “landed” on a substratum surface exerting a specific adhesion force. However, this suggestion has never been confirmed. Furthermore, adhesion force sensing has never been confirmed in other species than staphylococci. Adhesion to surfaces is a survival mechanism for streptococci in the oral cavity ( 14 ). Accordingly, Streptococcus mutans has the ability to adhere to oral hard and soft tissues, abiotic restorative dental materials, and other bacteria in the oral cavity ( 15 ). Frequently studied genes involved in S. mutans initial adhesion and biofilm formation are summarized in Table 1 . Based on the definition of “emergent” properties as given by Flemming et al. ( 6 ) and literature description of gene functions, a hypothetical distinction is made between genes whose expression prepares planktonic bacteria for adhesion to a substratum surface and genes relevant for the development of emergent properties in adhering bacteria. For instance, genes that regulate synthesis of specific ligands of planktonic streptococci for optimal initial adhesion to saliva-coated surfaces, such as ftf and gtfB ( 16 – 19 ), are not considered to be involved in the development of emergent properties that arise by definition in already adhering bacteria. Also, genes regulating bacteriocin production, cell death, and chemical stress responses ( comDE , virR , gbpB , and relA ), although vital in biofilm formation, may not bear direct relevance to EPS production, enforcing strong adhesion of biofilm inhabitants to a substratum surface ( 20 – 22 ). Autoinducer 2 in the S. mutans \n luxS quorum-sensing system (see also Table 1 ) coordinates communication in S. mutans biofilms ( 23 ) and may be expected to impact the extension of adhesion-force-sensitive genetic programming into a mature biofilm, as adhesion forces can only be directly sensed by initial colonizers ( 4 ). TABLE 1 Summary of genes involved in S. mutans UA159 initial adhesion and subsequent processes occurring during biofilm formation Gene a \n Function Reference(s) Genes relevant to prepare initial adhesion in planktonic S. mutans ftf Catalysis of sucrose cleavage to synthesize fructan to promote initial adhesion to salivary films 16 , 17 gtfB Synthesis of water-insoluble glucans (α-1,3-linked) to promote initial adhesion to saliva-coated tooth surfaces and establishment of microcolonies in biofilm 18 , 19 Genes relevant to develop emergent properties in adhering S. mutans brpA Regulation of cell wall stress responses, biofilm cohesiveness, and biofilm formation 24 , 33 , 34 comDE Persister cell formation, bacteriocin production 30 vicR Synthesis of EPS matrix components, regulation of bacteriocin production and cell death 44 , 45 gbpB Regulation of sensitivity to antibiotics, osmotic and oxidative stresses, cell wall construction and maintenance, cell shape, hydrophobicity, and sucrose-dependent biofilm formation 28 , 29 relA Regulation of stringent response, acid tolerance, and biofilm formation 46 , 47 luxS Coordination of collective behaviors and cohesiveness in biofilms 48 , 49 a A hypothetical distinction has been made with respect to genes relevant to prepare initial adhesion in planktonic streptococci and genes involved in the development of emergent properties in adhering bacteria. In order to further advance the concept of adhesion-force-induced gene expression in relation to emergent biofilm properties, the aim of this article is first to identify genes involved in biofilm formation by S. mutans and an isogenic, quorum-sensing-deficient mutant whose expression is controlled by environmental adhesion forces. This would confirm the hypothetical distinctions made in Table 1 between genes preparing planktonic bacteria for adhesion to a substratum surface and genes relevant for the development of emergent properties in adhering bacteria. To this end, biofilms of S. mutans UA159 and its ΔluxS isogenic mutant were grown on four substratum surfaces with different hydrophobicities, and single-bacterial contact probe atomic force microscopy (AFM) was applied to measure the forces by which both strains adhere to each substratum surface. Gene expression was evaluated using RT-qPCR. Up- or downregulation of selected genes upon adhesion was related to the forces by which the streptococci adhere to yield a new concept of “adhesion force sensitivity of gene expression.” Uniquely, the extension of adhesion-force-induced genetic programming over the height of the biofilms above a substratum surface was investigated in cryosections of the biofilms taken at different heights above a substratum surface. Herewith it can be determined to what extent quorum sensing controls adhesion-force-induced gene expression in later biofilm inhabitants, residing further away from the substratum surface and not in direct contact with the substratum surface. Whiteness analyses of optical coherence tomography (OCT) images of biofilms was employed to support the conclusions regarding height-dependent gene expression taken from cryosections of the S. mutans biofilms.",
"discussion": "DISCUSSION S. mutans is an avid sugar consumer in the oral cavity, allowing it to produce acids that make it one of the world’s most widespread pathogens, responsible for the decalcification of oral hard tissues. For its survival in the oral cavity, S. mutans needs to adhere ( 14 ). Once adhering, S. mutans enforces its adhesion to oral surfaces through the production of EPS ( 27 ) as a landmark, emergent biofilm property. In this article, we identified gbpB , brpA , and comDE as genes that became more strongly expressed upon adhesion of S. mutans UA159, compared with ftf , gtfB , vicR , and relA . This confirms our hypothetical distinction ( Table 1 ) of ftf and gtfB genes being more relevant for the preparation of planktonic streptococci for their initial adhesion to surfaces. Also, it justifies the classification of the gbpB , brpA , and comDE genes as more relevant for the development of emergent properties in adhering streptococci. The vicR and relA genes play roles with respect to diverse processes occurring during biofilm formation ( Table 1 ), but these are not exclusively involved in directly enforcing the initial adhesion of S. mutans to oral surfaces. Based on the differential expression of the gbpB , brpA , and comDE genes in streptococci adhering on different substratum surfaces and relating it to the adhesion forces experienced by adhering bacteria, a new concept of “adhesion force sensitivity of gene expression” is introduced. Adhesion force sensitivity reflects whether expression of a gene is more or less strongly influenced by the adhesion force sensed by bacteria upon their adhesion to a substratum surface. Among the three genes identified, gbpB had the weakest adhesion force sensitivity. However, gbpB is not only involved in enforcing initial streptococcal adhesion but also possesses an array of other pivotal functions in biofilm formation ( Table 1 ) ( 28 , 29 ). comDE is also weakly adhesion force sensitive and also possesses other functions than enforcing initial adhesion, including persister cell formation ( 30 ). However, persister cell formation usually involves bacteria closely associated with a substratum surface ( 31 ), and hence the weak control of adhesion forces over comDE expression as determined over the entire height of a biofilm is not surprising. Moreover, these weakly adhesion-force-sensitive genes as identified in this study have also been found to be upregulated in biofilm detached cells ( 32 ). Detachment is an important mechanism for bacterial survival, since it protects the biofilm from overpopulation, which is opposite from enforcing initial adhesion. Expression of brpA was by far several fold more sensitive to adhesion forces than gbpB and comDE , and its role in biofilm formation has been forcefully emphasized in the literature ( 24 , 33 , 34 ). When averaged over the entire height of relatively thin, 5-h biofilms of S. mutans UA159, biofilms demonstrated adhesion-force-controlled gene expression, but this was not observed in thicker, 24 h biofilms ( Table 2 ). In order to study the biofilm height above a substratum surface over which initially adhering streptococci in direct contact with a substratum surface can signal the news of being in an adhering state on a specific surface, 24-h biofilms on silicone rubber were sliced ( Fig. 3A) . Biofilm slices taken at different heights were examined for expression of the three adhesion-force-sensitive genes identified. In 24-h biofilms, slices taken closest to the substratum surface demonstrated higher expression of the three adhesion-force-sensitive genes than slices of biofilm taken more distant from the surface ( Fig. 3B) . Thus, adhesion-force-induced gene expression extended over at least half of the biofilm height above a surface, which represents a considerably larger distance than that over which adhesion forces arising from the substratum surface can range ( 2 , 3 ). In addition to this, most bacteria in a biofilm have never visited a substratum surface ( 35 ). This implies that quorum sensing must be responsible for the extension of adhesion-force-induced gene expression in biofilms. This conclusion is supported by the observation that adhesion-force-induced gene expression of quorum-sensing-deficient S. mutans UA159 Δ luxS was fully absent in both 5- and 24-h-old biofilms ( Table 2 ). Moreover, in quorum-sensing-deficient S. mutans UA159 Δ luxS , EPS production reflected by local back-scattered light intensities ( Fig. 4D ) showed identical distributions of soluble EPS over the height of biofilms on silicone rubber and glass ( Fig. 4D) . Alternatively, in biofilms of S. mutans UA159 with the ability of quorum sensing, soluble EPS production on hydrophobic silicone rubber was higher than on hydrophilic glass up to a distance of around 20 to 25 μm above the substratum surface. Thus, it can be concluded based on height-dependent gene expression and local EPS production that adhesion-force-induced expression of genes extends into a biofilm through quorum sensing over a height limited to 20 to 30 μm above the substratum surface, beyond which autoinducer concentrations become below their threshold concentrations required to invoke a response. “Calling” distances over which bacteria can communicate through quorum sensing have been reported between 5 μm ( 36 ) and 200 μm ( 37 ), which indicates that our estimate of 20 to 30 μm as the calling distance in streptococcal biofilms is reasonable. In summary, this work extends our understanding of emergent properties in streptococcal biofilms and the role of quorum sensing herein. Environmental adhesion forces have been identified to externally control expression of genes that are directly involved in the development of emergent biofilm properties in adhering S. mutans , leading to a new concept of “adhesion-force-induced gene expression in adhering bacteria.” brpA was the most adhesion-force-sensitive gene, as well as the most strongly expressed gene in adhering streptococci. Extension of its expression decreased with height above the substratum surface. Adhesion-force-induced gene expression was fully absent in a quorum-sensing-deficient isogenic streptococcal mutant. The concept of adhesion-force-induced gene expression and its extension through a biofilm through quorum-sensing mechanisms advance our understanding of why biofilms of the same strain or species may possess different properties when grown on different substrata, which is relevant in all environmental, industrial, and biomedical applications where biofilms develop."
} | 4,012 |
36009754 | PMC9405236 | pmc | 3,843 | {
"abstract": "Simple Summary The brain is an incredibly complex system possessing outstanding abilities to perform difficult tasks through a vast number of densely interconnected neurons. Aimed at discovering the underlying mechanisms of the brain, a number of spiking neural networks have been proposed to mimic biological neural dynamics. Subsequently, to perceive how the neural networks in the brain work, simulation and hardware realization of large-scale systems, similar to the brain, is an essential requirement. Behavior of a single neuron can be described by the mathematical equations in different levels of computing and biological accuracy. In this approach, a new modified ADEX model is presented based on sampling frequency by the nonlinear functions of the original model. This new model is capable for reproducing all aspects of the original model in low-error and high-degree of similarity conditions. Finally, the proposed model can be implemented on digital hardware platforms to have a real digital system. Digital results show the increase in system speed (frequency) and overall saving in hardware resources (compared by the original model and other similar works). This low-cost digital hardware is applied in large-scale neuronal networks. Abstract Design and implementation of biological neural networks is a vital research field in the neuromorphic engineering. This paper presents LUT-based modeling of the Adaptive Exponential integrate-and-fire (ADEX) model using Nyquist frequency method. In this approach, a continuous term is converted to a discrete term by sampling factor. This new modeling is called N-LUT-ADEX (Nyquist-Look Up Table-ADEX) and is based on accurate sampling of the original ADEX model. Since in this modeling, the high-accuracy matching is achieved, it can exactly reproduce the spiking patterns, which have the same behaviors of the original neuron model. To confirm the N-LUT-ADEX neuron, the proposed model is realized on Virtex-II Field-Programmable Gate Array (FPGA) board for validating the final hardware. Hardware implementation results show the high degree of similarity between the proposed and original models. Furthermore, low-cost and high-speed attributes of our proposed neuron model will be validated. Indeed, the proposed model is capable of reproducing the spiking patterns in terms of low overhead costs and higher frequencies in comparison with the original one. The properties of the proposed model cause can make it a suitable choice for neuromorphic network implementations with reduced-cost attributes.",
"conclusion": "7. Conclusions In this paper, an LUT-based modeling method of the ADEX model using Nyquist frequency method sampling is presented. This approach is a characteristic of a sampler, which converts a continuous function or signal into a discrete sequence. This new modeling is called N-LUT-ADEX (Nyquist-Look Up Table-ADEX) and is based on accurate sampling of the original ADEX model. The proposed approach exactly can follow the original ADEX model in a low error state and high degree of spiking pattern similarity. The time-domain and dynamical behaviors show that this new model is capable of reproducing all aspects of the original model. The main nonlinear terms of the original model are converted to LUT-based terms without any multipliers, dividers and exponential parts, which makes the proposed model an efficient approach. The implementation of this proposed model on an FPGA Virtex-II board shows that the new model is of low cost with high-speed attributes, which results in a real biological neural network. Two basic factors of the neuronal realization (speed-up and large-scale implementation) provide good results compared to the ADEX original neuronal model and other similar works. In this way, there is an overall saving of \n 97.61 % \n ; additionally, a speed-up to 212 MHz has been achieved. To test the population network, a system composed of 2000 randomly connected neurons is simulated. The proposed model in this simulation test is is highly similar to the original ADEX neuron. Our proposed model performs better than other works; Gomar et al. [ 5 , 7 ], Haghiri et al. [ 6 ] and Heidarpour et al. [ 8 ] obtained overall saving of \n 95 % \n , \n 94 % \n , \n 96.7 % \n and \n 70.76 % \n , respectively. Moreover, the frequencies in these works are 187, \n 187.5 \n , 196 and 134 MHz, respectively. Among all similar works, our proposed model obtains better results. Other similar papers present ADEX neuron implementation with different approaches. In these mentioned papers, in some cases, the accuracy of realization is reduced. In some implementation, this accuracy will be increased, but the overall saving in FPGA resources is decreased. In our proposed model (N-LUT-ADEX), since the original functions are sampled (without any approximation), a high degree of accuracy is achieved. On the other hand, the overall saving in FPGA will be increased when all nonlinear terms of the original model are removed. Moreover, because all multiplications are removed, the final frequency will be significantly increased.",
"introduction": "1. Introduction The Central Nervous System (CNS) is a basic biological system which includes three vital organs: neurons, synapses, and glias [ 1 , 2 ]. In this network, neurons are responsible for information processing and transformation of data in different parts of the human brain. By neurons switching, the essential informations are transferred to the brain parts [ 3 , 4 , 5 ]. Neurons are high-speed organs that have high number of different connections for transferring data [ 6 ]. On the other hand, synapses are the connections between two neurons. These connections causes the coupling of two connected neurons. These coupled neurons are basic submodules of the CNS that can be realized in real states [ 7 , 8 , 9 , 10 ]. Furthermore, glias are the protection-based cells for neurons and regulate the coupling behaviors between different neuron connections. One of the basic forms of the glias in the brain are astrocytes, which are responsible for protecting and regulating neurons behavior [ 3 , 4 , 9 ]. From this standpoint, investigating neurons dynamic can be a vital requirements in case of neuromorphic engineering. Thus, modeling of neural dynamics and spiking neural network mechanisms have been a strong tool in analyzing and processing behavior of biological neural networks. Interaction between these basic organs in the brain corresponds to transferring data, memory and learning [ 11 , 12 , 13 , 14 ]. In this approach, the important and basic block is neurons with a large number of connections. A large number of connected neurons with thousands of connection in the brain makes them vital organ in case of transferring data and informations. In this way, this important part must be investigated and analyzed to achieve an efficient system for designing real organs. These neurons have several functional roles, such as receiving, transmitting and analyzing data for producing voltage signals in all parts of the human brain [ 3 , 4 , 5 , 6 ]. These behaviors can be formulated by some mathematical Equations [ 15 , 16 , 17 , 18 ]. The basic signal behaviors of the biological neurons are modeled by these mathematical equations. In neuron modeling, two approaches are selected: first, the models which are based on the biological neurons behavior, such as Hodgkin-Huxley (HH) [ 2 ] or the ADEX neuron model [ 5 , 6 , 7 , 8 ], and the second is based on spike state modeling, such as Izhikevich [ 3 , 4 , 11 , 12 ]. Among the above list, the Izhikevich model is a computationally reasonable neuronal model that is capable of reproducing all spiking patterns. This model generates all patterns of brain spikes, which is a very significant issue in the neural networks [ 1 , 3 , 4 ]. On the other hand, the HH model is a full-biological neuron model with a high number of equations and terms and is a high-cost neuronal model [ 2 ]. It may be not acceptable for implementing in hardware form because of its high overhead costs. On the other hand, the ADEX model is a cost-effective and biological neuron model that has the both behaviors of the Izhikevich and HH models at the same time, and it is an appropriate choice in case of digital design and realization [ 5 , 6 ]. To realize the neuronal models in hardware state, there are different choices. To achieve a hardware implementation of the neuronal models, we have two cases: analog implementation and digital realization. For realizing in the analog case, CMOS Components are used for designing an analog approach in case of following the mechanisms of neuronal models. This process is high performance in case of speed, but long development time is its disadvantage. Moreover, two factors in digital implementation, high-silicon area and power are required, but this procedure is efficient compared by other approaches [ 1 , 2 , 3 , 4 , 5 , 6 ]. The advantage of digital approach is its flexible attributes, time-down in processing and power supply. Using reconfigurable digital boards, such as FPGAs cause the speed-up and flexibility in this case [ 8 , 9 , 10 , 11 , 12 ]. In the field of neural networks implementation, digital approach may be a suitable procedure [ 1 , 3 , 4 , 5 , 6 , 7 , 8 ]. Indeed, because of high-speed switching in the neuronal networks, it is required to have a accelerate hardware, which is capable of regenerating the same behavior and performance of the human neuronal cells. Different approaches are considered for implementing the neural networks in hardware state in the literature [ 10 , 11 , 12 , 13 , 14 ]. In the field of neuronal realization, since we have high-frequency networks, the digital way is a better solution. On the other hand, it is best suited for computing and digital electronics, and is less affected since noise response are analog in nature, digital hardware is flexible in implementation and digital instruments are free from observational errors such as parallax and approximation errors (compared to the analog approach). The implementation of the neuronal cells using configurable platforms such as FPGA is an attractive area in terms of low overhead costs and high-frequency digital systems [ 1 , 2 , 3 , 4 , 5 , 6 , 7 , 8 , 9 , 10 ]. FPGA is an integrated circuit designed to be configured by a customer or a designer after manufacturing, hence the term field-programmable. The FPGA configuration is generally specified using a Hardware Description Language (HDL), similar to that used for an Application-Specific Integrated Circuit (ASIC). Circuit diagrams were previously used to specify the configuration, but this is increasingly rare due to the advent of electronic design automation tools. FPGAs are flexible and high-speed platforms which are suitable approaches for the digital realization of neuronal networks. Different neuronal models have been realized on FPGA hardware boards. In this approach, Yaghini Bonabi et al. [ 2 ] proposed the FPGA implementation of the HH neuronal model using different methods. Moreover, Haghiri et al. [ 3 , 4 ], Nazari et al. [ 14 ] and Soleimani et al. [ 9 ] presented digital FPGA realization of coupled neurons with astrocyte cells. Furthermore, Soleimani et al. [ 12 ] proposed a digital effective realization of Izhikevich neuron model in low-cost and high-frequency state. In other words, the ADEX neuron model is implemented in the literature [ 5 , 6 , 7 , 8 ]. In detail, Gomar et al. [ 5 ] proposed power-2-based implementation of the ADEX model without any multiplications. Furthermore, Haghiri et al. [ 6 ] presented a novel realization of the ADEX neuronal model using full-matching approximation and low-error calculations. Gomar et al. [ 7 ] also proposed another approach for implementing this model using a CPG-based method and Heidarpour et al. [ 8 ] presented a CORDIC approach for ADEX realization. In all of these papers, the authors use an approximation method to realize the final hardware. This may causes different levels of errors of two original and proposed approaches and affect the final digital hardware. As a comparison between ADEX implementation, Gomar in [ 5 ] proposes a power-2-based function for approximating the exponential term. This conversion may cause error levels between the original and proposed models, and also based on Table 7, speed-up and overall saving are reduced. Furthermore, in this approximation, the accuracy of fraction part of modification is low and this affects on the FPGA realization. On the other hand, Gomar in [ 7 ] also implemented the ADEX model in a multiplierless state. As can be seen, in this approach, she also used the power-2-based method, which is low-accurate in case of fraction part calculation. In [ 6 ], Haghiri proposed a high-accurate approximation of the ADEX model using a power-2-based function with a high degree of similarity. In their approach, the accuracy of matching is increased, but the FPGA cost is also increased; additionally, some levels of errors between the original and proposed spiking patterns can be observed. Finally, Heidarpour in [ 8 ] presented the CORDIC algorithm for approximating the ADEX neuron model. In their way, although the accuracy will be increased, the overall saving in FPGA will be significantly decreased. This issue may affects on large number of implemented neurons on an FPGA core. The general technique for realizing the neuron models are clarified. Capable realization of biological neuronal networks are important. In biology approach, the experimental view of neuroscience are considered to have a basic inspiration of the brain architecture. Thus, studying hardware realization of these biological-like systems can be a necessary goal. At first, the neuronal modeling can be suited. Large number of neuron models are existed for SNNs based on different dynamical mechanisms. For example, the ADEX modeling is acceptable one and capable for duplicating different patterns of spikes in the brain. In the next step, the selected model can be validated in case of timing analysis and dynamics. Since the basic (original) models have high-cost functions, it is required that modification is done to achieve a low-area and cost modeling in hardware implementation. After that, the proposed model must be validated in terms of following the all aspects of original behaviors in MATLAB software. At the second step, to evaluate the proposed approach (in hardware consideration), the FPGA boards are applied in hardware case. Indeed, the Hardware Description Language (HDL) is considered for the proposed neuronal model in ModelSim and ISE Xilinx software’s. In this case, resource utilization and costs of the original and proposed models are compared in case of digital implementation. Two basic factors in this consideration (overall saving in FPGA resources and speed-up or frequency) are compared and the validated that the proposed model is in the better state of digital realization. This paper presents the N-LUT-ADEX (Nyquist-Look Up Table-ADEX) model, which is based on accurate sampling of the original model. Since in this model, high-accuracy matching is achieved, it can recreate a large number of spike patterns in terms of the high similarity state with the original model and also reducing the final overhead costs compared to other papers. Indeed, by accurate sampling of the nonlinear terms of the original ADEX model, the final computational error will be significantly reduced. In this approach, the Nyquist frequency converts a continuous term to a discrete one. When the bandwidth of a signal is lower than the Nyquist frequency of the sampling, the equivalent sample rate is overhead the Nyquist ratio for that certain signal. Using this procedure, the nonlinear term of the original ADEX neuron model (which is an exponential term with high-cost state for digital hardware realization) can be replaced by LUT-based memories. This model can be implemented in hardware state without using any multipliers and other nonlinear terms. In other words, these nonlinear functions are high-cost and low-frequency blocks and by replacing them with some LUT-based terms, the frequency (speed) of digital system will be increased. Then, the overhead costs are significantly reduced. Two basic factors in neural networks implementations are: large-scale approach and high-speed switching. Indeed, to achieve a real and nature-inspired neural system, these two factors must be taken into account. Consequently, using this new model, we have an efficient and modified model that can be considered for implementation in biological neural networks."
} | 4,168 |
29318199 | PMC5655382 | pmc | 3,844 | {
"abstract": "Polyhydroxyalkanoates (PHA) have been produced by some bacteria as bioplastics for many years. Yet their commercialization is still on the way. A few issues are related to the difficulty of PHA commercialization: namely, high cost and instabilities on molecular weights (Mw) and structures, thus instability on thermo-mechanical properties. The high cost is the result of complicated bioprocessing associated with sterilization, low conversion of carbon substrates to PHA products, and slow growth of microorganisms as well as difficulty of downstream separation. Future engineering on PHA producing microorganisms should be focused on contamination resistant bacteria especially extremophiles, developments of engineering approaches for the extremophiles, increase on carbon substrates to PHA conversion and controlling Mw of PHA. The concept proof studies could still be conducted on E. coli or Pseudomonas spp. that are easily used for molecular manipulations. In this review, we will use E. coli and halophiles as examples to show how to engineer bacteria for enhanced PHA biosynthesis and for increasing PHA competitiveness.",
"introduction": "1 Introduction Polyhydroxyalkanoates (PHA) have been produced since the 1980s with limited market success [1] , [2] , [3] . Many challenges are related to the limited PHA commercialization ( Table 1 ), especially the high production cost and instability on thermo-mechanical properties resulted from unstable molecular weights (Mw) and structures, that are also associated with unstable PHA synthase activity and monomer supplies [4] , [5] , [6] , [7] , [8] , [9] , [10] . Efforts have been made to meet these challenges [5] , [11] , [12] , [13] . Table 1 Challenges for producing cost competitive PHA. Table 1 Problems Reasons Solutions Reference High energy demands Sterilization and intensive aeration Unsterile and micro-aerobic processes [17] Low substrates to PHA conversions Substrates are consumed for other purposes Deletion or weakening PHA unrelated pathways [18] , [19] Unstable PHA structures Multiple pathways consuming PHA precursors Deletion or weakening PHA unrelated pathways [19] , [20] Unstable batch Mw Unstable PHA synthase activity Controlling PHA synthase activity [8] , [10] Slow growth Binary fission et al. Multiple fission et al. [21] Discontinuous processes Avoid possible contamination Use contamination resistant strains [16] , [17] , [22] Expensive downstream Complexity to extract and purify products Morphology engineering [23] , [24] The high cost is the result of high energy demand related to complicated sterilization and intensive aeration, low conversion of carbon substrates (S) to PHA products (P), slow growth of microorganisms, discontinuous processes and expensive downstream processing et al. ( Table 1 ) [14] , [15] . The use of extremophilic bacteria combined with metabolic engineering and synthetic biology could fully address these issues [16] , [17] . Future engineering on PHA producing microorganisms should be focused on contamination resistant bacteria especially extremophiles, developments of engineering approaches for the extremophiles (which is called “Next Generation Industry Biotechnology” or “NGIB”, which will be discussed in section 6 in this paper), increase on carbon substrates to PHA conversion and controlling Mw of PHA ( Table 1 ). The concept proof studies could still be conducted on E. coli or Pseudomonas spp. that are easily used for molecular manipulations. In this review, we will use E. coli, Pseudomonas spp., and halophiles as examples to show how to engineer bacteria for better PHA biosynthesis and for increased PHA application competitiveness."
} | 923 |
37746230 | PMC10512296 | pmc | 3,845 | {
"abstract": "Phosphorus (P) bioavailability affects plant nutrition. P can be present in soils in different chemical forms that are not available for direct plant uptake and have to be acquired by different mechanisms, representing different resource niches. These mechanisms, of which many seem to be attributed to mycorrhiza, likely influence the diversity and stability of plant communities in natural ecosystems, as they also might help to overcome a future shortage of P supply in agro-ecosystems. In order to understand the mechanisms of P acquisition, the associated carbon costs, and the resource partitioning by mycorrhizal fungi, the ecosystem situation has to be mimicked in smaller scaled experiments. Here, different experimental setups are evaluated using plantlets of Populus x canescens and its functional ectomycorrhizal (ECM) fungus Paxillus involututs strain MAJ. To investigate resource partitioning involving mycorrhizae, the protocols of this study describe preparation of an in vitro and a rhizotrone culture systems for studies under axenic conditions as well as a mesocosm culture system for greenhouse conditions. We also describe the construction of separate compartments containing nutrients and excluding plant roots as well as the progress that has been made in in vitro propagation of plant and ECM fungal material. The practical experience made in our study shows that the in vitro culture system is prone to desiccation and its construction and maintenance are more time consuming and complicated. In contrast, with the axenic rhizotrone culture system and the mesocosms we have created more robust and very versatile systems that are also suitable for greenhouse conditions.",
"conclusion": "Conclusion The present study revealed that the construction and maintenance of the axenic rhizotrone and the mesocosm culture systems are less complicated and time consuming compared with the in vitro culture system. But also, the in vitro culture system can be equipped with the external apparatus with sterile syringe filters (as reported for the axenic rhizotrone culture system) to supply the plants with a nutrient solution without the need to open the system in the time course of the experiment. Nevertheless, especially the mesocosms are robust and very versatile. But all presented culture systems enable the user to comprise additional tests, including labelling of the plant with 13 C or determination of P acquisition mechanisms. Also the separate compartments for nutrient supply are well adaptable to different experimental set-ups and enable the simulation of an ecosystem situation with an ECM plant having access to widely distributed P source patches with different bioavailabilities ( \n Figure 1 \n ) through mycorrhizal fungus, excluding the direct strife of plant roots and mycorrhizal hyphae for P. P. involutus is a long distance exploration type ECM fungus ( Gronbach, 1988 ) with few but highly differentiated rhizomorphs (review by Agerer, 2001 ). These type of ECM fungi were shown to transport efficiently water and higher rates of P. Moreover, the ectomycorrhizal fungus P. involutus (strain MAJ) is compatible with the poplar plant species P. x canescens (clone ‘Schleswig I’) and both are easy to maintain and propagate in vitro . Hence, these organisms provide valuable model systems for a more robust test of nutrient acquisition and exchange models ( Gafur et al., 2004 ; Müller et al., 2013 ). Therefore, the design of a compartmental culture system using these compatible ectomycorrhizal associates was a solid choice to down-scale the ecosystem situation of P source dependent host C exchange for mycorrhizal P as well as of mycorrhizal mediated P resource partitioning. Since the protocol described by Rygiewicz et al., 1988 , the present study was the first providing details on practical experience and evaluated protocols for the design and maintenance of the experimental set-ups to investigate such ecosystem situations. Moreover, these culture systems were designed not only for outdoor but also for controlled conditions excluding interferences with other micro-organisms, revealing the true capabilities of the mycorrhizal fungi in nutrient acquisition.",
"introduction": "Introduction More than 90% of phosphorus (P) in the soil is present in different chemical forms that are unavailable to plants ( Mengel et al., 2001 ). The association of plants with mycorrhizal fungi can increase the bioavailability of P ( Hinsinger, 2001 ; Hinsinger et al., 2011 ; Plassard et al., 2011 ) and so meliorate their nutritional state ( Gafur et al., 2004 ). Moreover, resource partitioning of P associated with mycorrhizal fungi could contribute to more efficient use of different P forms by plants, reducing competition for soil P ( Turner, 2008 ). It is hypothesised that mycorrhizal fungi are the key component in resource partitioning under P impoverished conditions in soil. Ectomycorrhizae (ECM) were shown to mine different chemical forms of P (reviewed by Plassard et al., 2011 ) in exchange for energy derived from hosts’ photosynthesis ( Buscot, 2015 ) and increase cost-efficiently the soil volume explored for nutrients via its extra-radical hyphae ( Jones et al., 1998 ). To understand if (i) ECM mobilizes P from differently accessible sources and (ii) consequences of higher energy costs for the plant development due to the acquisition of more complex P forms exist, adequate and innovative experimental systems consisting of compatible plants and mycorrhizal fungi have to be developed. The use of compatible plant and ectomycorrhizal (ECM) fungal species provide valuable model systems for a more robust test of nutrient acquisition and exchange models ( Gafur et al., 2004 ). In functional symbiosis, the mycorrhizal associates can penetrate the plant root, forming a Hartig net between the cortical cells ( Rousseau et al., 1994 ; Hampp et al., 1996 ; Gafur et al., 2004 ), where the exchange of nutrients between the symbionts is supposed to happen ( Landeweert et al., 2001 ). In contrast, incompetent ECM fail to penetrate the host roots, causing a defence reaction by thickening the cell wall of the epidermis ( Lei et al., 1990 ; Gafur et al., 2004 ). It was shown that only functional associates could increase plant P uptake in nature under impoverished nutrient conditions ( Rousseau et al., 1994 ; Smith and Read, 2008 ; Hoeksema et al., 2010 ). Poplar plant species Populus x canescens and the ECM fungus Paxillus involutus strain MAJ are the reassuring organisms on this matter ( Gafur et al., 2004 ; Müller et al., 2013 ). Under field conditions, poplars can form symbiotic associations with different mycorrhizal types such as arbuscular mycorrhizal (AM) and ECM fungi ( Khasa et al., 2002 ; Gherghel et al., 2014 ). The study of Gherghel et al. (2014) could show that P. involutus subsequent to Rhizophagus irregularis colonized poplar clones under various field conditions. Among poplars, P. involutus (Basidiomycetes) has a wide variety of hosts able to form ECM with many forest tree species belonging to gymnosperms and angiosperms ( Duddridge, 1987 ; Baum et al., 2000 ; Gafur et al., 2004 ) and be an appropriate help for trees in ‘bare-root’ conditions ( Jarosch and Bresinsky, 1999 ; Hönig et al., 2000 ). P. involutus was also shown to be able to exude oxalic acid to acquire P from mineral sources ( Lapeyrie et al., 1991 ) and to release surface-bound phosphatases that can mineralize organic P forms ( McElhinney and Mitchell, 1993 ; Alvarez et al., 2004 ). Moreover, P. involutus is easy to maintain and propagate in culture and is therefore increasingly used in ECM studies ( Wallander and Soderstrom, 1999 ; Gafur et al., 2004 ; Müller et al., 2013 ). The natural hybrid P. x canescens (grey poplar) result through pollination of P. alba (white poplar) by P. tremula (European aspen) ( Lexer et al., 2005 ). P. x canescens and P. alba occur sympatrically in European river valleys ( Rajora and Dancik, 1992 ; Fossati et al., 2004 ; Lexer et al., 2005 ), whereas P. tremula is an important pioneer tree species covering forests in the upland ( Adler et al., 1994 ; van Loo et al., 2008 ). Furthermore, the economic and scientific importance of Populus trees increased. Due to their natural distribution and genetic variability, they can be cultivated under polluted and degraded soil conditions ( Chen and Polle, 2010 ) and contribute to a site’s positive carbon balance. Thereby, the poplar supplies the industry with wood biomass, fibre, bioenergy and chemicals ( Klass, 1998 ). Furthermore, poplars are considered as beneficial model organisms due to the ease of micropropagation ( Confalonieri et al., 2003 ), which reduces stochastic variation by the use of clones instead of seedlings. Because of all these benefits, poplars gained importance in scientific fields, including biotechnology, molecular biology, and other areas related to nutrition, abiotic pressures, or the plant-soil interface (reviewed by Müller et al., 2013 ). The latest and first study aiming to investigate the ECM mediated resource partitioning for P was performed by Schreider et al. (2022) , revealing that the ECM P. involutus has the potential to occupy fundamental niches of various P sources, whereby the readily available phosphate was not as expected the most favourable P source for uptake within a mixed P pool. Previously, progress in studying partitioning for soil P has been achieved by examining plant responses to single P sources ( Steidinger et al., 2014 ) or a mixed pool of different P sources ( Liu et al., 2018 ), but the sole contribution of mycorrhizal fungi in resource partitioning was neglected in these studies. The studies of Andrino et al. (2019 ; Andrino et al., 2021 ) were the first to investigate the mycorrhizal mediated acquisition of differently available P sources using arbuscular mycorrhiza. Nevertheless, Andrino et al. (2019 ; Andrino et al., 2021 ) have determined that higher amounts of C were invested by the plant into the association with a mycorrhizal fungus that had access to more complex P sources, whereby the P sources were supplied as a single P source. By using separate compartments for the nutrients, it is possible to mimic the ecosystem situation with an mycorrhizal plant having access to widely distributed nutrient patches with different bioavailabilities through mycorrhizal fungus, excluding the direct competition of plant roots and mycorrhizal hyphae for P and depending on the experimental conditions revealing the true capabilities of mycorrhizal fungal associate in nutrient acquisition. The main aim of this study is to provide practical experience and to describe the learning curve in developing a culture system that is robust and versatile to investigate mycorrhizal mediated resource partitioning. Three different experimental setups using poplar plantlets of P. x canescens and its functional ECM fungus P. involututs strain MAJ for axenic and greenhouse conditions are presented. The present study covers also modified protocols ( Müller et al., 2013 ) of the in vitro multiplication and rooting of plant material and production of ECM fungal inoculum under axenic conditions that suited our experiments the most. Furthermore, the co-cultivation of poplar and ECM fungi and maintenance steps in different culture systems under axenic and greenhouse conditions are reported. Also, the construction of separate compartments for the nutrient supply that are adaptable into different experimental set-ups is described."
} | 2,903 |
22967313 | PMC3815895 | pmc | 3,846 | {
"abstract": "Summary Methanogenic community structure and dynamics were investigated in two different, replicated anaerobic wastewater treatment reactor configurations [inverted fluidized bed (IFB) and expanded granular sludge bed (EGSB)] treating synthetic dairy wastewater, during operating temperature transitions from 37°C to 25°C, and from 25°C to 15°C, over a 430‐day trial. Non‐metric multidimensional scaling (NMS) and moving‐window analyses, based on quantitative real‐time PCR data, along with denaturing gradient gel electrophoresis (DGGE) profiling, demonstrated that the methanogenic communities developed in a different manner in these reactor configurations. A comparable level of performance was recorded for both systems at 37°C and 25°C, but a more dynamic and diverse microbial community in the IFB reactors supported better stability and adaptative capacity towards low temperature operation. The emergence and maintenance of particular bacterial genotypes (phylum Firmicutes and Bacteroidetes ) was associated with efficient protein hydrolysis in the IFB, while protein hydrolysis was inefficient in the EGSB. A significant community shift from a Methanobacteriales and Methanosaetaceae towards a Methanomicrobiales ‐predominated community was demonstrated during operation at 15°C in both reactor configurations.",
"introduction": "Introduction Bioenergy production from waste streams is a key component in the global development of sustainable energy sources (Demirel et al ., 2010 ). Indeed, anaerobic digestion (AD) is poised to replace aerobic microbiological treatments as the core process of waste‐to‐energy technologies for enhanced sustainability in the coming decades (Verstraete and Gusseme, 2011 ). During AD, organic substrates are sequentially degraded by fermentative and acetogenic bacteria to simple precursor compounds, such as acetate, H 2 /CO 2 , formate and methanol, from which methanogenic Archaea produce a methane‐rich biogas. Temperature can influence the rate and path of carbon flow during methanogenesis by affecting the activity of particular microbial groups and the structure of the microbial consortia (O'Reilly et al ., 2009 ; McKeown et al ., 2009a ; Siggins et al ., 2011 ). Low‐temperature AD (LTAD) has emerged as an economically attractive waste treatment strategy, which confers considerable advantages over conventional mesophilic (∼ 30°C) and thermophilic (∼ 55°C) treatments, primarily due to the capacity to treat the wide variety of cool, dilute wastewaters, previously considered as not suitable for AD (McKeown et al ., 2012 ). LTAD has been successfully applied at laboratory‐ and pilot‐scale, using a variety of reactor types, for the treatment of a broad range of wastewaters (for example, Lettinga et al ., 2001 ; Collins et al ., 2003 ; 2006 ; McHugh et al ., 2004 ; Syutsubo et al ., 2008 ; Bergamo et al ., 2009 ; McKeown et al ., 2012 ). LTAD is an attractive technology because the process is stable, simple to operate and requires very low energy input. Improved reactor designs enable high rates of methanogenic conversion at low temperatures through a combination of: (i) high mixing intensities (i.e. facilitating high rates of mass transfer); and (ii) enhanced retention of psychro‐active biomass (Lettinga et al ., 2001 ; Alvarado‐Lassman et al ., 2008 ; McKeown et al ., 2012 ). However, failure of the bioreactors to retain granular sludge during LTAD may lead to severe hydraulic washout of psychro‐active sludge (Lettinga et al ., 1999 ). Therefore, non‐granule‐based systems using inert nuclei to promote re‐granulation could be of advantage during psychrophilic reactor operation (McKeown et al ., 2009b ). Knowledge gaps remain, however, regarding the nature and function of the microbial populations involved in LTAD, which are a deterrent to full‐scale applications (McKeown et al ., 2009b ; 2012 ). This information deficit is mainly due to the complex relationship between wastewater characteristics, process conditions and dynamics in microbial community structure. In an attempt to link microbial functional groups with process performance, we studied community dynamics in two different methanogenic anaerobic reactor configurations [i.e. an inverted fluidized bed (IFB) containing fixed fluidized biomass on the support particles and an expanded granular sludge bed (EGSB) containing crushed granular biomass], during operational temperature transitions from 37°C to 25°C, and from 25°C to 15°C. The present study is a continuation of the experiment described in Bialek and colleagues ( 2011 ), therefore a similar experimental approach has been employed. The methanogenic community structure and dynamics were examined qualitatively by DGGE and quantitatively by real‐time PCR, the results were then statistically analysed and compared. We hypothesized that the process performance and microbial community structure and dynamics can be influenced by the reactor configuration during transition from mesophilic to psychrophilic reactor operation and by changes applied to the loading rate and hydraulic retention time (HRT).",
"discussion": "Discussion Process performance Dairy wastewater is a complex substrate composed of easily degradable carbohydrates, mainly lactose, and less bioavailable proteins and lipids (Fang and Yu, 2000 ; Tommaso et al ., 2003 ). The latter are responsible for the typical problems associated with high‐rate AD of dairy waste effluents (Perle et al ., 1995 ). Hydrolysis of proteins and lipids is reported to strongly decline with decreasing temperature, especially approaching 15°C (Tommaso et al ., 2003 ). Given these reports, and considering the fact that skimmed dairy wastewater was used in the present study (Table 3 ), decreased protein degradation/hydrolysis can be assumed to be mainly responsible for the declining process performance in our digesters. Table 3 Composition of skimmed‐milk powder. Parameter % of COD COD (mg l −1 ) Proteins 40 1600 Sugars 55 2200 Fats 1 40 Others 4 160 Total 100 4000 Casein is the major protein in milk (up to 80% of the total proteins) and in dairy effluents. When fed to acclimated anaerobic reactors, degradation of casein is rapid, due to strong proteolytic activity, and the degradation products are non‐inhibitory (Perle et al ., 1995 ). This was likely the case in the systems investigated, with > 80% PRE recorded during acclimated mesophilic conditions of 37–25°C (Fig. 1 ). The fluctuations in effluent VFA concentration (Fig. 1 ) observed at mesophilic temperatures were therefore, most probably due to rapid hydrolysis and fermentation of carbohydrates and proteins into VFA. A gradual decrease in COD RE (Fig. 1 ), however, occurred immediately after the temperature reduction to 15°C, coupled with an instantaneous decrease in residual VFA concentrations and significant drop in PRE of c . 60% in the IFB and EGSB reactors. Since no VFA build‐up was observed at 15°C, it is considered that protein hydrolysis had become the rate‐limiting step. Following the immediate drop in PRE to 60%, the IFB system demonstrated a successful adaptation to low temperature operation and, after a brief temporal instability, > 78% COD RE and > 77% PRE was recorded at psychrophilic, steady‐state operation (Fig. 1 ). On the other hand, the EGSB system displayed a slower adaptation to low temperature operation, with a performance of 58% PRE 45 days after the temperature decrease, and long unstable performance, with minimum values of 13% PRE and < 60% COD RE (Fig. 1 ). We propose that the better flexibility and adaptability of the IFB biomass to low temperature might originate from the spatial arrangement of fixed fluidized biomass that developed in the IFB (Fig. S1) playing an important role in differences in transfer of intermediates and optimal degradation of substrates (Grotenhuis et al ., 1991 ). At low temperatures, the viscosity of effluents increases and, therefore, the diffusion of soluble compounds will drop, particularly in sludge bed reactors that become less easily mixed (Lettinga et al ., 2001 ). There was no evidence of granulation of the biomass in the EGSB reactor, which was seeded with crushed granular sludge as inoculum and the biomass remained a non‐granular floculant sludge throughout the trial (Fig. S1). Furthermore, the advantage of the IFB configuration over the EGSB in this experiment might arise from the use of floatable particles with a specific density lower than the liquid, thus particles were fluidized downward (Garcia‐Calderon et al ., 1998 ) and better substrate–biomass contact might be attained. Due to the large specific area, the support particles can retain more biomass (Alvarado‐Lassman et al ., 2008 ), which is especially crucial during transitional and permanent changes in operating conditions, like the temperature variation investigated in this study. Microbial population dynamics Understanding the impact of disturbances, such as temperature shocks or permanent temperature decrease, on process stability and performance will undoubtedly shed more light on the process of LTAD when placed in the context of microbial community dynamics. Together with increased knowledge on the impact of reactor configuration on the functional stability of microbial communities, informed decisions can be enabled regarding the optimal reactor type and process conditions for a given wastewater. A robust LTAD system must possess the ability to maintain process stability in response to disturbances and it has been reported that in general systems with more dynamic communities have greater functional and process stability (Hashsham et al ., 2000 ; Carballa et al ., 2011 ; Werner et al ., 2011 ). We thus considered our data in the context of the model proposed by Allison and Martiny ( 2008 ), which divides the population dynamics that maintain community function over time into three basic mechanisms (resistance, resilience or redundancy) to address the process performance of the IFB and EGSB reactors investigated. A microbial community is resistant if it is similar across a variety of environmental conditions and, therefore, it is difficult to perturb from an original state (Allison and Martiny, 2008 ). Initially, identical bacterial populations at mesophilic temperatures, as observed by the UPGMA cluster analysis, suggested resistance of the bacterial populations composition (Fig. 2 ): 100% similarity between IFB T1 (37°C) and IFB T2 (25°C) and > 95% similarity between EGSB T1 (37°C) and EGSB T2 (25°C). A similar trend was observed with the archaeal population behaviour between the two studied mesophilic temperatures where, for example, the EGSB and the IFB populations showed 80% similarity between T1 (37°C) and T2 (25°C) based on the NMS and moving window matrix (Fig. 5 A and B). The methanogenic community composition was thus resistant to the temperature change from 37°C to 25°C, indicating metabolic flexibility and physiological tolerance to the applied disturbance in this mesophilic temperature range. A further temperature decrease to 15°C showed that the community was, however, sensitive to the disturbance and resulted in an altered microbial composition (Fig. 5 A and B), indicating that the community responded to the second temperature disturbance using one of the mechanisms discussed below. The microbial composition of reactor biomass is resilient if it is sensitive to a disturbance and changes, but quickly recovers to its initial composition (Allison and Martiny, 2008 ). This mechanism was not observed in the systems and timescale investigated because after the disturbance the community did not return to its original composition. Werner and colleagues ( 2011 ) reported that resilience was important to maintain the function of syntrophic populations over time in mesophilic brewery wastewater treatment facilities. These authors concluded that syntrophic bacteria had very specialized metabolic functions within the overall trophic structure, which made them more likely to rebound after a disturbance, rather than undergoing competitive growth with different syntrophs that have a similar function in the microbial consortium. Whether the microbial composition rebounds, or not, is possibly determined by the severity of the disturbance and importance of the disturbance on the process stability and performance. It may be possible, for example, that the microbial composition of the anaerobic reactor subjected to a slight variation or short disturbance in environmental conditions returns to its original composition after such a disturbance. Madden and colleagues ( 2010 ) investigated the effect of transient (but severe) perturbations on the methanogenic community structure and process performance of replicate EGSB‐based reactors. Their cluster analyses of DGGE data suggested that temporal shifts in microbial community structure were predominantly independent of the applied perturbations (Madden et al ., 2010 ), although it is important to point out that community dynamics were monitored only via DGGE, where gel‐to‐gel variation and relatively low sensitivity (compared with qPCR) are limiting to ensure reproducibility and detection of minor populations or subtle population changes (Talbot et al ., 2008 ). Recent studies point out the limitations of studies based only on the DNA (Raes and Bork, 2008 ; Nelson et al ., 2011 ; Wendeberg et al ., 2012 ). Future studies should therefore focus on functional investigations to unravel metabolic activity of the microbial communities underpinning the AD processes. When the community composition is sensitive, and not resilient, it might produce process rates similar to the original community in case the members of the community are functionally redundant (Allison and Martiny, 2008 ). The highly dynamic community structure during well‐functioning periods (Fig. 5 A and B) may be explained by the functional redundancy among diverse phylogenetic groups, allowing oscillations of their populations, due to the presence of a reservoir of species able to perform the same ecological function with no effects on the reactor performance (Zumstein et al ., 2000 ; Briones and Raskin, 2003 ). The stability of reactor performance, especially that of the IFB reactors, after the temperature shift from 25°C to 15°C and consequent change in methanogenic community composition, could be explained by functional redundancy in one, or more, steps of the methanogenic pathway. Microbial community composition Methanogenesis can proceed through two pathways, where acetate and/or hydrogen and carbon dioxide are converted into methane, termed as aceticlastic methanogenesis and hydrogenotrophic methanogenesis respectively. Under certain conditions, homoacetogenic bacteria can compete with hydrogenotrophic methanogenesis for hydrogen (hydrogen is used to reduce carbon dioxide to acetate (Lovley and Klug, 1983 ). Homoacetogenesis has been observed under psychrophilic conditions, and some studies have reported that homoacetogens have a better ability to adapt to low temperatures than hydrogenotrophic methanogens (Kotsyurbenko et al ., 2001 ; Nozhevnikova et al ., 2003 ). No acetic acid accumulation (Fig. 1 presented as sum of VFA) was, however, observed at 15°C in this study. Competition between these groups did become apparent, when the temperature of the reactors was reduced to 15°C, as indicated by the bacterial DGGE results (Fig. 2 ). Sequence B21 was present in both systems, only after the transition to 15°C, and showed 99% similarity to Acetobacterium carbinolicum and Acetobacterium psammolithicum (phylum Firmicutes ; Table 1 ). Both organisms were formerly described as psychro‐active homoacetogenic bacteria, capable of growing at temperatures from 1°C to 35°C, with optimal growth between 20°C and 30°C (Conrad et al ., 1989 ; Simankova et al ., 2000 ). Enhanced activity of psychrotolerant homoacetogenic bacteria, followed by acetoclastic methanogenesis, has been previously reported to be important in the degradation of organic matter under low temperature conditions (Schulz and Conrad, 1996 ). There are also reports that homoacetogenic formation of acetate can bypass the formation of fatty acids and H 2 , which could explain suppressed VFA production in our systems at 15°C, similar to that described for the decomposition of organic matter in anoxic paddy soil at low temperature (Chin and Conrad, 1995 ), or for the acidic sediment of Knaack Lake (Conrad et al ., 1987 ). No increase in the maximum specific methanogenic activity with acetate as the substrate was noted with the biomass sampled from either reactors following the temperature decrease to 15°C (data not shown) and the quantitative analysis of methanogenic community structure did not record a large increase in acetoclastic methanogens (Fig. 4 ). Phylogenetic analyses of the archaeal DGGE bands A1–A3 (Fig. 3 B) indicated that hydrogen‐utilizing Methanocorpusculum ‐like organisms (order Methanomicrobiales ) were prominent after transition to low temperature. The Methanocorpusculum ‐like organisms deduced from A3 of the IFB system were likely to be the major hydrogenotrophic population in the low temperature reactor from day 365 (T3b, 15°C) onwards. In case of the EGSB reactor some part of the MMB population shifted towards Methanospirillum ‐like organisms deduced from A8 (T3a, 15°C) from day 298 onwards. Many studies have indeed documented that methanogenesis predominantly proceeded through the hydrogenotrophic route in low‐temperature anaerobic reactors (Syutsubo et al ., 2008 ; O'Reilly et al ., 2009 ; McKeown et al ., 2009a ). In these situations, conditions with low hydrogen availability and high biomass concentration seemed to favour hydrogenotrophic methanogens due to their higher affinity for H 2 and thus they out‐competed homoacetogens for hydrogen (Kotsyurbenko, 2005 ). It has been further proposed that hydrogen metabolism is thermodynamically and metabolically more favourable than acetate utilization; and that a higher level of hydrogen can be retained in the system (i.e. increased gas solubility) at low temperature (Lettinga et al ., 2001 ; Kotsyurbenko, 2005 ). Supporting this, other authors (O'Reilly et al ., 2009 ; McKeown et al ., 2009a ; Siggins et al ., 2011 ) clearly demonstrated the temporal methanogenic community shifts towards the dominance of hydrogenotrophs, especially the order Methanomicrobiales , in low‐temperature reactors compared with mesophilic systems. Hydrogenotrophic MMB showed a 1600‐ to 2200‐fold increase in the 16S rRNA gene concentration from its minimum to maximum, corresponding to mesophilic and psychrophilic reactor operation in the IFB and EGSB reactors respectively (Fig. 4 ). Indeed, given this remarkable rise, low temperature appeared to be the major factor facilitating the emergence to dominance of this group in our reactors."
} | 4,757 |
39679908 | PMC11809366 | pmc | 3,847 | {
"abstract": "Abstract Collagen fiber skeleton from animal skin is an ideal substrate for electronic skin (e‐skin). However, the interface mismatch between conductive materials and skeleton and the monotonicity of conductive network still hinder its creation. Herein, a novel collagen fiber‐based e‐skin with dual‐mode conduction of NaCl and conductive spheres (IECS) is accomplished by loading organohydrogel into the skeleton via “permeation and self‐assembly”. The resulting interpenetrating network produces a 3D continuous, conductive pathway and strong interface interaction with high‐density hydrogen bonding, thus exhibiting excellent strength (24.5 MPa), conductivity (14.82 S m −1 ), sensing performance (sensitivity of 16.64), and environmental stability. The physical structure (3D skeleton, interpenetrating network) and chemical interaction (interface interaction, salting‐out) achieve energy dissipation. Meanwhile, the sensitivity is enhanced by dual‐mode conduction, conductive sphere array, and deformation amplification induced by collagen fibers. Additionally, the strong bonding ability between glycerin and collagen fibers with water molecules provides anti‐freezing and moisture‐retention characteristics. Thus, the strategic synergy of compositional and structural design makes IECS a promising force‐sensing part of piezoresistive sensor for human movement, pulse frequency, cipher transmission, and pressure distribution. In short, IECS presents a multifunctional platform for the invention of high‐performance e‐skin with on‐demand property, which offers great application potential in wearable electronics.",
"conclusion": "3 Conclusion In conclusion, the multifunctional natural skin served as inspiration for the design and construction of a temperature‐induced “penetration and self‐assembly” strategy, resulting in an IECS‐based collagen fiber skeleton with a variety of biological functions and smart sensing capabilities. From a structural point of view, PGCB/PM/N in collagen fiber skeleton encountered uniform penetration and dynamic cross‐linking, enabling the two to interpenetrate and form multiple hydrogen bonding interactions in IECS. It ensured that PGCB/PM/N appeared in the form of encased collagen fibers while maintaining the naturally 3D hierarchical structure of the collagen fiber skeleton. From a performance angle, the collagen fiber skeleton was loaded with conductive materials (NaCl, PMMA@MXene spheres), organohydrogel network (PGCB), and an antifreeze agent (glycerin). The resultant IECS simultaneously maintained a variety of biological functions, including flexibility, tensile strength (24.5 MPa), conductivity (14.82 S m −1 ), sensing ability (GF was 16.64, 10.74, and 2.39 in the bending strain of 0–1.0, 1.0%–7.5%, and 7.5–10.0%, respectively), freezing resistance (−42.8 °C) and moisture‐retention (the residual weight rate reached ≈69.7% after storage at 20 °C and 50% RH for 7 days). IECS with both ionic and electronic dual conductive networks achieved high sensitivity over a wide working range by utilizing the directional migration of free ions under large deformation and the structural change of electron conductive network under small external force. Significantly, the external force sensing capability of IECS was also improved by the crack creation and propagation caused by PMMA@MXene spheres and the deformation amplification effect caused by collagen fibers. Besides, the glycerin/water binary solvent system and collagen fiber skeleton interacted with water molecules to form high‐density hydrogen bonds, which was the reason for its enhanced environmental stability. Based on a variety of fascinating characteristics, natural skin‐derived IECS was perfect for feasible employs like monitoring human movement, pulse frequency, cipher transmission, and pressure distribution. Wearable electronics, artificial intelligence, and human‐machine interfaces were anticipated to reap benefits greatly from the true and adaptable e‐skin platform of IECS.",
"introduction": "1 Introduction The growing need for people to take responsibility for their own medical problems is an advancement in the concept of health building. [ \n \n 1 \n , \n 2 \n \n ] The advanced flexible sensor has accomplished significant progress and is currently gaining widespread application in multiple fields of bio‐sensing, electronic skin (e‐skin), health monitoring and soft robots. It operates by mimicking the biological sensing mechanism of nerves, replicating the topological structure and the tactile sensing function similar to human skin. [ \n \n 3 \n \n – \n \n 5 \n \n ] \n The design and development of e‐skin has utilized an extensive selection of renewable resources and their synthetic counterparts (such as animal skin, silk, paper, etc) over the last several years. [ \n \n 6 \n , \n 7 \n \n ] Among them, collagen fiber is derived from natural animal skin, whose basic network is a 3D hierarchical structure. It has proven to be advantageous in biomedical materials, food, cosmetic, and traditional leather industries due to its high strength, good flexibility, bio‐compatibility, and simplicity in carrying other materials. Definitely, collagen fiber is a novel and fascinating substrate for the fabrication of e‐skin. [ \n \n 8 \n \n ] Therefore, the goal of this work is to create an innovative type of collagen fiber‐based e‐skin (also known as skin‐based e‐skin) by preserving the intact, 3D, and cross‐linked network of collagen fiber in natural skin as the skeleton structure and incorporating other molecules as the sensing material. In this construction strategy, collagen fiber serves as a load‐bearing and dissipative scaffold. By facilitating the slip and deformation of collagen fiber skeleton with a hierarchical structure, it dissipates external mechanical energy and provides the corresponding e‐skin with high mechanical strength and toughness. As well, it is also simple, quick, and scalable for creating collagen fiber‐based e‐skin utilizing natural animal skin. The primary purpose of collagen fiber‐based e‐skin is to recognize external signals and translate them into basic electrical signals. [ \n \n 9 \n \n ] In this regard, adequate conductivity is an essential prerequisite for sensing function. The present work on collagen fiber‐based e‐skin predominantly emphasizes inorganic conductive materials and conductive polymer materials. On the one hand, collagen fiber‐based e‐skin with outstanding conductive properties is primarily generated by vacuum filtration relying on the size advantage of inorganic conductive materials. However, since inorganic material is brittle, it is challenging to integrate it into a flexible collagen fiber skeleton throughout a broad area. Meanwhile, the inorganic material eventually falls off because of the weak interaction and the interface mismatch effect with the collagen fiber skeleton, resulting in the restricted sensing stability of collagen fiber‐based e‐skin. [ \n \n 10 \n , \n 11 \n \n ] On the other hand, collagen fiber‐based e‐skin is created by the reactive activation of conductive polymer precursor through in situ polymerization. The corresponding e‐skin possesses good stability thanks to the flexibility of conductive polymer and it is in situ production between collagen fiber skeleton, meeting the demands of large‐area manufacturing of flexible electronics. Unfortunately, this process depends on the penetration of conductive polymer precursor within the collagen fiber skeleton and the conditions of polymerization, which frequently ends in a small amount of conductive polymer and poor conductivity of collagen fiber‐based e‐skin. [ \n \n 12 \n , \n 13 \n \n ] The aforementioned evidence indicates that the research system of collagen fiber‐based e‐skin has been enriched by various designs and preparations, which have established the foundation for its generation and utilization. But there nevertheless remain some shortcomings. In addition, collagen fiber skeleton derived from animal skin is not resistant to high temperatures, strong acids, and strong alkalis, limiting its combination with sensing materials. Therefore, it remains a challenge to construct collagen fiber‐based e‐skin that simultaneously delivers good conductivity, sensing ability, mechanical property, and versatility, which requires the exploration of new approaches and better materials. Given that conductive hydrogel is a 3D soft material with high elasticity and good flexibility, many researchers have been particularly interested in it. [ \n \n 14 \n , \n 15 \n \n ] The utilization of conductive hydrogel as the sensing material offers another fascinating option for the construction of collagen fiber‐based e‐skin. [ \n \n 16 \n , \n 17 \n , \n 18 \n \n ] Specifically, the primary challenges of this strategy include the uniform distribution of conductive hydrogel solution through collagen fiber skeleton in the flow state and the in situ encapsulation of conductive hydrogel by transforming it from a flow to a solid state. [ \n \n 19 \n , \n 20 \n , \n 21 \n \n ] In current studies, the conductivity of hydrogel is primarily achieved through either ionic conduction or electronic conduction. The current conduction process of ion‐conducting hydrogel is achieved by facilitating the migration of ions in ionic liquids or inorganic salts. The resistance decreases with stretching and is directly proportional to the degree of deformation. Therefore, the ion‐conducting hydrogel has lower sensitivity at small strains but a wider sensing range. [ \n \n 22 \n \n ] Whereas, electron‐conducting hydrogel utilizes electrons and holes to transmit electrical signals. The percolation network generated by the contact of conductive materials determines its electrical behavior. Due to the complete separation of the percolation network under large strain, the electron‐conducting hydrogel has a limited working range but higher sensitivity under small strain. [ \n \n 23 \n \n ] In addition, the mechanical strength, anti‐freezing at low temperatures, and moisturization for a long time of conductive hydrogel are also the main issues that need to be focused on. [ \n \n 24 \n \n ] \n Based on the above background, a straightforward “permeation and self‐assembly” technique was employed to create a novel collagen fiber‐based e‐skin (IECS) with an interpenetrating network structure by introducing conductive materials, organohydrogel network, and anti‐freezing agent into collagen fiber skeleton. It overcame the drawbacks of the current collagen fiber‐based e‐skin. In addition, it enhanced the mechanical strength, anti‐freezing, and moisture‐retention performances of organohydrogel. The interpenetrating network structure between collagen fiber and organohydrogel was investigated to satisfy the requirements of synergistic deformation and continuous electrical signal transmission. As conductive media, the effects of polymethyl methacrylate@Ti 3 C 2 T X (PMMA@MXene, denoted as PM) spheres and inorganic ions (Na + , Cl − ) on the electrical and sensing properties of IECS were also studied systematically. In the meantime, IECS developed a dense hydrogen bonding structure upon the addition of glycerin, and its moisture‐retention, anti‐freezing, and self‐regenerating were also illustrated. It was worth noting that the effect of collagen fiber on the comprehensive properties of IECS was elucidated to confirm its necessity. The resulting IECS had excellent mechanical strength, conductivity, sensitivity, anti‐freezing properties, and moisture retention, which offered considerable advantages for precise acquisition of human movement, pulse frequency, cipher transmission, and pressure distribution. These advantages made IECS a promising candidate for wearable electronics, human‐machine interfaces, and artificial intelligence, providing a new platform for the realization of multifunctional flexible e‐skin.",
"discussion": "2 Results and Discussion 2.1 Design Principle and Synthesis of IECS Collagen fiber bundles with numerous binding sites and hierarchical structures, as the skeleton and carrier of IECS, offered an ideal structural foundation for the addition of functional materials. The final IECS with dual‐mode conduction, high mechanical strength, low‐temperature tolerance, and moisturization was constructed in this work by employing an innovative “permeation and self‐assembly” strategy based on stuffing and nano‐engineering, as illustrated in Figure \n \n 1 \n . The whole procedure was divided into four stages: (I) pre‐treatment, (II) permeation, (III) self‐assembly, and (IV) solvent replacement. First, the skin substrate (tanned leather) was prepared using a series of standard leather‐making techniques, such as fleshing, soaking, degreasing, deliming, unhairing, softening, and tanning. These procedures were created to optimize the retention of collagenous elements while effectively removing debris, hair, and non‐collagenous elements from the interior or exterior of the skin. At the same time, the collagen fiber skeleton with a 3D woven structure was obtained by splitting the skin to remove the surface layer (Step I). Then, after electrically conductive PMMA@MXene (PM) spheres were evenly distributed in polyvinyl alcohol (PVA)/gelatin (GEL)/carboxylated cellulose nanofibers (CNFs)/borax (named PGCB) solution, the obtained PGCB/PM pre‐gel solution was applied to collagen fiber skeleton. Utilizing the numerous binding sites and natural pore structure, PGCB/PM pre‐gel solution permeated uniformly and thoroughly in the hierarchical structure of the collagen fiber skeleton via vacuum filtration at room temperature (Step II). Subsequently, PGCB/PM pre‐gel solution, which was embedded in the skeleton of collagen fibers, underwent an in situ phase transition process at low temperature to form pre‐gel and fix the shape. Thanks to the interpenetrating network structure by interwoven pre‐gel and collagen fiber as well as the numerous hydrogen bonds between their active functional groups, PGCB/PM pre‐gel was firmly bonded to the collagen fiber skeleton (Step III). At last, PGCB/PM pre‐gel encapsulated collagen fiber structure was treated with a mixture of water, glycerin, and electrolyte NaCl through a solvent replacement process. The final IECS was obtained when PGCB/PM pre‐gel was transformed into PGCB/PM organohydrogel containing NaCl (PGCB/PM/N), which demonstrated excellent mechanical, conductive, sensing, anti‐freezing and moisture‐retention properties (Step IV). Figure 1 Schematic illustration of the preparation process of multi‐functional IECS. Collagen fiber skeleton and PGCB/PM/N were primary components of IECS. The former served as a flexible substrate and the latter acted as conductive material, both of which worked cooperatively to improve the performance of final products. On the one hand, the structure of the collagen fiber skeleton was incredibly loose and ranged from nanoscale to macroscale. In the meantime, the abundance of pores and active groups that existed between collagen fiber skeleton offered the structural foundation for the fabrication of sensing material PGCB/PM/N. On the other hand, PVA and GEL, as flexible polymers, formed independent macroscopic structures and complete 3D skeletons between collagen fibers in the presence of cross‐linking and reinforcing agents. An interpenetrating network and a continuous conductive network entangled with collagen fiber bundles were established as a result of multiple hydrogen bonds between polymer chains and collagen fibers. In this process, the mutual cooperation of various components provided the possibility for the multifunctional integration of IECS. The details were as follows. (I) PVA with excellent mechanical properties was selected as the main network material. Meanwhile, GEL acted as the secondary network and interacted with PVA to produce interpenetrating polymer networks and increase flexibility. (II) CNFs, as reinforcing agents, interspersed between polymer chains under the action of hydrogen bonds, which are generated between the active groups on CNFs (hydroxyl and carboxyl) and polymer molecules (hydroxyl, carboxyl, amino, and amide). (III) Borax containing tetraborate ions was used as a cross‐linking agent to form dynamic borate ester bonds with the vicinal diol of PVA chains, further enhancing the mechanical properties and maintaining the integrity of network structures. At the same time, the migration of ionized Na + provided ionic conductivity. (IV) Most notably, IECS utilized free ions (Na + , Cl − ), and PMMA@MXene spheres as conductive materials. The former reached ionic conductivity by feasible transport channels with pore structure and water, while the latter accomplished electron transport by creating an intact conductive network. The combined impact of the two contributed to significant changes in ionic and electronic transport pathways under deformation, thus leading to dual conductive capability and super sensitivity to the strain of IECS. (V) By forming high‐density hydrogen bonds with water molecules, IECS based on the mixed solution of water and glycerin lowered the temperature at which ice crystals form and slowed the rate at which water molecules evaporate. The favorable environmental stability of IECS was guaranteed due to its anti‐freezing and moisture‐retention characteristics, which laid a foundation for its comprehensive application in harsh environments. 2.2 Microstructure Characterization of IECS As illustrated in Figure \n 2 \n , the morphologies of prepared MXene sheets and PMMA@MXene spheres were characterized to confirm the successful self‐assembly of MXene sheets and PMMA spheres. Figure 2a displayed the corresponding schematic illustration. The first step involved selectively etching of Al layer in MAX precursor by using HCl and LiF to generate MXene sheets with sufficient surface groups (such as hydroxyl, oxygen, and fluorine groups). The XRD patterns in Figure 2b allowed for the clear identification of the transition from MAX to MXene. The characteristic peaks at 34.0° disappeared corresponding to the crystalline planes (104) for MAX, while the characteristic peaks at 7.3° appeared corresponding to the crystalline planes (002) for MXene. Additionally, MXene was revealed to be typical 2D layered structures with thin and transparent layers by SEM and TEM images in Figure 2c and Figure S1a (Supporting Information). Based on SAED pattern in Figure S1b (Supporting Information), the typical hexagonal crystal structure of MXene sheets was identified. The obtained MXene sheets were then ultrasonically stripped for subsequent operations. Next, PMMA spheres and delaminating MXene sheets were used as cores and shells to create PMMA@MXene spheres by template method. Delaminating MXene sheets were forced to be tightly encapsulated on the surface of PMMA spheres through hydrogen bonding. Figure S2 (Supporting Information) illustrated that PMMA spheres had an average diameter of 2.01 µm. As a result of the ample coating of delaminating MXene sheets on PMMA spheres, the prepared PMMA@MXene spheres in Figure 2d had slightly larger average diameters of 2.03–2.07 µm and rougher surfaces. The TEM image, corresponding element mappings, and HRTEM image in Figure 2e–g further revealed that the perfect core@shell structure formed by tightly encasing multiple layers of delaminating MXene sheets surrounding the surface of PMMA spheres. F and Ti elements primarily originated from delaminating MXene sheets, whereas C and O elements were found in both delaminating MXene sheets and PMMA spheres. It was evident that the original spherical structure of PMMA templates was perfectly preserved and the delaminating MXene sheets were equally distributed on their surfaces. The XRD pattern of PMMA@MXene spheres in Figure 2b also demonstrated the simultaneous existence of characteristic peaks belonging to MXene sheets and PMMA spheres, which further confirmed the success of their recombination. Figure 2 a) Schematic illustration of the preparation process of PMMA@MXene spheres. b) XRD patterns of MAX precursor, MXene sheets, PMMA spheres, and PMMA@MXene spheres. SEM images of c) MXene sheets and d) PMMA@MXene spheres. e) TEM image, f) element mappings, and g) HRTEM image of PMMA@MXene spheres. PGCB/PM/N assisted in the assembly of PMMA@MXene spheres into IECS. Notably, the successful construction of IECS was made feasible by temperature‐induced sol‐gel transition characteristics, as shown in Figure S3 (Supporting Information). PGCB/PM pre‐gel solution was discovered to be coated at room temperature on the surface of numerous substrates with arbitrary shapes. Afterward, the temperature was adjusted to achieve phase transformation and generate the composites of PGCB/PM pre‐gel and substrates. Also, the subsequent solvent displacement process did not change their macroscopic structure. Therefore, the above established a foundation for the construction of IECS. The microscopic morphology and bonding structure of IECS were studied to clarify its hierarchical structure and formation mechanism. The collagen fiber skeleton was well preserved despite PGCB/PM/N filling up the interstitial spaces between fiber structures, according to SEM images in Figure 3a–f . Concurrently, the spherical structure of PMMA@MXene spheres was observed on the surface of collagen fiber bundles. Furthermore, the distribution of different elements on the cross‐section of IECS in Figure 3g demonstrated the uniform penetration and assembly of PGCB/PM/N within the collagen fiber skeleton, which set the platform for the formation of the dual conductive path and interpenetrating network structure in IECS. Figure 3 SEM images of a–c) collagen fiber skeleton and d–f) IECS: (a, d) surface, (b, e) cross‐section, and c, f) partially enlarged images. g) Element mappings of IECS corresponding to Figure \n 3 e . h) XRD patterns and i) FT‐IR spectra of collagen fiber skeleton and IECS. As demonstrated in Figure 3h,i , the chemical interaction in IECS was further investigated by employing XRD and FT‐IR. In Figure 3h , the collagen fiber skeleton had two distinct diffraction peaks at 8.8° and 20.1°. The former, which came from the comparatively regular structure of collagen fibers, meant the distance between molecular chains of collagen fibers. The latter represented the diffuse reflection of hierarchical structure, resulting from the structure of amorphous regions in collagen fibers. When combined with PGCB/PM/N, the diffraction peak for IECS at 20.1° became stronger, while the diffraction peak at 8.8° weakened and shifted to a lower angle direction. These disclosed that the filled PGCB/PM/N increased the distance between molecular chains of collagen fibers and enhanced the amorphous structure of collagen by destroying the original van der Waals forces and hydrogen bonds between collagen fibers. [ \n \n 25 \n \n ] At the same time, there was no diffraction peak of titanium oxide in IECS, which indicated that PMMA@MXene spheres did not oxidize during the preparation of IECS. This result was also illustrated by the basically constant conductivity of PMMA@MXene spheres before and after treatment in Figure S4 (Supporting Information), and the corresponding treatment process was detailed in Section 4.4 (Fabrication of IECS). Furthermore, Figure 3i showed that the collagen fiber skeleton and IECS had comparable positions of characteristic absorption peaks in amide I, II, and III bands, demonstrating that the introduction of PGCB/PM/N did not change the three‐stranded helical structure of collagen fibers. Also, the typical absorption peak of IECS in the amide A band was located at 3305 cm −1 . It was shifted toward a lower wavenumber compared with the collagen fiber skeleton, suggesting the generation of hydrogen bonds between collagen fibers and polymer networks. [ \n \n 25 \n \n ] To clarify, the original hierarchical structure of the collagen fiber skeleton was not altered after the functional modification with PGCB/PM/N. This made IECS exhibit similar characteristics to natural collagen fiber skeleton. 2.3 Mechanical and Conductive Properties of IECS A number of samples, including S, PS, ICS, and ECS, were used as comparison models in order to better understand the effect of each component on the overall performance of IECS. Where, S, PS, ICS, and ECS represented the collagen fiber skeleton, collagen fiber skeleton combing with PGCB, ion‐conducting collagen fiber‐based e‐skin, and electron‐conducting collagen fiber‐based e‐skin, respectively. And these samples are listed in Table S2 (Supporting Information). The stability and durability of IECS were given support by its mechanical characteristics. IECS was particularly flexible and deformable, as demonstrated in Figure \n 4 a , enabling mechanical manipulation such as repeated twisting. The stress and strain of S, PS, ICS, ECS, and IECS were further measured to demonstrate their mechanical properties in Figure 4b . It was discovered that IECS possessed better mechanical properties than the other four samples, which was attributed to the contributions of network structure and conductive materials in Figure 4c . First, the formation of interpenetrating network structure was one of the reasons for the enhanced mechanical property of PS compared to S. The uniform permeation and dynamic cross‐linking of organohydrogel in the interstitial space of collagen fiber skeleton made them interpenetrate each other, thus forming interpenetrating network structure. The deformation process of the interpenetrating network involved greater energy dissipation. This was the reason why PS with an interpenetrating network structure had superior mechanical properties compared to S with a single network structure. Meanwhile, multiple hydrogen bonding between the collagen fiber skeleton and organohydrogel was produced with the assistance of hydroxyl, carboxyl, amino, amide, and other active groups, which was also valuable for the energy dissipation of PS. Second, the mechanical properties of ICS and ECS were further improved by the addition of NaCl and PMMA@MXene spheres, respectively. On the one hand, the process of salting‐out, symbolized by NaCl, boosted macromolecular aggregation and interaction by reducing electrostatic repulsion between polymer chains. The resulting large number of junctions prompted more cross‐linking of polymer chains, which generated porous 3D structures and dense physical cross‐linked networks. Furthermore, Na + and Cl − ions also possessed a tendency to immobilize polymer chains because of dipole interactions, triggering the development of positively and negatively charged chain segments. The improvement in the mechanical properties of ICS compared to PS was the consequence of physical cross‐linking generated by the entanglement of these segments. On the other hand, PMMA@MXene spheres formed multiple hydrogen bonds with polymer network and collagen fiber skeleton with the assistance of hydroxyl, oxygen, and fluorine groups, as well as borate ester bonds with PVA molecular chains by means of tetraborate ions. The rigidity of PMMA@MXene spheres and the formed dynamic sacrifice bonds enabled the polymer network to evenly distribute stress during deformation, promoting the energy dissipation process. So the mechanical properties of ECS were superior to PS. Finally, IECS featuring interpenetrating network structure and dual conductive paths was constructed upon these foundations. The resulting stress and strain of IECS were 24.51 MPa and 71.82%, respectively, proving significant improvement in mechanical properties attributed to the synergistic effect of each component. Notably, the developed IECS had higher mechanical strength and lower stretchability than PGCB/PM/N without a collagen fiber skeleton in Figure S4 (Supporting Information). This was because the 3D woven structure of the collagen fiber skeleton carried and transferred stress more effectively. As a result, IECS based on collagen fiber skeleton was difficult to collapse and exhibited excellent mechanical strength. In terms of strain, the stretchability of the collagen fiber skeleton was limited by the tightly woven 3D structure. As soft materials, PGCB/PM/N around collagen fiber bundles improved the deformation ability of IECS to a certain extent. Thus, it was better than the original collagen fiber skeleton but was still inferior to PGCB/PM/N. Figure 4 a) Visualized photos of IECS treated by manually hanging of ≈990 g weight and twisting. b) Mechanical properties of S, PS, ICS, ECS, and IECS. c) Schematic illustration of interpenetrating network structure and multiple hydrogen bonding of IECS. d) Conductivity of IECS prepared with different dosages of conductive materials. e) Conductive performances of S, PS, ICS, ECS, and IECS. f) Visualized photos of IECS as part of a circuit making LEDs glow. g) Schematic illustration of different conductive mechanisms based on face‐to‐face and point‐to‐point contacts. In addition to mechanical properties, good electrical conductivity was also crucial for IECS applications in the field of soft sensing, which was closely related to the dosage of conductive materials. From Figure 4d , it was found that the conductivity of IECS increased from 1.68 to 4.81 S m −1 with the increase of NaCl content. The reason for this was that the ion movement in IECS was enhanced and more ion migration channels were provided. Subsequently, the changes in conductivity tended to level off due to the restriction of the dense polymer network triggered by salting out. By changing the amount of PMMA@MXene spheres, the conductivity of IECS was further adjusted to 15.31 S m −1 . The reason was that the separated PMMA@MXene spheres gradually connected with each other and formed a complete electronic conductive path with the increase of PMMA@MXene sphere's dosage. However, the excess PMMA@MXene spheres made no contribution to the formation of a conductive pathway. After consideration, IECS with conductivity of 14.82 S m −1 was taken as the optimal sample for subsequent experiments, in which the dosages of NaCl and PMMA@MXene spheres were 2.0 mol L −1 and 3.0 wt.%, respectively. For comparison, the conductivity of S, PS, ICS, ECS, and IECS is shown in Figure 4e . The conductivity of S and PS was only 0.16 and 0.20 S m −1 in the absence of NaCl and PMMA@MXene spheres, respectively, which was due to the residual free ions of the tanning process and cross‐linking agent. With the addition of NaCl and PMMA@MXene spheres, the corresponding conductivity of ICS and ECS significantly increased to 4.79 and 10.02 S m −1 , respectively, which was closely related to the high concentration of free ion migration and the complete physical contact of electron conductive network. On these bases, the conductivity of IECS with dual‐mode conduction induced by NaCl and PMMA@MXene spheres was increased to 14.82 S m −1 . To further show its conductive capability, Figure 4f proves how IECS was utilized as a conductor to light LEDs. Remarkably, the unique microstructures of PMMA@MXene spheres inspired additional research on the conductive characteristics of IECS. By using the sample prepared with traditional MXene sheets as the control object (referred to as IECS‐MS), the relationship between the morphology of the electrically conductive network and the performance of the corresponding e‐skin was further clarified. As an electrically conductive network was essentially formed, the amount of PMMA@MXene spheres and traditional MXene sheets was 3.0 and 1.0 wt.%, respectively, according to the results in Figure 4d and Figure S5 (Supporting Information). For IECS‐MS, traditional MXene sheets were presented in a 2D form and electrically contacted by face‐to‐face stacking, [ \n \n 26 \n \n ] as illustrated in Figure 4g . For IECS, the mass ratio of PMMA spheres to MXene sheets was 10:1 during the preparation process of PMMA@MXene spheres, meaning that 3.0 wt.% of PMMA@MXene spheres was made up of roughly 2.73 wt.% of PMMA spheres and 0.27 wt% of MXene sheets. Compared to 1.0 wt.% of traditional MXene sheets, the actual amount of conductive materials in PMMA@MXene spheres was only ≈0.27 wt.% when an electrically conductive network was essentially formed. It demonstrated that PMMA spheres served as essential templates for isolating structures during the development of electrically conductive networks. The interaction between PMMA@MXene spheres was point‐to‐point contact, as shown in Figure 4g , which made it easier to produce a 3D conductive network and lower percolation threshold. Furthermore, in terms of sensing, PMMA@MXene spheres with point‐to‐point mode were more likely to separate under tensile strain, [ \n \n 27 \n \n ] resulting in a more pronounced change in the relative resistance of IECS. In addition, 0D materials such as silver nanoparticles (AgNPs) also had excellent electrical conductivity. Their disordered migration under external forces was conducive to the destruction and reconstruction of conductive networks, resulting in significant changes in the resistance of the sensor. However, the low aspect ratio of 0D materials made the threshold of conductive percolation usually high, which meant that a complete conductive network could only be formed in the polymer at a larger dosage. It was not conducive to the improvement of mechanical properties and the reduction of costs of the sensor. Therefore, PMMA@MXene spheres with spherical structures were more conducive to the improved overall performance of IECS. 2.4 Electronic Behavior and Sensing Mechanism of IECS Strain sensing behavior is an inherent property of smart materials, which makes them possible as strain sensors. [ \n \n 28 \n , \n 29 \n \n ] Various theoretical models have been developed to explain the mechanisms of self‐perception. [ \n \n 30 \n , \n 31 \n , \n 32 \n \n ] In general, the shape of conductive networks and the distance between conductive materials affect ion migrations and electronic transitions, resulting in changes in the output electrical signal of smart materials. [ \n \n 33 \n \n ] The above analysis showed that the dual conductive network proposed in this work formed a large number of conductive pathways in IECS. Meanwhile, the presence of NaCl‐induced free ions and PMMA@MXene spheres‐induced conductive sphere arrays provided different sources for signal variation. Therefore, the unique conductive network structures had a significant impact on the strain‐sensing behavior of IECS. As a conductor, IECS was connected to an electric circuit. And the performance of IECS in sensing for bending strain was visually evaluated by observing the change of LEDs brightness with bending angle. As illustrated in Figure \n 5 a , the brightness of LEDs connected to an electric circuit gradually decreased with an increase in bending angle. Then, LEDs instantly brightened as the bending strain decreased, which signified a quick resistance response of IECS to bending strain. Importantly, it was intuitively observed that IECS still had good conductivity during the entire bending process, indicating that the conductive network of IECS was stable even under a large bending angle. Sensitivity (GF), one of the most important parameters to evaluate sensing performance, was assessed by displaying relative resistance change (ΔR/R 0 ) under bending. This was determined using the formula of GF = (ΔR/R 0 )/ɛ. Where ΔR represented the resistance change before and after bending. R 0 represented the resistance without bending and ɛ represented the applied strain to IECS. The relationship between bending angle and strain was defined and quantified in terms of ε = ±h/2r and c = 2rsin(l/2r). Of them, the thickness (h), arc length (l), chord length (c), and radius of curvature (r) of IECS were associated with bending strain. Figure 5b illustrates the ΔR/R 0 curve of IECS at different bending strains. The ΔR/R 0 value of IECS increased with the increase of bending strain. This was because the deformation of IECS resulted in increased length and decreased cross‐sectional area, which hindered the migration process of ions and electrons and significantly increased its resistance. Furthermore, the sensing process of IECS was fitted and divided into three stages. When the bending strain of IECS was 0–1.0, 1.0%–7.5%, and 7.5–10.0%, the corresponding GF was 16.64, 10.74, and 2.39, respectively, demonstrating exceptional sensitivity to bending strain under the synergistic action of the dual conductive path and collagen fiber skeleton. For comparison, the ΔR/R 0 curves of ICS, ECS, and IECS‐MS at various bending strains were measured and presented in Figure S6 (Supporting Information). Significant differences were observed between different samples, illustrating the function of dual conductive paths and the construction of strain‐sensitive conductive networks derived from PMMA@MXene spheres in IECS. These were the main reasons affecting the change of conductive network structure and the sensing behavior of IECS under bending strain in terms of conductive materials. In addition, PGCB/PM/N and the sample of PGCB/PM/N formed only on the surface of the collagen fiber skeleton (denoted as IECS‐S) were subjected to the same sensing performance tests in Figure S7 (Supporting Information). As the results, both PGCB/PM/N and IECS‐S had lower ΔR/R 0 values than that of IECS under the same bending strain, which suggested the importance of the interpenetrating network structure between collagen fiber skeleton and PGCB/PM/N. The change in network structure and the process of signal transmission were facilitated by the sliding and separation of collagen fibers. Therefore, the distinct 3D network and hierarchical structure of the collagen fiber skeleton, as well as the interpenetrating network structure between the collagen fiber skeleton and PGCB/PM/N, were helpful in augmenting the structural alterations of the conductive network in IECS. Figure 5 a) Change in LEDs brightness with the strain of IECS connected in the electric circuit. b) Relative resistance variation of IECS at different strains. c) Schematic illustration of the strain deformation of IECS. d–f) Relative resistance variation of IECS at small strains (<1%), large strains (1–10%), and strain frequencies. g) Relative resistance change of IECS for 1000 loading‐unloading cycles at 6.7% strain. Based on the above analysis and results, the sensing mechanism of IECS was proposed in Figure 5c . The dual conductive networks of ions and electrons dominated the strain response behavior of IECS. The former was directly correlated with the degree of deformation, whereas the latter relied on the actual physical contact of PMMA@MXene spheres. Regarding ion‐induced sensing behavior, the electric field of the initial state caused Na + and Cl − in IECS to move toward negative and positive poles, respectively. During this period, the resistance of IECS was associated with its inherent characteristics, specifically the internal resistance of ionic movement, which has been explained in detail in our previous work. [ \n \n 34 \n \n ] The longer ion migration path and the narrower ion transport channel allowed IECS to have greater resistance as it strained, thus converting mechanical deformation into an electrical signal. To achieve electron‐induced sensing behavior, the 3D conductive pathway of IECS was created by connecting MXene sheets on the surface of PMMA@MXene spheres to one another. When no strain was applied, PMMA@MXene spheres in IECS were tightly assembled without cracks, and the inside electrons easily passed through. During the stretching process, the cracks usually developed and spread in the stress‐concentrated region of rigid PMMA@MXene spheres because PMMA@MXene spheres had higher Young's modulus and smaller elongation at break than collagen fiber skeleton and elastic polymer network. At this point, the resistance change was directly correlated with the formation, propagation, and disconnection of cracks between PMMA@MXene spheres. [ \n \n 35 \n \n ] A tiny strain on IECS triggered overlapping PMMA@MXene spheres to slide and eventually separate, creating a few small fractures that ran perpendicular to the direction of strain. In the meantime, PMMA@MXene spheres started to aggregate together due to the cracks, and the interconnected island‐like structures that sustained partial electron transport were formed. During this time, the contact area between PMMA@MXene spheres decreased, resulting in the narrowing of the conductive channel and the increase of resistance in IECS (Stage I). As strain increased, the cracks gradually widened and mainly concentrated in the region of the gap. When the island‐like PMMA@MXene spheres split from each other, the formed bridge‐like structure further slowed the electron transport of IECS (Stage II). With a large increase of strain, PMMA@MXene spheres with bridge‐like structures elongated and fractured, producing multiple wide penetrating cracks. Due to the complete separation of some bridge‐like structures from each other, the conduction process of electrons was further restricted and the resistance of IECS was increased (Stage III). Finally, when all the connections between PMMA@MXene spheres were severed, the resistance of IECS reached infinity and the corresponding working range of the electron‐conducting network reached its limit. Therefore, IECS used for ultra‐sensitive strain detection in the whole range was feasible due to the mechanism of the formation, propagation, and disconnection of cracks between PMMA@MXene spheres. [ \n \n 36 \n \n ] \n During the entire sensing procedure, the array structures formed by PMMA@MXene spheres concentrated external force on the contact points of the micro‐region in IECS, amplifying the structural alterations of an electron‐conducting network under strain. Meanwhile, the ability of IECS to detect strain was further enhanced by the hierarchical structure of collagen fibers and their interpenetrating network structure with polymer chains of PGCB/PM/N. In conclusion, IECS demonstrated a more significant ΔR/R 0 value and superior strain sensing ability at all stages owing to the dual structural changes of conductive networks with both ions and electrons and the deformation amplification effect induced by collagen fiber skeleton. Thus, the strategic synergy of the compositional and structural design of IECS made it a very promising force‐sensing part of an excellent piezoresistive sensor. Additionally, as demonstrated in Figure 5d–f , the ΔR/R 0 curves of IECS displayed continuous and stable signals without any discernible increase or decrease under dynamic bending tests at various small angles, large angles, and bending frequencies. The nearly constant curves showed that IECS was suitable for low hysteresis and good signal output stability when detecting bending deformations, indicating that it was versatile and could be used for a wide range of tasks. The durability of IECS was also evaluated in Figure 5g for widespread and practical application. After 1000 bending cycles, the electrical signals of IECS showed good amplitude and waveform without fluctuation, displaying exceptional durability and reliability to guarantee its long‐term availability. Besides, considering batch stability, the ΔR/R 0 curves of IECS from different batches were tested and shown in Figure S8 (Supporting Information). As a result, the corresponding ΔR/R 0 curves exhibited stable signals with only slight changes, which proved good batch stability of IECS. 2.5 Freezing Tolerance and Moisture‐Retention of IECS The key requirement for the practical use of IECS as a variety of sensor devices is good environmental stability, which includes anti‐freezing properties at low temperatures and long‐term moisture retention. Therefore, the existence state of water molecules in IECS was crucial, which was closely related to the water/glycerin binary solvent system, inorganic salt NaCl, and collagen fiber skeleton. Considering that the NaCl‐induced salting‐out effect was essential in sample preparation, its influence on performance has been discussed in our previous work. [ \n \n 34 \n \n ] Herein, PGCB/PM/N and IECS‐H 2 O were used as control samples to demonstrate the role of the solvent system and collagen fiber skeleton, especially the latter. To examine the anti‐freezing property of IECS at low temperatures, IECS‐H 2 O (without glycerin), PGCB/PM/N, and IECS were exposed to a low‐temperature environment (−40 °C) for 7 days. The flexibility of three samples was compared visually. IECS‐H 2 O hardened and froze, losing its initial flexibility. However, PGCB/PM/N and IECS were still able to withstand significant bending action, as seen in Figure \n 6 a . It was initially demonstrated that PGCB/PM/N and IECS performed well against freezing at low temperatures. In addition, DSC technology was used to confirm the freezing resistance of samples in Figure 6b . The freezing point of water molecules was significantly lowered by glycerin and collagen fiber skeleton, as demonstrated that IECS‐H 2 O, PGCB/PM/N, and IECS showed clear crystallization peaks at −21.0 °C, −40.3 °C and −42.8 °C, respectively. Figure 6g explained the reason for the anti‐freezing property of IECS. The hydroxyl group of glycerin and the hydroxyl, carboxyl, amino, and amide groups of collagen fibers formed high‐density hydrogen bonds with water molecules, thus preventing the formation of ice crystals. Despite the presence of collagen fibers in IECS, the freezing points of PCGB/PM/N and IECS showed minimal difference. This was because the physical structure and chemical composition of collagen fibers were three helices and collagen, respectively, and their molecules had strong interaction forces. In low low‐temperature environment, collagen fibers still showed high structural stability, so the flexibility of IECS might be affected rather than completely lost. In the meantime, it was clear from Figure 6c that the conductivity of IECS fluctuated only slightly over the temperature range of −40 °C to 20 °C, showing good temperature stability. These results made it possible for the application of IECS in low‐temperature environments. Figure 6 a) Visual comparison and b) DSC curves of IECS‐H 2 O, PGCB/PM/N and IECS. c) Conductivity of IECS at different temperatures. d) Visual comparison and e) moisture‐retention curves of IECS‐H 2 O, PGCB/PM/N, and IECS after placing for 7 days at 20 °C and 50% RH. f) Conductivity of IECS at different storage times. g) Schematic illustration of free water and bonded water of IECS. h) Moisture‐retention of IECS after vacuum drying at 60 °C for 8 h and regeneration at 20 °C and 50% RH for 16 h. i) Moisture‐retention and conductivity of IECS after 5 repeated drying and regeneration cycles. j) The comparison of comprehensive evaluation of IECS and other reported collagen fiber based e‐skins by conductive hydrogel. By comparing the water evaporation of IECS‐H 2 O, PGCB/PM/N, and IECS under the same conditions, their moisture‐retention capacity were assessed. The moisture‐retention of IECS‐H 2 O, PGCB/PM/N, and IECS after 7 days of storage at 20 °C and 50% RH was 39.8%, 57.6% and 69.7%, respectively, as shown in Figure 6d,e . The reason for the enhanced moisture‐retention performance of IECS was similar to its anti‐freezing mechanism. The moisture‐retention additive glycerin and collagen fiber skeleton formed strong hydrogen bonds with water molecules, lowering the saturated vapor pressure of water, as shown in Figure 6g . After being stored for 7 days, IECS continued to exhibit a high conductivity of roughly 13.50 S m −1 , which remained almost unchanged from its initial state in Figure 6f . Even with the addition of glycerin and collagen fiber skeleton, IECS might still dry out in harsh conditions. However, nearby water molecules could be effectively absorbed owing to the high hygroscopicity of glycerin. This led to the self‐regeneration capacity of IECS in Figure 6h . IECS underwent the pre‐treatment of vacuum drying for 8 h at 60 °C. And then, the dehydrated IECS regained 75.5% of its initial weight after 16 h at 20 °C and 50% RH. Furthermore, Figure 6i illustrated that IECS had good environmental adaptability even after 5 repeat cycles, demonstrating a similar self‐regeneration performance. The excellent anti‐freezing, moisture‐retention, and self‐regeneration capabilities of IECS were suitable for long‐term use in practical applications. In addition, IECS was compared to the previously reported functional collagen fiber‐based e‐skin in terms of mechanical, conductive, moisture‐retention, anti‐freezing, and sensing performances to illustrate its potential advantages in future intelligent flexible sensors, [ \n \n 10 \n , \n 13 \n , \n 16 \n , \n 18 \n , \n 21 \n , \n 36 \n , \n 37 \n , \n 38 \n , \n 39 \n , \n 40 \n , \n 41 \n , \n 42 \n , \n 43 \n , \n 44 \n , \n 45 \n \n ] as shown in Figure 6j and Table S1 (Supporting Information). It demonstrated that IECS offered an abundance of multifunctional features and smart sensing. 2.6 Real‐life Demonstration as Wearable Sensor of IECS Based on the aforementioned characteristics, IECS was employed as a motion sensor for the detection of human physiological signals in Figure \n 7 \n . Strain‐dependent resistive signals were generated by IECS for movements, related to various body parts, including joint bending (fingers, wrists, elbows, knees, etc.), throat activity (pronouncing, swallowing, etc.), and facial expression (smiling, frowning, etc.). The obtained multiple signals showed consistency, reliability, and repeatability under various deformations, as evidenced by the somatosensory monitoring data collected from various body parts. In particular, IECS was attached to joints, such as the fingers, wrists, elbows, and knees, to record resistance signals and monitor joint movement. Tensile action initiated the ΔR/R 0 value of IECS to increase when the joint of the human bent. Following the straightening of the joint, the ΔR/R 0 value gradually reverted to its initial state. Furthermore, the ΔR/R 0 value of IECS varied significantly with the bending angle of the joint. These phenomena were also observed in the resistance signals of other joint movements, which were explained by the piezoresistive effect. Under the action of applied strain, the physical deformation of IECS limited the transport path of ions and electrons, resulting in a change of resistance. Additionally, different response electrical signals were generated by IECS due to the movements of muscles and epidermis, when the throat was swallowing or pronouncing. The contraction and relaxation of facial muscles also elicited changes in response to electrical signals when the face displayed different expressions, such as frowning or smiling. Furthermore, substantial repeatability was indicated by the stable ΔR/R 0 curves of IECS under continuous joint bending, throat activity, and facial expression. Therefore, IECS was an advantageous device for monitoring human movement in real‐time, and it created new possibilities for the applications of collagen fiber‐based e‐skin in artificial intelligence, wearable electronics, and human‐machine interfaces. Figure 7 Real‐time monitoring of human movements by IECS and schematic diagram of sensing sites. Signals of relative resistance during vigorous activities (finger bending, wrist bending, elbow bending, and knee bending) and subtle motions (frowning, facial expressions such as smiling, swallowing, and speaking different phrases). Importantly, the wrist pulse is the key physiological signal to determine arterial blood pressure and heart rate. IECS was integrated with signal processors, conversion circuits, communication circuits, and filters to monitor pulse signals by attaching them to the wrist of the volunteer in Figure 8a . When IECS was subjected to the pressure induced by pulse, its resistivity changed, which resulted in the change of output and the acquisition of signal. By placing the arm of the volunteer in a homemade device, pulse signals were recorded by IECS at different temperatures, as shown in Figure S10 (Supporting Information). Under different temperatures (−20 °C, 25 °C, and 40 °C) and states (statics and motion), Figure 8b–d showed a typical radial artery pulse waveform, where clearly distinguishable peaks and a late augmentation shoulder were observed in one cycle. Due to the excellent environmental stability of IECS, these data remained stable at the test temperatures. Moreover, the state of the human body was closely related to pulse rate. So the strength of motion state was judged from Figure 8c . These above results held the promise of IECS for more detailed health diagnoses. Additionally, Morse code was known to be an effective, real‐time, and time‐honored way of communication in coded languages. For coded languages, different combinations of “dots” and “dashes” represented different letters in the alphabet, as shown in Figure 8e . In this work, electrical signals caused by finger pressure were converted to Morse code. Specifically, sharp peaks produced by short‐term stressing were designated as the meaning “dots”, while peaks with the plateau produced by longer‐term stressing were defined as being in the “dashe” state. Therefore, IECS was employed to build a mechanoresponsive sensor for information transfer. According to the Morse code table, the combination of the letters “HELP” and “SOS” was successfully detected. To further explore the practical sensing applications of IECS, a fully integrated array of wearable sensors was assembled as artificial electronic skin and was applied to detect spatial pressure distribution. In this case, 25 IECS were assembled into a sensing array and interconnected by copper wires. Each IECS acted as a sensor unit. When external pressure was applied to the surface of the sensor unit, the contact position was recognized immediately and the corresponding electrical signal was detected. Thus, the pressure distribution in the space was recorded. As shown in Figure 8g , when the sensor unit was pressed under glyphs “S”, “U”, “S” and “T”, different electrical signals were output. The results showed that the flexible sensor array constructed by IECS had great potential for application in the field of e‐skin and human‐computer interaction. Figure 8 Schematic diagram of pulse pressure by IECS in real‐time monitoring. Signals of relative resistance under different b) temperatures (−20 °C, 25 °C, and 40 °C) and c) states (statics and motion). d) The corresponding detailed pulse waveform in Figure \n 8 b,c . e) Morse code table. f) Morse code to type “HELP” and “SOS” by finger pressure on IECS. g) Schematic diagram of 5 × 5‐pixel sensor array composed by IECS under the applied force of “S”, “U”, “S” and “T” and the corresponding signal distribution."
} | 13,766 |
23940594 | PMC3734296 | pmc | 3,848 | {
"abstract": "Competition is a major driving force in freshwaters, especially given the cyclic nature and dynamics of pelagic food webs. Competition is especially important in the initial species assortment during colonization and re-colonization events, which depends strongly on the environmental context. Subtle changes, such as saline intrusion, may disrupt competitive relationships and, thus, influence community composition. Bearing this in mind, our objective was to assess whether low salinity levels (using NaCl as a proxy) alter the competitive outcome (measured as the rate of population biomass increase) of Daphnia-Simocephalus experimental microcosms, taking into account interactions with priority effects (sequential species arrival order). With this approach, we aimed to experimentally demonstrate a putative mechanism of differential species sorting in brackish environments or in freshwaters facing secondary salinization. Experiments considered three salinity levels, regarding NaCl added (0.00, 0.75 and 1.50 g L −1 ), crossed with three competition scenarios (no priority, priority of Daphnia over Simocephalus , and vice-versa). At lower NaCl concentrations (0.00 and 0.75 g L −1 ), Daphnia was a significantly superior competitor, irrespective of the species inoculation order, suggesting negligible priority effects. However, the strong decrease in Daphnia population growth at 1.50 g L −1 alleviated the competitive pressure on Simocephalus , causing an inversion of the competitive outcome in favour of Simocephalus . The intensity of this inversion depended on the competition scenario. This salinity-mediated disruption of the competitive outcome demonstrates that subtle environmental changes produce indirect effects in key ecological mechanisms, thus altering community composition, which may lead to serious implications in terms of ecosystem functioning (e.g. lake regime shifts due to reduced grazing) and biodiversity.",
"introduction": "Introduction Competition is a major driving force in freshwater systems, especially given the cyclic nature and dynamics of planktonic food webs [1] , [2] . While intra-specific competition is important in defining equilibrium in population dynamics, inter-specific competition tends to be destabilizing, causing ecological exclusion of one or the other competitor(s) [3] , [4] . Inter-specific competition generally translates into the mutual inhibition of growth rate among populations of different species that have common requirements for shared and limiting resources. Competition between populations of freshwater cladocerans can be responsible for shifts in competitor’s life-history, in terms of density, growth, juvenile survival and clutch-size [5] , leading to a co-existence scenario with different demographic cycles [5] , [6] . Regulation of cladoceran community structure is modulated by colonization and re-colonization events from the ephippial egg bank [7] , [8] . Competition is especially important in the initial species assortment [2] , [9] , [10] , which depends strongly on the initial species and gene pool (producing so-called founder effects [7] , [11] ), as well as the environmental conditions of the system. Under such scenarios, the order at which species appear in the system may configure priority effects, in which species that appear first have a competitive advantage over latecomers [8] , [12] . Priority effects are defined as the impact that a particular species can have on community development due to prior arrival (or hatching) at a site, and they usually result from resource and niche monopolization of early colonizers [8] , [13] . The environmental context is known to impact the strength of priority effects or even superimpose them (e.g. [8] ). Previous experiments with Daphnia \n [14] , [15] have shown that the environmental context influences the competitive outcome. Louette and De Meester [8] showed that predation may alter the competitive outcome of inter-specific relationships. Using plants as experimental subjects, several authors have shown that competitive ability or intensity is alleviated under environmental stress (e.g. [16] ). Also, Emery et al. [17] demonstrated that stress tolerators were consistently dominant competitors under some circumstances. The reasonable conclusion is that environmental stress, either abiogenic or biogenic, may alter radically the expected outcome of species sorting, a key process in the population dynamics of freshwater cladoceran populations. Salinity is an abiotic environmental stressor that can radically alter freshwater community structure (e.g. [18] , [19] ). In zooplankton, such community changes can occur at low salinity levels [20] , [21] , [22] , [23] . Salinization of freshwaters, which is a predicted consequence of global climate change and groundwater overexploitation [24] , represents serious implications for ecosystem functioning. For example, lake regime shifts from clear to turbid water may occur in brackish lakes [25] due to removal of large herbivores (either eliminated directly by salinity or via altered fish community composition – see [25] ). At lethal salinity levels (>2), sensitive species are purely eliminated or are unable to hatch. At lower levels, however, salinity could disrupt competitive relationships, with brackish conditions favouring different species composition than in freshwater conditions. So far, there is no experimental evidence for this in the literature. Bearing this in mind, our objective was to assess whether low salinity levels (using NaCl as a proxy) alter the competitive outcome of a Daphnia-Simocephalus experimental system, taking into account interactions with priority effects (sequential species arrival order). It is expected that salinity alters the competitive outcome of inter-specific relationships, provided that there are slight differences in halotolerance between competitor species; however, it is hypothesized that priority effects (inoculation order of the competitor species) may confer some protection to the less halotolerant species. With this approach, we aim to experimentally confirm the hypothesized mechanism of differential species sorting in brackish environments or in freshwaters facing secondary salinization.",
"discussion": "Discussion This experimental study demonstrated a salinity-mediated disruption of the competitive outcome in Daphnia - Simocephalus microcosms. A shift between a Daphnia -dominated and a Simocephalus -dominated community occurred along the NaCl gradient. A similar result was found in experimental zooplankton communities when an invertebrate predator was introduced [8] . Subtle environmental changes, such as low levels of salinity, produce indirect effects in key ecological mechanisms, namely species sorting. Thus, our results support the hypothesized mechanism of differential species sorting within zooplankton communities in brackish environments or in freshwaters facing secondary salinization. Also, this study demonstrates that salinity, even at low levels, was much stronger than priority effects, which – in this case – were negligible because either Daphnia or Simocephalus were superior competitors, depending on the NaCl concentration. Up to 0.75 g L −1 , Daphnia demonstrated to be a superior competitor, independently of the order of inoculation (see Figures 2 and 3 ). Although the order of inoculation contributed to the overall variation in the rate of population biomass increase (see Table 1 and Figure 3 ), priority effects were negligible: Daphnia always grew best. This was also the case when S. vetulus competed with two other Daphnia species [8] . However, the order of inoculation was important for the inferior competitor, Simocephalus , whose populations grew worse when Daphnia was the early colonizer and grew best when it ( Simocephalus ) arrived earlier (see Figures 2 and 3 ). This demonstrates priority effects, with the competitive pressure on Simocephalus being higher when its competitor arrived earlier in the communities. Nevertheless, even at its highest growth rate, Simocephalus was never a match for Daphnia at low salinity (<1.5 g L −1 ). Therefore, we can consider that priority effects were negligible, as they did not translate into contrasting or long-lasting differences in species dominance. Although biomass data suggest such contrasting differences, this was merely a product of the short duration of the experiment, as shown by the rate of population biomass increase (compare Figs. 1 and 2 ), which is a more suitable estimate of competitive outcome (see Statistical analyses). The superior competitor ability of Daphnia could be probably due to the successful establishment of its population through a rapid monopolization of resources [8] , [38] , ability to explore low levels of food [7] , [39] , [40] , and superior filtration rate relatively to Simocephalus \n [8] , [41] . Despite this, competitive exclusion [3] , [4] of Simocephalus was not observed here in any scenario. However, the experiment was of relatively short duration and food levels were not very limiting (see [31] ). Also, competition is not a force as radical as predation (see [8] ), which implies that active removal of individuals from populations occurs. Priority effect in these two species in the field could occur due to earlier arrival of propagules (ephippia), which depend strongly on dispersal vectors (such as aquatic birds) [8] , [9] , [11] . While this is true for temporary ponds [8] (also in amphibians, e.g. [42] ), it is not the case of lakes and reservoirs, which usually contain a large ephippial pool in the sediments [9] . In this case, priority effects could occur by differences in hatching time or in numbers ( Daphnia typically produces two resting eggs per ephippium, while Simocephalus only produces one [38] ). While both populations’ growth rate decreased with increasing NaCl concentration, D. galeata growth was much more affected at 1.5 g L −1 , and this alleviated S. vetulus from the pressure of a superior competitor. Consequently, priority effects were nullified, and Simocephalus experimental populations grew equally well in all species inoculation order scenarios at 1.5 g L −1 . The superior competitive ability of Simocephalus at 1.5 g L −1 may have resulted from its higher chronic halotolerance relatively to Daphnia . Preliminary laboratorial tests (unpublished data) showed that the two taxa have similar acute EC 50 values for NaCl –2.81 g L −1 (95% CI: 2.65–2.99 g L −1 ) for S. vetulus and 2.88 g L −1 (2.73–3.05 g L −1 ) for D. galeata – but the reproductive EC 50 for the S. vetulus clone was slightly higher than the D. galeata clone used in the experiments –1.28 g L −1 (95% CI: 1.22–1.33 g L −1 ) and 0.71 g L −1 (0.64–0.77 g L −1 ), respectively. Thus, our results do not support the hypothesis that priority effects confer some protection to the less tolerant species (in this case, Daphnia ). Similarly, a study with Microcystis populations in the presence of grazers also showed no protective effect of inoculation order in grazer-unprotected strains [12] . We conclude that the shift from a Daphnia -dominated (0.0 g L −1 ) to a Simocephalus -dominated assemblage (1.5 g L −1 ) was apparently mediated by their NaCl tolerance, resulting in depressed Daphnia growth at 1.50 g L −1 and consequent alleviation of competition pressure over Simocephalus , as seen by the lack of an effect of the inoculation order unlike in ≤0.75 g L −1 scenarios. These evidences support the theory that the competition between species can be alleviated under environmental stress [16] , favouring the inferior competitor or species, even if it arrives later to the community [6] , [8] , [12] . Consequently, as competitive strength is reduced, decreased impact of priority effects occurs in the presence of a stressor, such as predation [8] , limiting food resources [6] , or pesticides [42] . Similarly to our study, Louette and De Meester [8] showed that predation was responsible for an inversion of the dominant taxon in experimental communities. Although not as radical as predation (which lead to extinction of some species and hence negative growth rates in [8] ), low salinity levels (1.5 g L −1 ) inverted the competitive outcome in the Daphnia-Simocephalus experimental system. The salinity levels at which this occurred are in line with the predictions for community shifts of Schallenberg et al. [23] , as well as with the NaCl concentrations that elicit reproductive impairment in Daphnia \n [43] , [44] . This study indicates that the higher halotolerance of certain genotypes/taxa could contribute to their success in disturbed communities, being important in the dynamics of species succession in a progressive scenario of freshwater salinization. In brackish lakes, large filter-feeding herbivores (especially Daphnia spp.) tend to be eliminated [23] , [45] ; consequently, smaller or more tolerant species dominate [6] , [17] but their filtration efficiency is inferior, leading to lake regime shifts from clear to turbid water [25] . This rationale is applied here in the context of fish-populated lakes and reservoirs; it may not be true in brackish fishless ponds, where large-bodied Daphnia species that tolerate intermediate salinities occur (e.g. D. magna ; see discussion in [44] , [46] ). Although Simocephalus is a large cladoceran, it is usually restricted to littoral environments and has a sessile behaviour [38] . Consequently, its filtration rate at whole-lake scale may not be efficient in controlling phytoplankton growth [8] , [41] . Large cladoceran species, and particularly Daphnia , play a key role [25] , [47] in the regulation of primary production in freshwater ecosystems (PEG model; [2] ), because of their efficient algal filtration [39] , [40] , [47] . So, if the competitive ability of Daphnia species is compromised by external factors, such as demonstrated here for salinity, the dynamic of species succession could be modified, and the ecosystem services provided by these grazers (regulation of biogenic turbidity and prevention of cyanobacterial blooms, as well as nutrient cycling) would be nullified. We must recognize that, in a scenario of moderate to intense salinization, the levels of salinity used in this study are not ecologically relevant. However, saline intrusion may elicit a progressive scenario, particularly in coastal lakes [23] , [26] , [48] . In these systems, small increases in salinity may occur due to intermittent inputs of seawater [23] , [26] , [49] , but also via saline intrusion in groundwater, as the result of the conjugation of extended droughts [24] , [50] and overexploitation of aquifers [24] , [51] . Freshwater inland lakes can also suffer from salinization as result of extended drought and enhanced evaporation, especially in arid and semi-arid areas [46] . Thus, subtle or progressive changes in salinity may occur in freshwater systems, especially under a changing climate, and the potential impacts of small increases in salinity on biodiversity and trophic structure might be stronger than those of increased temperature per se [45] . This clearly justifies the need to assess the ecological consequences of such subtle changes in the resident assemblages, namely zooplankton, whose community structure is predicted to be highly sensitive to salinization (see [20] , [21] , [23] , [49] ). Although our experiments used a simplistic experimental design, they demonstrated that gradual salinization of freshwater may alter competitive interactions in freshwater zooplankton, thus affecting the initial assemblage structure in colonization or re-colonization events. This occurs via interference with species sorting and priority effects. Other studies have also shown that the structure of communities reflects the environmental conditions in the moment of species sorting [6] , [42] , [52] ; also, the environmental context is equally important in defining the community sensitivity to other stressors (e.g. pesticides) [53] . Future studies should therefore address the capacity of NaCl-altered zooplankton communities to cope with other stressors, as this could potentially compromise water quality (transparency, cyanobacterial blooms) and ecosystem functioning (e.g. primary productivity, nutrient cycling). Indeed, Wittebolle et al. [54] have shown that the initial assemblage structure is a key factor in preserving the resistance to environmental stress and functional stability of an ecosystem."
} | 4,183 |
30034320 | PMC6043813 | pmc | 3,849 | {
"abstract": "The structural organization of cortical areas is not random, with topographic maps commonplace in sensory processing centers. This topographical organization allows optimal wiring between neurons, multimodal sensory integration, and performs input dimensionality reduction. In this work, a model of topographic map formation is implemented on the SpiNNaker neuromorphic platform, running in realtime using point neurons, and making use of both synaptic rewiring and spike-timing dependent plasticity (STDP). In agreement with Bamford et al. ( 2010 ), we demonstrate that synaptic rewiring refines an initially rough topographic map over and beyond the ability of STDP, and that input selectivity learnt through STDP is embedded into the network connectivity through rewiring. Moreover, we show the presented model can be used to generate topographic maps between layers of neurons with minimal initial connectivity, and stabilize mappings which would otherwise be unstable through the inclusion of lateral inhibition.",
"introduction": "1. Introduction Ramón y Cajal postulated that: “In the adult centers, the nerve paths are something fixed, ended, and immutable. Everything may die, nothing may be regenerated” (Ramón y Cajal, 1928 ). We now know that not to be the case. Mammalian brains change their connectivity from early development and throughout adulthood. Perinatally, the neuromuscular junction sees neural competition for the innervation of muscle fibres resulting in their receiving inputs from single motoneurons (Buffelli et al., 2004 ; Favero et al., 2010 ). Postnatally, brains undergo a period of over-growth of synapses which is maintained until puberty when massive synaptic pruning occurs (Zecevic and Rakic, 1991 ). Connectivity changes to brains are not only limited to development, they occur throughout an adult's life. Synaptic rewiring occurs in ischemic areas to recover function (Butz and van Ooyen, 2013 ; Mascaro et al., 2016 ), in the reward center of the brain as a result of drug addiction (Robinson and Kolb, 1997 ; Russo et al., 2010 ), in neurogenic areas in order to functionally integrate newborn neurons (Lledo et al., 2006 ), during learning (Benson et al., 2001 ; Holtmaat et al., 2005 ; Xu et al., 2009 ), memory formation and long term storage, in which it plays a central role (Poirazi and Mel, 2001 ; Kleim et al., 2002 ; Lamprecht and LeDoux, 2004 ), and during enriched experiences (Van Ooyen and Butz-Ostendorf, 2017 ). There is tight interplay between structural changes in the connectivity between neurons and the efficacies of existing connections. When viewed at the microscopic level, the projections of cortical neurons are so crowded that they can essentially be viewed as a potential all-to-all connectivity; a potential connection is one in which the growth of a spine or terminal bouton could form a synapse (Hellwig, 2000 ; Kalisman et al., 2005 ). However, not all potential connections are formed; the local microcircuitry of the cortex is functionally highly selective and generally maintains a sparse connectivity (Stepanyants et al., 2002 ; Le Bé et al., 2006 ). Synaptic plasticity mechanisms such as spike-timing dependent plasticity, which cause long-term potentiation or depression (Markram et al., 1997 ; Bi and Poo, 1998 ), have been reported to be closely linked to structural changes (Le Bé et al., 2006 ; Holtmaat and Svoboda, 2009 ). In terms of structural plasticity, the research focus of computational neuroscience typically lies on synaptic rewiring, thus this paper will not discuss the creation of new neurons as a form of structural plasticity. In the remainder of the paper, “structural plasticity” is used in lieu of structural synaptic plasticity, and can be used interchangeably with “synaptic rewiring.” 1.1. Contributions In this paper we implement the model proposed by Bamford et al. ( 2010 ) (described in detail in section 2.2.2) using a novel structural plasticity framework designed for the SpiNNaker system (section 2.1). We make use of the speed-up achieved by running the model in real time on SpiNNaker to explore whether it is suitable for modeling developmental formation of topographic maps over longer time-scales (section 3.3). We explore the behavior of the network when excitatory lateral connections are replaced with inhibitory ones, revealing the stabilizing effect these have on the network (exploration of this effect beginning in section 3.4). Finally, we perform sensitivity analysis on the network to establish an operational range of various parameters. We show that: (1) SpiNNaker is well-suited for parameter sweeps and even hyperparameter optimization due to its massive parallelism, and (2) the network has low inter-trial variability, except in certain scenarios discussed in section 3.5. SpiNNaker is a general-purpose neuromorphic platform with a sizeable user base. The model described here has been implemented so as to allow use as an “off-the-shelf” learning mechanism. The Python scripts and data generated from executing the simulations on SpiNNaker are available online 1 . 1.2. Computational models of structural plasticity Structural synaptic plasticity is an omnipresent mechanism in mammalian brains, involved in learning, memory, and recovery from lesions. Structural plasticity in the form of synaptic rewiring is also a useful computational tool, used to automatically generate connectivity based on experimental activity data (Diaz-Pier et al., 2016 ), explore network states for Bayesian inference (Kappel et al., 2015 , 2018 ), assist synaptic plasticity rules to achieve better performance (Spiess et al., 2016 ), allocate limited computational resources optimally (George et al., 2017 ), and more efficiently and accurately encode patterns (Poirazi and Mel, 2001 ; Roy et al., 2014 ; Hawkins and Ahmad, 2016 ; Roy and Basu, 2017 ), to name a few. Diaz-Pier et al. ( 2016 ) view structural plasticity as a mechanism for network optimization. In a time when neural activity data is abundant, but connectivity data is sparse, and when network models are mostly hand crafted, they use the structural plasticity model proposed by Butz and van Ooyen ( 2013 ) to achieve a desired mean activity level for the network through automatic self-organization. The local homeostatic mechanism rewires neurons in a network based on their mean spiking activity, their available dendritic and axonal connection points, and the distance between them. This structural plasticity mechanism is able to account for cortical reorganization after deafferentiation and stroke. Both of these models were run on the NEST simulator (Bos et al., 2015 ). Structural plasticity models need not include both formation and deletion rules. Iglesias et al. ( 2005 ) simulates the effect of massive synaptic pruning resulting after an initial developmental overproduction of synapses, similar to those explored in vivo by Zecevic and Rakic ( 1991 ). They show that, even without continuous formation of new synapses, a network will stabilize at approximately 10% of the starting number of active synapses, regardless of network size. This relies heavily on the choice of deletion rule, which must be very conservative, i.e., only prune synapses with no chance of becoming active. Kappel et al. ( 2015 ) propose a mathematical view of structural plasticity from the perspective of Bayesian inference. Their networks sample parameters from a prior distribution to obtain desirable characteristics, such as sparse connectivity and heavy-tailed distributions of synaptic weights. Moreover, they explain that stochastic dynamics of network parameters, which cause trial-to-trial variability in experiments, should be viewed as a “functionally important component of the organization of network learning.” They subsequently they show that the underlying stochasticity of neuronal networks allows the exploration of various network configurations while maintaining its functionality. In a follow-up paper, Kappel et al. ( 2018 ) extend their framework with the introduction of reward driven reorganization, operating side-by-side with synaptic plasticity. They replicate experimental results which observe task-dependent reorganization of connections between the cortex and basal ganglia. Importantly, their model takes into account recent findings that a significant amount of synaptic and structural plasticity happens stochastically, not solely based on activity (Dvorkin and Ziv, 2016 ). Structural plasticity models need not include both formation and deletion rules. Iglesias et al. ( 2005 ) simulates the effect of massive synaptic pruning resulting after an initial developmental overproduction of synapses, similar to those explored in vivo by Zecevic and Rakic ( 1991 ). They show that, even without continuous formation of new synapses, a network will stabilize at approximately 10% of the starting number of active synapses, regardless of network size. This relies heavily on the choice of deletion rule, which must be very conservative, i.e., only prune synapses with no chance of becoming active. Structural plasticity has also been shown to aid in denoising the response of a network, and improve the learning speed of synaptic plasticity mechanisms. Spiess et al. ( 2016 ) designed a model of structural plasticity which strives to maintain a constant number of synapses. Overlaid on top is a “pruning” mechanism, which steadily decreases the target number of synapses. Within their regime, they see a halving of the required time for the network to learn the presented patterns. Modeling work by Poirazi and Mel ( 2001 ) showed that neurons with nonlinear dendrites have larger information storage capacity compared to those with linear dendrites. Moreover, they believe that long-term information storage rests in the connectivity to the dendritic sub-units, rather than in the weights of connections. This work has been extended in subsequent years by Roy et al. ( 2014 ) in the context of liquid state machines (LSM, Maass and Markram, 2004 ) to achieve efficient readout. Here, the gradient descent-driven synaptic rewiring rule is applied to the readout network to minimize error. They computed a weight modification for each cell, but rather than applying it to the weights it is used to drive rewiring: a low value of the fitness function results in the synapse being marked for replacement with a more performant one. The versatility of the two-compartment unit with the addition of a suitable structural plasticity rule was also proven within winner takes all (WTA) networks (Roy and Basu, 2017 ). Both of the previous experiments are characterized by the application of synaptic rewiring at the end of pattern presentations. The system of Hawkins and Ahmad ( 2016 ) (Hierarchical Temporal Memory) uses sparse encoding of sensory inputs to learn a sequence of patterns (Olshausen and Field, 2004 ). Neurons are attached to some number of coincidence detectors, relying on the existence of a set of co-located synapses that connect to a subset of the cells that are active in the pattern to be recognized. When attempting to learn a new pattern, new synapses are formed. These synapses have a permanence value which represents their growth; the weight of the synapse is binary and derived from the permanence. A high permanence means the synapse is active, otherwise the synapse is reassigned. Recent work by George et al. ( 2017 ) implementing synaptic rewiring in neuromorphic hardware showed that structural plasticity enables optimal computational resource allocation. Their rule requires the use of an FPGA co-processor to the Reconfigurable On-Line Learning Spiking device (ROLLS, Qiao et al., 2015 ); the co-processor drives the rewiring of synapses between point neurons by periodically inspecting the weights of connections against a threshold. In their model, synapses are pruned if they have been depressed, i.e., their weights are bellow a threshold, for a fixed amount of time; after pruning, a new connection is immediately formed. Bamford et al. ( 2010 ) used structural plasticity to reduce receptive fields of point neurons when modeling the formation of neuronal topographic maps. Their model also uses STDP alongside the formation and elimination of synapses. Their model performed a fixed number of rewiring attempts in each simulation time step with exactly one of three actions being performed: formation of a new synapse, removal of an older synapse or no action; synapses are not replaced instantaneously and all actions are probabilistic. They saw a reduction in the spatial variance of neuronal receptive fields, while also preserving the desired position of the centers of these receptive fields. Finally, the input selectivity learnt through STDP was embedded into the network connectivity by the structural plasticity mechanism. 1.3. SpiNNaker neuromorphic computing platform Furber et al. ( 2013 , 2014 ) designed and built the SpiNNaker neuromorphic computing platform. It is a many-core machine making use of 18 ARM968 processors per chip, operating at around 200 MHz and with 64 MBytes of local data memory, 32 MBytes of local instruction memory and 128 MBytes of SDRAM shared across all cores on the chip. Each chip connects to 6 of its neighboring chips via the SpiNNaker router. The SpiNNaker network communicates between the chips and cores using small packets which contain either 32 or 64 bits of data following the Address Event Representation (AER, Deiss et al., 1999 ). SpiNNaker chips are organized into boards of 48-chips which are then joined together, allowing up to 1 million processors to run and communicate in parallel in a single simulation, making it one of the largest neuromorphic platforms currently in existence (Furber, 2016 ). The SpiNNaker platform was designed to simulate networks of simple spiking neurons, with each core theoretically capable of simulating up to 1,000 neurons, each with 1,000 synaptic inputs. When simulated with a time step of 1 ms between updates of the neural state, the network can then run in real time. Thus, when a large number of cores are running simultaneously, large networks can be simulated that would otherwise run much slower on conventional computing systems. This makes the system ideal for neurorobotic applications where the systems need to respond to stimuli in real time. The SpiNNaker communication network allows spike packets to be transmitted to multiple cores simulating neurons near-simultaneously across the entire machine, guaranteeing that the messages arrive within less than the 1 ms simulation time step (Navaridas et al., 2009 ). This is made possible through the support of multicast within the router on each chip, which allows it to forward a received packet to any or all of the neighboring chips and local cores in a single clock-tick. The real-time constraint is one of two design drivers of SpiNNaker, the other being energy efficiency. SpiNNaker uses the RISC architecture of ARM processors with local, attached memory, coupled with protocols which turn off application cores when not in use. SpiNNaker makes use of an interrupt-driven computational system, which is very desirable when optimizing for energy efficiency. Additionally, SpiNNaker focuses on simulating “point neuron models,” morphological simplifications of a neuron wherein the details of their dendritic structure is ignored; synapses between neurons are simulated as being formed directly onto the soma. 1.4. Simulation software on spiNNaker SpiNNaker uses an event-driven computation model when computing synaptic and neuron updates (Brown et al., 2015 ). An incoming spike (packet event) is placed in a buffer, which triggers a direct memory access (DMA) read of the SDRAM memory region which contains synaptic information (weights, delays etc.). The synaptic processing is performed along with any spike-timing-dependent-plasticity updates, and the data is then written back the the SDRAM through another DMA. The neuron update is performed on a timer interrupt which occurs every simulation time step. PyNN is a simulator-independent language for describing spiking neural network models created by Davison et al. ( 2008 ). PyNN is used as a front-end to the SpiNNaker system, which make it easier for users to describe their networks without having to interact directly with the SpiNNaker hardware. The SpiNNaker-specific PyNN implementation (sPyNNaker, Stokes et al., 2016 ) controls how neuronal populations are partitioned and placed onto individual SpiNNaker cores and sets up the on-chip routers so as to allow multicast communication between chips (Mundy et al., 2016 ). The software then converts the neural network parameters and connectivity data into a form that the cores can make use of when performing their updates. 1.5. Current status of learning on spiNNaker Learning on SpiNNaker is currently implemented in the form of long-term potentiation/depression (LTP/LTD) instigated via spike-timing-dependent plasticity (STDP). From the perspective of a synapse, the relative timing of pre- and post-synaptic action potentials is used as a measure of causality and forms the basis of a synaptic weight change. Hebbian two-factor learning is employed (Song et al., 2000 ), where a pre-synaptic spike, shortly followed by a post-synaptic spike is identified as a causal relationship, and potentiated; conversely, a post-pre pairing is depressed (Markram et al., 1997 ; Bi and Poo, 1998 ). A range of STDP learning rules is implemented within the sPyNNaker API, with the implementation framework following the modular structure of PyNN. Individual timing and weight update rules are specified on a per projection basis, where a projection is defined as a directional link between populations of neurons. Timing rules operate on pre- and post-synaptic history traces, and include the classical spike pair rule, the triplet rule described by Pfister and Gerstner ( 2006 ), and the homeostatic rule designed by Vogels et al. ( 2011 ). Weight updates based on the results from timing assessment can be made either in additive or multiplicative form, following the methods summarized in Morrison et al. ( 2008 ). The implementation has been designed with chip architecture and performance in mind, with the goal of real-time simulation of neural networks containing plastic synapses. As the synapse input is processed on the post-synaptic core, a challenge presented by the hardware is the restricted information available with respect to the synapse structure and connectivity (Diehl and Cook, 2014 ). Whilst a neuron processing core stores neuron state variables in local memory, the comprehensive synaptic connectivity data is relatively large, and hence must be stored in shared memory. When a neuron receives a spike, the appropriate synaptic data is transferred from shared to local memory, enabling the appropriate weight contribution to be made to the neuron input. For efficiency and convenience, plastic weight updates are therefore limited to pre-synaptic events (i.e., receiving a spike), as it is at these times the synaptic weight has been transferred into local memory for processing. Pre-synaptic traces are stored alongside synaptic data in shared memory (and retrieved and updated on spike arrival at the post-synaptic neuron); whilst post-synaptic trace histories are maintained between pre-synaptic events in core-local memory. This framework, known as the deferred event driven (DED) model (Jin et al., 2010 ), provides all the information necessary to perform synaptic updates based on relative spike timing.",
"discussion": "4. Discussion We described the design and implementation of a model of synaptic rewiring simulated on the SpiNNaker neuromorphic platform using PyNN as the network description language. Moreover, we showed that the model can be simulated in real-time, allowing for longer runs modeling development or multiple runs to explore the sensitivity of the network. The provided validation and the open-source nature of the simulator and of the network description means that our effort can be easily interrogated or extended. We showed that continuously operating STDP and synaptic rewiring can refine a topographic map regardless of the initial mapping between neurons. Running simulations for longer revealed a possible source of instability in the network, namely the excitatory lateral connections. Under certain conditions, self-sustaining waves of activity occurred in the target layer, resulting in STDP favoring lateral connections over feedforward ones. We replaced the excitatory connections with inhibitory ones and immediately observed the homeostatic and stabilizing effect they had in conjunction with the sampling mechanism. Sensitivity analysis revealed operational ranges of STDP, rewiring and input parameters. In the following sections we discuss computational characteristics of SpiNNaker in relation to the model of synaptic rewiring at hand. Furthermore, we make an argument on the efficiency of the implemented structural plasticity framework. Finally, we present a suite of future extensions and uses of the work presented herein. 4.1. Performance This was the first attempt at structural plasticity on SpiNNaker. Moreover, this mechanism was never considered in the design process of the system, neither hardware, nor software. However, it has turned out that the complete neuromorphic package has been sufficiently flexible to accommodate synaptic rewiring in parallel with STDP in real time. Framework scalability can be interpreted from the point of view of individual processing cores (vertical scaling) or from the point of view of all available processing cores (horizontal scaling). Horizontal scaling is currently limited by the amount of additional metadata that needs writing in core-local memory and the entries present in the routing tables (router current assumes all-to-all. The exponential decay lookup tables discussed at the end of section 2.2.2 are dependent on (1) the size of the layers and (2) the magnitude of standard deviations used for formation σ form − ff and σ form − lat . The amount of metadata which needs to be added on chip could be reduced by pruning the information pertaining to impossible connections. Rinke et al. ( 2016 ) suggest an efficient algorithm for neuron selection during synaptic rewiring based on n -body problems, where pairs of bodies have to be considered for force calculations. Their approximation technique relies on observations that particles sufficiently far away from a target particle need not be considered individually. They apply this algorithm to the model proposed by Butz and van Ooyen ( 2013 ), but it could be applied here for pruning routes which will never materialise into synaptic connections. Vertical scalability depends on a few characteristics of the networks, but also on some of the model parameters. SpiNNaker processing cores operate at approximately 200 MHz and have small amounts of attached memory, capable, in theory, of simulating at most 1,000 neurons. Additional processes executing on a core, such as STDP, have the effect of reducing the number of simulated neurons significantly, especially when operating under the real time execution constraint. The requirement that rewiring be performed at a fixed rate requires additional retrieval of information from SDRAM than is otherwise necessary in normal operation. In the current implementation (detailed in section 2.2.2) a DMA read is required to retrieve the synaptic row of a pre-synaptic neuron to DTCM. This is an extra read operation in excess of the read operations performed whenever a spike is received. A spike-driven rewiring process could make use of the existing infrastructure and only operate on synaptic rows which are in core-local memory. 4.2. On efficiency The existing sPyNNaker framework for the implementation of neural networks on SpiNNaker supports user extensions in the form of new neuron models and new STDP plasticity weight update rules. The work presented in this paper extends this framework to support implementation of new structural plasticity rules. The use of frameworks when using neuromorphic systems such as SpiNNaker allows for efficient integration of new rules without requiring users first understand the existing code or even the underlying hardware and execution model. This is because the integration points are chosen specifically to execute in the most efficient way on the platform, and present interfaces to be filled in by the users which reflect this point of integration. For example, code which modifies synaptic connectivity information on SpiNNaker (such as changes in weight as performed by STDP) is best executed whilst the synaptic data is in the core local memory (DTCM); in normal operation of the software this occurs whilst processing an incoming spike. Thus STDP update rules are written to operate whilst an incoming spike is being processed, with appropriate information being cached until this is the case. In terms of structural plasticity, additional information is made available to allow updates to be performed on a time-step basis outside of the processing of incoming spikes. This information has been provided in a compact form that works well within the current software framework, and so can now be used by any other structural plasticity formation or deletion rule without having to be implemented a second time. This also ensures correctness as the synaptic information is copied from SDRAM during processing, and written back once processing is complete, the software must ensure that a second copy of SDRAM is not made until the first has been processed and written back. This is particularly important where STDP and structural plasticity may both wish to modify the same part of the synaptic matrix at the same time; the event-based nature of the execution of SpiNNaker code, and the use of the DMA engine to transfer data between DTCM and SDRAM in parallel to code execution on the CPU make it easy to get into an inconsistent state if the implementation is not carefully done. The use of the framework developed here helps avoid these mistakes in future structural plasticity implementations. 4.3. Future work An immediate imperative is addressing inherent scalability issues, as discussed in the previous section. Modeling larger neuronal layers could prove fruitful to investigate the behavior of the model with more realistic inputs, such as those originating from an event-based dynamic vision system and representing both natural and artificial scenes. Such models would necessarily also be simulated on longer time scales than presented herein so as to allow sufficient time for neurons to adapt to the statistics of the input. A move toward more natural input could consist of handwritten digits represented as spike trains (MNIST, LeCun et al., 1998 ; Liu et al., 2016 ). Classification could be performed using a network comprised of separate maps sensitized to different digits connected to a winner-takes-all circuit. The framework as implemented is a platform-dependent extension of the PyNN specification. More community input could drive the modification of the PyNN application programming interface (API) to natively support multiple models of structural plasticity. Following evidence of topographic projections being stabilized in the presence of lateral inhibition, an extension to the model including both inhibitory and excitatory lateral connectivity could similarly be implemented and analyzed. Moreover, all the aforementioned network architectures could be extended to include distance-based synaptic delays; spatio-temporal patterns could then be embedded into the network connectivity. An additional extension could see multiple formations or removals occurring each rewiring attempt, which could decrease the refinement time of the topographic mapping. This approach could be driven by a different type of event: the reception of an AER spike, rather than a timer interrupt because the synaptic information will be available in core-local memory. Finally, future work could also focus on choosing a different mechanism for selecting pairs of neurons as partners for formation. Preliminary results show little qualitative differences between a random selection of pre-synaptic partner for formation and the selection based on later spike times. However experiments have not been run for simulations longer than 300 s or when varying parameters."
} | 7,149 |
35951523 | PMC9371330 | pmc | 3,851 | {
"abstract": "Leaf fungal microbiomes can be fundamental drivers of host plant success, as they contain pathogens that devastate crop plants and taxa that enhance nutrient uptake, discourage herbivory, and antagonize pathogens. We measured leaf fungal diversity with amplicon sequencing across an entire growing season in a diversity panel of switchgrass ( Panicum virgatum ). We also sampled a replicated subset of genotypes across 3 additional sites to compare the importance of time, space, ecology, and genetics. We found a strong successional pattern in the microbiome shaped both by host genetics and environmental factors. Further, we used genome-wide association (GWA) mapping and RNA sequencing to show that 3 cysteine-rich receptor-like kinases (crRLKs) were linked to a genetic locus associated with microbiome structure. We confirmed GWAS results in an independent set of genotypes for both the internal transcribed spacer (ITS) and large subunit (LSU) ribosomal DNA markers. Fungal pathogens were central to microbial covariance networks, and genotypes susceptible to pathogens differed in their expression of the 3 crRLKs, suggesting that host immune genes are a principal means of controlling the entire leaf microbiome.",
"conclusion": "Conclusions Switchgrass leaf fungal communities are highly diverse, and are influenced by both host and environmental factors. Succession occurs each season as communities are assembled through stochastic, environmental, and host-determined processes. Pathogenic fungi play a critical role in the switchgrass leaf phyllosphere community, determining both the trajectory of microbial community development and acting as central nodes in community networks. Host immune genes such as receptor-like kinases control pathogens directly, and the prevalent mycoparasites that prey on them indirectly. The plant genes that control pathogens may therefore provide a principal means by which plants influence changes in their fungal microbiome.",
"introduction": "Introduction Microbial communities perform essential functions for their host organisms in all branches of life. In some systems, hosts can tightly control the microbes with which they form symbioses [ 1 , 2 ]. In others, the composition of the microbiome is more governed by ecological interactions such as the order of species arrival or abiotic conditions during colonization [ 3 , 4 ]. A key goal of microbial evolutionary ecology is to determine how both host and nonhost factors influence microbiome assembly [ 5 ], particularly in natural settings where host influence is more challenging to study. Communities that colonize available niches in the process of succession follow certain predictable ecological patterns. Early-arriving species are typically those with effective long-range dispersal, while the climax community is dominated by species that can more effectively use resources under competition [ 6 ]. While these broad patterns are generalizable, the composition of any particular successional community depends greatly on both the habitat colonized and interspecific interactions such as priority effects, where the order of arrival of taxa governs the success of later arrivals [ 7 , 8 ]. While most successional theory is based on studies in macro-scale organisms, the principles of succession are evident in microbial communities as well, but on a more rapid timescale [ 9 – 11 ]. In the case of microbiomes, host factors governing microbial succession must also be considered. Since the composition of the microbiome can greatly impact host fitness, it can be evolutionarily beneficial for the host to play a role in the successional process, encouraging mutualist colonization while dispelling pathogens as the community assembles. Hosts express genes that influence colonizing microbes through several means, including immunity, morphological adaptations [ 12 ], and chemical exudation [ 13 ]. While the immune system is often effective at preventing detrimental infections, immune receptors may recognize and exclude beneficial microbes if elicitors are structurally similar to a pathogen, so specific immunity can have wider impacts on the microbiome [ 14 ]. Hosts require finely calibrated mechanisms for attracting beneficial microbes without attracting pathogens in a constant coevolutionary push and pull. The phyllosphere microbiome, consisting of the microbes on and inside the plant leaf, comprises diverse taxa that impact plant health and productivity [ 15 – 18 ]. Leaf fungi in particular are common plant pathogens [ 19 ], but nonpathogenic taxa may perform beneficial functions for the host, including nutrient uptake and pathogen antagonism [ 20 – 25 ]. Since the phyllosphere microbiome of perennial plants is reassembled at the start of each growing season in freshly sprouted tissues, [ 26 , 27 ] it may show similar patterns to macro-scale secondary successional communities. Recent research has shown that host control of the leaf microbiome is often governed by numerous loci of small effect directly impacting relatively few microbes [ 28 – 30 ]. We hypothesized that the phyllosphere fungal microbiome develops seasonally as a successional community controlled by environmental factors, host genetics, and interspecific fungal–fungal associations. We used amplicon sequencing to compare the relative importance of these factors in the phyllosphere fungi of a replicated diversity panel of switchgrass ( Panicum virgatum [ 31 ]). We tested whether communities change directionally and whether the trajectory of succession differed across switchgrass genetic subpopulations and across different growing sites. Additionally, we sought to uncover whether specific genetic loci underlie host control of the microbiome through genome-wide association study (GWAS) and RNA sequencing analyses. Finally, we investigated the roles of specific fungal taxa in the microbiome through network analysis. Specifically, we aimed to determine whether known switchgrass leaf pathogens [ 32 ] covary with nonpathogenic symbionts, or are peripheral to microbial communities.",
"discussion": "Discussion Our results show strong support for the importance of time, geographic location and host genetics in influencing the switchgrass phyllosphere microbial succession over the growing season. We found evidence for clear successional dynamics that were consistent in direction across growing sites, but were distinct in community composition. Fungal communities were different across host genetic subpopulations, a pattern that may be driven by variation at 3 linked immune receptors. Leaf fungal communities are taxonomically diverse, but a few highly abundant pathogens and yeast species play a disproportionate role in shaping community progression. Viewing the switchgrass leaf microbial community through the lens of succession allowed us to delineate ecological patterns in these communities. Multidimensional scaling representations of the leaf communities at the focal site revealed a clear clustering by date of collection on the first NMDS axis ( Fig 1 ). This indicates that, as we predicted, date of collection is an important source of variation in the switchgrass leaf fungal community. Further, measuring the trajectories of these communities showed that succession is both directional and deterministic, since no samples showed negative trajectories (reversals of succession) by the end of the season, and most samples followed a similar trajectory ( Fig 1B ). We observed similar patterns to other studies that show early-season leaves as highly distinct from later time points, perhaps owing to greater influence from soil microbes [ 16 , 45 ]. While the overall shape of trajectories was similar among samples, the Midwestern population deviated from others, particularly in the late season. The Midwestern population is notable since we have previously shown that it is more susceptible to several fungal pathogens such as leaf rust ( Puccinia novopanici ) and leaf spot ( Bipolaris spp.; [ 32 ]; also see [ 46 ]) and has on average an earlier phenology than the other population groups [ 33 ]. Leaf microbiome relationships are consistently distinct in this population and may be linked to other traits such as cold tolerance that also differ [ 31 , 33 ]. In addition to temporal differences across subpopulations, the composition of fungal leaf communities differed markedly across geographic locations. This may be partially due to seasonality differences across the region we examined. The Kingsville, Texas site did not experience freezing temperatures between 1989 and 2020 (NOAA weather service), so perennial grasses in the region may have living aboveground tissue year-round. Growing season length has been shown as an important factor in governing the abundance and diversity of endophytic fungi [ 47 ], so it is unsurprising that we saw large differences across this latitudinal gradient. However, many other factors that influence fungal communities also differ across these sites, including precipitation regime, soil type, and surrounding vegetation, so further work is needed to determine if the growing season is truly the causal factor. We predicted that fungal communities would be impacted by host genetics as well as location. We found several lines of evidence for genetic control of the leaf microbiome. In addition to examining differing successional trajectories across subpopulations, we tested the covariance of genetic distance and fungal community differences using Mantel correlations. Genetic-fungal community correlations increased until DOY 260, then declined as host senescence began. Mantel tests are inappropriate for some ecological tests and often underestimate p -values, but can be useful for exploratory analysis of distance matrices [ 48 ]. While there was high variation in our pseudo-heritability estimates, the fact that they mirror temporal patterns in the Mantel tests strengthens the general trend of greater genetic associations in the late season. Previous studies have found similarly high variation in microbiome heritability estimates across time [ 29 ], so it is not surprising to see this in our case. Deng and colleagues calculated H 2 for individual OTUs, which ranged from 0 to 0.66 and a Mantel’s correlation of r = 0.13 between genetic and microbiome composition in sorghum rhizosphere. This value of r is lower than we saw in our study, possibly since it was based on a relatively small subset of samples. When selecting samples for this study, we randomly chose equal numbers of samples from the 2 major switchgrass morphological ecotypes, upland and lowland switchgrass [ 49 ] ( S2 Fig ). Lowland switchgrass, which is more highly represented in Gulf and Atlantic subpopulations, is more resistant to several leaf fungal pathogens [ 32 ], so subpopulation differences may be at least partially driven by differences in immunity across these genotypes. Since pathogens such as Microdochium and Alternaria were among the most abundant taxa in our samples, their differences across subpopulations may have driven overall community differences. In addition to immunity, however, subpopulations differ in other traits that may contribute to fungal colonization differences, such as leaf wax content [ 50 ], exudate concentration [ 51 ], and phenology [ 33 , 49 ], so microbiome differences may be responding to multiple host plant traits. A replicated receptor-like kinase is associated with fungal differences We found one outlier SNP associated with microbiome structure. While there were several peaks in the Manhattan plot ( Fig 5A ), our analysis showed a strongly skewed distribution of observed versus expected p -value ( S6 Fig ), indicating a risk of Type I errors. This is probably attributable to the low sample size in this GWAS. The influence of the identified locus is fairly strong, contributing to a clear decrease on NMDS axis 2 when the minor allele is present (MAF = 0.083; S9 Fig ). This SNP is not in Hardy–Weinberg equilibrium in switchgrass; we found only one minor-allele homozygote among our samples. This abnormal pattern may be attributable to structural variation at this locus. Switchgrass subpopulations vary widely in genome structure, which may result in alignment mismatches that resemble SNPs, particularly in regions with multiple gene copies [ 52 ]. Indeed, this region shows an elevated number of insertions and deletions compared to nearby sections of the 2N chromosome ( S10 Fig , data from [ 31 ]) and is adjacent to a region dense with repetitive long terminal repeat retroelements (positions 60960000–60980000). Given the confirmatory results for this locus as well as the RNA sequencing results, however, we expect that there is a true phenotypic association with the locus, but that it may be with a structural variant rather than a true SNP. The 3 nearby genes we identified were replicated variants of a cysteine-rich RLK whose function has not been experimentally verified in Panicum . RLKs are one of the largest plant gene families, including over 600 members in Arabidopsis [ 36 ]. The best studied of these is FLS2, which detects the bacterial flagellin protein and initiates an immune response cascade [ 53 ]. The 3 RLKs we identified show high sequence similarity to immune-related cysteine-rich RLKs in Arabidopsis and Oryza , and contain the “stress-antifungal domain” PF01657, which has been linked to salt stress as well as fungal responses when present in several proteins [ 54 , 55 ]. Arabidopsis CRK5 , for example, alters defense responses either through resistance to infection or programmed cell death, depending on how the gene is expressed [ 56 ], and CRK6 and CRK 14 are involved in the general non–self-response [ 57 ]. Related Arabidopsis genes may be the targets of immune repression by bacterial strains [ 58 ]. Similarly, the Oryza gene LIL1 (Os07g0488400) improves fungal rice blast resistance when overexpressed [ 59 ]. The pattern of these receptors being more highly expressed in pathogen-susceptible plants may seem counterintuitive given that many RLKs are immune receptors. However, this can often occur when pathogens produce effector proteins that target immune receptors [ 60 ]. Necrotrophic fungi in particular can benefit by over-inducing plant immune receptors to initiate programmed cell death, [ 61 , 62 ] then feeding on dead plant tissue. Since allelic variants at this locus have now been associated with variation in the fungal microbiome across several years in natural populations, it may represent a useful target for future research into genetic control of the leaf microbiome. Previous research has shown that microbiome control is often polygenic, with many contributing loci of small effect [ 28 – 30 ]. Uncovering only a single causal locus in this study may be a product of the relatively low sample size; there are more loci associated with microbiome community structure that did not meet the GWAS cutoff, but may contribute to a polygenic architecture for this trait. Pathogens and hyperparasites are important in succession We used several methods to identify important taxa in the phyllosphere community. In several other recent studies, genetic effects on microbiomes appear to be targeted toward particular microbes, with the effects permeating through the community through ecological effects [ 28 , 63 , 64 ]. We used “core” microbiome analysis to identify OTUs that show high occupancy (presence across multiple samples within a time point [ 16 ]). We found that core taxa overlapped well with important taxa identified by MTV-LMMs and trajectory analysis. We can therefore be confident that this group of taxa is influential in the switchgrass phyllosphere ( Fig 5 ). Within this group, we identified several as pathogens, including Alternaria , Mycosphaerella , Microdochium , and Taphrina . It is challenging to assign functional guilds to symbiotic fungi, since their benefit or detriment to the host may depend strongly on phenology, abiotic conditions, and ecological interactions [ 65 ]. For example, many endophytic fungi are commensal for most of the season, then shift to breaking down plant tissue as the host begins senescence [ 66 ]. Others may be weakly pathogenic, but may improve overall host fitness by enhancing nutrient uptake or preventing infection by more effective pathogens [ 23 , 67 ]. Yeasts and yeast-like fungi were also well represented in phyllosphere samples. Yeasts were historically thought to be dominant in the phyllosphere [ 68 ], but this may have been an artifact of methods used. Yeasts are more easily culturable than filamentous fungi, and are therefore overrepresented in studies using cultures to measure fungal diversity. The exact relationship between yeasts and plant hosts is not totally clear, but they are typically thought to be mostly commensal symbionts, feeding on small amounts of sugars on the leaf surface [ 69 ]. At the focal site, Tremellomycete yeasts and Dothideomycetes dominated the core microbiome and covaried negatively through time. This may be explained by different spatial distributions across samples; Tremellomyctes dominate some samples and Dothideomycetes other samples, but they rarely coexist. Priority effects, wherein early-arriving taxa gain advantage over late-arriving taxa, may therefore play a role in governing colonization in these taxa. Certain Tremellomycete yeasts have been shown to be potential biocontrol agents against pathogens, e.g., Papiliotrema spp. [ 70 ], and others have been shown to be “hub” taxa or negatively connected with leaf pathogens, e.g., Dioszegia spp. [ 63 ], both genera with high abundance in our focal site dataset. One unexpected finding of our taxon-specific analysis was that 2 mycoparasites were identified as important taxa, Epicoccum and Sphaerellopsis . Epicoccum is an ascomycete genus comprising several species with noted antifungal properties [ 71 , 72 ]. The species we identified in this study, Epicoccum dendrobii , is being investigated as a biocontrol agent of the pathogenic anthracnose fungus Colletotrichum gloeosporoides [ 73 ]. Similarly, Sphaerellopsis filum has been observed infecting multiple species of Puccinia rusts [ 74 , 75 ] and has been shown specifically to reduce switchgrass rust infection [ 76 ]. Another surprising finding was that switchgrass rust was not a core species, despite the fact that its disease symptoms are nearly omnipresent each year in the sites we studied [ 32 ]. Fungi in the Pucciniaceae family have an ITS sequence that differs substantially from general fungal primers used in this study, which we suspect resulted in reduced amplification of Puccinia rusts. We confirmed this suspicion by additionally sequencing the LSU for our confirmatory analysis; using ITS failed to identify any Puccinia rusts in these samples, but LSU identified 10 OTUs as Puccinia present in 18 of 20 samples. There were more than double the Puccinia OTU counts in individuals with the major allele at the focal outlier locus, but a large outlier obscures a reliable statistical pattern. The ubiquity of the Sphaerellopsis hyperparasite is a further indication that Puccinia may be more prevalent than our sequencing data show, a speculation that is supported by the fact that Sphaerellopsis was identified by indicator species analysis as clearly overrepresented in leaves with rust infection. The other OTU most closely associated with disease symptoms is OTU_4, Microdochium . While we could find little evidence of known associations between Puccinia and Microdochium pathogens in published studies, this result suggests that they may have a synergistic effect on host disease. Bacterial microbiomes may be just as important to leaf function as fungi [ 16 ], although ecological patterns may differ in some important ways. We used fungi in this study because they contain more known switchgrass pathogens and may be documented more clearly within leaf tissue without conflict by chloroplast DNA. However, interplay between microbial groups is an essential component to microbiome ecology. Interactions between fungi, bacteria, viruses, microfauna can all mediate impact on hosts. Bacteria [ 77 ] and viruses [ 78 ] have documented impacts on the functioning of host-dependent fungi in complex and fascinating multilevel interactions. Beyond individual interactions, functional microbiomes in soils require both diverse fungi and bacterial communities, so influence between these groups is impossible to connect to just one single microbe [ 79 ]."
} | 5,151 |
33781078 | null | s2 | 3,852 | {
"abstract": "Microbiomes are complex and ubiquitous networks of microorganisms whose seemingly limitless chemical transformations could be harnessed to benefit agriculture, medicine, and biotechnology. The spatial and temporal changes in microbiome composition and function are influenced by a multitude of molecular and ecological factors. This complexity yields both versatility and challenges in designing synthetic microbiomes and perturbing natural microbiomes in controlled, predictable ways. In this review, we describe factors that give rise to emergent spatial and temporal microbiome properties and the meta-omics and computational modeling tools that can be used to understand microbiomes at the cellular and system levels. We also describe strategies for designing and engineering microbiomes to enhance or build novel functions. Throughout the review, we discuss key knowledge and technology gaps for elucidating the networks and deciphering key control points for microbiome engineering, and highlight examples where multiple omics and modeling approaches can be integrated to address these gaps."
} | 274 |
25883680 | PMC4399083 | pmc | 3,854 | {
"abstract": "Background Lignocellulosic ethanol has a high potential as renewable energy source. In recent years, much research effort has been spent to optimize parameters involved in the production process. Despite that, there is still a lack of comprehensive studies on process integration. Single parameters and process configurations are, however, heavily interrelated and can affect the overall process efficiency in a multitude of ways. Here, we present an integrative approach for bioethanol production from wheat straw at a representative laboratory scale using a separate hydrolysis and co-fermentation (SHCF) process. The process does not rely on commercial (hemi-) cellulases but includes enzyme production through Hypocrea jecorina (formerly Trichoderma reesei ) on the pre-treated feedstock as key unit operation. Hydrolysis reactions are run with high solid loadings of 15% dry mass pre-treated wheat straw (DM WS), and hydrolyzates are utilized without detoxification for mixed glucose-xylose fermentation with the genetically and evolutionary engineered Saccharomyces cerevisiae strain IBB10B05. Results Process configurations of unit operations in the benchtop SHCF were varied and evaluated with respect to the overall process ethanol yield ( Y Ethanol-Process ). The highest Y Ethanol-Process of 71.2 g ethanol per kg raw material was reached when fungal fermentations were run as batch, and the hydrolysis reaction was done with an enzyme loading of 30 filter paper units (FPU)/g DM WS . 1.7 ± 0.1 FPU/mL were produced, glucose and xylose were released with a conversion efficiency of 67% and 95%, respectively, and strain IBB10B05 showed an ethanol yield of 0.4 g/g Glc + Xyl in 15% hydrolyzate fermentations. Based on the detailed process analysis, it was further possible to identify the enzyme yield, the glucose conversion efficiency, and the mass losses between the unit operations as key process parameters, exhibiting a major influence on Y Ethanol-Process . Conclusions Y Ethanol-Process is a measure for the efficiency of the lignocellulose-to-bioethanol process. Based on mass balance analysis, the correlations between single process parameters and Y Ethanol-Process were elucidated. The optimized laboratory scale SHCF process showed efficiencies similar to pilot scale plants. The herein presented process analysis can serve as effective and simple tool to identify key process parameters, bottlenecks, and future optimization targets. Electronic supplementary material The online version of this article (doi:10.1186/s13068-015-0232-0) contains supplementary material, which is available to authorized users.",
"conclusion": "Conclusions In this study, an integrative process analysis of a benchtop SHCF is presented. Based on mass balance analysis, the influence of varying process configurations on Y Ethanol-Process was analyzed. Thereby, a fundamental understanding of the complete process was established. This allowed for identification of the process parameters, which have the highest impact on Y Ethanol-Process , the enzyme yield, ηGlucose, and ηHandling. It was further shown that, under comparable process conditions (Conifg.5 - no enzyme production), Y Ethanol-Process of the benchtop SHCF process is equal to pilot scale plants. We therefore believe that the benchtop-scale analysis described herein presents an important and useful tool to identify bottlenecks and optimization targets within the process with reasonable effort and expenditure. Findings of this study stress the importance to establish a mass balance-based understanding of the lignocellulose-to-bioethanol process and to optimize unit operations or process parameters with respect to Y Ethanol-Process .",
"discussion": "Results and discussion The feedstock The substrate presented in this study was Austrian wheat straw pre-treated with steam explosion. Wheat straw has a high potential as sustainable biomass source in Europe based on its abundance and low cost [ 28 ]. Steam explosion, in combination with chemicals or alone, has been described as an efficient and cost-effective method for pre-treatment of wheat straw [ 9 ]. The raw material in this study was treated with a simple method based on steam explosion only. The pre-treated wheat straw had a dry mass (DM) content of approximately 90% and thereof the water insoluble fraction was 69%. The compositional analysis is depicted in Table 1 . The acid hydrolyzate contained majorly glucose and xylose. Other hemicellulose-derived sugars (e.g., L-arabinose, galactose) were only present in amounts below the detection limit of the high performance liquid chromatography (HPLC) system and are not mentioned in Table 1 . Although processed under the same conditions, the pre-treated wheat straw composition showed batch-to-batch variability. Thus, the glucose and xylose content varied by 5% and 13%, respectively (Table 1 ). Variation in the pre-treated feedstock composition, especially in xylose content, has been observed before [ 25 ]. The amorphous nature of the hemicellulose and different levels of degradation during the pre-treatment as well as seasonal variations are likely explanations. Since both batches of wheat straw were utilized throughout this study, analysis of unit operation and mass balanced process analysis were based on averaged values (Table 1 ). The raw material before pre-treatment had a xylose content of 24.9 ± 0.4% and a glucose content of 36.1 ± 0.1% dry matter (data not shown). Mass losses caused by pre-treatment were 10% on average. Table 1 \n Compositional analysis of the pre-treated wheat straw \n \n Components in dry matter \n \n Percentage [%] \n \n Batch 1 \n \n Batch 2 \n \n Mean \n Carbohydrates Glucose 43.7 48.7 46.2 ± 2.5 Xylose 17.2 13.1 15.2 ± 2.0 Non-carbohydrates Acid-soluble lignin 1.3 2.0 1.7 ± 0.3 Acid-insoluble lignin 27.7 30.0 28.9 ± 1.2 Ashes 4.5 4.4 4.5 ± 0.0 Others 5.6 4.3 4.9 ± 0.6 Experimental analysis of unit operations The three unit operations of the SHCF process, enzyme production, hydrolysis, and fermentation, were analyzed under varying process conditions. As shown in Figure 1 , the process streams between the unit operations were treated by centrifugation, filtration, and concentration. The resulting losses were included into the process analysis with the efficiency factor of conditioning steps (ηHandling), and it was determined to be 75% on average. Production of (hemi-) cellulases by T. reesei SVG17 T. reesei is the majorly applied organism for (hemi-) cellulase production, and it has been studied and continuously improved since the 1960s [ 29 - 33 ]. The herein presented T. reesei SVG17 is a mutant of the QM9414 strain, and previous studies have described it as useful enzyme producer at both laboratory and pilot scale [ 31 ]. Fungal cultivations were run as batch fermentation with a substrate loading of 30 g DM pre-treated wheat straw per L (g DM WS /L). In 7 days of fermentation, a total volumetric cellulolytic activity of 1.7 ± 0.1 FPU/mL was reached. The beta-glucosidase activity was determined to be 0.6 ± 0.1 U/mL. To increase the enzyme yields, fermentations were also run as fed-batch. Per 2 L of fermentation, 30 g DM WS was added three times after 66, 94, and 138 h of fermentation. The time course of the fed-batch fermentation is depicted in Figure 2 A. In 210 h of fermentation, the volumetric cellulase activity reached 2.7 ± 0.02 FPU/mL. Similar to the batch fermentation, the beta-glucosidase activity was approximately half of the FPU/mL value and it was determined to be 1.5 ± 0.02 U/mL. Addition of feedstock was described to prolong the phase in which enzyme production is most active by freshly inducing both biomass growth and enzyme expression [ 30 , 32 ]. Consequently, the cellulase activity is increasing after each addition of WS, and only towards the end (180 to 210 h), the FPU/mL-time curve is stagnating. The time courses of beta-glucosidase activity and protein concentration show a similar pattern with a less pronounced effect of the feed. In the fed-batch fermentation, it was possible to improve the volumetric cellulase activity 1.6-fold as compared to the batch fermentation. This increase, however, required a 2.5-fold higher substrate loading. The impact of both process configurations on Y Ethanol-Process was evaluated with mass balance analysis and will be discussed hereinafter. Figure 2 \n Time courses of the \n T. reesei \n SVG17 fed-batch fermentation (A) and the enzymatic hydrolysis (B). Thirty grams DM WS was added after 66, 94, and 138 h to the fungal fermentation. Hydrolysis reaction was run with 30 FPU/g DM WS . Both time courses represent mean values from two experiments. Symbols: (A) Total cellulase activity (empty circles), beta-glucosidase activity (filled squares), and protein concentration (empty triangles). (B) Glucose (empty circles), xylose (filled squares), and cellobiose (empty triangles). Despite the difficult substrate conditions, T. reesei SVG17 was able to grow on the pre-treated wheat straw and showed efficient and reproducible production of (hemi-) cellulolytic enzymes. A drawback of strain SVG17 is the relatively low beta-glucosidase activity in the enzyme solution which did not exceed 50% of the overall cellulase activity (measured in FPU). A beta-glucosidase activity to FPU ratio of 1 was found to be the lower limit for efficient biomass conversion [ 34 , 35 ]. Low beta-glucosidase activity is an often observed problem in the production of cellulolytic enzymes by T. reesei . To overcome this problem, genetically engineered strains overexpressing heterologous beta-glucosidase genes have been described in the literature [ 34 - 37 ]. Although the availability of a robust T. reesei strain producing a beta-glucosidase-boosted enzyme solution could be important in view of an optimized process output, the aim of this study was not primarily optimization itself, but rather provision of a basis for optimization through integrative mass balance analysis of a representative SHCF process. Therefore, the use of T. reesei strain SVG17 was fully in line with the concept of the study, and the limitation in beta-glucosidase noted was not considered to restrict the relevance of the current investigation. Moreover, the effects of enhanced beta-glucosidase activity on hydrolysis yield and process performance are described later in this manuscript. Enzymatic hydrolysis To render lignocellulose-to-bioethanol processes economically feasible, a high ethanol titer is crucial [ 17 ]. Consequently, hydrolysis reactions must aim for high solid loadings to increase the sugar content. This, in turn, does increase the content of compounds, which are potentially toxic for the fermentation organism. In a previous study, we have shown that the solid loading in the hydrolysis reaction can be increased from 5% to 15% DM WS without introducing severe inhibition effects on S. cerevisiae strain IBB10B05 [ 19 ]. Enzyme loadings were varied. Firstly, reactions were run with 25 FPU/g DM WS [ 19 , 27 ] and the final hydrolyzate contained 40.6 ± 5.7 g/L glucose and 18.0 ± 2.3 g/L xylose. Based on the compositional analysis of the feedstock (Table 1 ), this equals a conversion efficiency of 60% for glucose (ηGlucose) and 81% for xylose (ηXylose). To improve the conversion efficiencies, the enzyme loading was increased to 30 FPU/g DM WS , and ηGlucose of 67% and ηXylose of 95% were reached. The 15% hydrolyzate contained 46.1 ± 0.2 g/L glucose and 21.5 ± 0.2 g/L xylose. The time course of the 30 FPU/g DM WS hydrolysis reaction is depicted in Figure 2 B. In addition to glucose and xylose, a considerable amount of cellobiose (5.0 g/L, Figure 2 ) was released into the hydrolyzate. Cellobiose is known to have an inhibitory impact on cellulases (e.g., [ 38 , 39 ]). Accumulation of cellobiose during hydrolysis is caused by limitation in beta-glucosidase activity in the enzyme mixture used. To evaluate the impact of enhanced beta-glucosidase activity, hydrolysis reactions (30 FPU/g DM WS ) were additionally performed with supplemented Novozyme188. The FPU to beta-glucosidase activity was 1, which was chosen according to the literature [ 34 , 35 ]. The resulting hydrolyzate had a glucose and xylose concentration of 51.8 ± 1.1 g/L and 23.0 ± 1.0 g/L, respectively. The cellobiose concentration was below 1.5 g/L. Addition of beta-glucosidase increased the ηGlucose from 67% to 78%. The already high ηXylose was further increased and reached full conversion. Fermentation to ethanol with S. cerevisiae strain IBB10B05 S. cerevisiae IBB10B05 proved to be a sturdy and efficient fermentation strain for mixed glucose-xylose fermentation in spent sulfite liquor, wheat straw hydrolyzates, and a combination thereof [ 19 , 27 ]. Strain IBB10B05 performs excellently under most basic process and substrate conditions. Thus, fermentations were run in simple batch cultures without process monitoring and control (e.g., pH adjustment). The 15% hydrolyzate was applied without pre-treatment (e.g., detoxification) and yeast extract was added as a sole substrate supplement. An in-depth physiological characterization of strain IBB10B05 in fermentation of 15% hydrolyzates has been published recently [ 19 ]. The fermentation time course is depicted in Additional file 1 . Both glucose (~37 g/L) and xylose (~19 g/L) were depleted within 50 h of fermentation, and in total, 22 g/L ethanol was produced within this time frame (Additional file 1 , [ 19 ]). This represents an ethanol yield of ~0.4 g/g Glc + Xyl . Integration of unit operations and process analysis The single unit operations, enzyme production, hydrolysis, and fermentation, were integrated to one SHCF process as depicted in Figure 1 . To evaluate and compare the different process configurations in context of the complete SHCF process, mass balance analysis was performed. The different process configurations are described based on critical output parameters (FPU/mL, ηGlucose, ηXylose, ηHandling, and ethanol yield ( Y Ethanol )), and the corresponding mass balance analyses are summarized in Additional file 2 . A comparison of process configurations based on Y Ethanol-Process is depicted in Figure 3 . Note that to account for losses caused by pre-treatment and to facilitate comparison of the laboratory scale SHCF process with data from the literature, mass balance analyses were based on the raw material. It was assumed that the required input of raw material was 10% higher as the input calculated for the pre-treated wheat straw (Additional file 2 ). Throughout this study, Y Ethanol-Process is given in g ethanol produced per kg raw material (g Ethanol /kg DM RM ). Figure 3 \n The influence of different process configurations on \n Y \n Ethanol-Process \n . Detailed description of process parameters are summarized in Additional file 2 . Enzyme production, the first unit operation of the SHCF process, was run as batch or as fed-batch. The latter approach resulted in a higher volumetric activity but also required a higher substrate loading (‘ Production of (hemi-) cellulases by T. reesei SVG17 ’). When enzymes were produced in batch fermentations (Config.1), the resulting Y Ethanol-Process was 71.2 g Ethanol /kg DM RM . In fed-batch fermentations (Config.2), Y Ethanol-Process was 58.0 g Ethanol /kg DM RM , which is a 19% decrease as compared to Config.1. This clearly emphasizes the need for evaluating process parameters in context of the complete process. Despite the extensive research on cellulase production in T. reesei (e.g., [ 29 , 32 , 40 , 41 ]), the influence of substrate or process conditions on the success of the overall bioethanol production process is scarcely considered. Enzymatic hydrolysis, the second unit operation, was analyzed with two different enzyme loadings, 25 FPU/g DM WS (Config.3) and 30 FPU/g DM WS (Config.1). Config.3 resulted in an overall process ethanol yield of 67.5 g Ethanol /kg DM RM (Additional file 2 , Figure 3 ). This is 5% less as compared to the reaction with 30 FPU/g DM WS ( Y Ethanol-Process 71.2 g Ethanol /kg DM RM ). The last unit operation, the fermentation to ethanol, was accomplished with the xylose-fermenting S. cerevisiae strain IBB10B05 (Figure 1 ). To compare the efficiencies of a SHCF and a separate hydrolysis and fermentation (SHF) process, mass balance analysis was additionally performed with glucose conversion only (Config.4). With an ethanol on glucose yield of 0.45 g/g Glc [ 19 , 27 ], the Y Ethanol-Process was 54.6 g Ethanol /kg DM RM (Additional file 2 , Figure 3 ). This is 23% less as compared to Config.1. Although, genetically engineered xylose-fermenting S. cerevisiae strains have been described extensively in the literature (e.g., [ 10 , 15 ]), pilot plants often still operate with non-GM yeasts and rely on glucose conversion only [ 1 , 23 ]. This integrative process analysis, however, clearly highlights the importance of efficient conversion of all sugars in the hydrolyzates and strain IBB10B05 proved to be an excellent candidate. Integration of unit operations and analyses of the different process configurations (Config.1 to 3) showed that Config.1 has the highest Y Ethanol-Process (Figure 3 ) and the mass balance analysis is depicted in Table 2 . Thus, enzyme production in batch fermentation (working volume 4 L) and processing of the enzyme solution (ηHandling 75%) resulted in a total of 5,100 FPU. With an enzyme loading of 30 FPU/g DM WS , it was possible to hydrolyze 170 g DM WS. After treatment of the 15% hydrolyzate (ηHandling 75%), 39.5 g glucose and 18.4 g xylose were available for fermentation, which was converted to 23.1 g of ethanol. A total process ethanol yield of 71.2 g Ethanol /kg DM RM was reached (Table 2 ). Table 2 \n Mass balance analysis of the benchtop SHCF process \n \n Input \n \n Output \n \n 1st step: enzyme production ( \n T. reesei \n SVG17) \n Pre-cultures (WS pre-culture ) 5.6 g DM WS Batch cultivation (WS fungal ferm ) 120 g DM WS Total cellulolytic activity 1.7 FPU/mL ηHandling 75% 5,100 FPU total \n 2nd step: saccharification (enzymatic hydrolysis) \n Substrate loading (WS enz hydrolysis ) 170 g DM WS Glucose (ηGlucose 67%, ηHandling 75%) 39.5 g Xylose (ηXylose 95%, ηHandling 75%) 18.4 g \n 3rd step: ethanol production ( \n S. cerevisiae \n IBB10B05) \n Ethanol ( Y \n Ethanol 0.4 g/g Glc+Xyl ) 23.1 g \n Total \n Substrate loading - pre-treated 295.6 g DM WS Substrate loading - raw material 325.2 g DM RM Ethanol 23.1 g \n Y \n Ethanol-Process \n 71.2 g Ethanol /kg DM RM \n An overview of the process is depicted in Figure 1 . Boundary conditions: batch fermentations with 30 g DM WS /L, 15% DM WS, and 30 FPU/g DM WS. \n To assess the efficiency, the benchtop SHCF process was compared to currently available data from pilot (SEKAB, IBUS, BCyL) and commercial (Clariant, ‘SunLiquid’) scale plants. A summary of process configurations, plant capacities, and Y Ethanol-Process is depicted in Table 3 . The pilot scale plants summarized in Table 3 operate without ‘on-site’ production of (hemi-) cellulolytic enzymes. To still render a comparison of the process efficiencies possible, mass balance analysis of the laboratory scale SHCF process was performed excluding enzyme production (Config.5, Figure 3 and Additional file 2 ). Without the loss of feedstock required for the fungal cultivation, the Y Ethanol-Process was 123.7 g Ethanol /kg DM RM , which is 1.7-fold higher as compared to Config.1. This is already within the range of Y Ethanol-Process reported for the pilot scale plants (118 to 157.8 3 g Ethanol /kg DM RM ; Table 3 ), clearly highlighting the usefulness of the herein presented laboratory scale SHCF process as a model for establishing an integrative process analysis. However, direct comparison of Config.1 and Config.5, solely based on Y Ethanol-Process , is not sufficient. ‘On-site’ enzyme production has been described to entail several advantages, such as being cost-effective and efficient [ 26 , 42 ]. Thus, for further evaluation of the feasibility of the process configurations, a detailed economic analysis, as published for other process [ 2 , 14 , 22 ], must be conducted. Table 3 \n Y \n Ethanol-Process \n for commercial, pilot, and laboratory scale processes \n \n Substrate \n \n Capacity [1,000 t \n DM RM \n /year] \n \n Process \n \n Y \n Ethanol-Process \n [kg \n Ethanol \n /t \n DM RM \n ] \n \n Reference \n Cereal straw 150 SHCF + E a \n 222.2 Clariant ‘SunLiquid’ [ 42 ] Wheat straw 8.8 SSF 123 ‘IBUS’ [ 23 ] Forestry residues 0.7 SSF 118 to 157.8 ‘SEKAB’ [ 1 ] Wheat straw 25.6 SHF 154 ‘BCyL’ [ 1 ] Wheat straw - SHCF/+E a \n 123.7/71.2 This study \n a ‘On-site’ enzyme production. The by far most efficient process is the ‘SunLiquid’ process, which is operated as SHCF with implemented ‘on-site’ enzyme production. The reported Y Ethanol-Process is 222.2 g Ethanol /kg DM RM (Table 3 ) [ 42 ]. This is 18% higher as compared to the maximum theoretical yield of the benchtop SHCF process which was determined to be 183.1 g Ethanol /kg DM RM (Config.8, Figure 3 and Additional file 2 ). There are several factors that could explain the exceptionally high Y Ethanol-Process described for the ‘SunLiquid’ process. The first factor is the pre-treatment method applied. Ideally, pre-treatment of lignocellulosic biomass enriches the structural carbohydrate cellulose and hemicellulose by removal of the lignin and enhances the accessibility of the partially crystalline cellulose. It therefore has an impact on the efficiency of the fungal fermentation and the enzymatic hydrolysis and is directly affecting Y Ethanol-Process [ 7 , 11 ]. In this study, the steam explosion was performed as batch and running the process continuously could reduce mass losses caused by pre-treatment. However, evaluation of varying pre-treatment methods was beyond the scope of this study and is not considered in more detail hereinafter. The second factor influencing Y Ethanol-Process is the potential application of enzyme or solid recycling to boost the hydrolysis efficiency [ 43 , 44 ]. This process option was also analyzed in context of this study. With an estimated increase in glucose released per enzyme loading of 35% (e.g., [ 43 , 44 ]), the Y Ethanol-Process of the benchtop SHCF process would improve to 87.8 g Ethanol /kg DM RM (Config.6, Figure 3 and Additional file 2 ). The third factor is the organism employed for enzyme production. The importance of a T. reesei strain secreting a balanced enzyme mixture has been described in the literature [ 34 - 37 ]. In this study, the application of an enzyme solution with a beta-glucosidase to FPU ratio of 1 was investigated. The resulting Y Ethanol-Process was 80.3 g Ethanol /kg DM RM (Config.7, Figure 3 and Additional file 2 ). In comparison to Config.1, the altered process configurations resulted in an increase in Y Ethanol-Process of 19% (Config.6) and 11% (Config.7). However, the process yields did not exceed 50% of the reported Y Ethanol-Process of the ‘SunLiquid’ process. Therefore, we took the process analysis one step further and analyzed the benchtop SHCF process towards potential bottlenecks. Based on the detailed mass balance analysis of the integrated benchtop SHCF process and the resulting understanding of the process streams, it was possible to identify three key parameters which exhibit a significant influence on Y Ethanol-Process , the enzyme yield, ηGlucose, and ηHandling (Figure 1 ). The correlations between Y Ethanol-Process and these parameters are depicted in Figure 4 . Firstly, Y Ethanol-Process was plotted over the enzyme yield (FPU/mL) (Figure 4 A). To increase the Y Ethanol-Process of Config.1 to 100 g Ethanol /kg DM RM , for an instance, a 3.1-fold increase in total cellulolytic activity is required (from 1.7 to 5.3 FPU/mL, Figure 4 A). The importance of an efficient enzyme production is further highlighted, by comparing the maximum theoretic Y Ethanol-Process observed in this study with the ‘SunLiquid’ process. To enhance the Y Ethanol-Process of Config.8 to 222 g Ethanol /kg DM RM [ 42 ], the enzyme yield must be increased twofold, from 1.7 FPU/mL to 3.4 FPU/mL. In addition to the above discussed factors, a higher enzyme yield during fungal fermentation might therefore explain for the high Y Ethanol-Process of the ‘SunLiquid’ process. Figure 4 \n The influence of key parameters on \n Y \n Ethanol-Process \n . (A) \n Y \n Ethanol-Process in dependence of the total cellulolytic activities produced by fungal fermentation (boundary conditions: batch with 30 g DM WS /L, 30 FPU/g DM WS, ηGlucose 67%, ηXylose 95%, ηHandling 75%, and Y \n Ethanol 0.4 g/g Glc+Xyl ). (B) Influence of ηGlucose and ηHandling on Y \n Ethanol-Process (boundary conditions: batch with 30 g DM WS /L, 1.7 FPU/mL, 30 FPU/g DM WS , ηXylose 95%, and Y \n Ethanol 0.4 g/g Glc+Xyl ). As shown in Figure 4 B, the overall process efficiency is further influenced by the two parameters ηGlucose and ηHandling. Thus, Y Ethanol-Process can vary between 22.4 (ηGlucose 25% and ηHandling 50%) and 127.2 g Ethanol /kg DM RM (ηGlucose 95% and ηHandling 95%) under else same boundary conditions. Possible improvement of ηGlucose has already been described within this study. Data summarized in Table 3 suggest that Y Ethanol-Process increases with the scale of the plant (‘Capacity’). A correlation between plant capacity and overall yield is further supported by the literature [ 1 - 5 , 14 , 22 - 24 ], and it is suggested that processes are becoming more efficient with increasing scale [ 1 - 3 , 5 ]. Explicit information on ηHandling for pilot and commercial scale plants are not given. The fact that Y Ethanol-Process of Config.5 (“no enzyme production”) is within the range of to the pilot scale plants (Table 3 ), however, might indicate that the ηHandling are similar. To further increase Y Ethanol-Process , improvement of ηHandling will be a target of future optimization studies of the benchtop SHCF process."
} | 6,523 |
19287491 | PMC2654016 | pmc | 3,855 | {
"abstract": "Background There is accumulating evidence that in some marine environments aerobic bacteriochlorophyll a -producing bacteria represent a significant part of the microbial population. The interaction of photosynthesis and carbon metabolism in these interesting bacteria is still largely unknown and requires further investigation in order to estimate their contribution to the marine carbon cycle. Methodology/Principal Findings Here, we analyzed the structure, composition and regulation of the photosynthetic apparatus in the obligately aerobic marine gammaproteobacterium KT71 T . Photoheterotrophically grown cells were characterized by a poorly developed lamellar intracytoplasmic membrane system, a type 1 light-harvesting antenna complex and a photosynthetic reaction center associated with a tetraheme cytochrome c . The only photosynthetic pigments produced were bacteriochlorophyll a and spirilloxanthin. Under semiaerobic conditions KT71 T cells expressing a photosynthetic apparatus showed a light-dependent increase of growth yield in the range of 1.3–2.5 fold. The expression level of the photosynthetic apparatus depended largely on the utilized substrate, the intermediary carbon metabolism and oxygen tension. In addition, pigment synthesis was strongly influenced by light, with blue light exerting the most significant effect, implicating that proteins containing a BLUF domain may be involved in regulation of the photosynthetic apparatus. Several phenotypic traits in KT71 T could be identified that correlated with the assumed redox state of growing cells and thus could be used to monitor the cellular redox state under various incubation conditions. Conclusions/Significance In a hypothetical model that explains the regulation of the photosynthetic apparatus in strain KT71 T we propose that the expression of photosynthesis genes depends on the cellular redox state and is maximal under conditions that allow a balanced membrane redox state. So far, bacteria capable of an obligately aerobic, photosynthetic metabolism constitute a unique phenotype within the class Gammaproteobacteria, so that it is justified to propose a new genus and species, Congregibacter litoralis gen. nov, sp. nov., represented by the type strain KT71 T ( = DSM 17192 T = NBRC 104960 T ).",
"introduction": "Introduction The oceans harbor a huge population of diverse microorganisms that are involved in the global cycling of carbon. A study of the major players participating in the marine carbon cycle is important to estimate the evolving capacity of oceans as carbon dioxide sink. In the last few years numerous cultivation-independent studies have been published that illustrate the abundance, diversity and distribution of aerobic bacteriochlorophyll a -producing bacteria in marine environments [ e.g. 1] – [3] . However, many questions regarding their evolution and importance for the carbon cycle remain unanswered. For instance, the environmental factors that determine expression of bacteriochlorophyll genes in aerobic bacteria are only partly understood [4] , [5] . The impact of light on the growth yield of most species is also still unknown, although the number of aerobic bacteriochlorophyll-containing bacteria available in pure culture and thus accessible to a phenotypic characterization is steadily increasing [6] . On the other hand, it is widely accepted that environmental data solely indicating a presence of bacteriochlorophyll-containing bacteria in the aerobic euphotic zone of oceans are not sufficient to make reliable estimates of the extent of light-driven carbon assimilation. Furthermore, the available physiological studies on marine aerobic anoxygenic photosynthetic bacteria are restricted to isolates belonging to distinct clades of the Alphaproteobacteria \n [for review see 7] , so that the derived assumptions may be biased. Hence, a detailed analysis of diverse isolates representing abundant clades of marine aerobic bacteriochlorophyll-containing bacteria is essential and a prerequisite for a deeper understanding of this interesting phenotype. Only a profound knowledge of this type of metabolism will allow us to evaluate various evolutionary models that currently try to explain the origin and function of anoxygenic photosynthesis in aerobic marine bacteria [ e.g. 8] , [9] . We have chosen strain KT71 T as a model organism to study aerobic anoxygenic photosynthesis, because it represents a hitherto unrecognized group of marine bacteriochlorophyll-containing gammaproteobacteria [2] . A preliminary description of strain KT71 T based on a draft genome sequence has been published and it was shown that this isolate represents a novel taxon tentatively named “ Congregibacter litoralis ” [10] . Phylogenetically, KT71 T is affiliated to a major cluster of environmental 16S rRNA gene sequences that were retrieved mainly from marine habitats. This phylogenetic group has been designated OM60 by Rappé et al. [11] or NOR5 by Eilers et al. [12] and belongs to the class Gammaproteobacteria . The phylogenetic position of KT71 T among cultured representatives of non-photosynthetic and photosynthetic gammaproteobacteria is shown in Figure 1 . Currently, the most closely related classified species are Haliea salexigens \n [13] and Spongiibacter marinus \n [14] , which have both been described as non-photosynthetic bacteria. Approaches using in situ hybridization of whole cells with specific oligonucleotide probes revealed that members of this cluster are metabolically active in marine environments [12] , [15] and thus can play a major role in the littoral and euphotic zones of the oceans. The presence of bacteriochlorophyll genes in aerobic gammaproteobacteria other than KT71 T has been demonstrated by metagenomic [16] or cultivation-based studies (strains HTCC2080 [17] and EG19 [18] ). Thus, besides the well-known alphaproteobacterial Roseobacter clade [19] , [20] a second phylogenetic group of aerobic anoxygenic photosynthetic bacteria in marine environments is emerging, of which KT71 T is one cultured representative. In the course of a detailed characterization of this isolate, it was found that the synthesis of photosynthetic pigments was occasionally induced following a transfer from complex medium to defined medium containing only a single carbon source. Among various carbon sources tested malate had the most stimulating effect on the production of photosynthetic pigments. Based on this finding photoheterotrophically growing cultures of KT71 T could be obtained, which were used for an extensive analysis of the composition and regulation of the photosynthetic apparatus in this bacterium. The results obtained in this study allow the proposal of a metabolic model that helps to understand the function of anoxygenic photosynthesis in KT71 T . Furthermore, the proposed concept may provide some clues to the evolution, distribution and ecological importance of aerobic anoxygenic photosynthetic bacteria in ocean waters. 10.1371/journal.pone.0004866.g001 Figure 1 Phylogenetic dendrogram illustrating the position of Congregibacter litoralis KT71 T among cultured representatives of the class Gammaproteobacteria . Unless noted otherwise almost-complete 16S rRNA gene sequences of type strains were used for tree reconstruction. The sequence of Escherichia coli (X80725) was used as outgroup (not shown). Phylogenetic analyses were based on the alignment included in the SILVA database (release 95) [89] and trees were reconstructed using distance matrix (neighbor-joining), parsimony and maximum-likelihood algorithms using programs implemented in the ARB software package [90] . The dendrogram shown was reconstructed using a maximum-likelihood program (fastDNAml) and represents a branching order that was also frequently obtained when alternative programs were used. Representatives of anaerobic anoxygenic photosynthetic bacteria are emphasized with red lettering, whereas aerobic bacteriochlorophyll a -producing bacteria are shown in blue lettering. The availability of a genome sequence is indicated by an orange frame. Notes: $, several genome sequences of Pseudomonas aeruginosa strains are currently available, but none was obtained from the type strain. §, until now no 16S rRNA gene sequence of the type strain of Ectothiorhodospira mobilis is available so that the sequence of DSM 4180 was chosen.",
"discussion": "Discussion Correlation of Photosynthesis with the Cellular Redox State In facultatively anaerobic photosynthetic proteobacteria several redox balancing mechanisms are present that prevent a decrease of the cellular redox state to unfavorable levels. These systems are either based on enzymes of the electron transport chain like ubiquinol oxidase and fumarate reductase or metabolic pathways like the assimilation of CO 2 by the Calvin-cycle and fixation of nitrogen that require a large amount of reduction equivalents and ATP [60] . In KT71 T none of these mechanisms could be detected, neither by annotation of the draft genome sequence nor by physiological tests, so that no redox balancing system except for the universal thioredoxin buffer system [61] seems to be present. The identification of cellular redox monitors in KT71 T allowed us to estimate the cellular redox state of KT71 T cells growing on various substrates. It turned out that the intracellular redox state in this strain is variable and depends largely on the growth conditions. In the course of this study it was not possible to detect a principal correlation of the substrate type with an assumed cellular redox state, which would be based merely on theoretical considerations like a different utilization of biochemical pathways. However, most of the obtained results can be explained intuitively in the following way: We assume that the redox state in KT71 T is mainly controlled by the amount of reduction equivalents derived from the substrate and the respiratory activity. The utilization of highly reduced substrates like sucrose induces in cells an intracellular accumulation of NADH along with a high respiratory activity, whereas the utilization of substrates with a very low energy yield like acetate causes an inhibition of the respiration rate in order to avoid a harmful overoxidation of the cellular redox state, which in turn leads also to an intracellular accumulation of reduction equivalents. In some cases, if substrates with an intermediate energy yield like malate are utilized the redox state may drop to suboptimal levels, which, however, can still be tolerated without causing an inhibition of respiration. Most of the obtained data indicate that one of the main reasons for the repression of photosynthetic pigment production in KT71 T is a significant decrease or increase of the cellular redox state compared to the redox state in photoheterotrophically growing cells. Candela et al. [41] have demonstrated that the generation of ATP in photosynthetic membranes requires a balanced redox state that corresponds to a reduction of the membrane-bound quinone pool of about 50%, which is optimally for the promotion of a photosynthetic Q-cycle. In addition, they have shown that an overreduction or overoxidation of the membrane redox state leads not only to a prevention of the light-induced generation of energy in facultative anaerobic, but also in aerobic photosynthetic proteobacteria. Hence, the expression of photosynthesis genes in bacteria representing both types of metabolism only makes sense, if it is possible to establish a balanced membrane redox state for an extended period of time. This finding could provide an explanation for the assumed correlation of photosynthesis gene expression with the cellular redox state in KT71 T . Further support for this hypothesis was obtained from the observation that treatment of cultures with chemicals that prevent the oxidation of NADH (KCN) or induce oxidative stress (paraquat) resulted both in a strong repression of pigment production in KT71 T cells (unpublished results). It follows that any sensor-regulator protein that controls the expression of photosynthesis genes in this strain should be able to sense equally well an overreduction and overoxidation of the cellular redox state. We propose that the PpsR protein (annotated ORF: KT71_19353) is the most likely candidate to fulfill this function. The PpsR protein encoded in the genome of KT71 T contains two adjacent PAS (Per-Arnt-Sim) domains, a C-terminal HTH (helix-turn-helix) motif probably mediating DNA binding and two cysteine residues at positions 349 and 412, which are thought to play a key role in redox sensing along with the PAS domains [62] . DNA binding sites of PpsR can be recognized by the conserved palindromic sequence TGTcaN 8 tgACA [63] . In several species studied two adjacent palindromes are found overlapping with putative σ 70 – type promoters where the binding of PpsR causes a repression of photosynthesis genes located downstream of the operator sequences [63] . An analysis of the PGS nucleotide sequence of KT71 T revealed a total of 15 putative PpsR binding sites, of which 12 were arranged in a typical tandem orientation separated by seven or eight nucleotides ( Figure S5 ). The number of binding sites is comparable to that found in facultative anaerobic phototrophs [64] and indicates that PpsR plays a pivotal role in controlling the transcription level of photosynthesis genes in KT71 T . The following observations reported previously about the PpsR protein in other purple bacteria seem to support the view that PpsR can operate as “bidirectional” redox sensor: In Rhodobacter sphaeroides the PpsR protein was found to bind DNA not only in the oxidized state, but also under reducing conditions [65] , which is in clear contrast to its proposed function as an aerobic repressor [62] , but in line with our model of a protein sensing a distinct redox state. In addition, the PpsR1 protein in the strictly aerobic Bradyrhizobium strain ORS278 was found to have a higher DNA affinity in its reduced than in its oxidized form [66] . Hence, one intrinsic property of the PpsR protein architecture seems to be the ability to bind DNA under reducing and oxidizing conditions. It is known that the redox balancing system of facultative anaerobic purple bacteria prevents an overreduction of the cellular redox state under anaerobic conditions in the light, so that normally only the aerobic repression of photosynthesis genes is observed in laboratory cultures. Consequently, it could well be that the repression of photosynthesis genes due to an overreduction of the redox state was not realized in most studied photosynthetic proteobacteria and thus not studied in detail so far. In addition to the PpsR protein Rhodobacter species contain a two-component response regulator system (RegA/RegB or PrrA/PrrB), which is not present in KT71 T . In Rhodobacter sphaeroides the two-component activation system PrrBA is controlled by a cbb 3 -type terminal oxidase [67] . Increasing concentrations of oxygen enhance the electron flow through the cbb 3 oxidase, which generates a signal that prevents phosphorylation of the PrrA protein by the membrane-bound PrrB histidine kinase. This leads to a repression of photosynthesis, because the phosphorylated form of PrrA is required to stimulate transcription of photosynthesis genes [68] . We have found that in KT71 T the expression of a photosynthetic apparatus coincided with the expression of an active cbb 3 -type terminal oxidase. We speculate that this oxidase senses the electron flow under conditions of a balanced redox state and generates a signal that stimulates expression of photosynthesis genes. A potential signal could be the intracellular concentration of the second messenger c-di-GMP, because in the genome of KT71 T a large gene encoding GGDEF/EAL output domains (annotated ORF: KT71_16971) was found upstream close to the operon encoding a cbb 3 oxidase of the sensing type (annotated ORFs: KT71_16991 – KT71_17006). It is known that proteins containing GGDEF/EAL domains control the level of c-di-GMP in bacterial cells [69] . However, the mechanisms that control the activity of these proteins are still largely unknown and could for example involve an interaction with terminal oxidases. In the draft genome of KT71 T two genes were annotated that encode a BLUF (blue-light-using flavin adenine dinucleotide) domain [70] , [71] , which could have a function in the light-dependent control of pigment expression. One of both genes (annotated ORF: KT71_19323) is located close to the gene encoding PpsR. The negative effect of blue light on the expression level of the photosynthetic apparatus was quite independent from the initial oxygen tension, which indicates that probably no additional redox sensor is involved in the light-dependent regulation of pigment expression as it was shown in Rhodobacter sphaeroides . In this species repression of pigment synthesis in response to blue light depends on the interaction of the AppA protein that simultaneously senses oxygen and light with the PpsR repressor [45] , [72] . However, no open reading frame with significant similarity to the appA gene was detected in the draft genome sequence of KT71 T , so that this type of light-dependent regulation of pigment expression is not applicable to this bacterium. A protein with a single BLUF domain was also identified in the genome of the aerobic anoxygenic photosynthetic bacterium Roseobacter denitrificans in which pigment synthesis was repressed selectively by illumination with blue light [7] , although no gene with similarity to the appA gene was identified [64] . Hence, it may be speculated that in some photosynthetic proteobacteria the single domain BLUF protein could interact with PpsR in an unknown way in order to control pigment synthesis in response to the light intensity. In conclusion, it appears that the PpsR protein represents a terminal effector of a complex regulatory network that depends on the cellular redox state, activity of terminal oxidases, quinone reduction and light intensity. Consequently, the induction of photosynthesis genes would only occur if optimal conditions are sensed by various signal cascades, which could prevent a frequent and energy expensive change of the gene expression pattern. A summary of the proposed model for the regulation of photosynthesis in KT71 T is presented below: KT71 T represents a heterotrophic, obligately aerobic photosynthetic bacterium lacking a redox balancing system, which results in a variable cellular redox state that depends mainly on the type of utilized substrate, intermediary carbon metabolism and oxygen concentration. Some unidentified cellular signals that correlate with the redox state and/or the respiratory activity are sensed by the transcription regulator protein PpsR, which has the function to repress the transcription of photosynthesis genes under conditions that lead to an overreduced or overoxidized membrane redox state, because both states prevent an efficient generation of energy from light. A single domain BLUF protein acts as a sensor of blue light and modulates the activity of the PpsR protein in a yet unknown way. The pigment stoichiometry in photosynthetically active cells is controlled by the reduction of the ubiquinone pool, which depends mainly on the availability of oxygen and light. Photosynthesis in KT71 T may function to prevent an unfavorable overoxidation of the cellular redox state by photoreduction of NAD + involving an energy dependent reverse electron flow. In the future more advanced analyses of the gene expression in KT71 T are planned to verify the postulated metabolic model and to arrive eventually at a complete understanding of the molecular mechanisms that control the expression of photosynthesis genes in members of the abundant OM60/NOR5 clade within the Gammaproteobacteria . Taxonomic Status of Strain KT71 T \n In reconstructed phylogenetic trees strain KT71 T belongs to a branch comprising the recently described species Haliea salexigens and Spongiibacter marinus as well as numerous 16S rRNA gene sequences retrieved from uncultured marine bacteria ( Figure 1 ). An affiliation of this sequence cluster to any family described currently within the class Gammaproteobacteria is not supported by high bootstrap values in reconstructed maximum-parsimony or distance-matrix trees. Thus, members of this group represent most likely a novel family. Similarity values between the 16S rRNA gene sequences of KT71 T and the most closely related species Haliea salexigens and Spongiibacter marinus are only 95.4 and 91.0%, respectively, thereby excluding a relationship at species level [73] . In addition, the presented data clearly indicate that strain KT71 T can be distinguished phenotypically from related genera, so that we propose the formal classification in a novel genus and species. A comparison of distinguishing features of Haliea salexigens and the proposed species Congregibacter litoralis is shown in Table 3 and formal descriptions are presented below. 10.1371/journal.pone.0004866.t003 Table 3 Distinguishing phenotypic traits of Congregibacter litoralis KT71 T and Haliea salexigens 3X/A02/235 T . Characteristic KT71 T \n 3X/A02/235 T \n Isolation source North Sea (8 m water depth) Mediterranean Sea (water surface) Cell shape pleomorphic straight rods Cell size (µm) 0.5–4.5×0.4–0.7 1.3–1.9×0.3–0.7 Flagellation type 1–2 (sub)polar 1 polar \n Growth Characteristics \n NaCl range (%) 1–7 0.7–7 NaCl optimum (%) 2 4.2 Temp. range (°C) 9–33 10–37 Temp. optimum (°C) 28 25–30 pH range 6.5–9.0 5.0–9.0 pH optimum 7.5–8.0 8.0 Anaerobic Growth − − Growth on marine agar 2216 + + Photosynthetic pigments + n.d. Oxidase + + Catalase (+) + \n Carbon Source Utilization \n Cellobiose − − Galactose + − Glucose − − Sucrose + − Glycerol + + Acetate (+) − Citrate − − 3-Hydroxybutyrate + + Propionate (+) − Pyruvate + + Succinate + + Alanine + − Aspartate + + Glutamate + (+) Proline + + Serine + − \n Chemotaxonomy \n Major fatty acids 16∶0, 18∶1, 16∶1, (17∶1) 17∶1, 16∶1, 18∶1 Major 3-OH fatty acid 10∶0 3OH 11∶0 3OH Polar lipids PE, PG, PL DPG, PG, APL Quinone UQ8 UQ8 G+C content of DNA (mol%) 57.8 61.4 Data for H. salexigens were taken from Urios et al. [13] . +, positive; (+), weakly positive; −, negative; n.d., not detected; PE, phosphatidylethanolamine; PG, phosphatidylglycerol; PL; unknown phospholipid; DPG, diphosphatidylglycerol; APL, unknown aminophospholipid. Description of Congregibacter gen. nov \n Congregibacter (Con.gre.gi.bac'ter. L. adj. congregus -a -um , united in flocks; N.L. masc. n. bacter , a rod; N.L. masc. n. Congregibacter a rod that grows in flocks). Cells are Gram-negative, non-spore-forming and multiply by binary fission. Aggregates are frequently formed in liquid medium under suboptimal growth conditions, especially carbon starvation [10] . Mesophilic and moderately halophilic. Strictly aerobic, respiratory and heterotrophic metabolism. Tests for oxidase and catalase activity are positive. Cytochromes of the c -type are dominating in redox difference spectra. Bacteriochlorophyll a and carotenoids of the spirilloxanthin series are produced in photosynthetically active cells. In the presence of photosynthetic pigments light stimulates growth under semiaerobic conditions. The production of photosynthetic pigments is not repressed in aerobically growing cells by illumination with dim light, i.e. below 2000 lux of incandescent light (equivalent to 40 µE m −2 s −1 ). Under certain incubation conditions water-insoluble polar pigments with a pale yellow to orange-red color can be formed. Major cellular fatty acids are C 16∶0 , C 16∶1 and C 18∶1 . The dominating hydroxy fatty acid is C 10∶0 3OH. Ubiquinone 8 represents the sole respiratory lipoquinone. Phosphatidylethanolamine, phosphatidylglycerol and an unidentified phospholipid are the major polar lipids. Representatives can be found in sea water and the surface layer of littoral marine sediments. The type species is Congregibacter litoralis . Description of Congregibacter litoralis sp. nov \n Congregibacter litoralis (li'to.ra.lis. M.L. adj. litoralis from the shore, pertaining to the habitat from where the organism was isolated). In addition to traits noted for the genus the following characteristics were determined. Cells are pleomorphic and depending on the growth conditions either coccoid or irregular rod-shaped with rounded ends. The length of cells can vary between 0.5 and 4.5 µm and the width between 0.4 and 0.7 µm. Motility is conferred by one or two polar to subpolar flagella. The reserve polymers cyanophycin [10] and polyphosphate are stored in intracellular granules. The color of culture suspensions is variable and reflects mainly the production of photosynthetic pigments. Under conditions that allow maximal expression of BChl a and spirilloxanthin suspensions are pale-pink to deep-red depending on the cell density. Yellow-orange or cream-white suspensions are obtained, if the expression of pigments is repressed. In fully pigmented photoheterotrophically growing cells a rudimentary intracytoplasmic membrane system of the lamellar type is formed. Living cells of photoheterotrophically growing cultures showed absorption maxima in the near-infrared region of the spectrum at 802 and 876 nm indicating the presence of a reaction center and light-harvesting 1 complex. Colonies appear on plates of marine agar (DIFCO 2216) after 48 to 72 hours incubation at 28°C and can reach a size of 1–2 mm. They have a round shape with regular edges and are cream colored, thin, slightly convex and soft. Growth was observed between 9 and 33°C and at sea salts concentrations ranging from 1 to 15% (w/v). In media containing 10 mM MgSO 4 the range of suitable NaCl concentrations was 1–7% (w/v) with an optimum at 2% (w/v) NaCl. The requirement for salts is complex and sodium, chloride and either magnesium or calcium ions were needed for growth. No growth occurs below pH 6.0 or above pH 9.5. Optimal conditions for growth are 28°C, a sea salts concentration of around 4% (w/v) and a pH value between 7.5 and 8.0. The mean generation time under optimal growth conditions is 4.5 h. Facultatively microaerophilic; high oxygen concentrations inhibit growth under oligotrophic conditions. No growth under anaerobic conditions with fumarate, nitrate, ferric iron, sulfur, thiosulfate or sulfate as terminal electron acceptors or by fermentation. Excretion of a transparent slime, which causes an increase of medium viscosity, was frequently observed upon incubation in liquid nutrient-rich complex media under conditions of oxygen saturation. The polymers casein, gelatin, starch, cellulose, alginate, agar and DNA were not degraded. Growth depends on organic substrates that are utilized as carbon and energy source; no growth was observed with hydrogen, thiosulfate, sulfur or ferrous iron as electron donor and CO 2 as carbon source. The following compounds were used for growth: alkanes: decane (weak), dodecane (weak) and octane (weak); alcohols: glycerol and propanol (weak); carboxylic acids: acetate (weak), butyrate (weak), dodecanoate, heptanoate, DL-3-hydroxybutyrate, hexanoate (weak), DL-malate, oleate, oxaloacetate, 2-oxoglutarate, palmitate, pentanoate, pyruvate, propionate (weak) and succinate; amino acids: D-alanine, L-alanine, D-arginine (weak), L-arginine, L-asparagine, D-aspartate (weak), L-aspartate, L-cysteine (weak), L-glutamate, glutathione, L-proline and L-serine; carbohydrates: D-galactose and sucrose. The following compounds were tested, but not utilized: alkanes: hexane, hexadecane and tetradecane; alcohols: meso-erythritol, ethanol, myo-inositol, D-mannitol, methanol and resorcinol; carboxylic acids: acrylate, 2-aminobenzoate, benzoate, citrate, decanoate, formate, glycolate, DL-lactate and octanoate; amino acids: L-cysteate, DL-glycine, L-histidine, L-isoleucine, L-lysine, L-methionine, L-ornithine, L-phenylalanine, D-proline, D-serine, L-valine and taurine; carbohydrates: L-arabinose, cellobiose, D-fructose, D-glucose and melibiose. Tests were positive for tweenase and urease activity, but negative for tryptophanase, arginine dihydrolase and esculinase activity. Degradation of L-cysteine or glutathione did not result in the release of sulfide. Suitable nitrogen sources for growth were ammonium, urea, amino acids and purines, but not molecular nitrogen, nitrate, taurine or thiourea. Requirement for the vitamins thiamine, biotin and B-12. Sensitive to the antibiotics cefalotin, imipenem, chloramphenicol, gentamicin, neomycin, colistin and polymyxin B; resistant to oxacillin, tetracycline, doxycycline, vancomycin, lincomycin and bacitracin. The DNA G+C content of the type strain is 57.8 mol% (HPLC). The type strain is KT71 T ( = DSM 17192 T = NBRC 104960 T ). It was isolated from the water column (8 m depth) of the North Sea near Helgoland (Germany)."
} | 7,292 |
38954490 | PMC11264323 | pmc | 3,856 | {
"abstract": "Carbon dioxide emission and acidification during chemical\nbiosynthesis\nare critical challenges toward microbial cell factories’ sustainability\nand efficiency. Due to its acidophilic traits among workhorse lineages,\nthe probiotic Escherichia coli Nissle (EcN) has emerged\nas a promising chemical bioproducer. However, EcN lacks a CO 2 -fixing system. Herein, EcN was equipped with a simultaneous CO 2 fixation system and subsequently utilized to produce low-emission\n5-aminolevulinic acid (5-ALA). Two different artificial CO 2 -assimilating pathways were reconstructed: the novel ribose-1,5-bisphosphate\n(R15P) route and the conventional ribulose-5-phosphate (Ru5P) route.\nCRISPRi was employed to target the pfk AB and zwf genes in order to redirect the carbon flux. As expected,\nthe CRISPRi design successfully strengthened the CO 2 fixation.\nThe CO 2 -fixing route via R15P resulted in high biomass,\nwhile the engineered Ru5P route acquired the highest 5-ALA and suppressed\nthe CO 2 release by 77%. CO 2 fixation during\n5-ALA production in EcN was successfully synchronized through fine-tuning\nthe non-native pathways with CRISPRi.",
"introduction": "Introduction Growing concerns over environmental problems\nof traditional chemical\nproduction motivate the utilization of sustainable microbial cell\nfactories (MCFs). However, the biological process of some chemicals\ninvolves a decarboxylation reaction and generates a CO 2 byproduct. 1 Releasing CO 2 also\nwould decrease extracellular pH due to the dissociation of carbonic\nacid. 2 , 3 The alteration of acidic pH from pH 8.3\nto pH 4.5 even inhibited bacterial growth up to 5-fold. 4 , 5 For nonacidophilic strains, microorganisms require extra energy\nto maintain pH homeostasis by proton-translocating ATPase, thus reducing\nthe MCF efficiency. 2 , 3 Maintaining pH control during\nfermentation can mitigate acidification effects; however, dead zones\nin large-scale fermenters may hinder precise pH adjustment, thus posing\nchallenges to accuracy. 6 Therefore, incorporating\nan acidophilic strain with simultaneous CO 2 fixation is\na fantastic development to prevent and reduce CO 2 emissions\nduring chemical biosynthesis. 7 Photosynthetic\norganisms are known for naturally accommodating\nCO 2 assimilation. The Calvin–Benson–Bassham\n(CBB) cycle is an ancient CO 2 -fixing route that is highly\nregulated with ribulose-1,5-bisphosphate carboxylase-oxygenase (RuBisCO)\nand phosphoribulokinase (PRK). 8 , 9 Despite this, the workhorse Escherichia coli has shown distinct traits over photoautotrophs\nsuch as five times faster growth, light independence, comprehensive\ngenome databases, and sophisticated synthetic biology toolboxes for\nengineering metabolic pathways. 10 , 11 Former works of engineered E. coli have successfully assimilated 35% CO 2 into biomass via an artificial CBB cycle. The attempts were conducted\nby connecting the route from pentose phosphate (PP) into glycolysis\nalong with utilizing formate as the electron source. 12 Through adaptive evolution over 350 days, engineered E. coli utilized CO 2 as the sole carbon source. 13 Among E. coli lineages,\nNissle 1917 (EcN) is the\nonly probiotic E. coli that possesses acidophilic\nnature. The majority of EcN applications are related to living bacterial\ntherapeutics and bowel resistance. 14 Recently,\nEcN studies have been broadened as a safe chemical producer, thus\nleveraging its exploration for MCF. For instance, EcN has effectively\nproduced a high titer bulk chemical of itaconic acid and high-value\nchemicals of p -coumaric acid, heparosan, aminobutyric\nacid, and 5-aminolevulinic acid (5-ALA). 15 − 19 Of valuable chemicals and drugs, 5-ALA is a metabolic\nhub crucial for heme precursor and also has received approval from\nthe FDA as a second-generation photosensitizer and photodynamic drug\nfor the treatment of glioma. 20 − 22 Biosynthesis of 5-ALA is majorly\naccomplished via the C4 pathway by overexpressing heterologous ALA\nsynthase (ALAS), and CO 2 is a definitive side product. 20 , 21 Previous works of 5-ALA production from glucose have been integrated\nwith the CO 2 biomitigation system by co-overexpressing\nRuBisCO and PRK. 23 , 24 However, rerouting flux from\nglucose into Ru5P would release additional CO 2 , indicating\nthat CO 2 assimilation from chemical biosynthesis has not\nbeen effectively executed. 12 , 25 Moreover, as the strain\nused was E. coli BL21 or K-12, manipulating the acetate\npathway was necessary to hinder a high acetate accumulation and maintain\nstrain durability. 23 , 24 Hence, harnessing the acidophilic\nnature of EcN to express the CO 2 -fixing system and fine-tuning\nits artificial pathway are outright solutions for assimilating CO 2 emission during 5-ALA production. Taken together, this\nstudy exploited two different CO 2 assimilation pathways\nwithin the ribulose route via R15P and Ru5P\nand then connected them with a RuBisCO shunt. The common CO 2 fixation route from Ru5P into 3PG was directed using PRK and RuBisCO\nfrom Synechococcus elongatus PCC6301 (denoted as\nPR). 26 Another novel route between R15P\nand 3PG was bridged by synergizing R15P isomerase (R15PI) from Thermococcus kodakarensis ( 27 ) and\nRuBisCO from S. elongatus PCC6301 (denoted as RR).\nSubsequently, the function and efficiency of both CO 2 -fixing\nsystems were evaluated for assimilating the release of CO 2 during 5-ALA production. To strengthen the CO 2 -fixing\npathway, a carbon flux was derived from xylose instead of glucose\nand further manipulated by using CRISPRi. Different levels of CO 2 were also supplemented into the system. Finally, the efficiency\nof both CO 2 -fixing routes was examined in terms of biomass,\n5-ALA production, and the CO 2 assimilation capabilities.",
"discussion": "Results and Discussion Genetic Design of Artificial CO 2 -Assimilating Pathways\nvia Ribulose Route An artificial CO 2 -fixing pathway\nin EcN was reconstructed at the downstream of the ribulose route to\nprovide an additional carboxylation step, as in former works. 23 , 24 Two different routes, R15P and Ru5P, were added to connect the pentose\nphosphate (PP) pathway to 3-phosphoglycerate (3PG), which is an intermediate\nin glycolysis. The first and second schemes were designated as RR\nand PR, respectively ( Figure 1 A). To reveal whether EcN has a distinguished profile toward\nthe non-native CBB pathways, E. coli MG1655 was selected\nas a comparative host. EcN and MG1655 were equipped with T7 RNA polymerase\n(T7RNAP), yielding MT7 and ET7, for controlling CO 2 -fixing\ngenes under the T7 promoter. Six recombinant strains were cultured\nin the glucose-based medium at pH 5, 6, and 7. At pH 5, all recombinant\nET7 grew faster than MT7 strains even though the pH culture of ET7\nwas slightly lower than MT7 strains, confirming the acid tolerance\nof EcN ( Figure S1 ). At pH 6 and 7, recombinant\nMT7 and ET7 showed a contrast profile. ET7 would have rapid growth\nby expressing the RR plasmid, while MT7 showed a similar finding by\nharboring PR genes ( Figure 1 B). The natural mutations of the R15P-producing gene ( phn N) in EcN led to ET7 having better implementation toward\nthe RR route than the MT7 strain ( Figure S2 ). Figure 1 Reconstitution of two different non-native CO 2 assimilation\npathways in Escherichia coli . (A) Native pathway\nfrom glycolysis to the TCA cycle (left); artificial CO 2 -assimilating pathway in E. coli by harboring R15P\nisomerase (R15PI) from Thermococcus kodakarensis and\nRuBisCO ( rbcLS gene) from Synechococcus elongatus PCC7942 (i.e., RR plasmid, middle); phosphoribokinase ( prk gene) and RuBisCO ( rbcLS gene) from S.\nelongatus PCC7942 under T7 regulations (i.e., PR plasmid,\nright). Time course of (B) biomass, (C) residual glucose consumption,\nand (D) acetate formation during culture of recombinant E.\ncoli MT7 and ET7 at 8, 12, and 24 h. The strains were cultured\nin the glucose-based minimal medium (4 g/L) using a baffled flask\nin different pH cultures (pH 5, 6, 7) and 37 °C. Wild-type (WT)\nof T7RNAP-equipped MG1655 and EcN strains, denoted as MT7 and ET7.\nWhite, green, and orange backgrounds indicate MT7 or ET7 expressing\nempty plasmid, RR, and PR. (Metabolite abbreviations: G6P, glucose-6-phosphate;\nF6P, fructose-6-phosphate; FBP, fructose-1,6-biphosphate; GAP, glyceraldehyde\n3-phosphate; 1,3 BPG, 1,3-bisphosphoglycerate; 3PG, 3-phosphoglycerate;\nPEP, phosphoenolpyruvate; 6PG, 6-phosphogluconate; Ru5P, ribulose-5-phosphate;\nR5P, ribose-5-phosphate; PRPP, 5-phospho- d -ribosyl-α-1-pyrophosphate;\nR15P, ribose 1,5-bisphosphate, RuBP: ribulose 1,5-bisphosphate, CO 2 : carbon dioxide). In terms of glucose and acetate profiles, MT7 has\na slow glucose\nconsumption rate, which could be reasoned for its decelerated growth.\nMoreover, MT7 would accumulate a higher amount of acetate than ET7,\nexcept at pH 7. Such findings emphasized that this environment is\nunfavorable for EcN and limits the acetate-utilizing capability. Despite\nthis, the acetate accumulation would be utilized as a secondary carbon\nsource after depleting glucose ( Figure 1 C,D). 28 , 29 Overall, all strains acquired\nthe highest biomass at pH 7; yet, the effect of harboring CO 2 -fixing strains was intangibly observed from the growth profile.\nHence, further evaluation was directly conducted for assimilating\nthe release of CO 2 from 5-ALA production. Comparative Efficiency of CO 2 -Fixing Strains for\n5-ALA Production The functional application of an engineered\nCO 2 -fixing strain was examined for recycling the CO 2 emission from 5-ALA production. The CO 2 -fixing\nsystem in the upstream pathway was expected to take up CO 2 release and then increase the carbon flux in the downstream pathway.\nEventually, this scenario might impact in improving of either biomass\nor target product from the overexpressed design (5-ALA in this study).\nTo produce 5-ALA via the C4 pathway, a highly active ALAS from heterologous Rhodobacter capsulatus (Rc) was involved. 24 The coexpression of Rc and CO 2 -fixing genes\nwas designated into two systems of dual plasmids (DRc and DPc) and\nan all-in-one plasmid (ARc and APc) ( Figure 2 A). The evaluation was carried out in terms\nof biomass, 5-ALA titer, and metabolite profile. After 24 h of culture,\nboth E. coli strains harboring the all-in-one design\n(ARc and APc) exhibited lower biomass than that of the coupled dual-plasmid\nsystem (DRc and DPc). The highest biomass, reaching 2.23 and 2.14\ng/L, was acquired using DRc in MT7 and ET7 strains, respectively.\nAside from lower biomass, the all-in-one strains displayed a relatively\nhigher 5-ALA titer than the dual plasmid, with a remarkable titer\nof 1.52 g/L achieved using APc in ET7 ( Figure 2 B). Such behavior reflected a trade-off between\nbiomass and metabolites. Due to using two strong T7 promoters, expressing\nCO 2 -fixing genes on dual plasmids might hinder Rc capability,\nthereby limiting 5-ALA production. In contrast, the all-in-one design\ncontrolled the CO 2 -fixing genes under a single, later T7\npromoter, ensuring sufficient orthogonality of the T7 RNA polymerase\nfor Rc overexpression. However, compared to the use of a single Rc\ngene, coexpressing Rc with CO 2 -fixing genes acquired a\nlower 5-ALA titer than Rc alone (i.e., 1.64 g/L). A sole Rc expression\nalso achieved higher biomass up to 2.8 g/L, surpassing strains harboring\nRR or PR with a maximum biomass of 2.3 g/L, as shown in Figure S3 . From the metabolic analysis, the dual\nplasmids and all-in-one design of MT7 strains have comparable glucose\nutilization, while dual plasmids in ET7 completely consumed glucose\nwithin 18 h ( Figure 2 C). Acetate was also highly accumulated in the dual-plasmid system\n(i.e., DRc and DPc), especially in MT7, reflecting a reason for limited\n5-ALA production ( Figure 2 D). Figure 2 Functional comparison of overexpressing RR and PR in MT7 and ET7\nstrains for low-carbon-footprint 5-ALA production. (A) Metabolic pathway\nof low-carbon-footprint 5-ALA production in E. coli along with the genetic design of coupling CO 2 -fixing\ngene and ALA synthase (ALAS) from R. capsulatus .\nThe recombinant expression was regulated under the T7 promoter and\ndesignated in two different systems of dual plasmids (“D”)\nand all-in-one plasmid (“A”). DRc and ARc are dual plasmids\nand all-in-one designs of R15PI-RuBisCO-Rc genes, respectively, while\nDPc and APc are dual plasmids and all-in-one design of PRK-RuBisCO-Rc\ngenes, respectively. (B) Time course of biomass and ALA accumulation,\nMT7 on left and ET7 on right. (C) Residual glucose consumption and\n(D) acetate accumulation of recombinant MT7 (left) and ET7 (right)\nstrains. The cells were cultured in 10 g/L glucose-based minimal medium\npH 7 using a baffled flask at 37 °C. Glycine as cosubstrate was\nadded in vitro . (Metabolite abbreviations: PYR, pyruvate;\nAc-CoA, acetyl-CoA; a-KG, alpha-ketoglutarate; Suc-CoA, succinyl-CoA;\nOAA, oxaloacetate; 5-ALA, 5-aminolevulinic acid; TCA cycle: the citric\nacid cycle) Contrary to previous studies, the low 5-ALA titer\nand biomass in\nthe strain harboring CO 2 -fixing genes were somewhat unexpected.\nThis finding might correspond to the decarboxylation from 6PG to Ru5P,\nthus causing insufficient carbon flux in the RuBisCO pathway. Former\nstudies also have reported that carbon flux from glucose is primarily\ndirected into the main branch of glycolysis under both aerobic and\nanaerobic conditions. 25 , 30 , 31 Hence, the enrichment of carbon flux in the ribulose-related pathway\nis critical to reveal the significant function of the CO 2 -fixing genes. Fine-Tuning Carbon Flux Using CRISPRi Due to having\nan adjacent route with the ribulose-related pathway, 13 , 25 , 30 , 31 xylose was chosen to disclose the distinguished nature of RR and\nPR. Moreover, to enhance flux toward the RuBisCO pathway, zwf , pfk A, and pfk B genes\nwere knocked down, avoiding carbon loss (i.e., CO 2 release)\n( Figure 3 A). 25 , 29 CRISPRi-based gene regulation was chosen due to its high efficiency\nand flexibility in multiple targeting specific loci in the genome.\nAdditionally, CRISPRi requires only one protein, dCas9, to interfere\nwith gene transcription by allosterically binding to the target gene. 32 The effect of CRISPRi design was evaluated by\ncoexpressing the sgRNA plasmid with the CO 2 -fixing genes\nin glucose- and xylose-based mediums. In the glucose medium, strains\nimposed on the CRISPRi system showed strong cell burden and limited\ngrowth. Conversely, such a design could stimulate the growth in the\nxylose medium, indicating that flux redirection was successfully executed\n( Figure S4 ). The effect of CRISPRi on the\nCO 2 -fixing strains was also evaluated in a xylose-based\nmedium at 5% CO 2 supply. The result showed a consistent\nbehavior where coupling CRISPRi would have rapid growth ( Figure S5 ). Figure 3 Design of CRISPRi for redirecting carbon\nflux from xylose to the\nCO 2 -fixing pathways for strengthening CO 2 recycle and improving 5-ALA production. (A) Metabolic pathway\nof 5-ALA production using xylose and flux redirection through CRISPRi\ntargeting pfkAB and zwf genes (i.e.,\ndCas9–3PSG, denoted as “i”). Pink and yellow\nhighlights indicate the native and the extra artificial route from\nCO 2 -fixing genes toward xylose utilization. All recombinant E. coli ET7 strains harbor dual plasmids: one is CRISPRi,\nand the other is plasmid Rc for Rc-i, plasmid RR for ARc-i, or plasmid\nPR for APc-i. The cells were cultured in 10 g/L xylose-based minimal\nmedium pH 7 using a baffled flask as direct CO 2 capture\nfrom air (i.e., DAC) and bioreactor with 5% CO 2 supply\nat 37 °C, respectively. (B) Time course of biomass and 5-ALA\nduring culture with DAC and 5% CO 2 supply. Dashed red and\nblue lines indicate the highest biomass and 5-ALA acquired using ARc-i\nand APc-i, respectively. (C) CO 2 assimilation capability\nin terms of g-CO 2 /g-5-ALA and g-CO 2 /g-DCW at\n32 h. The percentage values represent remarkable CO 2 suppression\nin each corresponding term. (Metabolite abbreviations: Xlu, xylulose;\nS7P, seudoheptulose-7-phosphate; E4P, erythrose-4-phosphate; DHAP,\ndihydroxyacetone phosphate. Purple for knocked-down and blue for overexpressed\ngenes.) Taken together, the Rc gene, the CO 2 -fixing genes, and\nthe CRISPRi plasmid were incorporated, yielding Rc-i, ARc-i, and APc-i.\nThree recombinant ET7 strains were functionally evaluated for 5-ALA\nproduction from xylose with a direct air capture (DAC) of 420 ppm\nand 5% CO 2 supply. As expected, the current design showed\nmajor impacts of RR or PR presence toward cell growth and 5-ALA. The\nfastest growth and highest biomass were performed by the ARc-i strain.\nSupplying 5% CO 2 showed significant assistance that allowed\nthe engineered CO 2 -fixing strains to continue growing ( Figure 3 B). In terms of 5-ALA\nproduction, by supplying 5% CO 2 , the APc-i strain achieved\nthe highest 5-ALA of 1.66 g/L after 32 h. The findings presented trade-off\nphenomena in which obtaining a higher biomass would limit the strain\nin acquiring 5-ALA. At both CO 2 levels, a single Rc yielded\nthe lowest biomass and 5-ALA. It was caused by flux restriction from\nthe CRISPRi design and the lack of a CO 2 -fixing pathway,\nwhich corresponded to a higher remnant xylose ( Table S1 ). Furthermore, the CO 2 assimilation\ncapability of recombinant\nstrains under varied conditions was quantified according to the mass\nbalance analysis. The detailed calculations are presented in Tables S1–S3 . Coupling with CO 2 -fixing genes successfully suppressed CO 2 emissions in\nterms of either 5-ALA titer or biomass when compared with a sole Rc.\nCorresponding to 5-ALA production, the APc-i strain performed a notable\nsuppression of 77% by assimilating CO 2 release from −10.57\ng-CO 2 /g-5-ALA into −2.42 g-CO 2 /g-5-ALA).\nInterestingly, the CO 2 assimilation under 5% CO 2 inlet was detected to be higher than that in DAC. These results\nsuggested that both strains could capture CO 2 as a cosubstrate\nand correlated with higher remnant xylose ( Table S1 ). According to the dry cell weight (DCW), a remarkable suppression\nof 44% was acquired using the ARc-i strain, which could assimilate\nCO 2 from −10.0 g-CO 2 /g-DCW into −5.60\ng-CO 2 /g-DCW. A high CO 2 suppression toward 5-ALA\nusing APc-i might be related to the short CO 2 -fixing shunt\nfrom Ru5P into 3PG. An implicit assumption is that carbon flux via\nthe PR route could promptly coincide with overexpression of the Rc\ngene and the in vitro supply of glycine as a cosubstrate.\nMeanwhile, the RR design has a longer CO 2 -fixing shunt\nand interconnection with 3PG. Thus, it might retard the carbon flux,\nand thereby the directed flux into 5-ALA production did not equal\nthe glycine supply. To summarize, the reconstruction of two\ndifferent CO 2 -fixing pathways has been successfully applied\nin EcN and E. coli MG1655 as a reference. Natural\nmutations in phn N gene allowed EcN to perform a better\nCO 2 assimilation via the R15P route than MG1655. The functional\nefficiency\nof the CO 2 -fixing pathway via R15P resembled that of the\nconventional Ru5P route. Further optimizations focused on flux enhancement\ntoward the CO 2 -fixing pathway. Adopting CRISPRi for flux\nredirection from xylose to the CO 2 assimilation pathway\nwas the critical key to efficiently utilizing CO 2 -fixing\ngenes. Through this fine-tuned design, engineered EcN could achieve\nhigh biomass, 5-ALA, and CO 2 assimilation capability. This\nstudy attempted the comparative analysis of two non-native CO 2 -fixing routes from the ribulose-related pathway and extended\nthe promising potential of EcN as a low-carbon-featuring chemical\nproducer."
} | 4,912 |
39182147 | PMC11344931 | pmc | 3,858 | {
"abstract": "Background Magnetotactic bacteria (MTB) are a unique group of microorganisms that sense and navigate through the geomagnetic field by biomineralizing magnetic nanoparticles. MTB from the phylum Nitrospirota (previously known as Nitrospirae) thrive in diverse aquatic ecosystems. They are of great interest due to their production of hundreds of magnetite (Fe 3 O 4 ) magnetosome nanoparticles per cell, which far exceeds that of other MTB. The morphological, phylogenetic, and genomic diversity of Nitrospirota MTB have been extensively studied. However, the metabolism and ecophysiology of Nitrospirota MTB are largely unknown due to the lack of cultivation techniques. Methods Here, we established a method to link the morphological, genomic, and metabolic investigations of an uncultured Nitrospirota MTB population (named LHC-1) at the single-cell level using nanoscale secondary-ion mass spectrometry (NanoSIMS) in combination with rRNA-based in situ hybridization and target-specific mini-metagenomics. Results We magnetically separated LHC-1 from a freshwater lake and reconstructed the draft genome of LHC-1 using genome-resolved mini-metagenomics. We found that 10 LHC-1 cells were sufficient as a template to obtain a high-quality draft genome. Genomic analysis revealed that LHC-1 has the potential for CO 2 fixation and NO 3 − reduction, which was further characterized at the single-cell level by combining stable-isotope incubations and NanoSIMS analyses over time. Additionally, the NanoSIMS results revealed specific element distributions in LHC-1, and that the heterogeneity of CO 2 and NO 3 − metabolisms among different LHC-1 cells increased with incubation time. Conclusions To our knowledge, this study provides the first metabolic measurements of individual Nitrospirota MTB cells to decipher their ecophysiological traits. The procedure constructed in this study provides a promising strategy to simultaneously investigate the morphology, genome, and ecophysiology of uncultured microbes in natural environments. \n Video Abstract Supplementary Information The online version contains supplementary material available at 10.1186/s40168-024-01837-6.",
"conclusion": "Conclusion Linking microbial genomic potential to actual metabolism in the environment is challenging, especially at the single-cell level. In this study, we developed a workflow to simultaneously investigate the identity, morphology, genome, and metabolism of environmental MTB at the single-cell level. Our results show that the uncultured Nitrospirota MTB strain ‘ Ca. Magnetobacterium’ sp. LHC-1 can convert inorganic carbon and nitrogen into biomass and energy through CO 2 fixation and NO 3 − reduction. This indicates that LHC-1 is an autotroph and makes a contribution to the cycling of C and N in addition to Fe in the natural environment. We observed the temporal dynamics of C and N uptake in the LHC-1 population, which correlated with their growth status. Cultivation of Nitrospirota MTB cells would further help to quantify their contributions to C and N cycling. Together, the combination of different techniques (i.e., target-specific mini-metagenomics, FISH, FIB-SEM, and NanoSIMS) is a promising strategy to comprehensively study the mechanisms of microbe-environment interactions.",
"introduction": "Introduction With the development of microscopy and sequencing techniques, the morphological and (meta)genomic identification and characterization of environmental microbes have been greatly advanced. However, verification of metabolic features that are predicted from the obtained genomic data, as well as the quantitative information on the ecophysiology of the individual uncultured microbes, remains very challenging. Nanoscale secondary-ion mass spectrometry (NanoSIMS) is a powerful tool that can be used to measure the distribution of stable isotopes in the microbes at a single-cell resolution. It allows the uptake of stable-isotope labeled substrates to be monitored over time and directly linking individual cells to their phylogenies and metabolic activities in the environment by combining with molecular identification tools, such as fluorescence in situ hybridization (FISH) [ 1 – 4 ]. Magnetotactic bacteria (MTB) are a group of microorganisms that sense and navigate along the geomagnetic field. This unique ability of MTB is endowed by the production of intracellular magnetic nanoparticles of magnetite (Fe 3 O 4 ) and/or Greigite (Fe 3 S 4 ), defined as magnetosomes [ 5 , 6 ]. Magnetosome crystals are usually arranged in one or multiple linear chains within the cell, creating a permanent magnetic dipole moment and acting as a type of magnetoreceptor for the cell [ 7 ]. MTB are commonly found in a wide range of aquatic ecosystems and have been proposed to play important roles in the global biogeochemical cycling of Fe, C, N, S, P, etc. [ 8 ]. At least 16 different phyla of MTB have been identified thus far, although only a few have been cultivated in the laboratory [ 9 ]. Therefore, high-resolution information on the genetics, metabolism, and evolution of MTB populations remains mainly based on a few cultured MTB species, such as Magnetospirillum magneticum strain AMB-1 (AMB-1) and M . gryphiswaldense strain MSR-1 (MSR-1), which have been studied in detail using molecular biology approaches. The verified metabolic activities and ecophysiological roles of MTB within other MTB phyla are very limited. One of the most intriguing MTB phyla is the Nitrospirota phylum (previously known as Nitrospirae). In contrast to other MTB, they can produce many hundreds of (up to 1000) Fe 3 O 4 -type magnetosomes per cell, and the magnetosomes are normally arranged in multiple bundles of chains [ 10 , 11 ]. Nitrospirota MTB were originally considered to live in restricted environments with limited cell abundance. However, recent studies have shown that they are actually quite abundant in various aquatic ecosystems, including freshwater [ 12 – 15 ], estuaries [ 8 ], marine [ 16 , 17 ], hot springs [ 18 – 20 ], and acidic peatlands [ 21 ]. Therefore, Nitrospirota MTB could make important contributions to aquatic biogeochemical cycles and to the natural remanent magnetism of sediments. Cultivation of Nitrospirota MTB under controlled laboratory conditions has not yet been achieved, which could be due to a lack of critical information on their ecology, physiology, and key natural products for their growth. Previously, most diversity and ecology studies of Nitrospirota MTB have been based on 16S rRNA gene-based analyses; while more recently, omics-based studies provide an opportunity for predicting their metabolic potential [ 11 , 16 , 22 – 26 ]. Although the morphological and genomic investigations of Nitrospirota MTB have been greatly improved, to our knowledge, no literature has documented their verified metabolic activities and ecophysiological roles. In this study, we developed a correlative pipeline that combines electron microscopy, FISH, target-specific mini-metagenomics, and NanoSIMS-based stable-isotope analysis to characterize the morphology, phylogeny, genome, and metabolic activity of uncultured MTB at the single-cell level (Fig. 1 ). We applied this pipeline to characterize an uncultured Nitrospirota MTB population (named LHC-1). We recovered the high-quality draft genome of LHC-1 and revealed that LHC-1 had the potential to fix carbon dioxide (CO 2 ) and take up nitrate (NO 3 − ) as a nitrogen source to produce energy and biomass. Furthermore, we uncovered the distribution of C, N, O, and S elements in LHC-1 cells over time and observed cell-to-cell heterogeneity of carbon and nitrogen uptake within LHC-1 population. Moreover, the carbon and nitrogen uptake rates appeared to be related to the growth status of LHC-1 cells. Overall, our study provides the first experimental evidence of carbon and nitrogen uptake by Nitrospirota MTB at the single-cell level, shedding new light on their metabolism and ecology in the natural environment. Fig. 1 The pipeline developed in this study. For the target-specific cell sorting and mini-metagenomics, different numbers of magnetotactic bacteria (MTB) cells were sorted with a micromanipulation system (Step 1), which were then lysed and used as the template of whole genome amplification (Step 2). After sequencing, assembly, and binning, the draft genome of the MTB population was obtained (Step 3). Genome annotation and subsequent analysis (including phylogeny analysis, metabolism analysis) were then performed (Step 4). For the NanoSIMS-based isotopic analysis, the stable-isotope incubated MTB cells were first magnetically enriched (Step 1), then characterized by fluorescence in situ hybridization (FISH) (Step 2) and focused ion beam scanning electron microscope (FIB-SEM) (Step 3), and finally analyzed by NanoSIMS at the single-cell level (Step 4)",
"discussion": "Discussion Although cultivation-independent methods, such as 16S rRNA gene- and omics-based analyses, have provided general information on the diversity and metabolic potential of environmental Nitrospirota MTB, a deeper understanding of their metabolism and ecophysiology is still lacking. In this study, we developed a workflow to first obtain the high-quality genome of the uncultured Nitrospirota MTB LHC-1 by target-specific mini-metagenomics from a few cells and analyze the metabolic potential of LHC-1. Subsequently, we used NanoSIMS to test that LHC-1 could fix CO 2 and use NO 3 − as a nitrogen source. We also observed community dynamics and heterogeneity of C and N metabolism over time at the single-cell level. These studies have improved our understanding of the metabolism, ecophysiology, and biogeochemical dynamics of Nitrospirota MTB in the aquatic system. Combining target-specific mini-metagenomics and NanoSIMS By combining magnetic selection and microscopy-based single cell sorting, we showed that 10 LHC-1 cells as template of whole genome amplification are sufficient to obtain a high-quality draft genome (Table 1 ). Indeed, single-cell sorting by micromanipulation followed by whole genome amplification through MDA has been previously used to obtain draft genomes of several other Nitrospirota MTB populations from the environments. For example, Jogler and colleagues have identified the MGC of ‘ Ca. Magnetobacterium bavaricum’ (Mbav) using this approach together with PCR screening of metagenomic libraries [ 11 ]. Subsequently, Kolinko et al. [ 23 ] performed single-cell sequencing on individual cells of ‘ Ca . Omnitrophus magneticus’ SKK-01 (SKK-01), Mbav, and ‘ Ca . Magnetoovum chiemensis’ CS-04 (CS-04) and obtained single amplified genomes (SAGs) for each populations. They then combined six, six, and four SAGs from SKK-01, Mbav, and CS-04 cells to obtain the draft genomes with a completeness of 74%, 75%, and 87% for these three strains, respectively. These data, together with this study, demonstrate the great importance and efficiency of the target-specific mini-metagenomics technique in obtaining draft genomes of uncultured environmental MTB, especially those with low abundance. MGCs are the essential genes for magnetosome biosynthesis and their cellular localization. ANI analysis suggests that LHC-1 and XYR belong to the same species. Thus, the gene content and organization of MGCs between LHC-1 and XYR are highly conserved (Fig. 4 ), except that the MGC of LHC-1, but not of XYR, contains the mad29 and feoB genes, which are commonly identified in other MGCs of Nitrospirota MTB. For the MGC predicted in the LHC-1 draft genome, the mam genes of mamM , -Q , -B , -I , -E , and -O are conserved in almost all analyzed MTB strains and have been proposed to be essential membrane proteins for magnetosome membrane formation and growth during the early stage in the model alphaproteobacterial MTB strains AMB-1 and MSR-1 [ 79 – 83 ], indicating that the early stage magnetosome production of LHC-1 may be similar to that of AMB-1 and MSR-1, which produce cuboctahedral-shaped magnetic crystals. Little is known about the mad genes, which are probably involved in the production of bullet-shaped magnetite crystals [ 84 ]. One mystery of Nitrospirota MTB is why and how they produce such large amounts of magnetosomes and multiple bundles of magnetosome chains per cell. Our analysis suggests that the Man proteins may use enriched coiled-coil domains to organize the intricate multiple bundles of magnetosome chains. The exact functions of these man genes remain to be verified by genetic studies. A combination of fluorescence microscopy and electron microscopy have been previously performed on cultivated [ 85 ] and uncultivated [ 86 ] MTB to investigate their biomineralization, morphology, and phylogeny. The NanoSIMS technique has been increasingly applied to the metabolic analysis of various environmental microbes, including anaerobic phototrophic bacteria [ 4 ], N 2 -fixing bacteria [ 87 ], phytoplankton [ 88 ], and microbe-host interactions [ 89 – 91 ]. The FISH-NanoSIMS coupled techniques have been used to study the nitrogen metabolism of marine nitrite-oxidizing bacteria [ 92 ] and the carbon metabolism of an autotrophic, nitrate-reducing, Fe(II)-oxidizing enrichment culture [ 93 ]. For MTB analysis, NanoSIMS has been conducted to analyze the biomineralization of magnetosomes in cultured strains Desulfovibrio magneticus strain RS-1 [ 94 ] and MSR-1 [ 95 ]. Recently, a correlative approach of SIP (stable isotope probing)-FISH-Raman-SEM-NanoSIMS has been developed and further applied to characterize a population of uncultivated multicellular magnetotactic bacteria (MMB) producing Fe 3 S 4 magnetosomes belonging to the phylum Desulfobacterota [ 96 ]. The authors used NanoSIMS to characterize D 2 O uptake and the magnetosome distribution (localization of Fe and S) within the multicellular MTB [ 96 ]. More recently, individual MMB consortia were separated and sequenced, followed by SIP-FISH-NanoSIMS to test the genomic predictions, which simultaneously provides information on MMB diversity, ecology, genomics, and physiology [ 97 ]. In our case, we combined target-specific mini-metagenomics and stable-isotope analysis by cooperating single-cell sorting and sequencing, FISH, FIB-SEM, stable-isotope incubation, and NanoSIMS techniques (Fig. 1 ), to link the identity, morphology, genome, and verified metabolisms of an uncultured Nitrospirota MTB LHC-1 at the single-cell level. LHC-1 and many other MTB from different phyla are autotrophs Nitrospirota MTB are widespread and quite abundant in various aquatic ecosystems and have been proposed to play an important role in the biogeochemical cycles of C, N, S, P, Fe, etc. [ 8 , 12 – 14 , 16 – 19 , 21 ]. Several Nitrospirota MTB from different genera have been reported to contain the genetic potential for CO 2 fixation via the WL and/or rTCA pathways [ 16 , 22 , 23 , 58 ]. In this study, we predicted and verified the ability of CO 2 fixation in an uncultured Nitrospirota MTB LHC-1. Besides Nitrospirota MTB, the cultivated Pseudomonadota MTB strains Magnetovibrio blakemorei MV-1 (MV-1) [ 98 ], Magnetospira thiophila MMS-1 (MMS-1) [ 99 ], and Magnetococcus marinus MC-1 (MC-1) [ 100 ] are also capable of using CO 2 as a sole carbon source. The MV-1 strain possesses a type II ribulose-1,5-bisphosphate carboxylase/oxygenase (RubisCO) gene ( cbbM ) and uses the Calvin-Benson-Bassham (CBB) cycle for CO 2 fixation and autotrophy [ 98 ]. MMS-1 also uses the CBB cycle for CO 2 fixation and autotrophy [ 99 ]. During autotrophic growth, MC-1 relies on the rTCA cycle for CO 2 fixation [ 100 ]. Although LHC-1 has most of the genes involved in the CBB cycle, the predicted RubisCO is type IV, which could catalyze reactions other than RuBP carboxylation and may not be functional for carbon fixation [ 101 ], so LHC-1 may not use the CBB pathway for carbon fixation. Similarly, genes encoding type IV RubisCO were found in both the Mbav and Mcas genomes [ 13 , 22 , 23 ]. These data indicate that MTB from different phyla can use CO 2 as a carbon source through different carbon fixation pathways, suggesting an important role of MTB in the inorganic carbon cycling. The ability of nitrate reduction has also been proposed in several Nitrospirota MTB species by genomic prediction [ 16 , 22 , 58 ]. Based on the obtained draft genome of LHC-1, we predicted that LHC-1 could use NO 3 − as a nitrogen source by the denitrification pathway and/or dissimilatory nitrate reduction to ammonium, which was tested by NanoSIMS analysis, indicating that LHC-1 could use nitrate as a terminal electron acceptor. For the cultured Alphaproteobacteria Magnetospirillum magnetotacticum MS-1 (MS-1), when grown under conditions where NO 3 − is the sole nitrogen source, they simultaneously carry out denitrification to N 2 and dissimilatory nitrate reduction to ammonium [ 102 ]. Interestingly, magnetite biomineralization and anaerobic growth have been experimentally demonstrated to be closely related to the denitrification process in the cultured model Alphaproteobacteria strains MSR-1 and AMB-1 [ 103 , 104 ]. These data suggest that the ability to use NO 3 − as a nitrogen source may be conserved in many MTB groups and plays an important role in magnetosome biomineralization. The localization of C, N, O, and S is specific in LHC-1 Both the C and N signals are mainly distributed around the cell periphery of LHC-1 (Fig. 7 ). Much of the denitrification process of gram-negative bacteria has been found to be restricted to the periplasm [ 73 ]. The reduction of NO 3 − to NO 2 − is the first step in the utilization of NO 3 − and is thought to be catalyzed mainly by the periplasmic NO 3 − reductase complex NapAB and partly by a membrane-bound NO 3 − reductase NarGHI [ 105 ]. Then, a periplasmic NO 2 − reductase NirS or NirK catalyzes the reduction of NO 2 − to NO, which is further reduced to N 2 O by a NO reductase NorBC that is an integral membrane protein with its active site in the periplasm. Finally, a periplasmic N 2 O reductase (Nos) catalyzes the reduction of N 2 O to N 2 [ 105 ]. The genome of LHC-1 contains genes encoding NapAB, NarGHI, NirS, NorBC, and NosZ (Fig. 5 ), indicating that the denitrification process of LHC-1 most likely occurred in the periplasm, and the dissimilatory nitrate reduction pathway might occur close to the bacterial inner membrane, which was demonstrated in our study with the predominant location of nitrogen isotopes around the cell (Fig. 7 ). The WL pathway and the rTCA cycle are most likely to occur in the cytoplasm of the bacteria; thus far, the reason for the bacteria peripheral localization of the newly fixed carbon in LHC-1 is still unclear. One possible reason is that some stable-isotope-labeled carbon may passively diffuse through the bacteria outer and inner membrane into the cytoplasm. The colocalization of the major O signal with the magnetosome chain in the SEM images (Fig. 7 ) confirms that Nitrospirota MTB LHC-1 produces magnetite (Fe 3 O 4 ) magnetosomes. However, it is not clear why the intensity and distribution of the S signal are very similar to the 12 C 14 N − signal at different incubation times. Since no obvious sulfur granules were seen in these cells on the SEM images, the 32 S − signals may represent the distribution of sulfur-containing proteins and other sulfur-rich compounds in the LHC-1 cells. The heterogeneity and dynamics of C, N, and S metabolism in the LHC-1 cell population In this study, the LHC-1 strain displayed a physiological heterogeneity from cell to cell, including cell morphology and metabolism. As mentioned above, we observed morphological heterogeneity of sulfur granules in LHC-1 cells from SEM images (Fig. S2). Several uncultured Nitrospirota MTB species have been reported to accumulate sulfur granules in the cytoplasm and participate in the microbial sulfur cycling across the aquatic oxic-anoxic interface based on magnetotaxis [ 13 , 106 – 108 ]. The morphological heterogeneity of sulfur inclusions has been discovered in the Nitrospirota MTB strain Mbav [ 13 ] by electron microscopy observations. In other words, in the same Nitrospirota MTB population, the cells might contain different numbers of sulfur globules in the cytoplasm, probably due to the different metabolic status that the intracellular sulfur inclusions serve as a reservoir for further oxidation as previously proposed [ 10 , 13 , 22 ]. Moreover, the heterogeneity of C and N uptake in LHC-1 was also observed. The dots (each dot represents a single cell) on the scatter plot about 13 C − / 12 C − versus 15 N − / 14 N − (Fig. 8 c) were very concentrated and near to the natural values at 0 hpi. With the addition and incubation of H 13 CO 3 − and 15 NO 3 − , the dots on the scatter plot became more and more dispersed, indicating that the heterogeneity of C and N uptake in LHC-1 increased with incubation time. Interestingly, the dots spread more on the y -axis ( 15 N − / 14 N − ) than on the x -axis ( 13 C − / 12 C − ), especially at 17.5 hpi, which may represent a unique metabolic state or growth stage. These data indicate that the cell-to-cell heterogeneity of LHC-1 in N uptake is likely greater than in C uptake. The overall ratios of C and N uptake showed an increase in the trend at the beginning of the incubation period and a plateau phase towards the late incubation period. One striking time point is the 15 N − / 14 N − ratio at 17.5 hpi, which might due to the fact that the potential of the dissimilatory nitrate reduction pathway (NO 3− →NO 2− →NH 4+ ) is higher than that of the denitrification pathway (NO 3 − →NO 2 − →NO→N 2 O→N 2 ). The decrease in the 15 N − / 14 N − ratio from 17.5 to 23 hpi could be the opposite of what happened between 16.5 and 17.5 hpi. The decrease and increase in 13 C − / 12 C − and 15 N − / 14 N − uptake occurred at later time points, when the LHC-1 population has higher cell-to-cell heterogeneity and limited time point data, so the correlations between C and N metabolism and the different growth stages of LHC-1 remain to be explored."
} | 5,544 |
36572696 | PMC9792515 | pmc | 3,860 | {
"abstract": "Ever-growing demand for artificial intelligence has motivated research on unconventional computation based on physical devices. While such computation devices mimic brain-inspired analog information processing, the learning procedures still rely on methods optimized for digital processing such as backpropagation, which is not suitable for physical implementation. Here, we present physical deep learning by extending a biologically inspired training algorithm called direct feedback alignment. Unlike the original algorithm, the proposed method is based on random projection with alternative nonlinear activation. Thus, we can train a physical neural network without knowledge about the physical system and its gradient. In addition, we can emulate the computation for this training on scalable physical hardware. We demonstrate the proof-of-concept using an optoelectronic recurrent neural network called deep reservoir computer. We confirmed the potential for accelerated computation with competitive performance on benchmarks. Our results provide practical solutions for the training and acceleration of neuromorphic computation.",
"introduction": "Introduction Machine learning based on artificial neural networks (ANNs) has successfully demonstrated its excellent ability through record-breaking performance in image processing, speech recognition, game playing, and so on 1 – 3 . Although these algorithms resemble the workings of the human brain, they are basically implemented on a software level using conventional von Neumann computing hardware. However, such digital-computing-based ANNs are facing issues regarding energy consumption and processing speed 4 . These issues have motivated the implementation of ANNs using alternative physical platforms 5 , such as spintronic 6 – 8 , ferroelectric 9 , 10 , soft-body 11 , 12 , photonic hardware 13 – 18 , and so on 19 – 22 . Interestingly, even passive physical dynamics can be used as a computational resource in randomly connected ANNs. This framework is called a physical reservoir computer (RC) 21 – 23 or an extreme learning machine (ELM) 24 – 26 , whose ease of implementation has greatly expanded the choice of implementable materials and its application range. Such physically implemented neural networks (PNNs) enable the outsourcing of the computational load for specific tasks to a physical system such as a memory 27 , optical link 28 , 29 , sensor component 30 , 31 , or robotic body 32 . The experimental demonstrations of these unconventional computations have revealed performance competitive with that of conventional electronic computing 33 – 35 . Constructing deeper physical networks is one promising direction for further performance improvement because they can extend network expression ability exponentially 36 , 37 , as opposed to the polynomial relationship in wide (large-node-count) networks. This has motivated proposals of deep PNNs using various physical platforms 14 , 16 , 30 , 38 – 43 . Their training has basically relied on a method called backpropagation (BP), which has seen great success in the software-based ANN. However, BP is not suitable for PNNs in the following respects. First, the physical implementations of the BP operation are still complex and unscalable 40 – 43 . Thus, the calculation for BP for a PNN is typically executed on an external regular computer with a simulation model of a physical system 14 , 16 , 30 , 39 , 44 . This strategy results in a loss of any advantage in speed or energy associated with using the physical circuit in the training process. Thus, this method is not suitable for in-situ (online) training; it is only usable for “train once and infer many times” application. Second, BP requires accurate knowledge about the whole physical system. Thus, the performance of the PNNs entirely relies on the model representation or measurement accuracy of the physical system 45 . In addition, when we apply BP to RC, these requirements spoil the unique features of physical RC, i.e. we need to know and simulate a black-boxed physical random network accurately. Like BP in PNNs, the operational difficulty of BP in biological neural networks has also been pointed out in the brain science field; the plausibility of BP in the brain—the most successful analog physical computer—has been doubted 46 – 48 . These considerations have motivated the development of biologically plausible training algorithms 49 – 52 . One promising recent direction is direct feedback alignment (DFA) 53 – 55 . In this algorithm, fixed random linear transformations of the error signal at the final output layer are employed instead of the backward error signals. Thus, this approach does not require layer-by-layer propagation of error signals or knowledge of the weight. In addition, it has been reported that DFA scales to modern large-scale network models 54 . The success of such biologically motivated training suggests that there is a more suitable way than BP to train PNNs. However, DFA still requires the derivative f’ ( a ) of a nonlinear function f ( x ) for the training, which hinders the application of DFA methods to physical systems. Although previous studies on DFA training for spiking neural networks (SNNs) have reported that an approximated function can be used as an alternative 56 , this approach still requires modeling and simulation of the physical system. Thus, a more drastic extension of DFA is important for PNN applications. In this paper, we demonstrate physical deep learning by augmenting the DFA algorithm. In the augmented DFA, we replace the differential of physical nonlinear activation f ’( a ) in the standard DFA with arbitrary nonlinearity g ( a ) and show that the performance is robust to the choice of g ( a ). Owing to this augmentation, we no longer need to simulate f ’( a ) accurately. As the proposed method is based on parallel random projection with arbitrary nonlinear activation, we can execute the computation for the training on a physical system in the same manner as with the physical ELM or RC concept 21 – 23 . This enables the physical acceleration of both inference and training. To demonstrate the proof-of-concept, we constructed an FPGA-assisted optoelectronic deep physical RC as a benchtop. Although our benchtop is simple and easy to apply to various physical platforms with only software-level updates, we achieved performance comparable to that of large-scale complex state-of-the-art systems. Moreover, we compared the whole processing time, including that for digital processing, and found the possibility of physical acceleration of the training procedure. We also numerically found that the proposed augmented DFA is applicable to various network models, including more practical architecture and SNNs, suggesting the scalability of our approach. Our approach provides a practical alternative solution for the training and acceleration of neuromorphic physical computation.",
"discussion": "Discussion Augmentability to other physical systems In this study, we have verified the effectiveness of our approach through physical experimentations using an optoelectronic delay-based implementation. The remaining question is its applicability to other systems. To answer it, we performed numerical simulations using a widely investigated photonic neural network, and revealed the effectiveness of our approach even in complex-valued diffractive networks and nanophotonic unitary networks (see Supplementary Information S2 ). In addition, our experimentally demonstrated delay-based RC was shown to be highly suitable for various physical systems. The major difference from other physical systems is the nonlinearity in Eq. ( 4 ), which is sometimes difficult to identify accurately. However, as described above, our method is highly robust to g ( a ), which suggest the algorithm is effective for such cases. Regarding the scalability of the physical system, the major issue for constructing a deep network is its intrinsic noise. Here, we investigated the effect of noise by numerical simulation (see Supplementary Information S8 and previous works 74 , 75 ). We found the system to be robust to noise. Regarding to scalability of RC approach to more large-scale datasets, it has been reported that an RC-based transformer model (transformer with a fixed layer trained by BP) and a vision transformer-like RC works well 76 , 77 . As the transformer can be applied to many practical models, our deep RC scheme might scale to more advanced models. Further investigation will be performed in future. Scalability and limitation of proposed method Current physical implementations of neural networks are mainly focusing simple models such as an RC and MLP. We demonstrated the applicability of the augmented DFA to these models through the simulations and physical experiments described above. Here, we consider the scalability of the DFA-based approach to more modern models. One of the most commonly used models for practical deep learning is a deeply connected convolutional neural network (CNN). However, it has been reported that the DFA algorithm is difficult to apply to standard CNNs 78 .Thus, the proposed method may be difficult to apply to convolutional PNNs 33 , 67 , 70 in a simple manner. On the other hand, a recent study revealed that a full-connection network named MLP-Mixer can achieve state-of-the-art performance 79 . Although DFA-based training may be effective for such convolution-free models, the applicability of DFA for MLP-Mixer has not been investigated. In addition, it has also been reported that DFA can train modern network architectures without a convolution layer, including a graph neural network and transformer 54 . Those findings suggest that our algorithm might work on such a practical network structure. Considering analog hardware implementations, the applicability to SNNs is also an important topic. The suitability of DFA-based training for SNNs has been reported 56 , which implies that our proposed augmented DFA could make the training easier. Considering DFA for the CNN-based model, the investigation in a previous study was limited to the models without skip connections. It has been reported that the DFA angle increases with depth, which leads to failure of the training 78 . At the same time, it has been reported that the alignment angle in the convolution layer near the final layer is small enough even in CNNs, suggesting a shallow path to the final layer is one key to the success of the DFA-based training even in a CNN. Notably, it has been reported that forming skip connections is equivalent to forming an ensemble of deep and shallow networks. In addition, it has also been shown that most of the effective gradients in ResNet come from a shallow path. Thus, it is expected that ensembled shallow paths will have a positive impact on DFA training. In such a network, there remains the possibility of successful DFA training even for deep CNNs. While the DFA-based algorithm has the potential to scale to above more practical models beyond a simple MLP or RC, the effectiveness of applying DFA-based training to such networks is still unknown. Here, as additional work in this research, we investigated the scalability of DFA-based training (DFA itself and the augmented DFA) to the above-mentioned models (MLP-Mixer, Vision transformer (ViT), ResNet, and SNNs). The details are described in Supplementary Information S1 , and the main results for the MNIST, CIFAR-10 benchmarks for the examined results are summarized in Table 2 . We found that the DFA-based training is effective even for the explored practical models. While the achievable accuracy of DFA-based training is basically lower than that of BP training, some tuning of model and/or algorithm could improve the performance. Notably, the accuracies of DFA and the augmented DFA are comparable for all the explored experimental setups, suggesting that the further improvement of the DFA itself will directly contribute to improving the augmented DFA. The results suggest that our approach is scalable to future implementation of practical models to PNNs beyond simple MLP or RC model. Table 2 Applicability of augmented DFA to practical network models MNIST CIFAR-10 BP DFA (Ours) Augmented DFA (Ours) Train only final layer BP DFA (Ours) Augmented DFA (Ours) Train only final layer Custom MLP-mixer-3 98.90% 98.36% 98.49% 92.85% 76.50% 62.12% 62.01% 43.73% ResNet-18 99.51 99.41% 99.23% 81.70% 91.86% 81.68% 80.38% 34.92% ViT-3 (w/o fine tuning) 98.23% 98.04% 98.03% 47.6% 73.91% 59.00% (DFA+BP) 58.78% (DFA+BP) 32.50% SNN 97.90% (best) 86.13% (ave.) 96.73% (best) 93.19% (ave.) 98.14% (best) 98.05% (ave.) 92.21% (best) 91.95% (ave.) — — — — Scores for MNIST and CIFAR-10 for various ANN models trained by BP, standard DFA, and augmented DFA. As references, results for the models trained by the final layer only (the other layers are randomly fixed) are shown. Accuracies for DFA and augmented DFA higher than that for final layer-only training indicate the training works effectively. Since the training became unstable with BP and the DFA algorithm when using a derivative function of spiking neurons, both the best and average scores are shown for SNN setups for fair comparison (see Supplementary Information S1 .4 for the details). BP vs DFA in physical hardware In general, BP is extremely difficult to implement in physical hardware because it requires all the information in the computational graph. Thus, the training of physical hardware has been done by computational simulation, which incurs large computational costs. Also, the difference between the model and actual system leads the degradation of accuracy. In contrast, the augmented DFA does not require accurate prior knowledge about the physical system. Thus, in deep PNNs, our DFA-based approach is more effective even in terms of accuracy than the BP-based one. In addition, the computation is accelerable by using physical hardware as demonstrated in the Results section. While our first demonstration was slower than GPU implementations, it showed the potential to accelerate the computation of both inference and training on physical hardware. In addition, the DFA training does not require sequential error propagation with layer-by-layer computation, which means that the training of each layer can be executed in parallel. Therefore, a more optimized and parallel implementation of DFA could lead to more significant speed-up. These unique features suggest the effectiveness of our DFA-based approach, especially for physical hardware-based neural networks. On the other hand, the accuracy of the model trained by the augmented DFA was still inferior to one trained by BP. Further improvement of the accuracy for DFA-based training remains future work. One approach for the improvement (combination of DFA and BP) is described in Supplementary Information S1.2 . How to select alternative nonlinearity In this work, we introduced alternative activation for the training. Although g ( a ) is basically an arbitrary function, we should avoid it near η = 0. One simple way to do this is to use g ( a ) = sin( a + θ ). By scanning θ , we can sweep the η value for various functions and find a good solution. In addition, this nonlinearity is suitable for some physical implementations and, as shown in this article, we can accelerate the operation even in the training phase. Another approach is to use optimization problems such as a genetic algorithm (GA). Although a GA is hard to implement in a physical system, we can find a good solution for complex physical nonlinearity. An example of optimization is shown in Supplementary Information S5 . Further physical acceleration Our physical implementation confirmed the acceleration of recurrent processing for RC with a large-node count. However, its advantage is still limited, and further improvement is required. As mentioned in the Results section, the processing time of our current prototype is denoted as the data-transfer and memory allocation to the FPGA. Thus, integrating all the processes into the FPGA would improve the performance much more, with the sacrifice of experimental flexibility. In addition, in future, an on-board optics approach will reduce transfer cost drastically. Large-scale optical integration and on-chip integration will further improve the optical computing performance itself."
} | 4,107 |
33097742 | PMC7584638 | pmc | 3,861 | {
"abstract": "Recent years have witnessed tremendous progress of intelligent robots brought about by mimicking human intelligence. However, current robots are still far from being able to handle multiple tasks in a dynamic environment as efficiently as humans. To cope with complexity and variability, further progress toward scalability and adaptability are essential for intelligent robots. Here, we report a brain-inspired robotic platform implemented by an unmanned bicycle that exhibits scalability of network scale, quantity and diversity to handle the changing needs of different scenarios. The platform adopts rich coding schemes and a trainable and scalable neural state machine, enabling flexible cooperation of hybrid networks. In addition, an embedded system is developed using a cross-paradigm neuromorphic chip to facilitate the implementation of diverse neural networks in spike or non-spike form. The platform achieved various real-time tasks concurrently in different real-world scenarios, providing a new pathway to enhance robots’ intelligence.",
"introduction": "Introduction Humans have long aspired to develop an improved ability to handle multiple complex tasks in dynamic environments. Robots represent a physical manifestation of intelligence, particularly when placed in dynamic complex environments to make decisions and predictions. Although the operating principles of the human brain remain largely unknown, neuroscientific discoveries provide clues for designing intelligent robotic systems. Brain-inspired research has thus attracted widespread interest as a promising pathway for developing highly intelligent robotic platforms. Significant breakthroughs have been made in brain-inspired computing paradigms and hardware over the past decade 1 . Inspired by the human brain’s hierarchical topologies and parallel-processing networks, various artificial neural networks (ANNs), particularly deep neural networks, have achieved unprecedented success in numerous machine learning tasks 2 . For example, convolutional neural networks (CNNs) have surpassed human-level performance in image recognition and classification 3 , 4 . Inspired by the spike patterns of human brain activity, spiking neural networks (SNNs) exhibit high bio-fidelity, rich coding with spatiotemporal information, and event-driven peculiarity, emerging as a prominent neural computing paradigm in processing dynamic sequential information with high energy efficiency 5 , 6 . Meanwhile, there is currently a trend toward integrating deep learning and neuroscience, providing a highly promising pathway to develop artificial general intelligence (AGI) 7 , 8 . In parallel, rapid evolution in neural computing paradigms is also producing a proliferation of new types of computing hardware to accelerate computing. Distinct spike and non-spike computing paradigms have led to two developmental directions of computing hardware. Neural network accelerators are designed for optimizing operations in ANNs, such as ShiDianNao 9 , EIE 10 , and TPU 11 , which typically leverage parallel processing elements and efficient compression or data reuse. In contrast, neuromorphic chips support rich spatiotemporal bio-functionality, including Neurogrid 12 , TrueNorth 13 , SpiNNaker 14 , and Loihi 15 , providing high energy efficiency and event-driven representations. Some novel hybrid chip architectures have emerged, and Tianjic is currently the forerunner 16 , 17 . Continued progresses in brain-inspired computing algorithms and hardware have resulted in substantial advancements in intelligent robots 18 , 19 . The intersection of robotics and neuroscience are endowing robots with intelligent perception, flexible movement and natural interactions with environments 20 . Some real-world applications have been demonstrated, including humanoid platforms 21 , robotic arms 22 , medical robots 23 , robot navigation 24 , and automated driving 25 , 26 . These achievements have provided strategic opportunities to advance the design of intelligent robots. Most of these platforms, however, have been designed to be task-specific in simplified scenarios and have limited ability to perform multiple tasks simultaneously. Thus, a robotic platform with the capability to efficiently handle multiple complex tasks in a dynamic environment would be a valuable development. To promote robotic research by mimicking human intelligence, in the current study we developed a hybrid and scalable intelligent robotic platform based on an unmanned bicycle with primary modules including visual, auditory, motion and decision-making, which can deal with multimodal tasks simultaneously. Development of the platform involved three major challenges. First, because multimodal data-flows are constantly changeable and involve various information channels in the time and space, it is difficult to gather and handle different types of information from external environments. Second, because the integration of individual modules requires a high-level planner, determining how to dispatch them to accomplish comprehensive system-level behaviors is a challenge. Third, because evolution and continuous learning are important features of the human brain, intelligent robots require scalability for network scale, quantity and diversity. However, it is difficult for a computing system to achieve this scalability due to hardware restrictions. To overcome the abovementioned challenges, we proposed three design principles to develop the robot platform, inspired by the human brain (Fig. 1 ). First, inspired by the functional specialization of the cerebral cortex 27 and the rich coding schemes of biological neurons (rate, temporal, and population coding) 28 , we developed a hybrid architecture that can implement flexible inter-network cooperation and integrate different coding schemes efficiently. In this way, we can leverage the distinctive advantages of spiking and non-spiking neural networks in terms of energy efficiency and performance accuracy. Second, to adapt to dynamic environments, we developed a high-level decision-making module based on a hybrid neural state machine (HNSM) to integrate different modules flexibly, providing the capability to oversee and schedule different information flows, as well the capacity to be extended for dealing with increasing tasks during the implementation process. Third, inspired by neocortical regions organized with cortical columns 29 , we developed a scalable computing system based on our cross-paradigm neuromorphic chip, Tianjic, and a customized tool chain for hardware and software co-design 16 , 17 . The system has the potential to underpin brain-inspired system evolution and growth, similar to that exhibited in the human brain 30 . Figure 1 Intelligent architecture of the hybrid and scalable brain-inspired robotic platform. Software: Microsoft Visio 2019 MSO (16.0.10730.20102) 64-bit https://www.microsoft.com/en-us/microsoft-365/visio/flowchart-software . Adobe Photoshop version: 2015.0.0 20,150,529.r.88 2015/05/29:23:59:59 CL 1,024,429 × 64 https://www.adobe.com/products/photoshop.html . On the basis of these design principles, we developed a systematic solution for building a brain-inspired robotic platform. The system architecture consisted of full network-based modules to interact with the environment and a cross-paradigm neuromorphic chip to support seamless integration of different neural networks. A set of approaches were implemented to improve the system performance, such as seamless transform for blending rich coding schemes, and network-based state machines for module cooperation. We experimentally demonstrated that the unmanned bicycle accomplished various real-time tasks concurrently, including object detection, tracking, voice command recognition, riding over a speed bump, obstacle avoidance, balance control, and decision-making in complex dynamic environments. Collectively, the scalability in both algorithms and hardware in terms of network scale, quantity and diversity enables the system-level complexity and continuous evolution to cope with the complex and dynamic environment. Such a hybrid and scalable robotic platform could enhance the development of intelligent robots.",
"discussion": "Discussion This work reports a hybrid and scalable brain-inspired robotic platform that achieves multiple complex tasks simultaneously, involving multimodal perception, high-level decision-making and autonomous motion. The fundamental design principles of this platform are inspired by the human brain, including a hybrid architecture for integrating different coding schemes, a high-level decision-making module for network cooperation and a scalable computing system for evolution. Based on this platform, an unmanned bicycle was developed, which accomplished various tasks concurrently, including object tracking, obstacle avoidance, voice command recognition, balance control, and decision-making in various real-world environments. Our hybrid and scalable system can bring several unprecedented benefits and potentials. First, the hybrid architecture that integrates computer science and neuroscience-oriented approaches will benefit from the technological advances in these two fields, greatly promoting the development of brain-inspired robotic systems. Second, the excellent scalability in the platform, algorithms and computing capability will allow flexible integration of more sensors and functional modules to deal with complex scenarios. Third, the use of cross-paradigm neuromorphic computing system in the robot platform can not only support large-scale and diverse networks, but also promote the development of online learning. In summary, the system can serve as a general platform for a wide range of robotics research from fundamental theory to applications, including perception, cognition, auto-control, language comprehension, decision-making, learning and adaptation. In addition, the hybrid and scalable platform can be developed iteratively and continuously improved. For example, complex problems with high spatiotemporal information can be generated by randomly introducing new variables into the environment, such as different road condition, noises, weather factors, multiple languages, and more people. By studying the adaptation to these environmental changes, we can investigate some key challenges of AGI, such as generalization, robustness, and autonomous learning, promoting the development of AGI."
} | 2,614 |
28654054 | PMC5608424 | pmc | 3,863 | {
"abstract": "The optimal design and operation of photosynthetic bioreactors (PBRs) for microalgal cultivation is essential for improving the environmental and economic performance of microalgae-based biofuel production. Models that estimate microalgal growth under different conditions can help to optimize PBR design and operation. To be effective, the growth parameters used in these models must be accurately determined. Algal growth experiments are often constrained by the dynamic nature of the culture environment, and control systems are needed to accurately determine the kinetic parameters. The first step in setting up a controlled batch experiment is live data acquisition and monitoring. This protocol outlines a process for the assembly and operation of a bench-scale photosynthetic bioreactor that can be used to conduct microalgal growth experiments. This protocol describes how to size and assemble a flat-plate, bench-scale PBR from acrylic. It also details how to configure a PBR with continuous pH, light, and temperature monitoring using a data acquisition and control unit, analog sensors, and open-source data acquisition software.",
"introduction": "Introduction Due to growing concerns about global climate change and finite fossil fuel resources, governments have been developing policies to reduce fossil fuel consumption and to encourage the development of new, sustainable transportation fuels. The United States Environmental Protection Agency has developed the Renewable Fuel Standard (RFS), which requires that 36 of the annual 140 billion gallons of U.S. transportation fuel mix come from renewable fuel sources by 2022. Innovative and transformational technologies will be necessary to meet these and future renewable energy standards 1 . The use of microalgae-based biofuels has the potential to help meet the national RFS while reducing greenhouse gas emissions 2 . Microalgae-based biofuels have several advantages compared to first-generation biofuels based on terrestrial food crops, such as corn and soybeans. Unlike first-generation biofuels, algae-based biofuels consume fewer land, water, and food-related resources, since algae can be cultivated year-round and on barren land using saltwater or wastewater. Microalgae have high growth rates compared terrestrial crops and can accumulate high levels of lipids, which can be readily converted to biodiesel 3 . Currently, no industrial-scale algae-to-biofuel plants exist due to the high costs of the energy-intensive production processes, which consist of algal cultivation, lipid separation, and lipid refining into biodiesel. More research is needed to make these processes more efficient and sustainable. PBRs, which are optically clear, enclosed installations for the production of phototrophic microorganisms in an artificial environment, are considered one of the most promising cultivation methods 3 . However, current designs still lack the volumetric productivity necessary to make the algae-to-biofuel production process more efficient and economically attractive 4 . Powerful mathematical models that consider light irradiance and attenuation, the transport of nutrients and CO 2 , and the growth of the microalgae can greatly facilitate the optimization of PBR design and operation. Bench-scale growth experiments are required to determine species-specific growth parameters for these optimization models. Kinetic tests require the careful monitoring and control of experimental setups to prevent unintended inhibitors of growth. Given the photosynthetic nature of algae ( i.e., their consumption of CO 2 and absorption of light), maintaining controlled conditions is especially difficult in bench-scale PBRs. As depicted in Equation 1 , the amount of dissolved CO 2 in the growth medium, commonly denoted as ( Equation 2 ), will be, at minimum, a function of: 1) the CO 2 partial pressure and Henry's equilibrium constant, which dictates the amount of gas that will dissolve in solution ( Equation 3 ); 2) the initial chemical composition of the growth medium, which impacts the speciation and activity of the carbonate ions and pH ( Equations 4 and 5 ); and 3) the temperature, which impacts Equations 3-5 5 . \n \n \n \n \n \n The various phases and the chemical speciation of carbon create a challenge for measuring and maintaining a consistent dissolved carbon concentration within a PBR while holding other conditions constant ( e.g., the pH increases as the algae consume CO 2 , and increasing the dissolved CO 2 substrate can possibly lead to an acidic environment that inhibits growth) 6 . An additional layer of complexity for controlling conditions during algal kinetic tests involves the light intensity within the PBR. The average light intensity inside a PBR is a function of not only the incident light intensity, but also the design ( e.g., material, shape, depth, and mixing), the absorbance of algal biomass components (particularly chlorophyll), and the light-scattering properties of the algal cells. As the algae grow, the average light intensity will decrease. This change in light intensity, whether caused by an increase in total cells and biomass, an increase in chlorophyll content per cell, or both, can eventually induce a metabolic response, such as an increase in chlorophyll production per cell or the use of carbohydrate and lipid storage products for energy 7 . Continuous monitoring of the light intensity from within the reactor provides invaluable information. This data can help to ensure that conditions stay within a specified range and can be used to help estimate algal growth and absorbance parameters if combined with other measurements ( i.e., biomass, chlorophyll concentration, reactor depth, incident light, etc. ). Understanding how algae grow under a specified set of conditions requires that the pH, dissolved CO 2 , light intensity, and temperature be monitored in bench-scale kinetic experiments. Many algal growth setups are not equipped to monitor conditions to the extent required for calibrating kinetic models, making the modeling process extremely challenging 8 . Although many companies offer bench-scale PBRs with automation and control, these bench-scale setups can be extremely expensive (~$20,000) and might not accommodate all experimental considerations of a given research question. The first step in setting up a control-feedback system for a batch experiment is live data acquisition. This paper aims to demonstrate how to construct and set up a bench-scale PBR equipped with continuous light, pH, and temperature monitoring. This real-time monitoring setup can help to ensure that the experimental conditions stay within desired ranges, at the researcher's discretion. While this protocol does not detail specific control mechanisms, these step-by-step instructions provide a basic foundation for the data acquisition framework required before more sophisticated control feedbacks can be implemented.",
"discussion": "Discussion This PBR system offers the ability to monitor and control bench-scale algal kinetic growth experiments, allowing for more repeatable results from experimental assays used to quantify growth. However, an understanding of the limitations and uncertainties of sensor measurements is critical to ensure that the sensor readings accurately reflect reactor conditions. This understanding includes basic knowledge of the measurement principles involved with sensors, the process and frequency of calibration, the measurement uncertainty, and what the sensor can and cannot measure. For example, the electrical response for the light sensor described here is not equally distributed across the visible spectrum range, and certain correction factors may need to be applied to the sensor output, depending upon how this sensor data will be analyzed. Temperature levels and variations are also extremely important, as changes in temperature can drastically influence the sensor response. Understanding potential interferences that can impact the sensor readings is also critically important; this interference can be ambient electrical noise from the building or could stem from the measurement environment ( e.g., sodium ions can drastically impact pH readings at pH values over 10) 12 . Moreover, submerging multiple probes into a solution, especially a highly ionic and conductive salt solution, is also a potential source of interference. Electrodes that measure pH (or ionic strength, dissolved oxygen, dissolved CO 2 , etc. ) are especially sensitive to ambient electrical noise and can be easily perturbed. Signal conditioning used for protecting the electrode signal cannot guarantee that other factors will not interfere with the probe readings. As part of quality control, other laboratory equipment, such as a hand-held pH probe, a hand-held spectrometer, and a thermometer, should be used to verify the sensor readings and to ensure that the system is set up and running properly. Another limitation that must be addressed is the possible impact of the algae and/or culturing environment on the sensors. For example, if algal debris or bubbles cover the photodiode receptor of the light sensor, the readings will be affected. Similarly, pH electrodes are extremely sensitive and require extra care to ensure accurate readings. These electrodes work by measuring a voltage difference across an internal junction due to the buildup of H + ions; a hydrated buffer layer within the probe is required to maintain accurate measurements 12 . Depending upon the conditions within the reactor, this layer will wear off, and the response of the sensor may change over the course of the experiment while the probe is submerged. In preliminary tests, the pH voltage output did not drift by more than ~0.2 pH units over the course of a 20-day experiment, but further assessments should be performed to characterize this change in sensor response and to establish maximum experimental run times, especially if fine pH adjustments/quantifications are needed. Many current bench-scale PBR systems built to analyze algal growth do not monitor and control the internal culture environment as tightly as needed to discern how different factors impact algal growth, since setting up systems in this way can be challenging. This protocol can help facilitate more controlled experiments by giving step-by-step instructions for constructing a PBR with real-time monitoring. Moreover, this live data can be used not only to better control experimental conditions, but it can potentially be utilized to estimate growth kinetics ( e.g., optical density readings as reference for general growth rates). Controlled experimental systems can help to make algal research more reproducible. Bench-scale PBR setups that are monitored and controlled can increase experimental efficiency by minimizing unintended artifacts in experimental design and can help to advance efforts to make algal biofuels a sustainable, alternative fuel source."
} | 2,741 |
36354909 | PMC9697489 | pmc | 3,868 | {
"abstract": "Plants harbor a variety of fungal symbionts both above- and belowground, yet little is known about how these fungi interact within hosts, especially in a world where resource availability is changing due to human activities. Systemic vertically transmitted endophytes such as Epichloë spp. may have particularly strong effects on the diversity and composition of later-colonizing symbionts such as root fungal endophytes, especially in primary successional systems. We made use of a long-term field experiment in Great Lakes sand dunes to test whether Epichloë colonization of the dune-building grass, Ammophila breviligulata , could alter fungal root endophyte species richness or community composition in host plants. We also tested whether nitrogen addition intensified the effects of Epichlöe on the root endophyte community. We found that Epichloë increased richness of root endophytes in Ammophila by 17% overall, but only shifted community composition of root endophytes under nitrogen-enriched conditions. These results indicate that Epichlöe acts as a key species within Ammophila , changing richness and composition of the root mycobiome and integrating above- and belowground mycobiome interactions. Further, effects of Epichloë on root endophyte communities were enhanced by N addition, indicating that this fungal species may become even more important in future environments.",
"introduction": "1. Introduction The plant mycobiome includes a wide variety of fungal symbionts located within the above- and belowground tissues of host plants. Shifts in the mycobiome due to biotic interactions among fungal taxa can alter host plant growth, stress tolerance, and nutrient uptake [ 1 ]. Vertically transmitted, systemic Clavicipitaceous endophytes such as Epichlöe spp. may have particularly strong effects on the mycobiome of their host plants, as they have priority effects within a plant host and may influence assembly of later-colonizing, horizontally transmitted fungi in both above- and belowground tissues [ 2 ]. However, effects of Epichlöe on the plant mycobiome are not consistent. For example, Epichlöe has been shown both to reduce [ 3 ] and increase colonization of roots by arbuscular mycorrhizal fungi (AMF) [ 4 , 5 , 6 , 7 ]. Epichlöe has also had mixed effects on leaf endophytes in a variety of host plant species [ 8 , 9 , 10 ]. Epichlöe effects on root endophyte (non-AMF) communities are not well-understood, but seem minimal in the few studies that have examined them [ 11 , 12 , 13 ]. However, the magnitude of the effect of Epichlöe on mycobiome composition may depend on environmental conditions and resource availability (e.g., [ 14 , 15 ]), both of which are shifting in the Anthropocene [ 16 , 17 ]. Anthropogenic nitrogen (N) enrichment may be a particularly influential aspect of global change on fungal interactions and mycobiome community composition, as resource availability directly impacts plant-fungal associations [ 18 ]. Activities associated with agriculture, industry, wastewater, and fossil fuel combustion has more than doubled rates of nitrogen input into the terrestrial nitrogen cycle [ 19 ], and atmospheric N deposition is expected to increase globally by 250% over the next century [ 19 , 20 ]. Nitrogen enrichment is known to directly influence mycobiome composition within plant organs, for example by reducing root mycobiome diversity [ 21 , 22 , 23 ]. Aboveground, N addition is associated with decreased foliar endophyte diversity [ 24 , 25 ]. The few studies that have explored N enrichment effects on Epichlöe showed that N addition enhanced Epichlöe benefits to hosts [ 16 , 26 ]. For example, increasing soil N improved the alleviation of drought stress provided by endophytes to some grasses [ 27 ]. However, it is unclear whether N availability alters above-belowground fungal interactions within host plants. Changes in N availability may be particularly impactful on mycobiomes of plants in low nutrient primary successional ecosystems such as Great Lakes sand dunes. The U.S. Great Lakes coastal region has high N enrichment due to agricultural, atmospheric, and point-source inputs [ 28 ]. For example, atmospheric N deposition concentration levels, especially of NH 4 + , into Great Lakes ecosystems have increased 400% from historic levels [ 29 , 30 ] while dissolved inorganic N in Great Lakes coastal wetlands has risen as a direct result of row-crop agriculture in the region [ 28 ]. The dominant dune-building grass in this region, Ammophila breviligulata (hereafter Ammophila ) harbors a variety of fungal symbionts including the systemic endophyte, Epichlöe amarillans (hereafter Epichlöe ) [ 31 ], which is found in approximately one third of natural Ammophila populations in the Great Lakes [ 32 , 33 , 34 , 35 ] and in almost all nursery stock used for dune restoration work [ 36 ]. This provides an ideal system to examine effects of Epichlöe and N enrichment on other aspects of the plant mycobiome. Here, we evaluated the effects of Epichlöe and N addition on community composition of fungal root endophytes associated with Ammophila in a long-term experiment within the Great Lakes dunes. Specifically, we asked: Does colonization of the host grass by Epichlöe alter root fungal endophyte species richness or community composition? Furthermore, if so, does N addition intensify the effects of Epichlöe on the root endophyte community? While very little is known about how Epichlöe interacts with root endophyte communities in general, earlier work in this dune system showed that Epichlöe reduced diversity of other root-associated fungal communities (AMF) in Ammophila [ 14 ]. AMF and non-mycorrhizal root endophyte communities often show opposing responses to changing conditions [ 37 , 38 , 39 , 40 ] (although also see [ 41 ]), and so we expected that root endophyte richness would increase in response to Epichlöe presence . By enhancing the plant- Epichlöe symbiosis, we also expected that N addition would strengthen the effects of Epichlöe on root endophyte diversity and composition. Alternatively, N enrichment could act as an environmental filter that limits which fungal species can colonize plant hosts, or could enhance dominance of certain taxa at the expense of mycobiome biodiversity, independent of Epichlöe [ 21 , 22 , 23 , 42 ]. Our findings provide some of the first evidence that Epichlöe can increase root endophyte richness in host plants, and that nitrogen enrichment strengthens the effects of Epichlöe on root endophyte community composition.",
"discussion": "4. Discussion Epichloë influenced belowground mycobiome diversity and composition in this dune grass system, especially under N-enriched conditions, providing moderate evidence that Epichloë may act as a key species restructuring the above- and belowground mycobiome of host plants. Several potential explanations exist for why root fungal endophyte communities increased in richness in response to Epichloë infection. It might be expected that Epichloë would suppress root endophyte diversity due to its ability to produce systemic alkaloids [ 55 , 56 , 57 ] which directly inhibit growth of other fungal species, including pathogens [ 4 , 5 , 57 , 58 , 59 , 60 ]. However, indirect effects of Epichloë on its plant host may override direct mycobiome interactions to improve conditions for root endophyte communities. Within the harsh sand dune environment specific to our study, Epichloë acts as a mutualist, increasing Ammophila survival, vegetative growth, and belowground biomass [ 14 , 34 , 61 ]. In other systems, Epichloë can improve rhizosphere characteristics including soil fertility, root morphology, soil nutrients, and organic carbon [ 62 , 63 ]. By increasing habitat space in roots and improving belowground conditions, Epichloë may indirectly facilitate root endophyte diversity. The other studies that have demonstrated no effects of Epichloë on root endophyte communities were conducted in less extreme habitats such as agricultural fields and prairies [ 11 , 12 ], where indirect effects of Epichloë on host plants may be less important. Epichloë infection also altered root endophyte community composition in Ammophila , especially under N-enriched conditions; while under ambient conditions, the presence of Epichloë caused root fungal endophyte communities to converge across plots. These findings suggest that Epichloë acts a filter to restructure the endophyte communities that colonize Ammophila roots, possibly by altering the physical or chemical environment of host plants, such as root exudate chemistry [ 64 , 65 ]. Plant mycobiome composition can be strongly influenced by plant secondary metabolites [ 66 , 67 ], and Epichloë is known for enhancing alkaloid production inside host plants [ 68 ]. Alkaloids are nitrogen-rich secondary metabolites, so adding nitrogen could alter Epichloë’s ability to produce these chemicals within hosts, either positively [ 69 , 70 , 71 ] or negatively [ 72 ] (but see [ 73 ] where no effect was found), which could induce a strong filter on root endophyte species composition. We were able to identify several fungal taxa responsible for the shifts in root endophyte community composition. Both Microdochium bolleyi and Fusarium sp. were common taxa within root endophyte communities across all treatments. This Fusarium sp. best matched with Fusarium fujikuroi in the BLAST database (99.82% match), though we recognize that species designations within Fusarium usually require a tef1 sequence, which we did not have. Microdochium bolleyi is a common dark septate endophyte, primarily of grasses [ 74 , 75 , 76 ], including dune grasses of the Pacific Northwest [ 77 ]. A recent study found no effects of Epichloë exudates on M. bolleyi growth in an in vitro assay [ 6 ], which our field results support. However, F. fujikuroi (putative) was suppressed by Epichloë presence in N-enriched conditions, along with Sarocladium strictum , Leptosphaeria sp., and Cadophora sp. While the functions of many root endophytes are unknown, Fusarium fujikuroi has been classified in other systems as an asymptomatic nonobligate root symbiont best known for its gibberellin production [ 78 , 79 ] and has been found in other marine and coastal systems, along with S. strictum [ 80 ]. Since Epichloë is known to stimulate gibberellin production in both seeds and plants [ 81 ], any benefits that F. fujikuroi provides to hosts may become redundant. Sarocladium , Leptosphaeria and Cadophora are common root endophyte genera [ 82 , 83 ] with functions ranging from commensal to parasitic to saprophytic, making generalizations difficult. Two taxa increased in abundance in response to N addition for Epichloë colonized plants: Fusarium sporotrichioides (putative) and Acremonium sp. Fusarium sporotrichioides is a known pathogen of maize and an opportunistic pathogen of other cereal crops [ 84 , 85 ]. In agricultural systems, N fertilization often increases abundance of this and related Fusarium spp., possibly due to changes in plant N-metabolism (e.g., [ 86 ]). This may explain why increased N availability increased the occurrence of this species in Ammophila roots. Acremonium species may act as potential mutualists by increasing root growth in host plants [ 87 ], and so may hold a functionally similar role to Epichloë . While very little is known about root endophyte biology, especially in non-agricultural systems, these species-specific responses provide some insights into how Epichloë may filter mycobiome communities. In conclusion, by manipulating the presence of Epichloë in a long-term experiment, we found moderate evidence that this systemic endophyte is acting as a key species within Ammophila , changing diversity and composition of the root mycobiome and integrating above- and belowground mycobiome interactions. Further, effects of Epichloë on root endophyte communities were enhanced by N addition, indicating that this fungal species may become even more important in future environments. Ammophila is widely used in coastal dune restorations, and the importance of plant-fungal symbioses for restoration efforts are starting to be recognized [ 88 ]. Future work should address the consequences of such shifts in mycobiome communities for host plants, as intentional manipulation of mycobiome interactions may improve future restoration efforts in changing environments."
} | 3,138 |
36545892 | PMC10078402 | pmc | 3,869 | {
"abstract": "Abstract Facilitative interactions bind community species in intricate ecological networks, preserving species that would otherwise be lost. The traditional understanding of ecological networks as static components of biological communities overlooks the fact that species interactions in a network can fluctuate. Analyzing the patterns that cause those shifts can reveal the principles that govern the identity of pairwise interactions and whether they are predictable based on the traits of the interacting species and the local environmental contexts in which they occur. Here we explore how abiotic stress and phylogenetic and functional affinities constrain those shifts. Specifically, we hypothesize that rewiring the facilitative interactions is more limited in stressful than in mild environments. We present evidence of a distinct pattern in the rewiring of facilitation‐driven communities at different stress levels. In highly stressful environments with a firm reliance on facilitation, rewiring is limited to growing beneath nurse species with traits to overcome harsh stressful conditions. However, when environments are milder, rewiring is more flexible, although it is still constrained to nurses that are close relatives. Understanding the ability of species to rewire their interactions is crucial for predicting how communities may respond to the unprecedented rate of perturbations on Earth.",
"conclusion": "CONCLUSIONS We provide evidence for a distinct pattern in the rewiring of facilitation‐driven communities at different stress levels. In highly stressful environments where plants strongly depend on facilitation, rewiring is limited to growth beneath habitat‐specialized species. However, when environments are milder, rewiring, although of a lesser magnitude, is still limited to occur with nurse species relatively close to their preferred ones.",
"introduction": "INTRODUCTION The traditional view of ecological networks as static elements of biological communities neglects the fact that interactions between species in a network may shift in time and space (CaraDonna et al., 2021 ; Poisot et al., 2015 ; Trøjelsgaard & Olesen, 2016 ; Tylianakis & Morris, 2017 ). The realization of interactions is determined by the community's composition and a complex combination of environmental conditions and interacting species traits (Poisot et al., 2015 ). Shifts in species interactions may avoid coextinction cascades due to the loss of particular species, potentially increasing a community's robustness to face perturbation (Sheykhali et al., 2020 ; Vizentin‐Bugoni et al., 2020 ). However, partner switches (i.e., hereafter interactions turnover) are not unlimited (Montesinos‐Navarro et al., 2019 ; Raimundo et al., 2018 ) as certain species combinations may be unlikely. Therefore, it is essential to assess whether there are general patterns in the topological shifts of network interactions in order to predict how different communities may respond to perturbations. These shifts in interactions can be assessed through β‐diversity changes between networks (i.e., interactions turnover; Poisot et al., 2012 ). Interactions turnover between two networks can come from variations in species composition and thus their associated interactions (shifts due to species turnover) or changes in the interactivity patterns among the pool of species shared in the two networks (changes due to rewiring; Poisot et al., 2012 ). Assessing interactions turnover and whether it follows predictable patterns can be especially important in fragmented landscapes, where the inhospitable matrix may limit the recolonization of a locally extinct species (Corlett & Tomlinson, 2020 ). This scenario would be critical for those species that depend on disappeared species unless they can rewire with other community species to compensate for the interaction loss, thus alleviating extinction cascades caused by the extinction of key species. A framework to approach the β‐diversity of interactions has been assessed across ecological networks in different spatial and temporal contexts, unveiling patterns in structuring natural communities that cannot be inferred directly from changes in species composition (CaraDonna et al., 2021 ; Carstensen et al., 2014 ; Poisot et al., 2012 ). For instance, Carstensen et al. ( 2014 ) demonstrated that species involved in frequent pairwise interactions tended to rewire less often than species involved in rare interactions. Montesinos‐Navarro et al. ( 2019 ) showed that rewiring between mycorrhizal fungi and plants is not random but phylogenetically constrained. Analyzing the patterns driving rewiring interactions can shed light on the environmental factors and the species evolutionary history behind the establishment of pairwise interactions. Understanding these community dynamics is essential for correctly predicting shifts in the structure of communities facing an unprecedented rate of environmental changes. The role of rewiring can be critical for the survival of many species in communities governed by facilitation. Facilitation is defined as a biological interaction in which one (nurse) species alters the environment in a way that enhances the performance of a second (facilitated) species (Bronstein, 2009 ; Mcintire & Fajardo, 2014 ). These interactions can prevent the loss of species that require facilitation to survive (Bulleri et al., 2018 ). Mutualistic networks have traditionally been considered sensitive to the extinction of generalist species (i.e., species supporting multiple species), while they seem robust to the extinction of highly specialized species (i.e., species supporting few other species; Bascompte & Jordano, 2007 ), a pattern also observed for facilitation networks (Verdú & Valiente‐Banuet, 2008 ). However, this static vision of facilitative interactions does not consider the possibility of facilitated plants established beneath a new nurse when the positive association effect compensates for interspecific competition, a balance dependent on the environmental context and the interacting species traits (Qi et al., 2018 ). For example, it has been shown that some competitive interactions can turn facilitative under an increase in the severity of the stressful conditions (Bertness & Callaway, 1994 ; Qi et al., 2018 ), allowing the establishment of new facilitative interactions. However, this rewiring of partners is not unlimited since rewiring with highly competitive species could be very unlikely at any stress level. In contrast, facilitation rewiring would be easier with new species showing facilitative traits similar to those harbored by the original nurses. Traits driving facilitative interactions depend on the environmental context (Butterfield & Callaway, 2013 ; Navarro‐Cano et al., 2021 ). If particular traits are essential for the assembly of positive interactions, these traits could coerce the availability of nurses, so the more strict the requirements under specific stressful conditions, the more constrained the rewiring would be. There are several examples in restorations of stressful ecosystems where stress‐tolerant nurses harbor more facilitated species than stress‐sensitive nurses, suggesting a better amelioration derived from harboring specific traits to overcome harsh stressful conditions (Foronda et al., 2019 , 2020 ; Saiz et al., 2014 ). In contrast, in milder environments, facilitation could occur with a broader range of nurse species because the traits that limit facilitation in stressful contexts may be meaningless in milder conditions where other less specific traits can shape facilitative interactions (Catorci et al., 2016 ; Chen et al., 2015 ), allowing for more alternative facilitative interaction configurations. Phylogeny can provide clues about the relevant traits for the establishment of interactions when they are phylogenetically conserved, and therefore closely related species are expected to show similar patterns of responses to environmental pressures (Ackerly, 2003 ; Gómez et al., 2010 ; Webb et al., 2002 ). However, some selection pressures can lead to adaptive convergence in distantly related species that respond similarly to the selection pressure (Freeman et al., 2014 ; Webb et al., 2002 ). Convergent evolution has been revealed, for instance, in harsh edaphic environments where specialized traits appear in distantly related taxa across phylogeny (Moore et al., 2014 ). Understanding how abiotic stress and phylogenetic constraints modulate rewiring will help us to predict ecological community responses to different disturbance scenarios. Here, we hypothesize that the rewiring of facilitative interactions is more limited in stressful than in mild environments. Specifically, we propose that facilitated species rewire with nurses that are closely related to their preferred nurses in mild environments, while in harsh environments, facilitated species recruit preferentially beneath species with traits to overcome stressful conditions, which reduces the number of available nurses, and thus rewiring.",
"discussion": "DISCUSSION The observed dissimilarity pattern of interactions sheds light on how these communities subjected to different stress levels may respond to changes in community composition. In both environments, interaction dissimilarity was high, suggesting that the facilitative interactions can shift with high freedom across sites. Our results showed that interaction dissimilarity was slightly but significantly lower in stressful than in mild environments. However, much of this dissimilarity is explained by species turnover across communities, causing the appearance and disappearance of multiple interactions, a common pattern in our networks of stressful systems. Conversely, the dynamics reversed when we focused on the pool of shared species across sites, showing that rewiring was of less magnitude in stressful environments than in the mild ones. These spatial variations have implications for understanding how the structure of these communities will vary under a species loss scenario. On the one hand, we found a higher turnover in stressful environments even though species composition dissimilarities between mild and stressful environments did not differ. The higher reliance on facilitation of the recruiting species in stressful environments results in concomitant changes in interactions with the gain or loss of nurse species. Conversely, rewiring was low in stressful environments. Below we discuss that, in addition to stressful conditions, other phylogenetic and functional specific patterns may explain the limitations to rewiring patterns. Limitations on rewiring patterns Despite the high rewiring capacity found in mild environments, we still identified constraints to the establishment of interactions with nonpreferred nurse species. Specifically, we observed a pattern in which the recruiting species interacted with the nonpreferred nurse species that were close relatives of their preferred nurse species. That seems logical under the assumption that closely related species tend to share similar traits, so the traits that make a nurse suitable for certain species could also apply to similar species (Gómez et al., 2010 ). Indeed, this phylogenetic restriction to rewiring seems to apply to other mutualistic networks such as those established between plants and fungi (Montesinos‐Navarro et al., 2019 ), suggesting that it could be a pattern widely distributed in nature. Conversely, we found no functional limitations related to gypsum affinity for rewiring in mild environments, which is logical, as the specific traits that sustain facilitative interactions in stressful environmental contexts may be meaningless in these milder environments (Butterfield & Callaway, 2013 ). In contrast, we did not find any phylogenetic constraint to rewiring in stressful environments. In these systems, the harshness of the abiotic conditions has shaped a plant community harboring species with specific adaptations (Palacio et al., 2007 ). These specialized traits have emerged at different points in evolutionary history (Escudero et al., 2015 ; Moore et al., 2014 ), which prevents finding phylogenetic constraints to rewiring because species with varying gypsum affinity levels can be found within the same clades of the community (Appendix S1 , Figure S6 ). Indeed, in these systems, rewiring occurred with nonpreferred species showing higher gypsum affinity levels than expected independently of the phylogenetic relatedness. Among the potential mechanisms underlying this pattern might be the fact that some gypsum‐specialized species seem to be better nurses than stress‐sensitive species (Foronda et al., 2019 ; Saiz et al., 2014 ), which may promote the situation that many species depend on the presence of these nurses to survive (Verdú & Valiente‐Banuet, 2008 ). Ecological implications This paper emphasizes the critical role that well adapted species play in the stability of plant communities. When communities are subjected to high‐intensity stressors and recruitment on bare soil is highly undermined, facilitative interactions are established with those species that have mechanisms to deal adequately with this stressful condition (Foronda et al., 2019 ; Saiz et al., 2014 ). This fact fosters the appearance of benefactor species facilitating multiple species (Verdú & Valiente‐Banuet, 2008 ), a role that the gypsum specialists assumes as they are the only ones that can guarantee the survival of facilitated species. However, this limits rewiring with other less specialized species, as they may not provide the needed benefits. For this reason, retaining these specialist nurse species is vital for preserving the system's stability, as their disappearance could be catastrophic for the entire community (Valiente‐Banuet & Verdú, 2013 ). This is especially relevant in fragmented stressful edaphic environments in which many specialized species are endemisms whose isolation makes them more vulnerable to extinction (Corlett & Tomlinson, 2020 ). In contrast, in mild environments where the reliance on facilitation is less prominent, the rewiring capacity seems limited by nurse traits that must be phylogenetically conserved. It is worth noticing that the phylogenetically constrained rewiring found in these environments does not respond to recruits' necessity for particular traits as those found in the stressful environments, but to other general traits evolutionary conserved that are not homogeneous for all recruit species necessarily. These less stringent facilitation requirements could open up a gap for a more prominent contribution for rewiring in structuring these communities, as facilitative effects are not shaped by the traits of specific taxa. Nevertheless, these results should be interpreted with caution, as our results are based on a space‐by‐time substitution approach, using multiple site snapshots. Ideally, it would be interesting to monitor this process over time within each subsite. However, assessing biological interactions in the field requires intense sampling, so conducting these studies over the long term under field conditions might become unfeasible."
} | 3,810 |
25346175 | PMC4253120 | pmc | 3,870 | {
"abstract": "DNA has become a prime material for assembling complex three-dimensional objects that promise utility in various areas of application. However, achieving user-defined goals with DNA objects has been hampered by the difficulty to prepare them at arbitrary concentrations and in user-defined solution conditions. Here, we describe a method that solves this problem. The method is based on poly(ethylene glycol)-induced depletion of species with high molecular weight. We demonstrate that our method is applicable to a wide spectrum of DNA shapes and that it achieves excellent recovery yields of target objects up to 97 %, while providing efficient separation from non-integrated DNA strands. DNA objects may be prepared at concentrations up to the limit of solubility, including the possibility for bringing DNA objects into a solid phase. Due to the fidelity and simplicity of our method we anticipate that it will help to catalyze the development of new types of applications that use self-assembled DNA objects."
} | 253 |
26250403 | PMC4918357 | pmc | 3,873 | {
"abstract": "Droplet impacting on solid or liquid interfaces is a ubiquitous phenomenon in nature. Although complete rebound of droplets is widely observed on superhydrophobic surfaces, the bouncing of droplets on liquid is usually vulnerable due to easy collapse of entrapped air pocket underneath the impinging droplet. Here, we report a superhydrophobic-like bouncing regime on thin liquid film, characterized by the contact time, the spreading dynamics, and the restitution coefficient independent of underlying liquid film. Through experimental exploration and theoretical analysis, we demonstrate that the manifestation of such a superhydrophobic-like bouncing necessitates an intricate interplay between the Weber number, the thickness and viscosity of liquid film. Such insights allow us to tune the droplet behaviours in a well-controlled fashion. We anticipate that the combination of superhydrophobic-like bouncing with inherent advantages of emerging slippery liquid interfaces will find a wide range of applications.",
"discussion": "Discussion To elucidate how the collapse of air layer or the breakdown of the superhydrophobic-like bouncing affects droplet retraction dynamics, we compared the spread factor on samples with different μ o as a function of time at We=20. As shown in Fig. 5a , the contact line dynamics in the spreading stage is nearly the same, while it becomes closely reliant on the oil viscosity in the retraction stage, suggesting a substrate-dependent retraction behaviour. This is as evidenced by the contact line retraction velocity ν re measurement as shown in Fig. 5b . To analyze the effect of oil viscosity on droplet retraction dynamics, we calculated the Ohnesorge number Oh= μ o ( ρRγ oa ) −1/2 which measures the relative importance of inertia, viscous and surface tension forces, where γ oa is the surface energy of oil and R is the radius of water droplet. Since Oh for liquid interfaces with μ o =0.99 and 2.95 Pa s is 5.52 and ∼16.31 ( Fig. 5c inset), respectively, the droplet retraction process is primarily dominated by the viscous force. Indeed, measured retraction rates overlap well with the scaling of , where is the viscous time scale 60 . However, for Oh≈0.86 ( μ o =0.15 Pa s, Fig. 5c inset), the retraction rates diverge from the master curve because the inertia force and viscous force are comparable in this case. Finally, we demonstrated the utility of superhydrophobic-like bouncing for fast droplet shedding. Figure 6a shows time-lapsed images of water droplet collision on the composite interfaces under a tilt angle of ∼5° ( Supplementary Movies 6 and 7 ). To make the oil film with h =50 μm stable when it is tilted, the oil is locked in a thin porous Teflon membrane with thickness ∼20 μm. The droplet undergoes repetitive rebounding and falling for several cycles with trajectories indicated by dashed red line and finally rolls off the surfaces within ∼230 ms. In contrast, the droplet in the substrate-dependent retraction regime (large We) keeps intimate contact with underlying liquid and the sliding velocity is at least two orders of magnitude smaller than that in the superhydrophobic-like bouncing on the thin oil film ( Fig. 6b,c ). The enhanced droplet bouncing/sliding in the superhydrophobic-like bouncing regime is ascribed to the smooth nature of the composite interfaces, which helps to maintain a lubricating air layer even without the aid of external vibration. Thus, our results illustrate the important and previously unexplored effect of liquid substrate on the droplet impact. We expect that the occurrence of unexpected superhydrophobic-like bouncing and fast droplet removal on emerging slippery surfaces will extend their wide applicability in many processes 55 56 ."
} | 938 |
17462086 | PMC1868769 | pmc | 3,874 | {
"abstract": "Background Current methods for the automated generation of genome-scale metabolic networks focus on genome annotation and preliminary biochemical reaction network assembly, but do not adequately address the process of identifying and filling gaps in the reaction network, and verifying that the network is suitable for systems level analysis. Thus, current methods are only sufficient for generating draft-quality networks, and refinement of the reaction network is still largely a manual, labor-intensive process. Results We have developed a method for generating genome-scale metabolic networks that produces substantially complete reaction networks, suitable for systems level analysis. Our method partitions the reaction space of central and intermediary metabolism into discrete, interconnected components that can be assembled and verified in isolation from each other, and then integrated and verified at the level of their interconnectivity. We have developed a database of components that are common across organisms, and have created tools for automatically assembling appropriate components for a particular organism based on the metabolic pathways encoded in the organism's genome. This focuses manual efforts on that portion of an organism's metabolism that is not yet represented in the database. We have demonstrated the efficacy of our method by reverse-engineering and automatically regenerating the reaction network from a published genome-scale metabolic model for Staphylococcus aureus . Additionally, we have verified that our method capitalizes on the database of common reaction network components created for S. aureus , by using these components to generate substantially complete reconstructions of the reaction networks from three other published metabolic models ( Escherichia coli , Helicobacter pylori , and Lactococcus lactis ). We have implemented our tools and database within the SEED, an open-source software environment for comparative genome annotation and analysis. Conclusion Our method sets the stage for the automated generation of substantially complete metabolic networks for over 400 complete genome sequences currently in the SEED. With each genome that is processed using our tools, the database of common components grows to cover more of the diversity of metabolic pathways. This increases the likelihood that components of reaction networks for subsequently processed genomes can be retrieved from the database, rather than assembled and verified manually.",
"conclusion": "Conclusion We have described a method for automating the generation of substantially complete and coherent genome-scale metabolic reaction networks from annotated genomes. Our method builds on the subsystems approach to genome annotation and analysis embodied in the SEED. The SEED already provides well-curated genome annotations for central and intermediary metabolism across many organisms. We have extended the SEED by curating associations between reactions and functional roles in subsystems based on metabolic context. We have created tools for encoding components of reaction networks in subsystems, and verifying their coherence and interconnections. We have created tools for assembling these components into organism-specific complete and coherent reaction networks. We have demonstrated that our process can regenerate the reaction network from a published genome-scale metabolic model, and that it produces a cumulative effect supporting the subsequent generation of other reaction networks from published models. Our future work will focus on applying this process to generating reaction networks for new organisms, eventually extending to all organisms annotated in the SEED, thus producing a repository of organism-specific complete and coherent reaction networks. We envision that this repository will be useful for interpretation of large-scale data sets generated for metabolic genomics [ 38 ].",
"discussion": "Discussion Although our work to date has focused on regenerating reaction networks from published genome-scale metabolic models, our tools can be used to generate a preliminary reaction network for any organism in the SEED based solely on its genome annotation (Fig. 1B ). As with other methods for generating genome-scale metabolic reaction networks [ 12 - 15 , 18 , 19 ], the process of refining the reaction network until it is complete and coherent requires manual effort. The amount of manual effort required depends upon the quality of the preliminary reaction network, which in turn depends upon the quality of the annotation. An underlying goal behind our suite of tools is to narrow the gap between the preliminary reaction network and the complete and coherent reaction network as much as possible, so as to minimize the amount of manual effort required. We anticipate that the database of coherent reaction subnetworks will enable the generation of substantially complete and coherent preliminary reaction networks, focusing manual efforts on resolving gaps that are identified by the path-finding tool (Fig. 4B ), and on creating new reaction subnetworks for areas of metabolism not yet represented in the database. The collection of transport and biomass information from the published physiological literature and addition of corresponding reactions to the reaction network requires a significant manual effort for each organism, and is currently a bottleneck in generating a reaction network suitable for flux balance analysis. To date, this problem has not been adequately addressed in systems designed to facilitate genome-scale reaction network reconstruction [ 12 - 15 , 18 , 19 ], nor have we addressed this problem with our current set of tools. However, it may be possible to hypothesize computationally which transport reactions an organism uses, and which biomass components an organism synthesizes, based on the existence of paths through scenarios that represent catabolic pathways for common substrates, or anabolic pathways for common biomass components. Our approach to automated genome-scale metabolic reaction network generation represents two important advances when compared to previously published methods [ 12 - 15 , 18 , 19 ]. The first advance is a process for verifying the completeness and coherence of an overall reaction network by constructing a database of coherent reaction subnetworks that represent interconnected metabolic components. In contrast, most published methods for network generation do not explicitly provide support for the verification process. IdentiCS [ 15 ] and metaSHARK [ 14 ] provide automated annotation, assembly and visualization of a preliminary reaction network. The GEM System [ 12 ] additionally provides a heuristic for filling gaps in sets of consecutive reactions. However, these three systems do not provide tools for iterative refinement and verification of the preliminary network. The Pathway Tools software [ 13 ] takes an existing annotation and produces a pathway/genome database (PGDB), which includes predicted metabolic pathways and associated reactions for an organism. The software enables the visualization and manual refinement of the preliminary reaction network, but does not include tools for verification of the network. Segrè et al. [ 19 ] describe a process that builds upon Pathway Tools and includes an algorithm for the verification of the completeness and coherence of the overall reaction network, but do not describe a process for resolving gaps that are identified in the network. As discussed above, our approach enables resolving such gaps during the process of creating and assembling coherent reaction subnetworks (Fig. 1B ). The AUTOGRAPH-method [ 18 ] is unique among previously published methods in that it incorporates information from published genome-scale reaction networks to produce a preliminary reaction network for a particular organism. However, no process for verifying and refining the preliminary network is described. The second advance is the tight integration of our approach and our tools with a community-based genome annotation and analysis tool. From its inception, the SEED was designed to serve as a repository and clearinghouse for parallel annotation projects across all sequenced genomes. Because of the tight integration of our approach with the SEED, our tools can be used at all stages of every genome annotation project in the SEED, and reaction subnetworks created for one project are immediately available for all the other projects. This is in contrast with, e.g. , the Pathway Tools software [ 13 ], which is downloaded and installed locally by each genome annotation project, and creates a standalone PGDB for each project. Likewise, the published methods discussed above all focus on creating a reaction network for a single organism and, with the exception of the AUTOGRAPH-method, do not describe a process for reusing components of reaction networks previously developed for other organisms."
} | 2,241 |
37938635 | PMC9723742 | pmc | 3,876 | {
"abstract": "Taxonomic convergence is common in bacterial communities but its underlying molecular mechanism remains largely unknown. We thus conducted a time-series transcriptional analysis of a convergent two-species synthetic community that grew in a closed broth-culture system. By analyzing the gene expression and monitoring the community structure, we found that gene expression mainly changed in the early stage, whereas community structure significantly changed in the late stage. The significant change of gene expression occurred even at the very beginning, which was designated as “0 h effect”, suggesting the effect of species interaction on gene expression was inevitable. Besides, the effect of interaction on gene expression has a “population effect”, which means that majority species have greater impact on gene expressions of minority species than vice versa . Furthermore, gene set enrichment analysis revealed that among a total of 63 unique pathways (occupying about 50% of all the metabolic pathways in both species), 40 (63%) were consistently suppressed, 16 (25%) were conditionally expressed, and only 7 (11%) were consistently activated. Overall, they were strictly regulated by both time and initial structures. Therefore, we proposed that microorganism responses and the induced gene expression changes play important roles in the process of community succession."
} | 345 |
36460675 | PMC9718838 | pmc | 3,877 | {
"abstract": "Neuromorphic computing memristors are attractive to construct low-power- consumption electronic textiles due to the intrinsic interwoven architecture and promising applications in wearable electronics. Developing reconfigurable fiber-based memristors is an efficient method to realize electronic textiles that capable of neuromorphic computing function. However, the previously reported artificial synapse and neuron need different materials and configurations, making it difficult to realize multiple functions in a single device. Herein, a textile memristor network of Ag/MoS 2 /HfAlO x /carbon nanotube with reconfigurable characteristics was reported, which can achieve both nonvolatile synaptic plasticity and volatile neuron functions. In addition, a single reconfigurable memristor can realize integrate-and-fire function, exhibiting significant advantages in reducing the complexity of neuron circuits. The firing energy consumption of fiber-based memristive neuron is 1.9 fJ/spike (femtojoule-level), which is at least three orders of magnitude lower than that of the reported biological and artificial neuron (picojoule-level). The ultralow energy consumption makes it possible to create an electronic neural network that reduces the energy consumption compared to human brain. By integrating the reconfigurable synapse, neuron and heating resistor, a smart textile system is successfully constructed for warm fabric application, providing a unique functional reconfiguration pathway toward the next-generation in-memory computing textile system.",
"introduction": "Introduction As conventional complementary metal-oxide-semiconductor integrated circuits are approaching physical limits 1 , in-memory computing has emerged as an alternative low-power and high-efficiency technology 2 – 6 . Inspired by human brain, the basic elements of brain-inspired neuromorphic computing architectures, such as artificial neural network (ANN) and spiking neural network (SNN), are the electronic synapse and neuron 7 – 10 . These elements have specific nonvolatile synaptic and volatile integrate-and-fire functions, which are necessary for the system-level collaborative functioning of the neural network 11 – 14 . However, the fabrication processes and the materials of electronic synapses and neurons are mostly different 15 , thus causing difficulties in heterogeneous neurological integrations and limiting integration densities. Although separated artificial synapses and neurons with respective functions have been proposed for potential applications in a neuromorphic computing network, the performance mismatch between synaptic devices and neurons remains a problem to construct cooperative neural networks 16 . The need for high-efficiency cooperative neuromorphic electronics drives the development of reconfigurable memristors 17 – 19 . Developing reconfigurable memristor networks with functions of both artificial synapse and neurons is considered as an effective method to realize the next-generation neuromorphic electronics. Electronic textiles with functions of displaying, sensing, energy harvesting and energy storing showed great application prospects as new-generation wearable electronics 20 – 22 . Integrating neuromorphic computing memristors into electronic textiles in a seamless way is crucial to efficiently store and process signals from functional electronic components 23 – 25 . Low energy consumption is a critical feature for wearable electronic textiles to guarantee a long working life 26 . The energy consumption of a biological neuron is at a picojoule level (pJ) with range of 1–100 pJ/spike 27 – 30 , which ensures that the human brain consumes extremely low energy to complete daily activities. Therefore, the fabrication of fiber-shaped reconfigurable memristor with an energy consumption lower than biological neuron has great potential in constructing ultralow-power neuromorphic computing textiles. However, typical artificial neurons are usually based on a functional circuit consisting of three electronic components such as memristors, capacitors and resistors 31 , 32 , which increase the complexity and redundancy of the circuit. It remains an unmet need to achieve low-power neuron functions in simplified memristor circuits for efficient information processing in electronic textiles. Herein, a reconfigurable memristor textile network to function as both artificial synapse and neuron with ultralow energy consumption was fabricated. By designing the heterostructure of Ag/MoS 2 /HfAlO x /carbon nanotube (CNT), a three-dimensional reconfigurable memristor textile network with both synapse and neuron functions was constructed, exhibiting nonvolatile resistive switching and volatile threshold-switching characteristics based on regulation of conductive filaments. In the nonvolatile mode, synaptic weights could be modulated continuously with high long-term storage capability. In the volatile mode, action potential could be inspired by information integration from prior neurons. The artificial neuron based on reconfigurable memristor not only simplifies the circuit by using a single device, but also consumes ultralow power of 1.9 fJ/spike in integrate-and-fire function. Such a power consumption is three orders of magnitude lower than that of the biological neuron. The electronic synapses and neurons based on reconfigurable memristor textile networks were integrated to realize automatic heating function as a demonstration. The ultralow-power reconfigurable neuromorphic computing textile system may open up a new avenue for bio-inspired intelligent textile electronics.",
"discussion": "Discussion In summary, we present a functional textile network consisting of reconfigurable memristors, which was based on the structure of Ag/MoS 2 /HfAlO x /CNT with nonvolatile memory and volatile threshold-switching characteristics. Multi-level conductance modulation was achieved by the artificial synapse of top layer in the textile network. Integrate-and-fire function was simulated by the reconfigurable neuron of bottom layer in the textile network, which showed ultralow energy consumption of 1.9 fJ/spike, at least three orders of magnitude lower than that of biological neurons and reported artificial neurons. The artificial synapse, neuron and functional resistor were integrated into a heating textile system for intelligent temperature modulation. The ultralow-power textile memristor network could provide new directions in the development of brain-inspired reconfigurable and wearable neuromorphic computing electronics for intelligent Internet of Things applications."
} | 1,649 |
30631269 | PMC6315182 | pmc | 3,878 | {
"abstract": "The development of spiking neural network simulation software is a critical component enabling the modeling of neural systems and the development of biologically inspired algorithms. Existing software frameworks support a wide range of neural functionality, software abstraction levels, and hardware devices, yet are typically not suitable for rapid prototyping or application to problems in the domain of machine learning. In this paper, we describe a new Python package for the simulation of spiking neural networks, specifically geared toward machine learning and reinforcement learning. Our software, called BindsNET 1 , enables rapid building and simulation of spiking networks and features user-friendly, concise syntax. BindsNET is built on the PyTorch deep neural networks library, facilitating the implementation of spiking neural networks on fast CPU and GPU computational platforms. Moreover, the BindsNET framework can be adjusted to utilize other existing computing and hardware backends; e.g., TensorFlow and SpiNNaker . We provide an interface with the OpenAI gym library, allowing for training and evaluation of spiking networks on reinforcement learning environments. We argue that this package facilitates the use of spiking networks for large-scale machine learning problems and show some simple examples by using BindsNET in practice.",
"introduction": "1. Introduction The recent success of deep learning models in computer vision, natural language processing, and other domains (LeCun et al., 2015 ) have led to a proliferation of machine learning software packages (Jia et al., 2014 ; Abadi et al., 2015 ; Chen et al., 2015 ; Tokui et al., 2015 ; Al-Rfou et al., 2016 ; Paszke et al., 2017 ). GPU acceleration of deep learning primitives has been a major proponent of this success (Chetlur et al., 2014 ), as their massively parallel operation enables rapid processing of layers of independent nodes. Since the biological plausibility of deep neural networks is often disputed (Stork, 1989 ), interest in integrating the algorithms of deep learning with long-studied ideas in neuroscience has been mounting (Marblestone et al., 2016 ), both as a means to increase machine learning performance and to better model learning and decision-making in biological brains (Wang et al., 2018 ). Spiking neural networks (SNNs) (Maass, 1996 , 1997 ; Kistler and Gerstner, 2002 ) are sometimes referred to as the “third generation” of neural networks because of their potential to supersede deep learning methods in the fields of computational neuroscience (Wall and Glackin, 2013 ) and biologically plausible machine learning (ML) (Bengio et al., 2015 ). SNNs are also thought to be more practical for data-processing tasks in which the data has a temporal component since the neurons which comprise SNNs naturally integrate their inputs over time. Moreover, their binary (spiking or no spiking) operation lends itself well to fast and energy efficient simulation on hardware devices. Although spiking neural networks are not widely used as machine learning systems, recent work shows that they have the potential to be. SNNs are often trained with unsupervised learning rules to learn a useful representation of a dataset, which may then be used as features for supervised learning methods (Diehl and Cook, 2015 ; Kheradpisheh et al., 2016 ; Ferr et al., 2018 ; Hazan et al., 2018 ; Saunders et al., 2018 ). Trained deep neural networks may be converted to SNNs (Rueckauer et al., 2017 ; Rueckauer and Liu, 2018 ) and implemented in hardware while maintaining good image recognition performance (Diehl et al., 2015 ), demonstrating that SNNs can in principle compete with deep learning methods. In similar lines of work (Hunsberger and Eliasmith, 2015 ; Lee et al., 2016 ; O'Connor and Welling, 2016 ; Huh and Sejnowski, 2017 ; Mostafa, 2018 ; Wu et al., 2018 ), the popular back-propagation algorithm (or variants thereof) has been applied to differentiable versions of SNNs to achieve competitive performance on standard image classification datasets, providing additional evidence in support of the potential of spiking networks for ML problem solving. Finally, ideas from reinforcement learning can be used to efficiently train spiking neural networks for object classification or other tasks (Florian, 2007 ; Mozafari et al., 2018 ). The membrane potential (or voltage) of a spiking neuron is often described by ordinary differential equations. The membrane potential of the neuron is increased or decreased by presynaptic inputs, depending on their sign and strength. In the case of the leaky integrate-and-fire (LIF) model (Kistler and Gerstner, 2002 ) and several other models, the neuron is constantly decaying to a rest potential v rest . If a neuron integrates enough input and reaches its threshold voltage v thr , it emits a spike which travels to downstream neurons via synapses, its post-synaptic effect modulated by synaptic strengths, and its voltage is reset to some value v reset . Synapses between neurons can also have their own dynamics, which are modified by prescribed learning rules or external reward signals. Several software packages for the discrete-time simulation of SNNs exist, with varying levels of biological realism and support for hardware platforms. Many such solutions, however, were not developed to target ML applications, and often feature abstruse syntax resulting in steep learning curves for new users. Moreover, packages with a large degree of biological realism may not be appropriate for problems in ML, since they are computationally expensive to simulate and may require a large degree of hyper-parameter tuning. Real-time hardware implementations of SNNs exist as well, but cannot support the rapid prototyping that some software solutions can. Motivated by the foregoing shortcomings, we present the BindsNET spiking neural networks library, which is developed on top of the popular PyTorch deep learning library (Paszke et al., 2017 ). At its core, the software allows users to build, train, and evaluate SNNs composed of groups of neurons and their connections. The learning of connection weights is supported by various algorithms from the biological learning literature (Hebb, 1949 ; Markram et al., 1997 ). A separate module provides an interface to the OpenAI gym (Brockman et al., 2016 ) reinforcement learning (RL) environments library from BindsNET . A Pipeline object is used to streamline the interaction between spiking networks and RL environments, removing many of the messy details from the purview of the experimenter. Still other modules provide functions such as loading of ML datasets, encoding of raw data into spike train network inputs, plotting of network state variables and outputs, and evaluation of SNN as ML models. The paper is structured as follows: we begin in section 2 with an assessment of the existing SNN simulation software and hardware implementations. In section 3, the BindsNET library is described in details, emphasizing the motivation of creating each software module, describing their functionalities, and they way the inter-operate when solving a specific task. Code snippets and simple case studies are introduced in section 4 to demonstrate the breadth of possible BindsNET applications. Desirable directions and features of future developments are listed in 5, while potential research impacts are assessed in section 6.",
"discussion": "6. Discussion We have presented the BindsNET open source package for rapid biologically inspired prototyping of spiking neural networks with a machine learning-oriented approach. BindsNET is developed entirely in Python and is built on top of other mature Python libraries that lend their power to utilize multi-CPU or multi-GPU hardware configurations. Specifically, the ML tools and powerful data structures of PyTorch are a central part of BindsNET 's operation. BindsNET may also interface with the gym library to connect spiking neural networks to reinforcement learning environments. In sum, BindsNET represents an additional and attractive alternative for the research community for the purpose of developing faster and more flexible tools for SNN experimentation. BindsNET comprises a spiking neural network simulation framework that is easy to use, flexible, and efficient. Our library is set apart from other solutions by its ML and RL focus; complex details of the biological neuron are eschewed in favor of high-level functionality. Computationally inclined researchers may be familiar with the underlying PyTorch functions and syntax, and excited by the potential of the third generation of neural networks for ML problems, driving adoption in both ML and computational neuroscience communities. This combination of ML programming tools and neuroscientific ideas may facilitate the further integration of biological neural networks and machine learning. To date, spiking neural networks have not been widely applied in ML and RL problems; having a library aimed at such is a promising step toward exciting new lines of research. Researchers interested in developing spiking neural networks for use in ML or RL applications will find that BindsNET is a powerful and easy tool to develop their ideas. To that end, the biological complexity of neural components has been kept to a minimum, and high-level, qualitative functionality has been emphasized. However, the experimenter still has access to and control over groups of neurons at the level of membrane potentials and spikes, and connections at the level of synapse strengths, constituting a relatively low level of abstraction. Even with such details included, it is straightforward to build large and flexible network structures and apply them to real data. We believe that the ease with which our framework allows researchers to reason about spiking neural networks as ML models, or as RL agents, will enable advancements in biologically plausible machine learning, or further fusion of ML with neuroscientific concepts. Although BindsNET is similar in spirit to the Nengo (Bekolay et al., 2014 ) neural and brain modeling software in that both packages can utilize a deep learning library as a “backend” for computation, Nengo optionally uses Tensorflow in a limited fashion while BindsNET uses PyTorch by default, for all network simulation functionality (with the torch.Tensor object). Additionally, for users that prefer the flexibility and the imperative execution of PyTorch , BindsNET inherits these features and is developed with many of the same design principles in mind. BindsNET has advantages with respect to other simulation libraries using GPU computation, which require costly compilation steps between network building and deployment. BindsNET does not need these expensive intermediate steps as it uses “eager” execution of PyTorch regardless of the actual simulation hardware. Hardware platforms for spiking neural network computations have advantages over software simulations in terms of performance and power consumption. For example, SpiNNaker (Plana et al., 2011 ) combines cheap, generic, yet dedicated CPU boards together to create a powerful SNN simulation framework in hardware. Other platforms (e.g., TrueNorth Akopyan et al., 2015 , HRL, and Braindrop) involve the design of a new chip. A novel development is Intel's Loihi platform for spike-based computation, outperforming all known conventional solutions (Davies et al., 2018 ). Other solutions are based on programmable hardware, like FPGAs which transform neural equations to configurations of electronic gates in order to speed up computation. More specialized hardware such as ASIC and DSP can be used to parallelize and therefore accelerate the calculations. In order to conduct experiments in the hardware domain, one must usually learn a specific programming language targeted to the hardware platform, or carefully adapt an existing experiment to the unique hardware environment under the constraints as enforced by chip designers. In either case, this is not an ideal situation for researchers who want rapid prototyping and testing. BindsNET platform introduces a flexibility, which can be exploited in future hardware developments, in particuliar in machine learning problems. BindsNET is a simple yet attractive option for those looking to quickly build flexible SNN prototypes backed by an easy-to-use yet powerful deep learning library. It encourages the conception of spiking networks as machine learning models or reinforcement learning agents, and is one of the first of its kind to provide a seamless interface with machine learning and reinforcement learning environments. The library is supported by several mature and feature-full open source software projects, and benefits from their growth and continuous improvements. Considered as an extension of the PyTorch library, BindsNET represents a natural progression from second generation neural networks to third generation SNNs."
} | 3,271 |
36341023 | null | s2 | 3,879 | {
"abstract": "Reinforcement learning is a powerful framework for modelling the cognitive and neural substrates of learning and decision making. Contemporary research in cognitive neuroscience and neuroeconomics typically uses value-based reinforcement-learning models, which assume that decision-makers choose by comparing learned values for different actions. However, another possibility is suggested by a simpler family of models, called "
} | 106 |
22905032 | PMC3408237 | pmc | 3,881 | {
"abstract": "Nitrite-dependent anaerobic methane oxidation (n-damo), which couples the anaerobic oxidation of methane to denitrification, is a recently discovered process mediated by “ Candidatus Methylomirabilis oxyfera.” M. oxyfera is affiliated with the “NC10” phylum, a phylum having no members in pure culture. Based on the isotopic labeling experiments, it is hypothesized that M. oxyfera has an unusual intra-aerobic pathway for the production of oxygen via the dismutation of nitric oxide into dinitrogen gas and oxygen. In addition, the bacterial species has a unique ultrastructure that is distinct from that of other previously described microorganisms. M. oxyfera -like sequences have been recovered from different natural habitats, suggesting that the n-damo process potentially contributes to global carbon and nitrogen cycles. The n-damo process is a process that can reduce the greenhouse effect, as methane is more effective in heat-trapping than carbon dioxide. The n-damo process, which uses methane instead of organic matter to drive denitrification, is also an economical nitrogen removal process because methane is a relatively inexpensive electron donor. This mini-review summarizes the peculiar microbiology of M. oxyfera and discusses the potential ecological importance and engineering application of the n-damo process.",
"introduction": "INTRODUCTION Methane (CH 4 ) is an important greenhouse gas, which has, so far, contributed an estimated 20% to global warming ( Knittel and Boetius, 2009 ). In the past, the oxidation of methane was believed to be restricted to oxic environments. This view gradually changed with the discovery of anaerobic methane oxidation (AMO) coupled to sulfate reduction in anoxic marine sediments and water columns ( Martens and Berner, 1974 ; Reeburgh, 1976 ; Valentine and Reeburgh, 2000 ). In 2006, a new AMO process, nitrite-dependent anaerobic methane oxidation (n-damo), which couples AMO to denitrification, was discovered in an enrichment culture ( Raghoebarsing et al., 2006 ). Several peculiar properties of the n-damo process have been discovered that make AMO of particular interest to those interested in microbiology, ecology, and environmental engineering: (i) the discovery of a new species (“ Candidatus Methylomirabilis oxyfera”) linking the carbon and nitrogen cycles ( Ettwig et al., 2010 ; Wu et al., 2011a ), (ii) the potential contribution of the n-damo process to reduction of global warming via oxidization of methane to carbon dioxide (CO 2 ), and (iii) the potential application of n-damo for nitrogen removal from wastewater by using methane instead of organic matter as an electron donor to drive denitrification. This mini-review summarizes the microbiology of the n-damo process, including the phylogenetic affiliations, physiological and ultrastructural properties of M. oxyfera , and the molecular mechanisms of its intra-aerobic metabolism. In addition, this mini-review discusses the potential ecological importance of the n-damo process in natural ecosystems and the application of the process for wastewater nitrogen removal."
} | 774 |
31117280 | PMC6566420 | pmc | 3,883 | {
"abstract": "A tactile sensor is an indispensable component for electronic skin, mimicking the sensing function of organism skin. Various sensing materials and microstructures have been adopted in the fabrication of tactile sensors. Herein, we propose a highly sensitive flexible tactile sensor composed of nanocomposites with pyramid and irregularly rough microstructures and implement a comparison of piezoresistive properties of nanocomposites with varying weight proportions of multi-wall nanotubes and carbon black particles. In addition to the simple and low-cost fabrication method, the tactile sensor can reach high sensitivity of 3.2 kPa −1 in the range of <1 kPa and fast dynamic response of 217 ms (loading) and 81 ms (recovery) at 40 kPa pressure. Moreover, body movement monitoring applications have been carried out utilizing the flexible tactile sensor. A sound monitoring application further indicates the potential for applications in electronic skin, human–computer interaction, and physiological detection.",
"conclusion": "4. Conclusions In this work, a highly sensitive flexible tactile sensor was proposed by combining a pyramid structure and a double-sided rough structure with a simple and low-cost fabrication method. Based on the comparison results of the piezoresistance of nanocomposites with different proportions of MWCNTs and CB particles, 6% content of MWCNTs in the nanocomposite was chosen with the best performance. Moreover, different structure combinations were compared and the tactile sensor with 50 μm DPRS showed the highest sensitivity of 3.2 kPa −1 in the range of <1 kPa, and 0.11 kPa −1 exceeding 1 kPa. The tactile sensor possessed a short response time under a large pressure and has good repeatability. What is more, body movement monitoring applications were carried out which demonstrate great potential of the tactile sensor in applications of electronic skin, human–computer interaction, physiological detection, and sounding recognition.",
"introduction": "1. Introduction Bionic electronic skin has attracted much research attention because of its great potential in medical treatment [ 1 ], human–computer interaction [ 2 ], robotics [ 3 ], and visual display [ 4 ]. Flexible tactile sensors play an important role in research of artificial skin. Based on the study of the perception mechanism of human skin for external pressure [ 5 ], good flexibility and high sensitivity are necessary characteristics for artificial tactile pressure sensors to achieve the desired sensing effect. In order to achieve this goal, various processes and materials are utilized in the manufacturing process. According to the working principle, tactile sensors are divided into piezoresistive [ 6 ], piezoelectric [ 7 ], capacitive [ 8 ], triboelectric [ 9 ], optical [ 10 ], and electromagnetic [ 11 ]. Among them, piezoresistive tactile sensors have attracted much attention because of their simple fabrication process, good robustness, and stability. Contact pressure can be perceived according to changes of sensor element resistances. In piezoresistive sensing elements, conductive paths are realized by various nanoscale conductive materials including carbon nanotubes (CNTs) [ 12 , 13 ], graphene [ 14 , 15 ], and carbon black (CB) [ 16 ], and the spatial structure of piezoresistive elements also determines the sensitivity of the sensor. For instance, Zhang et al. [ 17 ] proposed a highly sensitive flexible tactile sensor with three-axis force sensing capacity by combining the microstructured pyramids’ polydimethylsiloxane (PDMS) arrays and reduced graphene oxide (rGO) film. The deformation of microstructured rGO/PDMS results in the change of contact area between the rGO film and electrode, and the adoption of the pyramid structure increases the sensitivity of the sensor. Li et al. [ 18 ] developed a type of high-performance flexible capacitive tactile sensor utilizing bionic microstructures on natural lotus leaves. Taking advantage of the unique surface micro-pattern of lotus leaves as the template for electrodes and using polystyrene microspheres as the dielectric layer, the proposed devices present stable and high sensing performance. Wang et al. [ 12 ] presented a simple and low-cost method for fabrication of a large-area tactile sensor with patterned PDMS conducting thin films. By using the microscale surface texture of silk as the mold, they fabricated micro-patterned PDMS thin film and the sensing device demonstrated superior sensitivity. Chun et al. [ 19 ] suggested a highly sensitive tactile sensor using a conductive polyurethane sponge where graphene flakes were self-assembled into the porous structure of the sponge. The special porous structure in sponge provides a conductive path for the piezoresistive element and increases the sensitivity of the sensor at the same time. Nevertheless, there are still two major drawbacks for these tactile sensors. The first is that sensors with a high micro-pressure measurement limit hardly possess a large detection range, which results in saturation of the sensing device under a large tactile pressure and limits the application scenario. Another disadvantage is that high-sensitivity sensors are often accompanied by complex and high-cost fabrication processes, which will limit the large-scale application of sensors and introduce more uncertainties in the fabrication process. Therefore, a method that can balance a large sensing range, high sensitivity, and cost of fabrication needs to be proposed to make the tactile sensor have good comprehensive characteristics including large-scale fabrication capacity, high sensitivity, and good flexibility. Herein, we propose a highly sensitive flexible tactile sensor by combining two types of microstructures in a stack structure with a simple and low-cost fabrication method. A pyramid structure and a double-sided rough structure are utilized to improve the sensitivity of the sensor and comparison of different structure distributions are implemented. Multi-walled carbon nanotubes (MWCNTs) and carbon black (CB) are filled into polydimethylsiloxane (PDMS) matrix as nanoscale conductive materials to obtain nanocomposites, and systematic comparison on proportions of different conductive materials and the piezoresistive property of a sensing device are carried out to determine the optimum filling ratio of CNTs. The tactile sensor in this work possesses good reliability, flexibility, a simple fabrication process, and shows short response times even under large contact pressure. What is more, the tactile sensor can detect muscle movement and help recognize different words from output signals by attaching to the human throat, which has the potential of language recognition.",
"discussion": "3. Results and Discussion Figure 6 a,b illustrates the flexibility of the tactile sensor. Fabricated from fully flexible material layers, the tactile sensor can bend at wide angles, revealing the ability of the sensor to attach on irregular surfaces in practical applications. The plane side of the pyramid structural layer contacts the middle rough structure layer, and the other pyramid structural side point contacts the electrode in the initial state, as shown in Figure 6 c. When external pressure is applied to the sensor, the rough structure layer and pyramid layers suffer from compressive strain, seen in Figure 6 d. The contact area between pyramids and the electrode surface and the contact area between the rough layer and the pyramid layer increase simultaneously. The resistance of the sensor can be written as: (1) R = ρ × l A \nwhere ρ is resistivity of the nanocomposite, A is the cross area between nanocomposite layers and electrodes, and l is the total length of nanocomposite layers. According to Equation (1), the increase of contact area and decrease of length of nanocomposite layers caused by compressive strain will lead to the decrease of resistance. In this work, the influence of different pyramid structures on the tactile sensor performance was compared and Figure 7 reveals the SEM images of different pyramid structures. Dense and sparse pyramid arrays with side lengths of 50 and 20 μm were fabricated, respectively. The spacing between adjacent pyramids in the dense array was 10 μm and the distribution density of the sparse array was half of that in the dense array, and individual pyramids in adjacent rows stagger with each other, which are shown in Figure 7 b,d. The details in the SEM images illustrate good consistency and structural integrity of the pyramid structure. In addition to the comparison of different distributions of pyramid structure, different combination types of structure layers were also compared, and the current-voltage curves were measured, as shown in Figure 8 a. The linear I–V curves reveal stable ohmic performance of the tactile sensor under varying pressures. Figure 8 b shows the resistance variations of different combinations between pyramid structure layers with a 50 μm side length of the individual pyramid and rough structure layer. The combination types include DPRS (dense pyramid structure and rough structure), dense pyramids only (two dense pyramid layers contacted to each other directly back-to-back without a middle rough structure layer), and reversed DPRS (two pyramid layers placed in reverse to make the pyramid structure get interlocked with the rough structure). The results reveal that DPRS has the highest sensitivity of 3.2 kPa −1 in the pressure range of <1 kPa, and the dense pyramids only have the sensitivity of 3.0 kPa −1 . As a contrast, the reversed DPRS has the lowest sensitivity of 1.53 kPa −1 . According to Equation (1), the change rate of resistance can be written as: (2) ∆ R = ∆ ρ × l A + ∆ l × ρ A − ∆ A × ρ × l A 2 . The definition of sensitivity is: (3) S = | ∆ R / R ∆ P | . When the nanocomposite layer is subjected to compressive stress, the change of resistivity ∆ ρ is negligible comparing to the change of sensor length. Additionally, the relative strain under unit pressure can be expressed as Young’s modulus by definition. Thus, the sensitivity of the tactile sensor can be reorganized as: (4) S = | ∆ l | / l ∆ P − | ∆ A | / A ∆ P = 1 E − | ∆ A | / A ∆ P . As displayed in Figure 8 b, the resistance variation of the tactile sensor can be divided into two stages obviously. In the ultralow pressure range, the sensor shows pretty high sensitivity and the sensitivity decreases significantly as the pressure exceeds 1 kPa. To explain the phenomenon theoretically, the low pressure results in smaller overall elastic modulus of the tactile sensor as the deformation of the microstructure layers is relatively small. According to Equation (4), the sensitivity of the sensor at this stage becomes higher. With the increase of the compressive strain of nanocomposite structure layers, Young’s modulus E of the sensor increases, which results in a lower sensitivity. Figure 8 c,d illustrates resistance variations of sensors with different pyramid structure layers in the low and high pressure range, respectively. It can be seen that the sensitivity of DPRS is higher than that of the sparse pyramid structure and rough structure (SPRS) with the same side length. Meanwhile, DPRS and SPRS with 50 μm side length possess higher sensitivity than DPRS and SPRS with 20 μm side length. In the pressure range exceeding 2 kPa, DPRS and SPRS with 50 μm side length show a similar sensitivity of 0.11 kPa −1 . The characteristic of high sensitivity in the low pressure range and low sensitivity in the high pressure range guarantees the tactile sensor suitable for practical application scenarios on simultaneous detection of small contact forces with high sensitivity and larger contact forces with unsaturated states in the large pressure range. Table 2 lists the comparison of this work with several recent reports. The test result indicates good compatibility and balance between pressure sensitivity and detection range, and good properties of our flexible tactile sensor. As shown in Figure 9 a, the flexible tactile sensor performs with good response stability and repeatability under the dynamic pressure loading and unloading process. The corresponding pressure was applied to the sensor through the pressure gauge cyclically and the output voltage of the sensor in a constant voltage circuit was recorded by the oscilloscope in real time. Benefiting from the good elasticity of polymer in the nanocomposite, the tactile sensor device has a short response time even under ultrahigh pressure, which is displayed in Figure 9 b. A 20 kPa pressure was applied to the sensor and released quickly, and the response time under 20 kPa was 217 ms and the recovery time was 81 ms. Figure 9 c shows the resistance variation according to wrist bending in different angles. The resistances of the sensor element under bending angles of 5°, 15°, 30°, and 45° are 81, 23, 14.6, and 10.5 kΩ respectively. Figure 9 d illustrates the image of application of monitoring human sounding by gently attaching the sensor to the neck near the throat. The real time monitoring of repeated signals of speaking a word are displayed in Figure 9 e,f, respectively. The high sensitivity, fast response time, good stability, and flexibility of the tactile sensor indicate wide applications in electronic skin and medical monitoring. The sensor shows good angle resolution in the measurement of wrist bending, providing the possibility of sensitively detecting the movement of human muscles and showing its potential application in human–computer interaction and physiological detection. The resistance of the tactile sensor changes steadily and consistently when the human body speaks different words. The detection of different human sounds demonstrates the potential of speech recognition of the sensor from the perspective of tactile perception and judging the voice status in real time by combining with a machine learning method."
} | 3,487 |
26760315 | null | s2 | 3,885 | {
"abstract": "In this study, we examine how the physical properties of cross-linking molecules affect the bulk response of bio-filament networks, an outstanding question in the study of biological gels and the cytoskeleton. We show that the stress-strain relationship of such networks typically undergoes linear increase - strain hardening - stress serration - total fracture transitions due to the interplay between the bending and stretching of individual filaments and the deformation and breakage of cross-linkers. Interestingly, the apparent network modulus is found to scale with the linear and rotational stiffness of the crosslinks to a power exponent of 0.78 and 0.13, respectively. In addition, the network fracture energy will reach its minimum at intermediate rotational compliance values, reflecting the fact that most of the strain energy will be stored in the distorted filaments with rigid cross-linkers while the imposed deformation will be \"evenly\" distributed among significantly more crosslinking molecules with high rotational compliance."
} | 261 |
33324497 | PMC7116476 | pmc | 3,886 | {
"abstract": "Asynchronous fluctuations of populations are essential for maintaining stable levels of bio-mass and ecosystem function in landscapes. Yet, understanding the stabilization of metacommunities by asynchrony is complicated by the existence of multiple forms of asynchrony that are typically studied independently: Community ecologists, for instance, focus on asynchrony within and among local communities, while population ecologists emphasize asynchrony of populations in metapopulations. Still, other forms of asynchrony, such as that which underlies the spatial insurance effect, are not captured by any existing analytical frameworks. We therefore developed a framework that would in one analysis unmask the stabilizing roles of local communities and metapopulations and so unify these perspectives. Our framework shows that metacommunity stabilization arises from one local and two regional forms of asynchrony: (1) asynchrony among species of a local community, (2) asynchrony among populations of a metapopulation, and (3) cross-community asynchrony, which is between different species in different local communities and underlies spatial insurance. For each type of stabilization, we derived links to diversity indices and associated diversity-stability relationships. We deployed this framework in a set of rock pool invertebrate metacommunities in Discovery Bay, Jamaica, to partition sources of stabilization and test their dependence on diversity. Cross-community asynchrony was the dominant form of stabilization, accounting for >60% of total metacommunity stabilization despite being undetectable with existing frameworks. Environmental variation influenced types of stabilization through different mechanisms. pH and dissolved oxygen, for example, increased asynchrony by decorrelating local species, while salinity did so by changing the abundance structure of metapopulations. Lastly, all types of asynchrony depended strongly on different types of diversity (alpha, metapopulation, and beta diversity drove local, metapopulation, and cross-community asynchrony, respectively) to produce multiple diversity-stability relationships within metacommunities. Our new partition of metacommunity dynamics highlights how different elements—from local communities to metapopulations—combine to stabilize metacommunities and depend critically on contrasting environmental regimes and diversities. Understanding and balancing these sources of stability in dynamic landscapes is a looming challenge for the future. We suggest that synthetic frameworks which merge ecological perspectives will be essential for grasping and safeguarding the stability of natural systems.",
"conclusion": "Conclusions Metacommunity dynamics defy simple analysis and management, at least in part, because they are not tractable by the local community or metapopulation perspective alone. Our novel partition unifies these organizational hierarchies to show how asynchrony arises through multiple local and regional pathways of environmental variation. A more complete view of metacommunity stability will come from recognizing the multiple forms of asynchrony that stabilize metacommunities, gauging their relative importance and studying the diversity-stability relationships which underlie them. We anticipate that highly resolved approaches like ours will prove powerful for disentangling stabilizing mechanisms that span the range of ecological hierarchies (e.g., subpopulations and functional groups).",
"introduction": "Introduction Community-level biomass or abundance varies over time and governs the rise and fall of ecosystem functions in landscapes. Such system-level fluctuations are stabilized when components (e.g., species) fluctuate asynchronously so that declines in one component are compensated by increases in another ( Doak et al. 1998 , Yachi and Loreau 1999 , Schindler et al. 2015 ). Because asynchrony reduces variation of community or ecosystem properties, it is important for ensuring their reliability. Alaskan salmon returns, for example, are stabilized by the existence of hundreds of uncoupled populations ( Schindler et al. 2010 ). Tallgrass prairie biomass is similarly stabilized where fire and grazing create a mosaic of asynchronous patches ( McGranahan et al. 2016 ). In turn, biomass stabilization can be crucial for stabilizing ecosystem functions like net primary production ( Wilcox et al. 2017 ). Recent work has isolated the mechanisms by which asynchrony stabilizes natural systems. Support for the insurance hypothesis ( Yachi and Loreau 1999 ) highlights the stabilizing effect of asynchronous species responses to environmental fluctuations ( Leary and Petchey 2009 , Hector et al. 2010 , Loreau 2010 ). Confirmed portfolio effects (sensu Doak et al. 1998 , Tilman 1999 ), meanwhile, demonstrate the power of diversity to stabilize communities or functional groups when species dynamics are weakly correlated ( Bai et al. 2004 , Cardinale et al. 2012 ). But while stabilization by asynchrony is well understood in local communities ( Thibaut and Connolly 2013 ), there is an urgent conservation need to scale that understanding up to metacommunities ( Wang and Loreau 2014 ). In a recent advance, Wang and Loreau (2014) partitioned the variability of total metacommunity biomass or abundance—gamma variability (γ CV ) —into local and regional components representing the variability of local communities (α CV ) and asynchrony among those communities (β). Their approach has rapidly become the most common in metacommunity asynchrony research and has underscored the importance of spatial heterogeneity in stabilizing metacommunity biomass and ecosystem function ( McGranahan et al. 2016 , Wilcox et al. 2017 ). But despite this progress, two barriers—one analytical and the other conceptual— prevent a deeper understanding of stabilization at the metacommunity scale. The analytical barrier is that the main local community framework used to date ( Wang and Loreau 2014 ) does not capture some forms of regional asynchrony that interest ecologists. Asyn-chrony among populations of a metapopulation, for example, helps to stabilize overall metacom-munity biomass ( Wilcox et al. 2017 ) and is critical for species persistence in landscapes ( Anderson et al. 2015 , Schindler et al. 2015 ). But this form of asynchrony is only implicit in the local community framework ( Wilcox et al. 2017 ), leaving its contribution to stability at the metacommunity scale unquantified. Another form of asynchrony overlooked by current frameworks is that which underlies the spatial insurance hypothesis ( Yachi and Loreau 1999 ), wherein different species occupying different patches fluctuate asynchronously and disperse to maintain ecosystem function ( Gonzalez et al. 2009 ). The above gaps may be seen to result from a conceptual problem: The form of asynchrony measured depends on the organizational hierarchy used to conceptualize and study a metacom-munity ( Fig. 1 ). Viewed as a set of local communities ( Fig. 1A ), for instance, the meta-community is stabilized by asynchrony among local communities (which we call type I asyn-chrony) and asynchrony of species within those local communities (type II; Wang and Loreau 2014 ). But if viewed (equally validly) as a set of metapopulations ( Fig. 1B ), it is stabilized by asynchrony among species metapopulations (type III) and asynchrony of populations within those metapopulations (type IV). Progress in stability research depends on bringing these overlapping metacommunity perspectives together in a single frame of reference. Wang et al. (2019) made an important step in this direction by relating the local community and metapopulation hierarchies in an analytical framework. However, the approach does not reconcile local communities and metapopulations in a single analysis to give their independent contributions to metacommunity stability. Nor does it capture the fifth form of asynchrony (type V)— among different species in different local communities—that is the generative mechanism for spatial insurance ( Loreau et al. 2003 , Gonzalez et al. 2009 ). Here, we present a new perspective on meta-community stabilization that overcomes the analytical and conceptual barriers left unad-dressed by past approaches ( Fig. 1C ). Viewing the metacommunity as a set of asynchronous local populations (i.e., population of species i in local community k ) allows a highly resolved view of metacommunity dynamics (e.g., Gouhier et al. 2010 ). Moreover, it lets us partition asynchronies that would be hidden if the meta-community was analyzed as a set of local communities or a set of metapopulations. On the conceptual front, the approach unifies the local community and metapopulation hierarchies in a single analytical partition by including elements of each. The resulting framework exposes how metacommunities are stabilized by one local-scale and two regional-scale forms of asynchrony—among local species (type II), among populations of a metapopulation (type IV), and among different species in different communities (type V), which we call cross-community asynchrony. Notably, these forms are wholly consistent with the definition of metacommunity dynamics (cf. Holyoak et al. 2005 :9) as including a local community component (e.g., type II), a spatial component (e.g., type IV), and a community x spatial component (e.g., type V). A further advantage is that because the framework works at the resolution of species populations, it offers ties to biological mechanism and diversity that other frameworks do not (see Box 1 for details). The framework thus has strong potential for synthesizing community and population ecology as well as exposing the stabilizing roles of diversity. Moreover, its application to empirical data— rock pool metacommunities here—should help to resolve the complex stabilization of ecosystems that emerges over many lower levels of organization ( Proulx et al. 2010 ). It therefore offers a tantalizing step toward a full accounting of temporal stability at the metacommunity scale. Analytical framework: Disentangling stabilization by local communities, metapopulations, and more Stabilization here is the reduction of variability at the metacommunity scale due to asynchrony. Analyzing the metacommunity as a set of local populations is the key to partitioning stabilizing asynchrony from both the local and metapopulation hierarchies. We define a local population ik as the individuals of species i living in a sampled local community k , though we recognize that these may not constitute a population in the demographic sense. As shown in Fig. 1C , focusing on local populations is the only approach that avoids an intermediate hierarchical level (e.g., local communities which are aggregates of local species) to expose all intra-and interspecific stabilization occurring at the population level. Using local populations as the unit of analysis, we can quantify total stabilization from local population asynchrony, x. This is the degree to which variability of metacommunity biomass (gamma) is reduced by asynchrony among all local populations in the metacommunity (i.e., between all populations living in all local communities or, equivalently, between all populations of all species; see Table 1 for formulae and Appendix S1 for derivations). Total stabilization reduces variability of meta-community biomass or abundance as (1) γ CV = ι CV − ω , \n where γ \n CV is the squared coefficient of variation (CV 2 ) of metacommunity biomass or abundance ( Wang and Loreau 2014 ). ι CV is a weighted and squared average variability of all populations in the meta-community. It is also the value of γ \n CV when all local populations are perfectly synchronized ( Appendix S1 : Eq. S6). There are just three forms of asynchrony that can occur among local populations to stabilize the metacommunity ( Fig. 1C ). Total stabilization (x) thus splits into three components corresponding to the different pairings of populations and covariances possible ( Appendix S1 : Fig. S1): (2) ω = δ + β mp + β CC , \n δ measures local stabilization or stabilization due to type I asynchrony among local species (species i with j in local community k ). It is equivalent to within-community stabilization in the local community hierarchy ( Fig. 1A ) and Wang and Lor-eau’s (2014) additive partition (see Appendix S2 ). β mp measures metapopulation stabilization or stabilization from type II asynchrony among populations in metapopulations (species i in local communities k and l ). It is equivalent to within-species stabilization in the metapopulation hierarchy ( Fig. 1B ) and an additive version of Wang et al.’s (2019) partition (see Appendix S2 ). Lastly, β \n CC quantifies cross-community stabilization from type V asynchrony between different species in different local communities (species i in local community k with species j in local community l ). This source of stability reflects the degree to which contrasting dynamics of species spread across the landscape reduces metacommunity variability and is notably masked in existing hierarchical frameworks ( Appendix S2 : Fig. S1). Because the same asynchrony mechanism underlies the spatial insurance hypothesis, β cc sheds new light on how species diversity and environmental heterogeneity interact to stabilize metacommunities. Metacommunity biomass or abundance is stabilized whenever there is asynchrony among local populations in the metacommunity. We can express this asynchrony as 1 – φ pop , where φ is Loreau and de Mazancourts (2008) dimensionless measure of synchrony. Doing so, we find that stabilization (ω) depends on the average variability of local populations (ι CV ) and their asynchrony in the metacommunity ( Appendix S3 ): (3) ω = ( 1 − φ pop ) ι CV . \n Ecologists often study asynchrony as opposed to the resulting reduction of metacommunity variability ( Thibaut and Connolly 2013 , Hautier et al. 2014 ). For these applications, we can apply the same additive partition of stabilization ( Eq. 2 ) to partition asynchrony ( Appendix S3 ). We find population asynchrony (1 – φ pop ) to be a composite of asynchrony from local (δ/ι CV ), metapopulation (β mp /ι CV ), and cross-community pairs of populations (β cc /ι CV ): (4) 1 − φ pop = δ ι CV + β mp ι CV + β cc ι CV . \n These components change with the degree of correlation between populations and enable deeper analysis of asynchrony in metacommunities. Stabilization of rock pool metacommunities We illustrate our analytical framework in a set of tropical rock pool metacommunities. This system has been well-studied and has many positive attributes for testing metacommunity theory, such as high species diversity, discrete, identifiable local communities, and relatively independent annual samples of community composition and structure ( Kolasa and Romanuk 2005 ). It thus offers a clear and well-resolved system for testing our framework. Our main goal is to understand how forms of stabilizing asynchrony that were previously overlooked or considered separately combine to stabilize metacommunities. In specific terms, stabilized metacommunity biomass or abundance has implications for sustaining generalist predators in rock pools (e.g., crab larvae Sesarma miersii Rathburn 1897) and smoothing ecosystem processes (e.g., primary productivity; Wilcox et al. 2017 ). But as a broader exploration of stabilization pathways, we ask: Is metacommunity abundance most stabilized by local communities, metapopulations, or cross-community combinations of species? What environmental factors drive the various forms of stabilization? How do different forms of diversity influence stabilization at the metacommunity scale? \n Through a unifying approach, our findings highlight multiple paths by which diversity stabilizes metacommunities from local to regional scales.",
"discussion": "Discussion We presented a solution for partitioning forms of asynchrony that are partially or wholly hidden when metacommunities are analyzed as a hierarchy of local communities or metapopulations. By taking a metacommunity as a set of asynchronous local populations, our analytical framework reveals how metacommunities are stabilized by one local and two regional forms of asynchrony (local, metapopulation, and cross-community). Not only is this perspective consistent with the classical conception of metacommunity dynamics (see Introduction; Holyoak et al. 2005 :9), but it also unifies the local community and metapopulation approaches to studying metacommunities by capturing stabilization from each (see also Wang et al. 2019 ). Our empirical results further underscore how diversity and environmental variation support the wide range of stabilizing mechanisms in natural metacom-munities. Cross-community stabilization: a hidden source of stability Cross-community stabilization, while a core of our framework, is seldom recognized as a force smoothing metacommunity variability. This omission is in spite of being deemed necessary for spatial insurance ( Gonzalez et al. 2009 ) and being implicit in finely resolved descriptions of metacommunity dynamics ( Gouhier et al. 2010 ). Yet, this particular form of asynchrony dominated over all others ( Fig. 2A ), highlighting its importance in reducing variation at the meta-community scale. Since this source of stability is not evident when gamma variability is decomposed as a local community or metapopulation hierarchy ( Fig. 1C ), studies using these frameworks may underestimate stability arising from asynchrony and miss a unique (spatial 9 species) component of metacommunity dynamics. In turn, recognizing this component will strengthen theoretical and empirical understanding of how spatial heterogeneity and species richness interact to stabilize landscapes, such as through spatial insurance effects. TTwo factors—one biological and one numeric —may make cross-community stabilization a widespread and potent force in natural ecosystems. First, and biologically, cross-community populations were the least correlated ( Fig. 3B ) likely due to stronger differential responses to environment. Since Jamaican metacommunities are strongly forced by environmental variation, weak correlation among populations probably owes to differential responses of populations to environmental changes. Thus, the observed weak correlation within local communities ( Figs. 2C , 3 ) is consistent with the local insurance hypothesis ( Yachi and Loreau 1999 ) in which species respond differently to local environmental cues ( Leary and Petchey 2009 , Thibaut et al. 2012 ). Weak correlation within metapopulations, in turn, likely reflects the tracking of different environmental regimes by local populations ( Ringsby et al. 2002 ). And the very weak correlations we found among cross-community populations likely stem from differential responses of species across space—the same mechanism of compensatory dynamics in the spatial insurance hypothesis ( Loreau et al. 2003 ). The strength of this stabilizing effect may owe to a doubling up of differential responses: Different species have contrasting responses to the environment and, by living in different local communities, experience different environmental fluctuations. Notably, this effect increased with scale as cross-community asynchrony, pairwise decorrelation and stabilization were accentuated at the whole landscape level ( Figs. 2A , 3A, B ), presumably and in part as more spatial heterogeneity, and species turnover was included in the sampled area. Second, the biological causes of high cross-community stabilization are likely to be compounded by the numerical dominance of cross-community populations. We found that the number of cross-community population pairs outstripped the number within local communities or metapopulations ( Fig. 2B ), with each additional pair adding stabilizing potential akin to a portfolio effect ( Blüthgen et al. 2016 ). Our calculations further suggest that cross-community pairs will dominate in all but the smallest metacommunities (those with less than three local communities and regional species; see Appendix S9: Fig. S1 ). We propose that further exploration of the numeric and biological causes of cross-community stabilization will bring important insights about when and where cross-community pairs will contribute most to metacommunity stability. Our framework may also be profitably extended to include functional groups and their specific contributions to stabilization in the local, metapopulation, and cross-community context. We further note that asynchrony specific to ecologically important interactions—such as between predator and prey or plants and pollinators—may also be obscured in current metacommunity frameworks. Future and targeted incorporation of these into partitions will bring ecology closer to a full accounting of stabilizing forces in metacommunities. An integrated view of metacommunity stabilization Our approach allowed for an integrated view of stabilization from local communities and metapopulations. Though cross-community stabilization dominated the metacommunities, stabilization from within local communities and metapopulations was still indispensable and together accounted for nearly half of all stabilization ( Fig. 2A ). This observation promotes the unifying view that metacommunities are meaningfully stabilized by several lower levels of organization and ecological entities. Thus, sorting out the relative impacts of local communities, metapopulations, and more will be crucial to understanding and managing the stability of landscapes. With multiple forms of stabilization or asynchrony to balance comes the potential for tradeoffs. Most simplistically, this is because forms of stabilization or asynchrony are collectively exhaustive (Eqs. 2, 4). This property means that given a fixed amount of total stabilization or asynchrony, an increase in one type (e.g., local stabilization) comes at the expense of another (e.g., metapopulation stabilization). Some real-world trade-offs indeed seem possible. Gouhier et al. (2010) , for instance, report differential effects of environmental variation on local community and metapopulation asynchrony at certain levels of dispersal. Similarly, managing for one type of asynchrony may unwittingly modify other types. Species-based management, for example, encourages habitat heterogeneity to stabilize metapopulations (e.g., of butterflies; Oliver et al. 2010 ) but could promote habitats with factors that synchronize local species (e.g., generalist predators; Raimondo et al. 2004 ). Conversely, community-based management may prioritize species with asynchronous dynamics (e.g., in forests; Morin et al. 2014 ), but these could include species with easily synchronized local populations (e.g., masting species; Koenig and Knops 2013 ). The relative balance of asynchrony forms will also likely be relevant to the maintenance of resilience in metacommunities. Disturbances or management actions that dampen local species asynchrony, for instance, may weaken local insurance effects ( Yachi and Loreau 1999 ). Rescue effects ( Brown and Kodric-Brown 1977 ) similarly depend on metapopulation asynchrony for vigorous local populations to subsidize moribund ones via dispersal. Disruption of cross-community asynchrony, finally, may impair spatial insurance effects ( Loreau et al. 2003 ) in which ecosystem functions are buffered by asynchrony within functional groups (e.g., primary producers; Symons and Arnott 2013 ). Because rescue, local, and spatial insurance effects depend on different types of asynchrony, an important future research question is how these can be optimized in managed landscapes. Our framework might prove useful for connecting underlying patterns of asynchrony with their associated ecological effects (i.e., rescue, local insurance, and spatial insurance effects). The multifaceted nature of metacommunity stabilization was also apparent in the variety of environmental controls over stabilization ( Appendix S7 ). Notably, stabilization could be variously promoted or impaired by environmental forcing of the correlation, relative abundance, and variability components of stabilization. If such complex causation is the norm, ecologists will need to move beyond single causes of stabilization ( Downing et al. 2014 ) and elucidate how multiple environmental drivers impact different forms of stabilization and asynchrony. A complete picture of metacommunity stabilization—similar to the local community case ( Thibaut and Connolly 2013 )—will include understanding how environmental variation differentially affects each statistical component of stabilization (e.g., evenness, correlation, and variability). Absent this detailed understanding, our analysis suggests that preserving biodiversity may be the most viable route to maintaining asynchrony and stability in changing environments (cf. Anderson et al. 2015 ). Multiple paths from diversity to metacommunity stability Diversity–stability research asks how much and what kind of diversity is needed to support stable ecosystems. Our results show that population diversity increases asynchrony ( Fig. 4A ), indicating that large metacommunities buffer change in the same way that large financial portfolios enable diversification and variance reduction ( Doak et al. 1998 , Anderson et al. 2015 ). Looking deeper, we find that population diversity—and its stabilizing effect—is a composite of other known types of diversity ( Eq. B4 ). Strikingly, at least three diversity–asynchrony relationships stabilize metacommunities and depend on how populations are distributed across local communities and metapopulations. Alpha diversity, for instance, predicted the amount of asynchrony generated within local communities. This is consistent with previous work showing that a diverse local species pool often buffers community-level variation ( Cardinale et al. 2012 , Wang and Loreau 2016 ). Similarly, metapopulation asynchrony grew with the diversity of constituent populations and agreed with studies showing the variance-reducing effects of large metapopulations ( Anderson et al. 2015 ). Cross-community asynchrony, lastly, increased with additive beta diversity. From its equation in Table 2 , we see why: As beta diversity grows, so too does the weight of cross-community population pairs and thus their potential contribution to spatial asynchrony. This, combined with the numerical dominance of cross-community pairs, suggests that preserving beta diversity may be of paramount importance for metacommunity stability—a position supported by positive beta diversity–stability relationships in the literature ( Mellin et al. 2014 , Wang and Loreau 2016 ). The diversity–asynchrony relationships we found can be considered portfolio effects because rock pool populations were very weakly correlated (see Figs. 3 , 4 )—a common assumption of portfolio theory ( Doak et al. 1998 , Tilman et al. 1998 ). They may therefore be expected in similarly stochastic metacommunities. But equations and simulations show they may also emerge in more deterministic systems where environmental fluctuations synchronize dynamics. First, Eqs. B1–B3 and our null models predict that the main condition for a positive diversity–asynchrony relationship is simply that an added population has a unique response to environmental fluctuations. Second, diversity–asynchrony relationships are robust to varying levels of interpopulation correlation and only weaken and disappear as populations approach perfect correlation, as predicted by theory ( Appendix S8 ; see also Fig. 5.3 in Loreau 2010 ). Given this robustness, the smoothing of metacommunity variability by multiple diversity–stability relationships may be a widespread phenomenon. If so, the critical challenge will be to recognize and conserve, not just species diversity (e.g., Leary and Petchey 2009 ) or patch diversity ( Wilcox et al. 2017 ), but the suite of local community, metapopulation, and cross-community diversities that collectively stabilize landscapes."
} | 7,048 |
33855280 | PMC8024921 | pmc | 3,888 | {
"abstract": "Summary The collective motion of swarms depends on adaptations at the individual level. We explored these and their effects on swarm formation and maintenance in locusts. The walking kinematics of individual insects were monitored under laboratory settings, before, as well as during collective motion in a group, and again after separation from the group. It was found that taking part in collective motion induced in the individual unique behavioral kinematics, suggesting the existence of a distinct behavioral mode that we term a “collective-motion-state.” This state, characterized by behavioral adaptation to the social context, is long lasting, not induced by crowding per se, but only by experiencing collective motion. Utilizing computational models, we show that this adaptability increases the robustness of the swarm. Overall, our findings suggest that collective motion is not only an emergent property of the group but also depends on a behavioral mode, rooted in endogenous mechanisms of the individual.",
"introduction": "Introduction The ability to form groups that move collectively is a key behavioral feature of many species ( Sumpter, 2006 ; Ward and Webster, 2016 ), assumed to increase the survival of both individuals and groups ( Be'er and Ariel, 2019 ; Yang and Schmickl, 2019 ). Collectively moving organisms, however, differ in the levels of peer-to-peer interactions, ranging from minimal cooperation to complex social behaviors ( Attanasi et al., 2014 ; Cavagna et al., 2010 ). Furthermore, endogenous differences among individuals, heterogenic environments, and variability in the interactions between the individual and its direct environment are all sources of variance that may affect the coordinated behavior of the collective. Accordingly, it is not clear how synchronized collective motion constitutes such a robust phenomenon, maintaining its form across various group sizes and densities, and under heterogeneous and unpredictable environmental conditions. One of the most interesting, albeit disastrous, examples of collective motion is that of the marching of locusts. These insects swarm in groups of millions, migrating in mass across large distances, devastating vegetation, and agriculture ( Ayali, 2019 ; Cullen et al., 2017 ; Zhang et al., 2019 ). In the context of social interactions, locust swarming is characterized by a minimal level of cooperation between individuals: collectivity, which is based on local interactions, is mostly manifested in alignment among neighboring individuals and in maintaining the overall movement in the same general direction (e.g., Ariel et al., 2014a ; Bazazi et al., 2008 ). Nonetheless, the locust swarming phenomenon is extremely robust, with huge swarms demonstrating moderate to high collectivity on huge scales (up to 6–7 orders of magnitudes), in terms of both the number of animals and their spatiotemporal distribution ( Ellis and Ashall, 1957 ; Uvarov, 1977 ; see further references in Ariel and Ayali, 2015 ). Thus, locusts exhibit a considerable disparity between little local cooperation and large-scale collectivity. What is the key to this ability of locust swarms to maintain their integrity? Here, we show by a series of carefully controlled behavioral experiments that collective movement induces an internal switch in the individual gregarious locust, activating a behavioral mode we refer to as a “collective-motion-state.” In this state, the kinematic behavior of individuals notably differs from that during a non-collective-motion-state. It is important to emphasize that both the “collective-motion-state” and the “non-collective-motion-state” are internal states of swarming-gregarious locusts. We are not referring to the well-known solitarious-gregarious phase transition in locusts ( Ayali, 2019 ; Cullen et al., 2017 ). How, then, does the collective-motion-state affect the formation and robustness of the swarm? Interestingly, the switch into this state seems to occur rapidly, and in response to coordinated walking. In particular, our experiments indicate that aggregation alone is not sufficient. Switching out of the collective-motion-state occurs over a longer timescale—significantly longer than the typical timescale of normal fluctuations around the swarm typical dynamics. Hence, stochastic fluctuations, typical to swarming behavior ( Algar et al., 2019 ; Ariel and Ayali, 2015 ; Escaff et al., 2018 ), are “smoothed-out,” leading to highly robust dynamics of the swarm collective behavior, which is in turn beneficial for the swarm integrity. Using a simplified computer model, we simulated the swarming properties of locust-like agents with different kinematic parameters, representing the different behavioral states. The results support the functional advantages of the collective-motion-state, allowing us to conclude that the collective-motion-state provides an individual-based mechanism that increases the stability of swarms in the presence of fluctuations, preventing the swarm from collapsing.",
"discussion": "Discussion Our findings reported here suggest that, in locusts, the sensorimotor act of collective motion is accompanied by an internal state of the individual locust—a collective-motion-state, which is manifested in specific behavioral kinematics. This state is induced by the experience of synchronous, collective marching. In turn, it has an important role in maintaining the integrity and consistency of the swarm. Next, we discuss several key aspects and implications of this finding. It should be stressed once again that the current study focused on gregarious, crowded-reared locusts only. The described behavioral states should not be confused, therefore, with the well-known and much researched locust density-dependent phase polyphenism ( Ayali, 2019 ; Cullen et al., 2017 ). Collective motion is limited to the gregarious, swarming, and migrating phase. Accordingly, all our experimental animals were taken from our gregarious (crowded-reared) breeding colony, maintained in crowded conditions for many consecutive generations. In their breeding cages, mostly due to the physical constraints and abundance of food, despite experiencing high density, the locusts very rarely, if at all, demonstrate collective motion. Thus, they adopted the collective-motion-state only upon experiencing, and taking part in, collective marching within the experimental arena. In a recent study ( Knebel et al., 2019 ), we have introduced a comparison between the walking behavior kinematics of individual gregarious locusts in different social (density) contexts. Our reported findings are reconfirmed and further elucidated here by the results of the initial isolation and grouping stages in our experiments. The novel idea posited here is that these differences represent not only the spatially and temporally immediate social environment and the instantaneous local interactions among locusts but also are dictated by the effects of an internal state induced by the general experience of collective motion. A fundamental aspect of the concept of the collective-motion-state arises from our findings related to its persistent effect in time: upon re-isolation, the individual locust adopted behavioral kinematics that critically differed from that in the first experimental stage (initial isolation). We also showed that, as expected, the collective-motion-state is transient. If the locust does not experience collective motion for some time, and is then isolated once more, it loses the unique walking-related kinematics it previously adopted in response to the collective motion, i.e., the internal collective-motion-state. The dynamics of this decay were not explored, but are likely to be affected by many external factors, such as the availability of food and the day-night cycle. The individual locusts in our experiments retained the variability demonstrated in our previous report ( Knebel et al., 2019 ), while demonstrating a second layer of variability or plasticity upon experiencing collective motion, when entering the collective-motion-state. Considerable research has been devoted in recent years to understanding the effect of variability among individuals on the group's collective behavior, both experimentally—ranging from bacteria to primates ( Benisty et al., 2015 ; Brown and Irving, 2014 ; Crall et al., 2016 ; Dyer et al., 2009 ; Farine et al., 2017 ; Fürtbauer and Fry, 2018 ; Herbert-Read et al., 2013 ; Jolles et al., 2018 ; Planas-Sitjà et al., 2015 )—and theoretically ( Aplin et al., 2014 ; Ariel et al., 2014b ; Calovi et al., 2015 ; Copenhagen et al., 2016 ; Guisandez et al., 2017 ; Jolles et al., 2017 ; Menzel, 2012 ; Mishra et al., 2012 ; see Mar Delgado et al., 2018 ; Modlmeier et al., 2015 ; Webster and Ward, 2011 for recent reviews). The interactions between variability in specific aspects of the individuals' behavior and group-level processes were found to be complex and, moreover, bidirectional (e.g., Knebel et al., 2019 ). Variability among individual animals was found to have important consequences for the collective behavior of the group (e.g., O'shea-Wheller et al., 2017 ; Szorkovszky et al., 2018 ). However, beyond the variability among the individuals composing a group, variability is also expected in the behavior of the individual animal over time, as it experiences changes in environmental and social conditions. The swarm (or flock, shoal, herd, etc.) is a heterogeneous entity, moving in a heterogeneous environment. The individual is bound on occasion to find itself in different locations within the swarm (e.g., leading edge, at the outskirts, trailing), and it may also find itself separated from the group by natural obstacles (vegetation, rocks, and boulders). It is essential for the robustness and consistency of the swarm that throughout these changing conditions the behavior of the individual will adapt accordingly, such as to be appropriate for the changing context. For example, if temporarily separated from the core of the swarm, a locust's walking kinematics should change to support rapid reunion with the group, as reported in both our experimental and simulation findings (e.g., increased fraction of walking and duration of walking bouts). If previously naive to collective motion, that individual's kinematics would, however, be disadvantageous, or even hinder the formation of a swarm. In Bazazi et al., 2012 , the authors suggest that behavioral variability can be explained by the existence of two internal states. Studying single locusts in isolation for 8 consecutive hours, they have observed changes in behavioral kinematics that were suggested to result from “internal state behavioral modulation.” The observed variations, however, were merely attributed to changes in “starvation/satiation state,” i.e., as the locust becomes starved, it changes its walking behavior, searching more vigorously for food. Moreover, they conclude that animals continually switch between the two states on a scale of minutes. The collective-motion-state reported here is, of course, a very different type of internal behavioral state, which is strongly involved with the locust past and current social environment. It may be viewed as a form, or a manifestation of a social carryover effect ( Niemelä and Santostefano, 2015 ), where a social environment experienced by a focal individual affects aspects of its locomotion behavior at a later, non-social context. As noted, however, the change in behavioral state described here is induced by collective marching, i.e., a particular mode of social interaction, rather than by aggregation or being around other conspecifics per se. Moreover, as our simulations show, the enhanced marching displayed in the re-isolation stage is advantageous for maintaining collective swarming—it is still much related to the social context rather than carried over to a non-social one. The reported collective-motion-state is also in accord with the overall daily behavioral changes of marching locust swarms. The swarm will spend the night (as well as times of low temperature or other unfavorable climatic conditions) roosting among the vegetation. Upon suitable conditions, after a period of feeding, the locusts will initiate marching—highly synchronized, collective motion. Frequently, when temperature becomes too high around noon, or when dusk arrives, the swarm will again switch to feeding and roosting. These daily patterns call for corresponding changes in the internal behavioral states of the individual locusts and mostly a dedicated collective-motion-state. In the current work we are cautious in discussing the underlying mechanisms of the behavioral states reported. Although this is beyond the scope of this study, it is clear that these behavioral states represent physiological states. With some confidence, we can speculate about the nature or the physiological mechanisms involved in the demonstrated behavioral states. Behavioral plasticity in locust behavior has been attributed to various second messengers or neuromodulators, or to the balance among them. Most notable are the biogenic amines (e.g., serotonin, a prominent bio-amine, was recently reported to inhibit walking behavior in Drosophila ; Howard et al., 2019 ). Hence, it may well be that the (spatial and temporal) immediate social environment affects biogenic amine levels, and these in turn modulate the walking-related behavioral kinematics manifested in the different behavioral states. Another candidate that may be involved in the collective-motion-state is the locust adipokinetic hormone (AKH). AKH is a metabolic neuropeptide principally known for its mobilization of energy substrates, notably lipid and trehalose, during energy-requiring activities such as flight and locomotion, and also during stress (e.g., Perić-Mataruga et al., 2006 ). It is well accepted that the metabolic state affects the level of general activity of an organism, and AKHs are reported to stimulate locomotor activity, either directly by way of their activity within the central nervous system (e.g., Wicher, 2007 ) or via octopamine—a biogenic amine with ample behavioral effects ( Verlinden et al., 2010 ; Yang et al., 2015 ). Furthermore, as noted, we have demonstrated here an extended effect of the experience of collective motion. Hence, learning and memory-related mechanisms would also seem to be involved. Again, previous work may suggest some candidate molecules and pathways, including cGMP-dependent protein kinase (PKG) and protein kinase A (PKA) ( Geva et al., 2010 ; Lucas et al., 2010 ; Ott et al., 2012 ). Last, as noted, solitarious phase locusts lack the capacity to demonstrate collective motion, and thus also the collective-motion-state. Accordingly, they differ from gregarious locusts in all the above-mentioned physiological pathways (bioamines: e.g., Alessi et al., 2014 ; Cullen et al., 2017 ; Ma et al., 2015 ; AKH: Ayali and Pener, 1992 ; Pener et al., 1997 ; PKG: Lucas et al., 2010 ; PKA: Ott et al., 2012 ). An in-depth investigation of the development of gregarious-like states in solitary locusts should prove to be very enlightening. A central question is whether a collective (herd, flock, or swarm) is merely a sum of its parts, or a new entity. Most related studies have perceived collectivity as a self-emergent phenomenon, suggesting that new dynamics and behavior are the result of intricate, multi-body, typically non-linear interactions (e.g., Cucker and Smale, 2007 ; Vicsek and Zafeiris, 2012 ). One hidden assumption underlying this perception is that individuals remain inherently unchanged when isolated or in a crowd. Even studies of heterogeneous swarms, in which conspecifics may differ from each other, still assume consistency in the properties of the individual over time. This is essentially a physical point of view, in the sense that agents/individuals possess certain properties that determine their behavior across a range of situations. Thus, the collective motion is an emergent property that builds up in particular contexts, such as a sufficiently high local density of animals. This point of view allows, among others, extrapolation from experiments with one, two, or a few animals to large swarms (e.g., Calovi et al., 2015 ). Our findings reported here suggest a fundamentally different point of view. We perceive the sensorimotor act of collective motion as accompanied by an internal state—a collective-motion-state that is manifested in specific behavioral kinematics. This state is induced by the experience of synchronous, collective motion. Most importantly, it is not induced by spatial aggregation alone. Collectivity, therefore, is not just self-emerging. Rather, the collective-motion-state has an important role in maintaining the integrity and consistency of the swarm. The robustness of the swarm is also a major challenge and requirement in swarming robotics, making the current novel insights applicable and even important also to this emerging field. In the case of locusts, our far-from-complete understanding of the swarming phenomenon is also proving crucial for human well-being and survival, as evident from the current devastating locust situation in large parts of Africa and Asia ( FAO, 2020 ). Much scientific attention has been dedicated to the perception, decision-making, and individual kinematics of locusts in a swarm. These efforts have led to various models that attempt to explain the collective behavior on the basis of local interactions among the individual locusts (see Ariel and Ayali, 2015 for review). The current study is, to the best of our knowledge, the first to include the internal state of the individual locust as an important factor in dictating its behavior, and in turn affecting the maintenance and the properties of the swarm. Limitations of the study The study presented here outlines a post-swarming behavioral state of individuals. Clearly, as noted, this state is induced by neurochemical changes such as secretion of neuromodulators and/or hormones. Yet, it was beyond this research to pinpoint the exact neuronal mechanisms involved. Furthermore, the presented model is simplified and ignores various aspects of locust swarming that might be critical. However, the simplicity is also a virtue of the model, which can be easily generalized to other systems. In addition, although we show that the collective-motion-state is transient, we did not explore its temporal materialization and decline. Data and code availability The data will be made available upon request."
} | 4,658 |
25801303 | PMC4382991 | pmc | 3,891 | {
"abstract": "Neural networks are currently implemented on digital Von Neumann machines, which do not fully leverage their intrinsic parallelism. We demonstrate how to use a novel class of reconfigurable dynamical systems for analogue information processing, mitigating this problem. Our generic hardware platform for dynamic, analogue computing consists of a reciprocal linear dynamical system with nonlinear feedback. Thanks to reciprocity, a ubiquitous property of many physical phenomena like the propagation of light and sound, the error backpropagation—a crucial step for tuning such systems towards a specific task—can happen in hardware. This can potentially speed up the optimization process significantly, offering important benefits for the scalability of neuro-inspired hardware. In this paper, we show, using one experimentally validated and one conceptual example, that such systems may provide a straightforward mechanism for constructing highly scalable, fully dynamical analogue computers.",
"discussion": "Discussion In this paper, we have proposed a framework for using reciprocal physical dynamical systems with nonlinear feedback as analogue RNNs. We have demonstrated that the error backpropagation algorithm, which efficiently optimizes an RNN, can be implemented physically on the same system, thus greatly reducing the necessary computations required for the optimization process. This in turn paves the way to faster, more scalable analogue computing. We have experimentally verified the proposed system using a real-world acoustic set-up as a proof of concept, as well as a simulated electro-optical set-up. In this second, more complex set-up, we explored the impact of expected sources of non-ideal behaviour on the performance of a real-world task and where we demonstrated that good performance does not require a very precise physical implementation of the backpropagation algorithm, and can tolerate reasonable levels of noise and nonlinearity. We have also included a short discussion on what processing speed may be obtained for an electro-optical set-up in the Supplementary Discussion . The concepts presented in this paper provide a sizeable step forward towards a novel form of processing, which relies far less on software implementations and comes much closer to brain-like computation, especially in the sense that the system can partially internalize its own training procedure. By using analogue physical processes to compute, we may benefit from inherent massively parallel computing, great potential speed benefits and low-power usage, the properties that were the initial motivation for research into physical RC systems. Specifically, the obtainable speed does not depend on the dimensionality (number of sources and receivers) of the system, offering inherent scalability. The possibility to fully optimize all internal system parameters using the backpropagation algorithm offers great performance improvements, and makes the application domain of the proposed set of systems far greater (as has been evidenced in ref. 20 ). This also means that such physical set-ups can potentially become competitive with digitally implemented neural networks, which are currently the state-of-the-art for several important signal processing problems. The training process too benefits from being implemented physically, meaning that there is only limited need for external processing. If the physical system under consideration has speed benefits compared with a digitally implemented neural architecture, these benefits are also present for training the system. Finally, while physically implementing the training process comes at an additional complexity cost (modulation of the feedback with the jacobian), the benefits over optimization of the parameters in simulation is paramount. Optimization in simulation might require a very precise model, and it is hard to predict how model-reality discrepancies would manifest themselves once the parameters obtained in simulation are applied to the physical set-up. When instead optimizing on the physical system there can be imperfections like the ones we explored in Supplementary Note 4 but one is certain that the measurements used to base the training on are those from the real system, and no additional system characterization is needed to perform the training. Important challenges for the large-scale execution of this scheme still remain. The current set-up still requires some level of external processing for computing the gradients. If the analogue part of the system is sufficiently fast, gradient computations may become the bottleneck of training, though this may be partially redeemed by the fact that they are solely matrix–matrix multiplications (without a sequential part), which means that it is fully parallelizable. Recording the signals in the forward and backward pass still requires digitization. Currently, this is the most important hurdle to scaling up the system in practice, as analogue-digital conversion at high speeds is expensive and consumes a lot of power. This limits the number of sources and receivers that can be practically applied. Time multiplexing of the input, as used in both examples of this paper, partially solves this problem, but at the cost of reducing the obtainable speed of the system. Larger numbers of sources and receivers would also benefit more from the spatial parallelism that is offered by acoustic or optic systems. With currently available hardware, one could potentially build physical systems that are competitive with digitally implemented neural networks, as we demonstrated using a simulated electro-optical set-up. Truly exploiting the full potential of analogue physical computation, however, very likely requires the design of novel hardware that internalizes all necessary elements into a single device. In particular, future research into this topic should explore ways to develop hardware which has impulse responses determined by large amounts of controllable parameters. This would increase the number of trainable parameters and hence the representational power of the systems. Finally, it is also of importance to relate the results found in this paper to developments in neural network research. For one, it was recently found 31 that random feedback weights for the backpropagation phase can also be used to train feedforward networks. It should be investigated if this has implications for the recurrent systems under considerations in this paper."
} | 1,622 |
29491915 | PMC5829439 | pmc | 3,893 | {
"abstract": "Abstract Ant colonies self-organize to solve complex problems despite the simplicity of an individual ant’s brain. Pavement ant Tetramorium caespitum colonies must solve the problem of defending the territory that they patrol in search of energetically rich forage. When members of 2 colonies randomly interact at the territory boundary a decision to fight occurs when: 1) there is a mismatch in nestmate recognition cues and 2) each ant has a recent history of high interaction rates with nestmate ants. Instead of fighting, some ants will decide to recruit more workers from the nest to the fighting location, and in this way a positive feedback mediates the development of colony wide wars. In ants, the monoamines serotonin (5-HT) and octopamine (OA) modulate many behaviors associated with colony organization and in particular behaviors associated with nestmate recognition and aggression. In this article, we develop and explore an agent-based model that conceptualizes how individual changes in brain concentrations of 5-HT and OA, paired with a simple threshold-based decision rule, can lead to the development of colony wide warfare. Model simulations do lead to the development of warfare with 91% of ants fighting at the end of 1 h. When conducting a sensitivity analysis, we determined that uncertainty in monoamine concentration signal decay influences the behavior of the model more than uncertainty in the decision-making rule or density. We conclude that pavement ant behavior is consistent with the detection of interaction rate through a single timed interval rather than integration of multiple interactions.",
"discussion": "Discussion Pavement ant colonies self-organize to accomplish complex and nuanced tasks despite the relatively simple brains of the individual ants. The aggregation of simple deterministic rules in response to external stimuli can lead to subtle and wide-ranging societal behavioral changes at the colony level. Here, we use a model to demonstrate how changes in brain concentrations of the monoamines 5-HT and OA after interactions with nestmate ants could cause a brain state that leads to the engagement of fighting behavior of conspecific non-nestmate ants. In particular, we examined the behaviors associated with nestmate recognition and the ritualized, aggressive exclusion of conspecific non-nestmates at the boundary between territories. Although it is known that pavement ants only engage in ritualized combat when they had sufficiently interacted with other nestmates (Greene MJ, unpublished data), it was assumed that the rate or abundance of interactions with nestmates over time was being integrated by the worker ants (Greene MJ, unpublished data). However, increased density and, therefore increased interaction rate, did not result in changes in brain monoamine concentrations ( Figure 3 ; Greene MJ, unpublished data). Based on this result, we hypothesize that the ant’s decision making observes a hysteretic effect of their most recent interactions based on the return of brain monoamines to baseline concentrations and thus determine interaction rate through a single interval measurement instead of integration over multiple interactions. This conceptualization of decision making via a “monoamine clock” timed by the rate of return of monoamines to baseline concentrations reconciles the apparent requirements of integrating complex information with the relatively simple organization of an individual ant brain. A key implication of this finding is that the density sensing apparatus of the pavement ant worker observes the mathematical forgetfulness or memoryless property ( Leemis and McQueston 2008 ), that is, there is no difference between an ants 1st interaction and its 20th, the probability of having concentration levels above the decision threshold at some time ( t ) after each interaction is the same. This model, consistent with the proposed single interval measurement, demonstrates a way in which interaction rate would not lead to differences in monoamine concentrations but still affect decision making in ants. It further implies that the decision-making processes attributed to rate of interaction is more accurately attributed to the amount of time a worker spends primed for a decision after an increase in monoamine concentrations. The proposed decision mechanism is further supported by our agent-based model as uncertainty in density had minimal effects on the system behavior compared with changes due to uncertainty in the rate of monoamine concentration decay. In the terms of a “monoamine clock,” this would indicate that the rate at which the clock ticks (decay rate) is more important than how often the clock is reset (interaction rate). To date, the dynamics for monoamine concentrations decay in the ant brain are unknown, and therefore estimates of this key parameter for decision making are by necessity highly uncertain. In order to advance our understanding of how individual, simple, and physiologically driven decision making can lead to the development of complex, nuanced, and colonial responses, a better understanding of both the timescale and function of this “monoamine clock” must be established. Future directions Due to the small size of ant brains, HPLC analysis requires an aggregation of 2 brains per a sample. This course grain data is unable to resolve enough information to fully propose a mechanistic explanation of aggressive behavior in ants. To that end future research should look at specific aminergic pathways within the ant brain and their dynamics under the different social context that pavement ant workers experience in their life. Having at length explored the necessary conditions for the engagement of aggressive conflicts between neighboring pavement ant colonies, we look forward to studying the processes that sustain and ultimately conclude this behavior. Although this article explores the interactions of ants along the territorial border, it is not well known why some ants decide to engage in active recruitment of sisters by returning to the nest instead of fighting. This critical positive feedback is a necessary component of the observed escalation in wars starting from dozens of ants to many thousand. Additionally, it has been suggested that in societal conflicts group size performs a similar function as body size does in conflicts between solitary organisms. In this way, recruitment of a colonies workforce is an important indicator of strength, and recruitment rate becomes a reliable indicator of victory in an engagement between colonies. Pavement ant wars last over 10 h but then disengage in under 30 min. The source of this synchronized withdrawal is not well understood. Preliminary data demonstrate that dyads of non-nestmate ants engaged in ritualized fighting experience a marked increase in dopamine both 3 min and 2 h after the start of the fight (Greene MJ, unpublished data). By extending the logic from this article, we propose that the length of fighting behavior is modulated by the rate at which this elevated concentration decreases to basal levels. We suspect that a superthreshold level of dopamine sustains fighting and the drop to subthreshold levels will signal a cessation of hostilities. Finally, although this behavior has been attributed to territory defense there are at this time no studies that empirically study the cost and benefit of this behavior in the terms of colony fitness. Colonies will often return to sites of previous wars within 24 h, and a new war will begin in approximately the same location with no clearly demarked change in territory size. This pattern can last for 3–5 days, and it is not clear what benefit this prolonged dedication of colony resources achieves."
} | 1,938 |
29152587 | PMC5686522 | pmc | 3,895 | {
"abstract": "Understanding microbial interactions is a fundamental objective in microbiology and ecology. The synthetic community system described here can set into motion a range of research to investigate how the diversity of a microbiome and interactions among its members impact its function, where function can be measured as exometabolites. The system allows for community exometabolite profiling to be coupled with genome mining, transcript analysis, and measurements of member productivity and population size. It can also facilitate discovery of natural products that are only produced within microbial consortia. Thus, this synthetic community system has utility to address fundamental questions about a diversity of possible microbial interactions that occur in both natural and engineered ecosystems.",
"introduction": "INTRODUCTION There is modest knowledge about how microorganisms interact with each other in their native habitats and whether these microbial interactions have implications for emergent community or ecosystem properties. Microorganisms can communicate chemically, and these chemical interactions underlay a range of relationships from commensalism to antagonism ( 1 – 3 ). Because of the specificity of many microbe-microbe relationships, it is thought that most microorganisms produce certain chemical products only within a particular consortium ( 1 , 4 – 7 ). However, understanding of relatively well-described microbial interactions often is incomplete. For example, sensitive mass spectrometry was employed to discover new components of an interaction between Bacillus subtilis and Streptomyces coelicolor ( 7 ), which suggested that knowledge of this interaction was limited despite having been studied previously. Investigations of microbial exometabolite production have been predominantly focused on the analysis of a single taxon or pairs ( 8 – 10 ) of microbial taxa rather than on multimember profiling ( 11 ). However, the collective abilities of microbiomes to produce and exploit extracellular enzymes have been hypothesized to be key in discriminating situations in which microbial community structure has implications for ecosystem processes like carbon and nitrogen cycling ( 12 , 13 ). These studies and others suggest that most microbial interactions remain obscure and that improved understanding of some of these interactions likely will provide important insights into microbial community functions. Synthetic microbial systems recently have garnered reinvigorated interest because of their potential to address fundamental unknowns in microbial ecology, engineering, and systems and synthetic biology ( 4 , 7 , 14 ). Synthetic microbial systems are a key approach used in microbial ecology to understand how microbial interactions lead to emergent properties of communities, such as resistance and resilience ( 7 ). For example, synthetic communities have been assembled from marine waters onto artificial particles to observe community primary succession and the resulting functional changes in model heterotrophic particles ( 15 ) and phototrophic biofilms ( 16 ), spatially constrained synthetic communities have been used to investigate reciprocal syntrophy ( 17 ), and computationally modeled synthetic communities have been applied to predict coculture growth given the metabolic needs of the members ( 18 ). Other recent work used a combination of metabolic flux analysis and multimember coculture to determine that the net outcome of complex interactions between an antagonist and a syntroph was not necessarily the sum of all expected pairwise outcomes, especially given particular spatial arrangements of the members ( 19 ). Other synthetic microbial systems are engineered to control and manipulate genetic circuitry toward required functions ( 20 ). These studies and others demonstrate that synthetic microbial communities can be applied in diverse and creative ways to provide insights into the dynamic biological and ecological interactions of microbiomes ( 21 , 22 ), with the anticipation that these insights then can be applied to manage these communities toward desired outcomes. We have developed a simple synthetic community system to interrogate exometabolite interactions among microbial community members. This synthetic community system permits direct investigation of chemical interactions among microorganisms via secondary metabolites, signaling molecules, and other exometabolites and allows for observation of behaviors that only occur when those microorganisms exist as part of a particular consortium. This system combines concepts and tools from systems biology, microbiology, biochemistry, genomics, and ecology and can provide both top-down and bottom-up approaches to investigate key questions in synthetic microbial ecology ( 7 ). Thus, it can advance understanding of microbial interactions within diverse natural and artificial microbiomes. It can also facilitate discovery of novel microbial products that are made given certain community memberships.",
"discussion": "DISCUSSION Application to the fields of systems microbiology and microbial ecology. The synthetic community system introduced here can be applied to address a variety of timely and compelling questions in systems and community microbiology. First, the synthetic system can be used to address fundamental questions about the consequences of community diversity. Membership manipulations can be performed to address the importance of community richness (total number of taxa) and structure (relative contributions of members) on member interactions and emergent community properties. For example, it has been suggested that microbial community structure matters most for function in the production of exometabolites such as enzymes and polysaccharides, which have implications for biogeochemical processes like carbon cycling ( 12 ) and N fixation ( 13 ). The system could be used to interrogate these exometabolites directly. Similarly, the system can be used to investigate temporal changes in member interactions, to determine how member interactions change in response to perturbations, and to experimentally evolve microbial interactions within communities. These and similar lines of inquiry will allow researchers to ask how microbial interactions and products change as the result of controlled and specific environmental cues. Finally, the system can facilitate discovery of natural products. Member genomes can be mined for cryptic metabolic pathways from bioinformatic predictions, and this information could then be coupled to synthetic community manipulations to observe regulation. Novel or unknown exometabolites are likely to be discovered in an untargeted analysis of the community exometabolites, and their chemical structures and activities can be pursued subsequently. Advantages and limitations. This synthetic community system offers several experimental advantages. First, the synthetic system offers an opportunity to interrogate a relatively simple community within well-defined experimental conditions ( 7 ). It allows researchers to focus specifically on community outcomes driven by exometabolites and not by physical contact, as well as the causes and consequences of those outcomes. These interactions can be challenging to target in other mixed batch or bioreactor systems. The system also is versatile because it can be used with microbial consortia from any ecosystem and adjusted to simulate environmental conditions of interest. The system also is scalable, not only for increasing the overall community diversity, but also for moving toward higher-throughput and high-content screens for molecules and community outcomes of interest. Finally, the system is accessible. The filter plates are available to any lab, and manual manipulation of the system without specialized equipment is feasible. We suggest that the most limiting factor is access to mass spectrometers and expertise in mass spectral analysis, which if not available locally is accessible via research support facilities. As is true for any laboratory-scale experimental system, this synthetic community system also has several limitations. First, all possible types of microbial interactions are not observable using this system. Some exceptions include interactions that are contact dependent or in which chemical exchanges and physical contact are not clearly distinguishable or independent. The system instead allows for control of the influence of physical spatial structure on microbial interactions, which has been shown to be important for stabilizing some communities, especially for highly structured environments like biofilms or soil matrices (e.g., 17 , 19 , 31 , 32 ). However, because members are spatially partitioned yet permitted to interact chemically, this system allows the researcher to control for spatial effects without a typical limitation of homogenous coculture: overgrowth of one member that prevents long-term observations of community interactions ( 17 ). Not all relevant microbial exometabolites and community outcomes will be observable in this system. Specifically, exometabolites that have a rapid turnover or that are concentration dependent in ranges outside the system’s physical constraints and imposed experimental conditions will be inaccessible (e.g., total medium volume or experiment duration, respectively). The ability to observe a molecule and its community outcomes also depends on the sensitivity of the microbial sensing/signaling systems involved, which will depend on the members’ capabilities. Also, this system may not be ideal for situations in which the local accumulation of an exometabolite inhibits the reaction generating the exometabolite, but this will depend on the duration of the experiment relative to the expected rate of exometabolite accumulation in the reservoir. Finally, interactions that are reliant on volatiles that may off-gas during plate shaking will not be observable. Only cultivable organisms can be applied easily to the system, and so if the most functionally important or prevalent members of a community are yet uncultivable, their interactions will be difficult to observe with this system. However, cultivation methods are improving, in part because cultivation conditions can be informed by metagenome and (meta)transcriptome data ( 33 ), and there is evidence that growing microbial community members in cohorts from the environment can improve isolate recovery ( 34 ). Thus, this synthetic community system could be used to provide insights into the precise memberships and molecules required to bring new isolates into laboratory culture. A general limitation of any system used to observe exometabolites is that many microbial exometabolites are unknown and difficult to identify. We anticipate that this limitation will be overcome as technology and infrastructure for exometabolite identification advances. Analysis pipelines to integrate exometabolite data with other microbial omics approaches, such as transcripts and metabolic flux analysis, are also in active development ( 35 , 36 ). Therefore, the first experiments using the synthetic community system will face necessary challenges in spearheading analysis and integration approaches. There are general criticisms offered for using model or laboratory-scale systems in microbial ecology, and a common concern is that any model cannot mimic natural conditions and therefore is not biologically relevant. The synthetic community system described here is an artificial, simplified model. However, it is a model that offers many advantages specifically for understanding the chemical feedbacks on community ecology driven by microbial interactions, which is a key goal of synthetic microbial ecology ( 7 ). These interactions have the potential to occur in nature, especially when thoughtful experimental designs are employed to (i) include organisms that are naturally cooccurring or have evidence of interactions and (ii) manipulate the pertinent primary drivers of natural ecosystems. Furthermore, important advances in ecology and evolution have been made using model systems ( 37 – 40 ). Microbial synthetic systems especially have offered insights because of their malleable communities and molecular tools for understanding population dynamics ( 4 , 41 – 43 ). Thus, researchers continue to use model systems because they can inform as to both biological potential and constraints in nature. When complemented with careful studies in situ , the synthetic community system described here can serve to discover and interrogate microbial interactions, the signatures of which may otherwise be unobservable within the complexity of natural systems."
} | 3,181 |
38507447 | PMC10990125 | pmc | 3,899 | {
"abstract": "Significance There is great interest in investigating the genetic basis for adaptation in microbes, yet few studies reveal both the genes and evolutionary dynamics that allow microbes to adapt to natural environmental variation. We identify genes associated with the ability to tolerate stressful soil conditions in wild symbiotic bacteria and demonstrate that these genes drive replicated patterns of adaptation across the landscape. Phylogenetic evidence indicates that these adaptive genes are transferred among otherwise distinct lineages and drive parallel molecular solutions to stress among populations across large spatial scales. These findings reveal molecular processes of adaptation in wild microbes across the landscape and are widely applicable to efforts to understand the evolutionary origins of microbial diversity.",
"conclusion": "Conclusions Remarkably little is known about the impacts of horizontally transmitted loci on evolutionary dynamics and outcomes of adaptation in wild microbial communities across the landscape. This work elucidates the molecular genetics of environmental adaptation to stress in wild soil bacteria across variable selection in natural environments. We observe a repeated pattern along the western United States whereby wild Mesorhizobium bacteria adapt to the presence of nickel in their native soil conditions by acquiring nre, an operon that increases fitness in the presence of nickel. This key adaptive locus is harbored by diverse Mesorhizobium clades and resides in putative MGEs exchanged horizontally among lineages. Our results highlight the importance of HGT gene modules and GIs to microbial adaptation to spatially varying selection, analogous to inversions in organisms like plants, in that they are transferred by recombination but protected from being broken up by their genomic architecture.",
"discussion": "Results and Discussion Genetic Basis for Microbial Adaptation to Serpentine Soils. Rhizobial adaptation to spatially variable heavy-metal enrichment. We sampled sites to span a transect through most of the latitudinal extent of the geographic range of A. wrangelianus and A. brachycarpus , which occur from Northern Mexico to Southern Oregon, USA ( SI Appendix, Table S1 ). Variation in soil chemistry among these sites is largely driven by two classic bioindicators of serpentine soil, high Ni and low calcium (Ca) ( 17 , 19 ). We measured soil chemistry, including Ni, macro-, and micronutrients, at 55 grassland sites hosting populations of A. wrangelianus and/or A. brachycarpus, that were in 18 reserves located in California and Oregon, USA. Principal component analysis revealed two major axes of variation, PC1 and PC2, which explain 25% and 23% of soil chemistry variation, respectively ( Fig. 1 A ). The top loadings for PC2 were calcium (−0.50) and Ni (0.39) and for PC1 are soil organic matter (0.43) and potassium (0.40). Soil Ni plays a key role in driving adaptation across soil variation in many taxa ( 17 , 18 ), so to investigate adaptation to local soil conditions, we focused on this major axis of natural environmental variation. Fig. 1. Mesorhizobium from heavy-metal-enriched serpentine soil adapts to nickel (Ni) and shows a polyphyletic distribution of Ni tolerance candidate genes in a GWAS. ( A ) Ni enrichment reflects an important axis of soil niche space. Principal components analysis of 11 soil chemistry parameters across 55 sites inhabited by host legumes reveals that Ni (ppm) is a top loading for PC2. ( B ) The growth of Mesorhizobium strains from serpentine soils (n = 204) is insensitive to Ni; however, strains from non-serpentine soils (n = 92) grow slowly in the presence of Ni ( Dataset S1 ). Shown is an interaction plot of the estimated marginal mean growth at 72 h ±SE based on a general linear model. SE; standard error. ( C ) Serpentine soils (serp) contain more Ni than non-serpentine soils (non-serp). Soils with a Ca:Mg < 1 are classified as serpentine. Bars indicate means ±1 SEM. Points indicate individual soil samples (n = 114). ( D ) A phylogeny based on amino acid sequences of 1,542 single copy core genes built with RAxML. The presence (colored tile) or absence (gray tile) of genes in each strain associated with Ni tolerance (MIC for Ni and/or growth in nickel media). Candidate genes are grouped by whether they were significant in both Clades, Clade 1 only, or Clade 2 only (FDR-adjusted P -value < 0.05). Predicted function of candidate genes is indicated by tile color. A strain’s MIC is indicated by the shade of a green tile. Clades 1 to 13 are designated by background colors in the phylogeny (scale bar indicates substitutions per site). Branches with bootstrap support values of less than 70 are collapsed to polytomies. For strain identities and additional strain information, see SI Appendix, Fig. S3 . Mesorhizobium from replicated sites within serpentine soil adapts to local Ni concentrations. We isolated rhizobia from 669 field-collected nodules from both Acmispon species from serpentine and non-serpentine soils across our 55 sites. We also included 46 Mesorhizobium strains from a similar study ( 27 ). We then measured each strain’s tolerance to Ni via the minimum inhibitory concentration (MIC) of Ni grown on tryptone yeast agar supplemented with 0 to 5 mM Ni. For strains selected for genomic sequencing (see below), we also measured growth in Ni media as the OD 600 after 72 h in tryptone yeast broth supplemented with 1 mM NiCl 2 . Strains from soil with higher levels of Ni (and higher PC2 values) show higher Ni tolerance ( Fig. 1 A and B and SI Appendix , Fig. S1 A – C and Table S2 ). This repeated pattern of adaptation was indistinguishable for strains isolated from different natural reserves and host species ( SI Appendix, Table S2 ). These environmentally adapted rhizobia comprise multiple clades of Mesorhizobium. We characterized the pangenome of the collected Mesorhizobium strains using PacBio (n = 16) and Illumina (n = 286) sequencing to provide complete (n = 3) or draft genomes (n = 299) for 302 strains spanning the geographical breadth of the collection ( SI Appendix , Table S3 and Dataset S1 ). The pangenome was obtained through Markov clustering (MCL) ( 36 ) of protein sequences into orthologous clusters using Anvi’o ( 37 ). This Mesorhizobium pangenome comprises a core genome that includes 2,423 genes (orthologous clusters) conserved across the strains and a flexible genome that includes 55,613 genes ( SI Appendix, Fig. S2 A and B and Table S4 ). A phylogeny built using a concatenated alignment of single copy core genes ( Fig. 1 D and SI Appendix , Fig. S3 ) and pairwise average nucleotide identity (ANI) distances ( 38 , 39 ) of marker genes at a 95% threshold indicate that the majority of these genomes comprise two major phylogenetically distinct clades, Clade 1 (n = 89) and Clade 2 (n = 144), with the remaining strains comprising 11 smaller clades ( Fig. 1 D and SI Appendix, Fig. S4 A and Table S3 ). This clade structuring was recaptured in an additional phylogeny based on marker genes universally conserved in bacteria ( 40 – 42 ) ( SI Appendix , Fig. S5 A and B ). Within our Mesorhizobium marker genes, 15.5% of the variation in amino acids is attributed to variation among reserves (PERMANOVA, P < 0.01), while host species and serpentine or non-serpentine soil type accounted for 4.9% and 2.3% ( P < 0.01) ( SI Appendix, Table S5 ), respectively. Therefore, we observe some regional differentiation among Mesorhizobium, but weak differentiation across soil types. However, dispersion, the variation within a group, differs among reserves ( df = 14, F = 7.652, P < 0.01), and host species ( df = 1, F = 29.497, P < 0.01), but not between soil types ( 40 ). Thus, differences in dispersion within these groups may contribute to measures of Mesorhizobium differentiation among reserves and hosts ( 41 ). To identify genetic variants that underlie environmental adaptation, we sought locally abundant genetic variants associated with high fitness under conditions that reflect an important axis of local niche space ( 5 ). We used a genome-wide association study (GWAS) approach to associate genetic variants to Ni tolerance, an adaptation to an important axis of environmental variation across the landscape. We tested both patterns of gene presence/absence in the flexible genome and single nucleotide polymorphisms (SNPs) as sources of genetic variation. Since different Mesorhizobium clades could differ in which genetic variants are correlated with Ni tolerance, GWAS was performed separately for Clade 1 and Clade 2. To account for a clonal background among strains, which can reduce power due to large amounts of shared DNA ( 42 ), we implemented a linear mixed model (FaST-LMM) with phylogenetic population structure correction through pyseer ( 43 – 45 ) using the marker gene phylogeny. To add additional support, we repeated the pyseer GWAS analyses using first a FaST-LMM with population structure correction based on the single-copy core gene phylogeny and then again under a fixed-effects model with population structure correction using Mash-computed whole-genome sequence distances ( 46 ). These three methods for accounting for population structure yielded largely overlapping results ( Dataset S2 ), and below, we discuss the candidate loci that were significant in all three analyses. Our findings are consistent with a scenario in which the gain or loss of genes contributes to Ni adaptation in Mesorhizobium . After performing GWAS through pyseer and applying multiple test corrections, we found that the presence of 32 and 30 genes in Clades 1 and 2, respectively, were associated with Ni MIC, growth in Ni media, or both at the 5% significance level ( Dataset S2 ). Twenty-one of these genes associate with Ni tolerance for both Clade 1 and Clade 2, consistent with a scenario in which genetic strategies for Ni tolerance are shared across divergent Mesorhizobium clades ( Fig. 1 D ). Patterns of presence/absence of these Ni tolerance candidate loci in other clades showed these genes are not limited to Clades 1 and 2 ( Fig. 1 D ). Our findings are also consistent with a model in which allelic variation in the core genome does not contribute substantially to Ni adaptation. After testing 197,600 and 38,665 SNPs for Clade 1 and Clade 2, respectively, and applying Benjamin–Hochberg FDR-correction for multiple testing, there were no SNPs associated with either Ni tolerance phenotype at the 5% significance level. Our findings suggest SNP alleles of large effect that segregate at intermediate frequencies in core genes are unlikely to underlie Ni adaptation. However, this association genetics approach could fail to detect small effect or rare alleles ( 47 ). Many of the genes whose presence is associated with Ni tolerance phenotypes reside in close physical proximity in the genome. One strain in which the genetic proximity of these GWAS candidates is especially evident is C089B, in which 30 candidate genes are closely associated along the chromosome ( SI Appendix , Fig. S6 C and D ) in a region bearing similarities to a MGE ( Fig. 2 A ). Further inspection in other genomes shows a similar pattern ( Fig. 1 D ), indicating the possibility of a small genomic island (GI) bearing P4 integration motifs, with multiple groups of genes contributing to Ni tolerance. To determine whether genetic variants associated with a phenotype via GWAS that are co-localized contribute to the phenotype or are spuriously associated via physical linkage, we tested for a causal link to the phenotype for each locus ( 48 ) and used experimental genetic evidence to validate whether we observe an impact of genotype on the Ni adaptation phenotype. Fig. 2. The candidate gene island contains genes necessary and sufficient for Mesorhizobium nickel tolerance. ( A ) A Mesorhizobium GI from strain C089B enriched in Ni tolerance candidate genes from GWAS. Brackets group genes into 20 probable TUs for plasmid-based testing. ( B ) Spot dilution testing of A. fabrum clones expressing Mesorhizobium candidate Ni tolerance genes. A plasmid for expressing TU8 was not successfully constructed. ( C ) Mesorhizobium genes residing in TU12 ( nreX , nreY , and dedA ) were individually introduced into A. fabrum on an expression plasmid and the resulting strains were spotted onto medium with 0 and 2 mM NiCl 2 . ( D ) Deletion of the two-gene nreXY region was accomplished in the Mesorhizobium C089B background and the resulting strain (compared to nreXY + ) was tested for Ni tolerance by spot dilution test. ( E ) Growth of nreXY + C089B and the isogenic Δ nreXY mutant in the presence and absence of Ni. Shown is an interaction plot of the estimated marginal mean growth at 48 h ±SE from a general linear model. SE; standard error. Functional validation of the nre operon for Ni tolerance. To identify genes that contribute to Ni tolerance in the genomic region enriched for candidate loci, we conducted a functional analysis of individual genes and multi-gene transcription units (TUs) within this region, using C089B as a reference strain. Twenty regions (excluding some genes for DNA processing and transcriptional regulation) were defined within the C089B Ni tolerance gene cluster ( Fig. 2 A ). These TUs were ligated into a broad host-range plasmid for expression in the fast-growing Rhizobiaceae strain Agrobacterium fabrum UBAPF2, which is sensitive to Ni. One transcription unit (#8) could not be cloned despite several attempts. Of the remaining 19 clones, only one (#12) conferred robust Ni tolerance to the A. fabrum test strain ( Fig. 2 B and SI Appendix , Fig. S7 ). This clone consisted of four putatively co-transcribed genes, designated nreA , nreX , nreY , and dedA . The protein encoded by nreA belongs to the CsoR family of transcriptional regulators and shares similarity to the nreA gene of Cupriavidus metallidurans 31A, which contributes to Ni tolerance in that organism ( 49 ). The proteins encoded by nreX and nreY are predicted to be proton:cation antiporters of the cation diffusion facilitator and major facilitator superfamily families, respectively. The dedA gene encodes a putative transmembrane transporter protein associated with resistance to various metal and organic compounds, either indirectly through its effects on bacterial surface chemistry, or directly via efflux of the chemical stressor ( 50 , 51 ). Ni tolerance gene clusters in Clades 1 and 2 vary in nreAX ( Y ) -dedA gene organization. Of 246 occurrences of the nre gene cluster in 208 strains (some strains have more than one copy), 65% of loci had the nreAXYdedA organization, 19% had an nreAXdedA organization, 9% had nreAX occurring without nreY or dedA , 5% had an organization with nreAXY separated from dedA by 3 to 4 hypothetical genes, and 2% have undeterminable organization due to contig breaks. The nreX , nreY , and dedA genes were tested individually for their ability to confer Ni tolerance in A. fabrum ( Fig. 2 C ). A. fabrum expressing nreX or nreY supported strain growth at 2 mM NiCl 2 , a Ni concentration that did not allow growth of the vector-only control strain, while A. fabrum expressing dedA did not exhibit Ni tolerance above background. The nreX and nreY genes are therefore each sufficient to confer Ni tolerance. Having demonstrated the sufficiency of nreX or nreY to confer Ni tolerance in the heterologous system, we investigated how crucial this gene pair is for Ni tolerance in the Mesorhizobium strain C089B, which is Ni tolerant. We created a precise deletion of the nreX - nreY gene pair in the C089B strain background using an allele exchange strategy. This deletion mutant lost considerable tolerance to Ni compared to the wild-type parent strain background ( Fig. 2 D and E ), highlighting the role of nre as necessary for high Ni tolerance in this serpentine soil-derived strain. Genes in the nre operon are thus primary determinants of C089B nickel tolerance, though additional untested loci could contribute to the phenotype as well. Evolutionary History of the nre Gene Cluster. Origins of nre . To investigate the origins of adaptive loci, we examined whether lineages tend to inherit adaptive loci vertically or acquire them from distantly related lineages. We find that many Ni tolerance candidate genes are shared across highly divergent Mesorhizobium clades ( Fig. 1 D ), consistent with globally shared sweeps of adaptive loci, rather than lineage-specific innovation followed by clonal sweeps in response to environmental Ni. The phylogeny based on the nreAX locus is incongruent with the Mesorhizobium core and marker gene phylogenies, and strains that share high sequence similarity, genome-wide, differ in the presence/absence of the nreAXY locus ( SI Appendix, Fig. S8 A and B ). This incongruence, combined with the fitness advantage nreAXY confers in the presence of Ni, is consistent with a scenario in which diverse Mesorhizobium lineages acquire this niche-specific locus where it is locally advantageous via HGT-driven recombination that has freed nreAXY from tight linkage with the core genome ( 6 , 10 – 13 ). To explore the evolutionary history of the nre operon (and to test whether certain forms of the nreAXY genes associate with higher Ni tolerance, below), we clustered nreA , nreX , and nreY protein-coding sequences with a sequence similarity approach (referred to as “alleles” hereafter). Comparisons of the distributions of these alleles with respect to the phylogeny showed that although some alleles have a higher frequency in one clade or the other, there is imperfect sorting of alleles across clades ( Fig. 3 A ). This is consistent with HGT of adaptive alleles primarily among close relatives, but also occasionally across deeply diverged Mesorhizobium clades. Fig. 3. nre Ni efflux genes reside within a variety of putative GIs and display allelic diversity. ( A ) Single copy core gene phylogeny of Mesorhizobium isolates with nre allele plotted adjacent showing a polyphyletic distribution of nreA , nreX , and nreY alleles. Clades 1 and 2 are highlighted. Allele counts per clade are displayed in the key (presence/absence). ( B – D ) Estimated marginal mean MIC ±SE of strains with nre ( B : nreA , C : nreX , and D : nreY ) allele number designations. Allele 0 refers to the absence of that gene. Alleles 7 and 8 of nreY were removed because of low abundance. ( E ) Diagram comparing GI-like regions containing Ni tolerance candidates from different genomes. Regions of similarity as computed by the Mauve aligner ( 52 ) connected by gray. Ni tolerance candidates and putative HGT-related genes are color coded and labeled with predicted function. Segments cut due to length in islands of C432A and M1142 are symbolized with diagonal lines. ( F ) Natural allelic variants of the nreAXY operon inserted into A. fabrum confer different growth in the presence of nickel. Bars represent mean growth ±SE after 24 h for low tolerance, reference strain, and high tolerance alleles. Inside brackets are the strain ID and clade the alleles were derived from. Letters above bars indicate significant differences at P < 0.05. SE; standard error. The nre gene cluster often resides in a putative transmissible element. The mechanisms by which adaptive loci transfer between cells and integrate into the genome are predicted to shape biogeographic and phylogenetic distributions of the adaptations they encode ( 5 ). We find that candidate genes for Mesorhizobium Ni tolerance tend to be physically clustered in various putative GIs, such as putative MGEs, segments of DNA that carry genes for their own excision and transfer ( 53 , 54 ). For example, the nre operon, and linked candidate genes, reside within an inferred mobile element region in all eight of the complete or nearly complete PacBio-sequenced Mesorhizobium genomes that contain the nre operon ( SI Appendix , Table S7 ). This physical organization is recapitulated in eight draft genomes with contigs large enough to enable us to detect MGE-related genes in regions neighboring the nre operon and linked candidate genes ( SI Appendix , Table S7 ). Introduction of DNA from a MGE can result in multiple gene modules that provide novel functions and potentially contribute to adaptation to a specific niche ( 5 , 53 , 55 ). In C089B, the putative mobile element containing the nre operon and candidate Ni tolerance genes resides 2.4 Mb from the symbiosis island ( SI Appendix, Fig. S9 ) and is inserted at the 3′ end of a tRNA-Met gene with a 14-bp direct repeat 54 kbp downstream. This region contains an open reading frame (ORF) coding for a serine recombinase as well as ORFs with similarity to the traACD mobilization locus of the integrative mobile element IME Ml R88B of Mesorhizobium sp. R88B ( 56 , 57 ). ( Fig. 3 E ). We consider similar regions in other Mesorhizobium that share attributes such as insertion at a tRNA-met and att sites, and bear multiple Ni tolerance candidate genes, to be “C089B-type Ni islands.” Out of the 29 strains with this organization that we examined, 27 belonged to Clade 1, one to Clade 2, and one to Clade 10. Among intact C089B-type Ni islands (i.e., those that lack contig breaks), island size ranges widely, from 53,319 bp in strain M0626 to 172,320 bp in strain M0025. While the Ni island of C089B does not contain genes with predicted conjugative functions, C089B-type Ni islands in several other genomes contained putative conjugative type IV secretion systems which could function to transfer the Ni island among lineages ( 58 , 59 ). HGT of these putative GIs thus appears to rely on mobile element features and insertion sites that are broadly conserved among Mesorhizobium strains and present a mechanism by which an environmentally adaptive locus is distributed across phylogenetically distant and geographically widespread strains. Across Mesorhizobium strains, nre loci occur in various putative mobile genomic contexts. Clusters of Ni tolerance genes localize in the same genomic region, but with variable gene order across divergent strains, even when a distinct GI-like region is not identifiable ( Fig. 3 E ). The genomic context of nreAXYdedA and other Ni-associated gene clusters is not limited to C089B-type Ni islands, but also includes putative GIs with alternative insertion sites and features consistent with integrative and conjugative elements ( 60 ) ( Fig. 3 E and SI Appendix, Table S7 ). The variable co-occurrence and mosaic structure of Ni candidate gene modules in different MGE-like systems could result from recombination between MGEs, which can occur in Mesorhizobium ( 56 ). Additionally, insertion sequence (IS) transposases were sometimes found located near Ni candidate genes and IS-mediated transposition can result in gain and loss of genes ( 61 ). Thus, HGT of nre, which confers fitness benefits in Ni-enriched habitats, could drive gene transfer of variable numbers and identities of other loci in diverse mobile contexts and contribute to genomic diversity. Of the 41 genes associated with Ni tolerance, only nreX and nreY contribute to functional Ni tolerance in vitro. The fitness impacts of the loci co-transferred as components of putative Ni GIs remain unknown. On one hand, these genes could provide little to no fitness benefit ( 5 , 62 ). Their absence from some putative Ni GIs across different strain backgrounds could result from genome streamlining whereby selection has purged lineages with extra neutral DNA or a process by which shorter mobile elements have had a greater probability of integrating into the chromosome ( 5 , 63 , 64 ). Alternatively, co-transferred genes may confer tolerance to other stressors in serpentine soil. Beyond Ni enrichment, serpentine soil can impose stress due to excess Cu and Mg, dry conditions, and low levels of macronutrients ( 17 , 19 ), The fact that gene clusters with potential roles in the transport of other metals occur within some Ni islands is consistent with the latter possibility ( SI Appendix , Table S8 ). We note that while metal tolerance systems in bacteria sometimes confer tolerance to multiple metals ( 65 , 66 ), Ni tolerance is uncorrelated with Co or Cr tolerance across 94 haphazardly selected strains we examined ( SI Appendix , Fig. S10 A and B ). Often the serpentine soils we examined contained low levels of Co or Cr, so selection for metal cross-tolerance may be weak in our system ( SI Appendix , Table S9 ). Allelic variation in nre genes may contribute to environmental adaptation. The nre alleles we identified associate with different levels of Ni tolerance. Based on patterns of amino acid divergence at each gene, we delineated 6 nreA alleles, 3 nreX alleles, and 8 nreY alleles. While some alleles were more common in one Mesorhizobium clade over another, other alleles occurred at similar frequencies in both Clade 1 and Clade 2 strains ( Fig. 3 A ). General linear mixed models (GLMMs) indicate that the allelic identity of nreA (X 2 = 132.69, P < 0.001), nreX (X 2 = 114.56, P < 0.001), and nreY (X 2 = 100.78, P < 0.001) predicts the Ni minimal inhibitory concentration (MIC) of strains, which was confirmed with estimated marginal means post hoc analysis ( Fig. 3 B – D ). Naturally variant nreAXY alleles that were expressed in A. fabrum conferred distinct degrees of Ni tolerance to elevated levels of nickel. We used nreAXY alleles that were predicted in the model above to confer high Ni tolerance or low Ni tolerance, from clade 1 and 2 strains ( Fig. 3 A – C ) as well as reference strain C089B, which we find to have intermediate Ni tolerance. While nreAXY alleles confer growth indistinguishable from that of the vector-only control, in the absence of Ni, all nreAXY alleles confer Ni resistance, compared to the vector-only control, in 1.5 mM NiCl 2 (F 5,66 = 27, P < 0.001; Fig. 3 F ). However, at higher Ni levels, the predicted high tolerance alleles confer greater tolerance to Ni than the predicted low tolerance alleles (2 mM nickel: F 5,66 = 1044.7, P < 0.001; 2.5 mM nickel: F 5,66 = 1342.8, P < 0.001; 3 mM nickel: F 5,66 = 14.28, P < 0.001; Fig. 3 F ). Genetic variation conferred by different alleles occurs across both clades 1 and 2. These findings are consistent with a scenario in which nre alleles that confer a level of Ni tolerance just sufficient to withstand local levels of Ni enrichment could accumulate in a population. Thus, allelic variation at the adaptive nre locus contributes genetic variation that has the potential to fine-tune environmental adaptation driven by the gain or loss of the Ni GI. nre across the Landscape. Replicated clines in nre . Our findings document large-scale observations of repeated clinal variation in the frequency of an adaptive genetic variant in soil bacteria. Across the western United States, Mesorhizobium strains from sites with higher levels of Ni enrichment are more likely to contain nreAX or nreY (X 2 (1) = 48.2, P < 0.0001; X 2 (1) = 36.8, P < 0.0001; respectively; SI Appendix, Fig. S11 A and B and Tables S10 and S11 ). Even at fine spatial scales, including those separated by only hundreds of meters, frequencies of strains possessing nre were dramatically elevated on serpentine soils, often up to 100%, while strains from adjacent sites with non-serpentine soils rarely possessed nre ( Fig. 4 and SI Appendix , Table S1 ). In fact, for every 1 ppm increase in soil Ni, the log-odds of an isolate bearing nreAX increased by 1.15, and the log-odds of an isolate bearing nreY increased by 1.11. Similarly, soil PC2, a major axis of soil chemistry variation heavily influenced by Ni enrichment, also had a positive relationship with the presence of nreAX [X 2 (1) = 21.3, P < 0.0001] and nreY [X 2 (1) = 21.4, P < 0.0001]. Our findings are consistent for symbiont populations collected from both Acmispon host species ( SI Appendix, Tables S10 and S11 ). Thus, despite the potential for high rates of soil microbial migration via transport in wind, water flow, or on mobile animals ( 33 ), spatial variation in natural selection due to soil chemistry appeared to be sufficient to maintain the repeated evolution of environmental adaptation to Ni via clines in the frequency of the nre locus. Fig. 4. Across natural Mesorhizobium populations in Oregon and California, USA, nreAXY presence is higher at sites with higher soil Ni concentrations. Pies indicate sites within a reserve. Reserve name is labeled by each cluster of sites. Colors within the pie indicate the proportion of sequenced strains with the nre locus (defined as genome containing nreAX and/or nreAXY locus) present (dark green) or absent (light green). A total of 56 sites with strains sequenced for this study are shown: 50 from the from this study (2019 sampling); and 6 from a similar study from 2009 ( 27 ). The outer circle around each pie indicates the level of Ni enrichment in soil at a site (darker green indicates higher Ni). Serpentine sites (defined as soil Ca:Mg < 1; Right ) and non-serpentine soil ( Left ) are shown clustered by soil type. SI Appendix , Table S1 gives the number of strains sequenced per site and reserve. Conditional neutrality of nre . While adaptive loci that are transmitted horizontally can be key to persistence across environments, their fitness effects are rarely measured across environmental conditions ( 5 ). To understand the evolutionary dynamics that cause nre to segregate at high frequency (90%) in Ni enriched serpentine soil, but low frequency (23%) in low Ni non-serpentine soil, we tested whether nre imposes a detectable cost in the absence of Ni. Under antagonistic pleiotropy, a variant beneficial in one environment can be costly elsewhere and this genetic trade-off maintains the spatial distribution of adaptive variants. However, trade-offs are not ubiquitous. Under conditional neutrality, variants beneficial in one environment have no impact on fitness elsewhere, and here, barriers to gene flow can maintain the non-random spatial distribution of adaptive variants ( 67 ). We found that while the C089B Ni tolerant streptomycin-resistant strain, which bears the nre locus, grows more rapidly in the presence of Ni than does the isogenic nreXY deletion strain (F 3,60 = 51.3, P < 0.001), these strains show indistinguishable growth at 48 h in TYC media supplemented with streptomycin in the absence of Ni ( Fig. 2 E ). Thus, while the nre locus increases fitness in the presence of Ni, it has no detectable cost in the absence of Ni under our experimental conditions, consistent with a conditionally neutral adaptive variant. In fact, across the set of 296 Mesorhizobium strains we tested, Ni tolerance and the presence/absence of nre do not predict a strain’s growth rate in media that lacks Ni enrichment ( SI Appendix , Fig. S12 A – L and Table S12 ), in contrast to findings from a smaller previous study ( 27 ). Thus, our present findings are consistent with growing evidence that horizontally transmitted adaptive loci can segregate across habitat types despite imposing no detectable costs ( 68 ). However, future experiments could test for other pleiotropic costs of metal efflux proteins like nre that might be missed in in vitro assays, such as a reduction of survival under natural biotic and abiotic conditions, due to factors such as increased susceptibility to biotic antagonists present in non-serpentine soil ( 29 , 69 ). The lack of a detected fitness trade-off at the single locus level ( 70 ) is consistent with findings from other systems in which low expression of gene modules in the absence of an environmental cue results in very weak costs for bearing inducible adaptations ( 29 ). Given the absence of detectable costs to bearing nre, it is possible that its low frequency in strains at sites with low levels of Ni could be explained by genomic streamlining, whereby genetic variants that are neutral in a given environment tend to be lost due to random genetic drift leading to the reduction of excess DNA ( 71 , 72 ). The variable gene content in the putative GI in which nre resides may arise from HGT combined with “soft” selective sweeps, which preserve genomic diversity surrounding a sweeping locus that confers an environment-specific selective advantage ( 10 , 73 ). Neutral variants, such as nre in a low-Ni context, can persist in populations long enough to recombine or undergo HGT into different genomic backgrounds. Despite the lack of benefit nre confers in the absence of Ni, nre operons present in strains from low-Ni non-serpentine soil still appear to confer functional Ni tolerance to the strains that bear them ( SI Appendix , Fig. S13 ; P < 0.001; Wilcoxon two-sided test). If these lineages subsequently migrate onto high-Ni serpentine soil, the selective advantage nre would confer could cause them to rise to high frequency ( 10 ). This could cause different Ni island variants to rise to high frequency as strains bearing different nre alleles migrate onto Ni-enriched patches of serpentine soil, a process akin to adaptation via standing genetic variation in organisms like plants where HGT is uncommon."
} | 8,404 |
34960831 | PMC8705576 | pmc | 3,900 | {
"abstract": "The Conducting of polymers belongs to the class of polymers exhibiting excellence in electrical performances because of their intrinsic delocalized π- electrons and their tunability ranges from semi-conductive to metallic conductive regime. Conducting polymers and their composites serve greater functionality in the application of strain and pressure sensors, especially in yielding a better figure of merits, such as improved sensitivity, sensing range, durability, and mechanical robustness. The electrospinning process allows the formation of micro to nano-dimensional fibers with solution-processing attributes and offers an exciting aspect ratio by forming ultra-long fibrous structures. This review comprehensively covers the fundamentals of conducting polymers, sensor fabrication, working modes, and recent trends in achieving the sensitivity, wide-sensing range, reduced hysteresis, and durability of thin film, porous, and nanofibrous sensors. Furthermore, nanofiber and textile-based sensory device importance and its growth towards futuristic wearable electronics in a technological era was systematically reviewed to overcome the existing challenges.",
"conclusion": "12. Conclusions and Future Perspectives From the above progress, it is perceptible that strain and pressure sensing studies have significantly grown, which has led to obvious developments in terms of e-skin, soft-robots, actuators, synaptic devices, and smart textiles. Self-powered energy-generating systems capable of harvesting biomechanical motions and other environmental stimuli including noise, wind, water, etc. Wearable electronics have captivated futuristic plots of research, and it is anticipated to achieve scalability and industrialization. Wearable sensors alone cannot fulfill the facile durable monitoring, wearable displays and LEDs can drive the futuristic view of integrating self-powered wearable sensors as point-of-care diagnostics to trace finer health details. Recently, following the green chemistry aspects, one pot synthetic conjugated block copolymer based touch a responsive LED working intelligently in achieving the wearable electronics [ 247 ]. Novel biomass-based smart orthopedics have been working intelligently with the concept of low melting polyester materials [ 248 ]. Antibacterial, breathable, and the wound healing nature of the e-skin devices can possibly grant the foundation for better wearable and disposable/degradable electronics. Smart, wearable technology has also been significant in monitoring human health such as body temperature. In compliance, Kuo et al. designed colorimetric sensors [ 1 ], and stretchable thermochromic heaters capable of producing instant visible color transformation in response to temperature [ 249 ]. Very recently, underwater self-healable electronics and its demonstration with perovskite optoelectronic device fabrication has evidenced the advancement of next generation wearable electronics [ 250 ]. Although several strategies worked satisfactorily in achieving the better figure of merit with strain and pressure sensory devices, there are still other plausible routes to explore and accomplish in near future. We have herein presented various fields of interest, demonstrating the reliability of thin film and porous and nanofibrous sensory systems. Biocompatible, eco-friendly elastomers and conducting networks that can sustain higher strains and fatigue can empower the field of strain and pressure sensors. Recently, wearable devices with good elasticity, recovery ratio, toughness, stability, hysteresis free, self-healable, adhesiveness, and breathability factors have been achieved with better balances. It is of greater importance to balance the mechanical, electrical, and ambient condition’s stability of the wearable devices to reduce the probability of mechanical and electrical failure. Fabrication methodology and simple architectures can drastically reduce the fabrication cost and time factors. By integrating greener composites, biodegradable and self-healable characters can significantly contribute to the development of next-generation sensory devices. Conducting polymer composites have a promising potential for use as stretchable and wearable sensors, field effect transistors, memory devices, LEDs, and human interactive devices to strengthen and develop the internet of things’ technological era.",
"introduction": "1. Introduction Recent trends have evolved to numerous stimuli-based electronic appliances, especially in terms of understanding human biological and physiological fitness. Numerous credible establishments have achieved this via a spectrum of factors that include materials, design, fabrication, processing, and stability conditions. An emerging demonstration of efficient high resolution sensors in recent years has caused significant progress in elucidating the chemical, biochemical, biological, toxic, organic effluents, inorganic metals, strain, and pressure signals through chemical bonding interactions, aggregation-induced, optical, colorimetric, and structural color responses [ 1 , 2 , 3 , 4 ]. Among which, strain and pressure sensors are highly desirable for assessing human health status, object detection, heartbeat, muscle, body motion, respiratory detection, and infant–elderly comfort monitors [ 5 , 6 , 7 , 8 ]. Although the evolution of strain and pressure sensors has been rapid, the increasing progression towards the achievement of 5S, i.e., sensitivity, selectivity, size ability, stability, and scalability is evident. The upcoming surge in elevating the eco-friendliness and bio-compatibility remains as another crucial task for the researchers to benefit the human society. Material choice ranging from rigid, flexible, and stretchable polymers, along with conductive nanostructures, including nanoparticles, nanowires, nanorods, nanoflakes, nanosheets, and nanofibers, pave the way to upgrades in the sensors’ real-time suitability [ 9 , 10 , 11 ]. Polymer composites not only limit their arms to sensors and energy generators, but they also govern breakthroughs in fields such as light-emitting diodes (LEDs), lasers, solar cells, field effect transistors, memory devices, and soft robots [ 12 , 13 , 14 ]. For instance, polymeric composite light emissive layers empower the efficiency and stability of the LEDs [ 15 , 16 ]. Recently, polymeric interlayers proved to be effective in brightening LED device performance. Kuo et al. designed the stretchable perovskite LEDs composed of tri-composite perovskite polymeric emissive layers that can retain its luminance even under strained conditions [ 17 ]. Following this, polymeric interface-assisted grain control process sorts out the efficiency and the air and humidity’s environmental stability issue synchronously to obtain the bright, emissive layers [ 18 ]. The performance of solar cells has flourished significantly, achieving better stability and power conversion efficiency, as it curtails defect states and promotes grain growth [ 19 , 20 ]. Recent reviews pose a better insight on how such polymeric composites have brought enlightenment in the field of light emitters and solar cells [ 21 , 22 ]. Persistent efforts and attempts dedicated to developing highly efficient optoelectronic devices and several breakthroughs have been convincingly made to leverage the polymer composites [ 23 , 24 ]. Apart from optoelectronics, polymer composites and their influence in generating gamut of strain and pressure sensors is appreciable. Energy generators play a critical role in enhancing the sensors’ utilities, as they reduce the power consumption, i.e., self-powered complexities, in coupling the energy source or other energy storage components [ 25 ]. Both in terms of sensing and energy generators, it is inexorable to neglect the role of nanofibers and the functionality inclusion into smart textiles, experiencing explosive outgrowth in wearable electronics [ 26 , 27 , 28 ]. This review highlights the importance of polymer composites and nanofibers in the fabrication of strain and pressure sensors. We herein review the challenging aspects of designing highly responsive sensors through facile synthetic strategies, sensor fabrication, mechanistic aspects, various sensing modes, and their operational ranges. Recently, reviews on conjugated copolymers and nanofibers in the field of sensing various factors including metal ions, pH, temperature, and humidity imparts knowledge on colorimetric and fluorometric optical visible responses [ 29 , 30 ]. Conducting polymers playing a credulous role in the design of biosensors and recent reviews highlights their significance in the preparation and properties, of bioapplications [ 31 , 32 ]. This review details the classification of strain and pressure sensor fabrication and its importance in overcoming existing sensory challenges such as sensitivity, operating range, durability, response time, stability, and their adaptability to sensing environments. The basic mechanism for operating the strain sensors falls within the fabrication of conductive networks and their regulated conductive responses with respect to applied strains [ 33 , 34 ]. Several conducting polymer composites, conductive hybrid networks, functionalized composites, co-polymeric systems, 2D nanostructures, structural modifications, and nanofibrous architectural strategies are moving progressively towards the attainment of crucial characteristics in strain and pressure sensors [ 35 , 36 , 37 , 38 , 39 ]."
} | 2,373 |
37915106 | PMC10621202 | pmc | 3,902 | {
"abstract": "Background As a cost-effective and eco-friendly approach, biocatalysis has great potential for the transformation of 5-hydroxymethylfurfural (HMF) into 2,5-furandicarboxylic acid (FDCA). However, the compatibility of each enzyme in the cascade reaction limits the transformation efficiency of HMF to FDCA. Results Coupled with an alcohol oxidase from Colletotrichum gloeosporioides ( Cgl AlcOx), this study aims to study the potential of bacterial laccase from Bacillus pumilus ( Bp Lac) in an enzymatic cascade for 2,5-furandicarboxylic acid (FDCA) biosynthesis from 5-hydroxymethylfurfural (HMF). Bp Lac showed 100% selectivity for HMF oxidation and generated 5-hydroxymethyl-2-furancarboxylic acid (HMFCA). Cgl AlcOx was capable of oxidizing HMFCA to 2-formyl-5-furancarboxylic acid (FFCA). Both Bp Lac and Cgl AlcOx could oxidize FFCA to FDCA. At the 5 mM scale, a complete transformation of HMF with a 97.5% yield of FDCA was achieved by coupling Bp Lac with Cgl AlcOx in the cascade reaction. The FDCA productivity in the reaction was 5.3 mg/L/h. Notably, Bp Lac could alleviate the inhibitory effect of FFCA on Cgl AlcOx activity and boost the transformation efficiency of HMF to FDCA. Moreover, the reaction was scaled up to 40 times the volume, and FDCA titer reached 2.6 mM with a yield of 58.77% at 168 h. Conclusions This work provides a candidate and novel insight for better design of an enzymatic cascade in FDCA production. Supplementary Information The online version contains supplementary material available at 10.1186/s13068-023-02406-z.",
"conclusion": "Conclusions The present study established an effective bi-enzymatic cascade system for FDCA synthesis using Bp Lac and Cgl AlcOx. These two oxidases played a complementary role in catalyzing HMF to form FDCA. Bp Lac showed good selectivity for the oxidation of HMF without the addition of redox mediators. Bp Lac could alleviate the inhibitory effect of FFCA on Cgl AlcOx activity and then boost the synthesis of FDCA in the cascade reaction. The key residues of Bp Lac and Cgl AlcOx that are involved in binding to the substrates were identified. The synthesis of FDCA in cascade reaction also succeeded when the volume scaled up to 0.2 L.",
"discussion": "Results and discussion Biotransformation of HMF and its oxidized derivatives by Bp Lac The purified Bp Lac was obtained through Ni-chelating affinity chromatography. The purity and the molecular weight of Bp Lac were confirmed via SDS‒PAGE (Additional file 1 : Figure S2). The catalytic potential of Bp Lac toward HMF and its oxidized derivatives was investigated. As shown in Fig. 1 , Bp Lac exhibited good catalytic performance for the oxidation of HMF and FFCA. When the concentration of HMF was 5 mM, HMF was completely converted to HMFCA at 100% selectivity after being treated by Bp Lac for 24 h (Fig. 1 A). When DFF was used as the substrate in the reaction, the yield of FFCA reached only 2.9% after being treated by Bp Lac for 120 h (Fig. 1 B). Similarly, when HMFCA was used as the substrate, no FFCA was detected and the yield of FDCA reached only 1.17% after being treated by Bp Lac for 120 h (Fig. 1 C). We speculated that the oxidation of HMFCA to FFCA occurred spontaneously. The trace of FFCA might further transformed into FDCA by Bp Lac transiently. Notably, Bp Lac showed a good capacity for FDCA synthesis using 5 mM FFCA as the substrate (Fig. 1 D). After being treated by Bp Lac for 120 h, the titer and yield of FDCA reached 4.69 mM with a yield of 94.94%. Fig. 1 Analysis of the oxidation capacity of Bp Lac towards HMF and its oxidized derivatives. Time-course analysis of the oxidation potential of Bp Lac to A 5 mM HMF, B 5 mM DFF, C 5 mM HMFCA, and D 5 mM FFCA, respectively. E Schematic illustration of the catalyzing potential of Bp Lac to HMF and its oxidized derivatives. means the reaction could not proceed, means the reaction could proceed successfully To date, both bacterial laccases and fungal laccases show the potential in FDCA synthesis by oxidizing HMF and its derivatives [ 26 , 33 – 35 ]. Laccases from bacteria generally have low redox potential (340–470 mV), while fungal laccases show high redox potential (490–790 mV) [ 36 ]. Relatively low redox potential limits the catalytic activity of bacterial laccases for many substrates. Previous study reveals that the redox mediators can enhance the capacity of laccases for oxidizing various substrates, including HMF [ 26 , 27 , 29 , 37 ]. The typical mediator among the synthetic mediators, 2,2,6,6-tetramethyl-1-piperidinyloxy (TEMPO), is widely used in LMS for HMF oxidation by bacterial laccase [ 34 , 35 ]. Nevertheless, the synthetic mediators are generally expensive and toxic [ 27 ]. In our study, without the redox mediators, Bp Lac was capable of oxidizing HMF to HMFCA with high selectivity and also oxidizing FFCA effectively. This characteristic of Bp Lac not only reduces the potential cost and health risk during HMF oxidation but may also reduce by-product formation in cascade reactions. Biotransformation of HMF and its oxidized derivatives by Cgl AlcOx Cleveland and co-workers characterized Cgl AlcOx systematically and considered it an alcohol oxidase [ 16 ]. They also found that Cgl AlcOx can completely oxidize HMF to form 91% of FFCA and 9% of FDCA in the presence of catalase and horseradish peroxidase [ 16 ]. The potential of Cgl AlcOx on HMF oxidation without the addition of other enzymes is unknown. Here, we aimed to assess the catalytic activity of Cgl AlcOx for the oxidation of HMF and its derivatives. The purified Cgl AlcOx was obtained through heterologous expression using P. pastoris X33 as the host. As expected, the molecular weight of recombinant Cgl AlcOx was 51.3 kDa, which was confirmed using SDS‒PAGE analysis (Additional file 1 : Fig. S2). Cgl AlcOx showed a different selectivity in HMF oxidation compared with Bp Lac. When 5 mM of HMF served as the substrate, Cgl AlcOx catalyzed HMF to form DFF. The DFF yield reached 45.35% rapidly within 24 h and the final yield of DFF reached 50.14% after being treated by Cgl AlcOx for 120 h (Fig. 2 A). Cgl AlcOx showed 100% selectivity for HMF oxidation to DFF because no HMFCA was detected in the reaction. As shown in Fig. 2 B, when DFF was used as the substrate in the reaction, the yield of FFCA reached only 2.91% after being treated by Cgl AlcOx for 120 h. Cgl AlcOx was able to oxidize HMFCA and then generated FFCA and FDCA (Fig. 2 C). With the decrease in HMFCA, the yield of the intermediate FFCA reached 15.57% at 24 h. After the reaction proceeded for 72 h, the final product FDCA appeared along with the disappearance of the intermediate FFCA. The yield of FDCA reached 12.07% with respect to HMFCA after being treated by Cgl AlcOx for 120 h. As shown in Fig. 2 D, when FFCA served as the substrate, the targeted product FDCA formed effectively after being treated by Cgl AlcOx for 120 h. The yield of FDCA reached approximately 89.12% after the reaction proceeded for 120 h. Fig. 2 Analysis of the oxidation capacity of Cgl AlcOx towards HMF and its oxidized derivatives. Time-course analysis of the catalyzing capacity of Cgl AlcOx to each substrate, including A 5 mM HMF, B 5 mM DFF, C 5 mM HMFCA, and D 5 mM FFCA, respectively. E Schematic illustration of the catalyzing potential of Cgl AlcOx to HMF and its oxidized derivatives. means the reaction could not proceed, means the reaction could proceed successfully As previous literature mentioned, Cgl AlcOx is classified as a member of AA5_2 subfamily and exhibits high specific activities toward a wide range of substrates, such as diols, primary alcohols, and aryl alcohols [ 16 ]. The results presented in this study were slightly different from those of previous research. In the previous study, Cgl AlcOx was capable of oxidizing HMF and HMFCA while being disabled to oxidize DFF and FFCA [ 16 ]. In the present study, we found that Cgl AlcOx could oxidize HMF, HMFCA, and FFCA, but not DFF. Therefore, Cgl AlcOx not only showed a potential to catalyze HMFCA to FFCA but also transformed FFCA to FDCA effectively, implying that Cgl AlcOx can be involved in the last two steps of the oxidation of HMF to FDCA. Establishment of the enzymatic cascade reaction for FDCA synthesis by Bp Lac and Cgl AlcOx As we described above, neither Bp Lac nor Cgl AlcOx could convert DFF to FFCA. Although Cgl AlcOx is capable of oxidizing HMFCA and FFCA, the selective oxidation of HMF by Cgl AlcOx to form DFF limited the accumulation of HMFCA at the first step. Bp Lac is well suited to overcome this issue because HMF could be selectively oxidized to HMFCA with 100% selectivity. Thus, the enzymatic cascade reaction for FDCA synthesis by Bp Lac and Cgl AlcOx was established and divided into two steps. First, we added only Bp Lac into the reaction using HMF as the substrate. After HMF was completely transformed into HMFCA by Bp Lac, the second step was initiated by adding Cgl AlcOx into the reaction for HMFCA oxidation. Then the intermediate FFCA could be transformed into FDCA with the help of both Bp Lac and Cgl AlcOx (Fig. 3 A). As shown in Fig. 3 B, the conversion of HMF to FDCA on a 5 mM scale was achieved with good yield. When the reaction proceeded at first step, HMF was completely transformed to HMFCA within 24 h. Thereafter, Cgl AlcOx was supplemented to the reaction. The HMFCA amount decreased to 4.02% along with the FFCA amount increased to 78.14% at 48 h, indicating that Cgl AlcOx could effectively convert HMFCA to FFCA in the cascade reaction. The yield of FDCA increased constantly along with the decrease of FFCA when the reaction proceeded after 48 h. After the reaction continuously proceeded for 168 h, the FDCA titer reached 4.88 mM with a yield of 97.5%. The FDCA productivity in the reaction was 5.3 mg/L/h. Additionally, for the characteristic signals of each chemical, the disappearance of HMF and the formation of HMFCA were observed when Bp Lac completely converted HMF to HMFCA. When Cgl AlcOx was added to the reaction mixtures, the intermediate HMFCA was rapidly transformed into FDCA. The 1 H-NMR analysis confirmed that only the characteristic signal of FDCA was detected in the final product (Fig. 3 D–E). Therefore, 1 H-NMR spectra confirmed that HMF could be almost completely transformed into FDCA using enzymatic cascade that was established in our study. Fig. 3 The enzymatic cascade reaction for FDCA biosynthesis in the present study. A Schematic representation of the proposed enzymatic cascade reaction for FDCA synthesis from HMF using Bp Lac and Cgl AlcOx. B Analysis of the FDCA synthesis catalyzed by Bp Lac and Cgl AlcOx using HMF as the substrate. Bp Lac was initially added to the reactions at 0 h. Cgl AlcOx was added into the reactions at 24 h when the HMF was completely transformed into HMFCA by Bp Lac. C \n 1 H-NMR spectra for initial HMF in the reaction mixture. D \n 1 H-NMR spectra for the complete formation of HMFCA from HMF catalyzed by Bp Lac. E \n 1 H-NMR spectra for identification of the final product FDCA in the enzymatic cascade. means the reaction could proceed successfully Compared with fungal laccases, bacterial laccases recently show great potential in cascade reactions for FDCA synthesis. Chang and co-workers designed a tandem biocatalysis strategy using immobilized laccase from Bacillus subtilis TJ-102 and Novozym 435 (immobilized lipase B from Candida Antarctica ), which provides a 94.2% FDCA yield from HMF [ 38 ]. Coupling with catalase, TEMPO, and H 2 O 2 , laccase from B. subtilis 168 can effectively utilize HMF as the substrate and yield 97.1% of FDCA [ 35 ]. As the above literature reported, the redox mediator, TEMPO, is used in enzymatic cascade reactions. Compared with the studies that added the mediators, we found that the complete transformation of HMF with a 97.5% yield of FDCA was achieved by coupling laccase from B. pumilus ZB1 and only alcohol oxidase. A comparison of the results of different enzymatic cascades is presented in Additional file 1 : Table S2. An increase in substrate HMF concentration and higher reaction efficiency for FDCA synthesis are desired in enzymatic cascade reactions. Notably, most reported cascades generally need three or more enzymes to catalyze HMF into FDCA. In our study, two oxidases catalyzed HMF to FDCA with high efficiency, implying that reducing the enzyme dosage might lower the cost during the biosynthesis of FDCA. Previous studies generally establish the enzymatic cascade reactions by eliminating the generated H 2 O 2 from alcohol oxidase catalysis [ 10 , 17 ]. In the present work, the transformation efficiency in cascade reaction is much higher than that of the single reactions, implying that laccase from B. pumilus ZB1 may be tolerant H 2 O 2 generated by Cgl AlcOx oxidation. Notably, Cgl AlcOx showed much higher oxidation capacity toward HMFCA in enzymatic cascade reaction compared with its single reaction. Here, we speculated that the intermediate FFCA and final product FDCA might inhibit the catalytic activity of Cgl AlcOx toward HMFCA. Bacterial laccase might alleviate the inhibitory effect of FFCA and FDCA on Cgl AlcOx. Analysis of the inhibitory effect of FFCA and FDCA on Cgl AlcOx activity To confirm whether FFCA and FDCA could inhibit Cgl AlcOx activity during the oxidation of HMFCA, we mixed the intermediate FFCA or the product FDCA to the reaction mixtures containing HMFCA and Cgl AlcOx, respectively. FFCA might have an inhibitory effect on Cgl AlcOx activity during HMFCA oxidation (Fig. 4 A). When the reaction mixture contained both HMFCA and FFCA, the amount of FFCA increased to 3.6 mM within 12 h while only a trace amount of FDCA was detected. The amount of HMFCA was stable when the reaction proceeded for 24 h. Interestingly, when the reaction mixture contained Bp Lac, HMFCA was almost completely transformed into FFCA within 12 h (Fig. 4 B). After being treated with Cgl AlcOx and Bp Lac for 24 h, FFCA amount decreased to 3.23 mM and FDCA amount increased to 2.55 mM. Fig. 4 Analysis of the potential inhibitory effect of FFCA/FDCA on Cgl AlcOx activity for HMFCA oxidation. A Measurement of the oxidation capacity of Cgl AlcOx using mixed substrates containing 2.5 mM FFCA and 2.5 mM HMFCA. B Verification of the role of Bp Lac in alleviating the inhibitory effect of FFCA on Cgl AlcOx in cascade reactions containing 2.5 mM FFCA and 2.5 mM HMFCA C . Measurement of the oxidation capacity of Cgl AlcOx using mixed substrates containing 2.5 mM FDCA and 2.5 mM HMFCA. D Analysis of the oxidation capacity of mixed enzymes ( Bp Lac and Cgl AlcOx) to the mixed chemicals (2.5 mM HMFCA and 2.5 mM FDCA) As shown in Fig. 4 C, FDCA probably also inhibited Cgl AlcOx activity for HMFCA oxidation due to HMFCA amount had no obvious change after the reaction proceeded for 24 h. However, when the reaction mixture containing HMFCA, FDCA, Cgl AlcOx, and Bp Lac, the disappearance of HMFCA and the formation of FFCA have occurred simultaneously within 12 h (Fig. 4 D). This result revealed that the presence of FDCA would not inhibit HMFCA oxidation by Cgl AlcOx. The continuously decreasing FFCA of 0.64 mM and increasing FDCA of 5.03 mM were detected at 24 h. This result also confirmed that the presence of Bp Lac in the reaction could improve the transform efficiency of HMFCA by Cgl AlcOx. In our study, we found that the FFCA has an inhibitory effect on alcohol oxidase activity toward HMFCA. Interestingly, Bp Lac could alleviate the inhibitory effect of FFCA on Cgl AlcOx, and boost the efficiency of the enzymatic cascade for FDCA synthesis. H 2 O 2 generated by alcohol oxidase is the limiting factor to enzymatic activity during cascade reactions [ 8 ]. Bp Lac might have a good tolerance towards H 2 O 2 generated from Cgl AlcOx catalysis. Therefore, Bp Lac is suitable for the high-efficiency production of FDCA from HMF by coupling with Cgl AlcOx. Additionally, alleviating the inhibitory effect could be achieved by increasing the relative amount of Cgl AlcOx in the reaction mixtures. Different from the original reaction mixtures containing 5 mM HMFCA and 1 U/mL Cgl AlcOx, lowering the amount of HMFCA to 1 mM or increasing the Cgl AlcOx loading to 5 U/mL benefit the transformation of HMFCA proceeded effectively (Additional file 1 : Fig. S3). These results provided a potential strategy that may improve the efficiency of FDCA biosynthesis in future studies. Prediction of the key residues of Cgl AlcOx and Bp Lac for binding substrates via molecular docking To deeply understand the enzymatic cascade system for FDCA synthesis, we investigated the biotransformation of HMFCA and FFCA by Cgl AlcOx via molecular docking, as well as the biotransformation of FFCA by Bp Lac. The prediction of the binding site between enzyme and substrate is presented in Fig. 5 . Subsequently, the molecular interaction of the enzyme–substrate complex was analyzed in detail (Table 1 ). Cgl AlcOx-HMFCA has a docking score of − 6.34 kcal/mol which formed six hydrogen bonds with Phe303, Ser304, Asp305, Pro331, Asn333, and Tyr334, and one salt bridge with His362. Cgl AlcOx-FFCA has a docking score of − 5.52 kcal/mol which formed five hydrogen bonds with Phe303, Ser304, Asn333, Tyr334, and Gly352, and one hydrophobic interaction with Glu360. Bp Lac formed four hydrogen bonds (Ser364, Gln427, Arg429, and Arg480), two hydrophobic interactions (Val406 and Ile478), and one salt bridge (Arg362) with FFCA, which resulted in the minimum binding energy of − 4.63 kcal/mol. The lowest binding energy represents highly stable conformation of substrate to enzyme [ 28 ]. In our study, both HMFCA and FFCA could form stable complex with Cgl AlcOx. Four amino acid residues, including Phe303, Ser304, Asn333, and Tyr334, are involved in binding to both HMFCA and FFCA. Cgl AlcOx showed a salt bridge with HMFCA, while forming one hydrophobic interaction with FFCA. The hydrogen bonds and hydrophobic interactions contribute to the stability of enzyme–substrate complex [ 39 ]. Compared with hydrogen bonds and hydrophobic interactions, the salt bridge has proven to be the strongest interaction between enzyme and substrate [ 40 ]. Therefore, the formation of one salt bridge with His362 may contribute to lower binding energy of HMFCA to Cgl AlcOx. In addition, arginine in the substrate binding pocket of bacterial laccase is considered to be important for substrate oxidation [ 41 ]. In the present study, three arginine residues from the active site of Bp Lac were found to be involved in binding to FFCA. Thus, the presence of Bp Lac in the cascade reaction may reduce the binding of FFCA to Cgl AlcOx and then oxidize FFCA to FDCA. Fig. 5 Representation of molecular docking results of the enzyme–substrate complex. A \n Cgl AlcOx-HMFCA. B \n Cgl AlcOx-FFCA. C \n Bp Lac-FFCA Table 1 Molecular docking analysis of the enzyme–substrate complex and the key amino acid residues Enzyme–substrate complex Binding amino acid residues Binding energy (kcal/mol) Hydrogen bonds Hydrophobic interactions Salt bridges Cgl AlcOx-HMFCA Phe303, Ser304, Asp305, Pro331, Asn333, Tyr334 His362 − 6.34 Cgl AlcOx-FFCA Phe303, Ser304, Asn333, Tyr334, Gly352 Glu360 − 5.52 Bp Lac-FFCA Ser364, Gln427, Arg429, Arg480 Val406, Ile478 Arg362 − 4.63 Scale-up experiment The enzymatic cascade reaction was scaled up to 0.2 L in a 2 L flask. As shown in Fig. 6 , the bi-enzymatic cascade system was successfully scaled up to 40 times the volume. With HMF as the substrate, the yield of HMFCA reached 92.45% after being treated by Bp Lac for 72 h. After Cgl AlcOx was added and the reaction proceeded for 96 h, the yield of FFCA and FDCA reached 52.59% and 28.75%, respectively. When the cascade reaction proceeded for 168 h, FDCA titer reached 2.6 mM with a yield of 58.77%. Scale-up brings the challenge to the effective biotransformation of HMF to FDCA. In scale-up reaction, Bp Lac took a longer time for the complete oxidation of HMF. Compared with the 5 mL reaction system, the yield of FDCA in the scaled-up system decreased from 97.5% to 58.77%. Due to oxygen transfer and mixing may be negatively affected, the catalytic efficiency decrease in scale-up system is expected [ 42 ]. An improvement of the transformation efficiency of the bi-enzymatic cascade system in scale-up process is our effort in the future. Fig. 6 Time-course analysis of FDCA biosynthesis using bi-enzymatic cascade system at 0.2 L scale-up reaction"
} | 5,164 |
35519042 | PMC9056729 | pmc | 3,903 | {
"abstract": "Superhydrophobic/superoleophilic materials have shown great potential for applications in oil/water separation. However, practical applications of these materials are restricted due to their toxicity and complicated, expensive, and non-eco-friendly fabrication procedures. Here, we have successfully developed an easy, simple, cost-effective, and environmentally friendly strategy towards the synthesis of superhydrophobic and superoleophilic porous polypyrrole nanotubes. Such wettability has been introduced into polypyrrole by co-doping with sodium dodecylbenzenesulfonate, a surfactant for lowering surface energy and controlling the morphology of the nanotubes. These non toxic and environment friendly polymer nanotubes exhibit oil absorption capability from oil/water mixtures with a reasonable efficiency with good reusability.",
"conclusion": "Conclusions In conclusion, we have successfully demonstrated a strategy for synthesizing superhydrophobic and superoleophilic conjugated polymers by tuning the dopants for oil–water separation and absorption. For this purpose, we have synthesized superhydrophobic polypyrrole nanotubes co-doped with SDBS, a surfactant using an easy and cost-effective soft template method. It is proposed that the wettability properties of the polypyrrole arise due to SDBS, which acts as a stabilizer for nanotubes and lowers their surface energy. This environment-friendly and non-toxic polymer was tested for the oil/water separation and showed a reasonable efficiency with good reusability. Thus, in this paper, the potential of PPy nanotube doped with surfactants is demonstrated as a probable contender for the application in oil/water separation. These results will also inspire synthesis of conjugated polymer nanofibers doped with different surfactants using soft template methods for oil/water separation.",
"introduction": "Introduction Accidental oil spillages constitute a significant source to pollute and imbalance the marine ecosystem. 1,2 Oil spill or leakage frequently occurs during exploration in the sea, transportation through pipelines or tankers by the sea, storage of oil, and from industrial waste-oil discharge. 3 The gravity of oil spills to marine water could be understood from the 1989 Exxon Valdez oil spill (EVOS) which had disastrous consequences. 4 Therefore, effective and efficient oil remediation is needed to save marine ecology. 5–10 Many conventional techniques are used for oil/water separation, including booms, skimming, oil burning etc. 5,6,12 These energy intensive and time-consuming techniques have issues of inefficient separation and secondary pollutants. For example, use of oil booms requires a water velocity of less than 1 knot. 1,6,9,12 Similarly, the skimming of oil from water requires vacuum pumps, relatively calm water, and chemical surfactants that are harmful to the marine ecosystem. 6,9,12 Therefore, to overcome these shortcomings, many new materials or techniques have been explored. Recently, bioinspired superhydrophobic/superoleophilic (SHSO) materials have attracted significant interest in oil/water separation owing to their unique surface wettability properties of having an extremely high water contact angle (CA) > 150°, and very low oil CA < 10°. 5,6,12 When SHSO materials are dipped in an oil–water mixture, oil permeates easily through these but water is rejected at the surface. In addition, these materials also have advantages of self-cleaning, corrosion resistance, anti-fouling, anti-bacterial, and UV resistance etc. 5–16 These special materials are synthesized by making hierarchal micro or nano structures such as nanotubes/nanofibers. The air could get trapped in such structures, which contributes to increasing water contact angle or hydrophobicity. 17,18 The contact angle can further be increased by coating with low surface energy materials. 17 Based on these properties, filters or oil absorbers have been developed using various porous materials, which include membranes, polymer fibers, metal mess cloths, cotton hydrogels/aerogels, and sponges, etc. 5,6,12,13,19,20 The membranes are used to separate oil from oil/water mixture by filtration, making the process efficient, simple, and cost-effective. 5,12 On the other hand, polymer nanofibers are used as sorbents which separate water by selectively absorbing/adsorbing oil from the oil/water mixture. These sorption characteristics of polymer nanofibers arise from: (a) superoleophilic and superhydrophobic properties of the fibers, (b) the porosity of the fibers (for retaining oil), and (c) capillary effect/adsorption by the fibers (for oil absorption). 13 The absorbed oil can be collected by squeezing, solvent extraction, or centrifugation. For oil spillage in open water systems where collection of contaminated water is difficult, such oil sorbents are applied by hand on the polluted water. 7,12,13,20 The fabrication processes of these superhydrophobic materials are complex, time-consuming, and require sophisticated instruments. 13,16,20 Therefore, quest for a cheaper, efficient, reusable, and environment friendly oil–water separator is a need of the time, and new strategies need to be devised for fabrication of SHSO materials. 7,16 Synthesis of nano/microtubes or fibers of conjugated polymer materials could be a prospective option. These nanofibers can be cost-effectively synthesized by chemical methods using predesigned templates (porous alumina) or soft template method, where self-assembled nanostructures of organic molecules, surfactants, and dyes act as templates. 21–23 The specialty of this class of polymers is that the positive charges are delocalized in the conjugated system counterbalanced by anions called dopants. The wettability properties of such polymers can be tuned by dopants, for illustration, perfluorinated doped polypyrrole (PPy) is hydrophobic whereas ClO 4 − doped PPy is hydrophilic. 24 Since these polymers are conducting polymers, studies of their hydrophobic properties were mainly focused for electrical, opto-electrical, and sensors applications. 24–27 However, conjugated polymers such as polyaniline, polythiophene, and polypyrrole are hardly explored exclusively for oil–water separation despite their potential for the same. Here, we are reporting the strategy for the synthesis of superhydrophobic/superoleophilic polypyrrole (PPy) nanotubes network for separating oil from water and the collection of the separated oil efficiently. The objectives of present work are: (1) to develop a cost-effective environmentally friendly method for synthesizing conjugated polymer nanotube networks similar to polymer nanofibers, (2) synthesized polymer should be self-standing, which could be spread in the open area and collected easily, and (3) the polymer should be environment friendly. For this purpose, polypyrrole is a suitable candidate as it is biocompatible and can be easily synthesized from monomers with varying morphologies by chemical oxidation. 22,23,28–32 We followed the ferric chloride-methyl orange soft template method for synthesizing polypyrrole (PPy) nanotubes. 22 The PPy was co-doped with a surfactant, sodium dodecylbenzenesulfonate (SDBS) to enhance hydrophobicity and control its morphology. The surface wettability properties of these PPy–SDBS co-doped nanotubes networks were studied. These nanotubes were utilized for oil/water separation and exhibited a robust absorption capacity, reusability, and retrievability of absorbate.",
"discussion": "Results and discussion Synthesis and characterization of the polypyrrole nanotubes Our strategy is inspired by superhydrophobic properties shown by polymer nanofibers or carbon nanotubes functionalized with low surface energy materials such as fluorinated silane, PTFE. 17,20 Thus, in present work, we replicated this carbon nanotube/nanofiber design by choosing polypyrrole nanotubes doped with SDBS. The choice of SDBS, a non-fluoro compound with a long hydrophobic alkyl chain, was made to use it as a low surface energy material to increase the hydrophobicity of PPy. 33 Moreover, SDBS is one of the important surfactants used in oil/water separation. 34 For synthesizing polypyrrole nanotubes, we adopted a soft template method in which ferric chloride–methyl orange (FeCl 3 –MO) template was used. The formation of PPy nanotubes using this soft template is well reported. 22,31,32 Our group also reported the synthesis of polypyrrole acicular nanorods using ferric chloride–methyl orange template. 30 Thus, the mechanism of the formation of nanotubes is briefly described in two steps. Under first step, FeCl 3 –MO nanorods are prepared for templates, and in second step, pyrrole gets polymerized by ferric chloride present in the template during which ferric ions convert to ferrous ions resulting in the decomposition of the template. Therefore, the polymerized pyrrole replicates the template structure resulting in polypyrrole nanotubes, which is obtained after washing the decomposed templates. The absence of the template in PPy is confirmed from the missing –N \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 N– stretching vibration at 1608 cm −1 in the FTIR spectra of PPy, as shown in Fig. 2 . 35 The formation of PPy is confirmed from its characteristics vibrations such as the anti-symmetric and symmetric pyrrole ring C C stretching vibrations at 1581 cm −1 and 1482 cm −1 , C–N stretching vibration at 1448 cm −1 , C–H deformation vibrations at 1240 cm −1 and 1059 cm −1 , and ring bending vibrations at 948cm −1 and 979 cm −1 . 36–38 The band from 1300 cm −1 to 1440 cm −1 is assigned to C–N and C–H in-plane deformation vibration modes. 39 The band at 1604 cm −1 and 1490 cm −1 are attributed to aryl C C anti-symmetric and symmetric stretching and 693 cm −1 to –SO 3 − bending vibration modes of dodecylbenzenesulfonate and corresponded to doping state of PPy. 36,37,40 Fig. 2 FTIR spectra of polypyrrole showing the characteristics vibration bands in the spectra. Surface morphology The surface morphologies of PPy doped with SDBS were studied by field emission scanning electron microscope (FESEM). Fig. 3(a) shows nanotube-like structures that are uniformly distributed and are interconnected to make porous structures. These tubes have an average length of 4.5 ± 0.2 μm and diameters of 200 ± 50 nm. The high-magnified SEM images show small branches or initiation of branching as dots that increase the roughness of the tubes (inset of Fig. 3(a) ). Such 3D nanoporous structures are reported for polymer nanofibers synthesized by electrospinning technique, which we could get using a simple soft template method. 13,20 We also studied the effect of concentration of SDBS on the morphology of the PPy. Broadly, the morphology has tubular-structure for PPy with or without SDBS as expected due to the template effect ( Fig. 3(a)–(c) ). Besides, the variation of the concentration of SDBS from 1 mM to 5 mM does not affect the morphology significantly, as shown in Fig. 3(a) and (b) , respectively. However, the variation in morphology is observed for PPy without SDBS, where the length of nanotubes (10 ± 0.1 μm) and their diameters (500 ± 50 nm) get doubled as compared to SDBS doped PPy ( Fig. 3(c) ).Consequently, its morphology appears more compact and less porous than that for the SDBS doped PPy. Fig. 3 SEM images of the PPy flakes with different doping concentration (a) 1 mM, (b) 5 mm of SDBS, and (c) without SDBS. The insets show the magnified images. These morphological studies suggest that the presence of SDBS affects the extent of polymerization and agglomeration of PPy nanotubes. Thus, the SDBS acts as a stabilizer for the PPy nanotubes, which could be explained as follows. During doping of SDBS, its anionic part (sulphonate group) along with Cl − generated during the reduction of FeCl 3 act as counter ions for polarons present in the oxidized PPy, and the tail part (alkyl chain) lies away from the polymer chain as schematically shown in Fig. 4(a) . After a certain doping concentration of SDBS, the presence of the long alkyl chains (C12) around PPy becomes prominent. Then these SDBS chains prevent the monomer from approaching the polymerization sites. Thus, this causes the termination of chain propagation, which leads to small polymer tubes ( Fig. 4(a) ). Similarly, the presence of long alkyl chain prevents nanotubes from getting agglomerated ( Fig. 4(b) ). Such steric effects are not present in the PPy without SDBS causing chain propagation and agglomeration, as shown schematically in Fig. 4(c) . In this case, only Cl − acts as the counter ion for polaron present in the PPy. Moreover, the reduced diameter for PPy polymerized tubes in the presence of SDBS may be attributed to partial dissolution of FeCl 3 –MO nanorod template on the addition of SDBS. 41 Fig. 4 Schematic representation of the effect of concentration of SDBS on morphology of PPy with SDBS causes (a) termination of chain propagation and (b) agglomeration. (c) Without SDBS causing chain propagation and agglomeration. In situ chloride (Cl − ) doping arises from the reduction ferric chloride to ferrous chloride during oxidation of pyrrole. Contact angle measurement The surface wettability of synthesized PPy film was evaluated by the water contact angle (WCA) and oil contact angle (OCA). As shown in Fig. 5 , the water drops on PPy are perfectly spherical with water contact angle (WCA) 157° ± 5°confirming its superhydrophobic nature. When a drop of hexane was placed on PPy, the organic drop was immediately absorbed, exhibiting excellent superoleophilicity. We also carried out a control experiment in which PPy was synthesized without SDBS as dopants. Such polymer was found to be hydrophilic possessing low WCA, although they have nanotube structure. The effect of concentration of SDBS (1 mM and 5 mM) was also studied on the wettability of PPy, and found to be superhydrophobic. These results suggest that the presence of SDBS alters the morphology of PPy as discussed above and also lowers the surface energy of the PPy. Since these qualities of the PPy nanotube doped SDBS match with superhydrophobic oil/water separators, this polymer should be useful for oil recovery. Fig. 5 Optical images of (a) water droplet on PPy flakes doped with SDBS and (b) water contact angles (161.8°) measurement. Oil/water separation and absorption For exploring its actual potential in oil/water separation, a model system of hexane/water mixture was chosen. Owing to lower density, the hexane layer floats over the water layer. The hexane was also dyed with BODIPY so that two distinct layers of water and oil are visible ( Fig. 6(a) ). When the powder of SDBS co-doped PPy was added in the liquid mixture, the colored layer was absorbed by PPy powder within a few seconds, as seen in Fig. 6(b) and (c) . The mixture was then filtered, and the filtrate did not show any coloration, confirming the total separation of hexane ( Fig. 6(d) ). Since surfactants are generally used to remove slick oil from the water, we explored further the possibility to remove the thin oil layer from water using PPy nanotubes. For this, in a similar fashion a few drops of hexane were spread on the surface of the water, and then the nanotubes were added ( Fig. 7(a)–(c) ). Within a few seconds, this polymer absorbed the colored layer (hexane) leaving behind the colorless water, as shown in the photographs ( Fig. 7 ). No polymer powder was observed in the water layer, owing to its superphydrophobic nature. These results confirm the good oil absorption and separation properties of SDBS co-doped PPy. Fig. 6 Photographs of (a) water hexane mixture, (b) and (c) the addition PPy nanotubes and PPy nanotubes + hexane layer, and (d) filtration. The BODIPY dye was added for coloration of hexane layer. Fig. 7 Photographic images of thin layer of hexane (dyed with BODIPY) on water (a) side view and (b) top view; (c) PPy naotubes on the hexane layer, and (d) after absorption hexane layer (disappearance of color) by PPy. The absorption capacity of PPy was calculated using eqn (1) , 1 where, m 0 and m sat represent the polymer before and after saturation by oil. The absorption capacity was found to be 8 g g −1 . The measured absorption capacity of PPy is comparable to other reported absorbents for hexane viz. polypropylene sponges (∼8g g −1 ), 42 polypropylene aerogel material (3–5 g g −1 ). 28 Further, absorbate retrievability and recyclability were also studied as these are important attributes for an ideal absorber. 7,12,15 Fig. 8 shows the retrieved hexane in a tube, obtained from hexane absorbed PPy nanotubes powder after heating at 70 °C. This low retrieval temperature indicates very weak interactions between hexane and PPy network, making it a low energy process. The powder was reused to absorb hexane for 10 cycles and exhibited consistent repeatability. As far as hexane retrievability is concerned, very little difference (<0.5%) was observed between 1 st and 10 th cycle. These results clearly demonstrate that the hexane was stored in voids and nanotubes. The morphology of PPy after retrieving the hexane was also investigated. However, surface morphographs of the used PPy were found the same as before the oil/water separation (Fig. S1 † ). This study confirmed robust reversibility in absorption/retrieval and, therefore a good reusability of PPy for oil/water separation applications. Fig. 8 The photographs showing the collection of hexane by heating the absorbed PPy powder. The empty tube was perforated to collect the condensed hexane. Although the absorption capacity of PPy is less than the commonly used oil absorbers like sponges, hydrogel aerogel, polypropylene/PVC nanofibers, these absorbers require complex, expensive fabrication process and also these are not environmentally friendly. 7,10 In contrast, PPy is freestanding and does not require any additional support material. The preparation method discussed in this paper is simple and easy, and the raw as well as product materials, are environmentally friendly and cheaper, making the process cost-effective. Therefore, these many advantages of PPy compensate for its limitation in the present absorption capacity and place it in the league of many options available for cleaning oil spills."
} | 4,691 |
35422973 | PMC8976094 | pmc | 3,904 | {
"abstract": "Highlights • KEMET allows evaluating and expanding microbial genome annotations. • KEMET can identify unannotated genes having partial sequence truncations. • Annotation expansion can improve contextual draft genome-scale metabolic model reconstruction.",
"introduction": "1 Introduction Metagenomics investigates environmental, engineered, and host-associated microbiomes, stimulating new fast-growing applications in biomedicine and biotechnology [1] , [2] . The shift towards a holistic approach in microbiome studies can uncover biological activities emerging from synergistic cooperation of microorganisms [3] . Many environments are now being inspected to decipher inhabiting microbial communities, with the aim of predicting their functions and interactions. Thanks to recent improvements in genome-resolved metagenomics, the recovery of metagenome-assembled genomes (MAGs) of high quality is becoming accessible and fast [4] . Functional analysis of genomes derived from metagenomic approaches allows estimating the metabolic potential of species present in a given microbiota. Several dedicated databases, such as the Kyoto Encyclopedia of Genes and Genomes (KEGG), are used as knowledgebases for metabolic pathway inference and reconstruction [5] , while tools such as KEGG Mapper [6] and eggNOG-Mapper [7] can assign open reading frames to their function and predict metabolic capabilities at the genome level. However, newly generated metagenomes contain a large number of poorly characterized species, which can be hardly annotated exhaustively with traditional tools [8] . Moreover, genome-scale metabolic models (GSMMs) are now starting to be applied on a metagenome scale [3] , [9] . GSMM are directly informed by annotation databases and can be automatically reconstructed using tools like CarveMe [10] or gapseq [11] . Such models are useful to infer interactions among microbial species, but the application to uncultured and non-model species can be challenging. In fact, MAG-based GSMMs are especially prone to reconstruction errors due to the gapped nature of metagenomic assemblies. Starting from GSMM reconstructions, several algorithms for network gap-fill enable in silico growth simulation and phenotype data fitting. Nevertheless, reactions added this way are not always supported by genomic evidence [12] , possibly resulting in erroneous predictions. To obtain a more exhaustive functional annotation of microbial genomes and improve associated GSMMs, we present KEMET. KEMET - KEgg Module Evaluation Tool - is a command-line, open-source Python toolbox aiming at summarizing and expanding KEGG annotation by comparing microbial sequences to orthologs with curated annotations. With KEMET, annotation recovery from trusted knowledgebases can strengthen the biological fidelity and phenotype prediction in GSMMs and lower the manual refinement effort.",
"discussion": "3 Results and discussion To validate KEMET, we first compared its KEGG Module partitioning with those performed by KEGG Mapper and METABOLIC v4.0 [19] across all the KEGG Modules present at the time of the tests. As shown by Fig. 2 A, the three tools interpret the Module block structure in a largely consistent way. However, KEMET is able to capture more Modules in the evaluation and has a block structure that more closely resembles that of KEGG as compared to METABOLIC. Fig. 2 Results of KEMET quality tests. (A) Comparison between KEMET and METABOLIC in terms of KEGG Module block structure with respect to the original KEGG Modules obtained through KEGG Mapper. The plot shows the intersections among the Module datasets for the three tools, together with the total number of Modules evaluated by each of them. (B) True positive rate for gene sequence identification by HMMs. Results for both isolated genomes (red) and MAGs (blue) are reported. Gene deletions of different extents were performed prior to running KEMET. When deletions were performed, gene annotation recovery was evaluated both with the gene prediction resulting from the original sequences and from those truncated, in order to account for the impact of deletions on gene prediction. (C) Fraction of correct metabolic phenotypes predicted by GSMMs reconstructed from microbial MAGs (green), the same MAGs with an expanded annotation through KEMET (orange), and the corresponding genomes from isolates (purple), based on the literature. The lines track the performance of individual GSMMs corresponding to the same strain. For readability purposes, only lines between points having performance differences across the datasets were drawn. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.) Next, we validated KEMET annotation expansion by two different approaches: (a) an annotation removal strategy to test its ability to identify known KO annotations, and (b) a draft GSMM reconstruction strategy to verify that newly identified annotations produce more sound quantitative models of microbial metabolism, and thus reflect correctly identified functions. Strategy (a) was used to test kemet.py --hmm_mode capability to retrieve the proper annotated sequences when either the original annotation was removed or the sequence was truncated. The rationale was to simulate misassembly-derived gene disruptions and other problems impairing functional prediction in MAGs. KEMET was tested on 12 MAGs derived from a contig-level assembly resulting from a previous work [20] as well as 5 complete genomes downloaded from NCBI (details in Supplementary Data ). In terms of taxonomic “novelty”, the MAGs were highly different and included species assigned at different levels (spanning from class to species) using GTDB-tk v1.5.0 [21] . The gene calling was performed using Prodigal v2.6.3 [22] with default options. Functional annotations of predicted genes were performed using eggNOG-mapper v2 [7] with default parameters. While in principle alternative gene predictions can impact the subsequent functional annotation, previous empirical investigations found negligible performance variation among different tools [23] , [24] . For this reason, our tests focused on benchmarking functional annotation prediction by using a single state-of-the-art gene prediction tool. The test consisted in the removal of three KO annotations from the input set of each genome (i.e. from eggNOG results) before running KEMET with the --hmm_mode onebm option. The selected KOs were annotated once per genome, only on a single gene. Moreover, removed KOs were chosen from different Modules marked “Complete” by KEMET, among the mandatory orthologs for a given biochemical step. In this way, removing them would result in the change of Module completeness to “1 block missing”. Altogether, 20/36 and 8/12 KO mock removals (55% and 67% true positive rate) resulted in the correct gene and annotation recovery for MAGs and complete genomes, respectively ( Fig. 2 B and Supplementary Data ). To model MAG construction issues more closely, the removal strategy was repeated two more times by simulating the deletion of tested KO-annotated gene sequences, either by 30% or 70% of their original length. This was done to mimic the typical scenario of a highly fragmented assembly where gene sequences can be split between two different contigs, resulting in a missed gene prediction or improper functional annotation. These additional tests resulted in a decreased performance using both the complete genomes and the MAGs dataset, as expected, but nonetheless gave a significant annotation recovery rate for gene truncations shorter than 50%. Specifically, an annotation recovery between 20% and 33% was achieved when accounting for the impact of sequence truncation on the gene prediction step, whereas a recovery rate of 42% was obtained assuming an unbiased gene prediction. This interval therefore captures KEMET performance in the presence of minor gene deletions. Similarly, for 70% gene truncations the annotation recovery rate further decreases, more clearly for the MAG dataset, as it is sensible with most of the gene sequence lacking. Hence, these results provide a proof-of-principle of KEMET annotation recovery in the occurrence of gene sequence disruption. Detailed results are included in the GitHub page at https://github.com/Matteopaluh/KEMET/blob/main/tests/README.md . Strategy (b) was implemented to assess the impact of recovering missing KO annotation on downstream metabolic analyses, i.e. via GSMM reconstruction. Specifically, we compared microbial phenotypes recovered from the literature (indicated in Supplementary Data ) in terms of metabolite production or consumption capabilities, to their corresponding in silico model predictions. This analysis was performed starting from MAGs and their corresponding complete genomes recovered from the NCBI or from the PATRIC database (as pointed by https://github.com/snayfach/IGGdb ), by selecting species collected from the anaerobic digestion microbiome [20] . MAG quality metadata were recovered and included genome completeness and contamination. If more than one MAG per species was present in the database, those with ≥ 90% completeness and ≤ 5% contamination were considered for the subsequent analysis. Both MAGs and the complete genomes of isolates were used to check the Module completeness. MAGs were also used to search for missing KOs by using kemet.py -- hmm_mode onebm . GSMMs were reconstructed from complete genomes and MAGs using CarveMe v1.4.1 [10] with the options --fbc2 -u , using as input both the MAG original gene calling and this same data added with the translated nucleotide sequences identified with the HMM via KEMET using the --gsmm_mode denovo parameter. Moreover, KEMET performance times were monitored and are included in Supplementary Data . To benchmark how the addition of newly identified sequences affects GSMM ability to describe in silico microbial physiology, metabolic capabilities retrieved from the literature were compared with predictions obtained starting from three types of input for GSMM reconstruction: MAG annotation, MAG annotation expanded with KEMET, and complete genome annotation. Flux variability analysis (FVA) was performed on the obtained GSMMs for assessing such metabolic capabilities, as follows. For each metabolite export reaction, it was determined whether the range of possible fluxes was directed towards metabolite consumption or production (respectively, having flux ranges consisting only of negative or positive values), while maintaining a fixed maximal growth rate. FVA results showing blocked reactions or flux ranging both positive and negative values were considered as incorrect predictions. The results show a nearly 10% improvement in the ability of MAG-derived GSMMs to produce and consume metabolites predicted from wet lab experiments, with an acquired accuracy comparable with the accuracy of GSMMs reconstructed from the genomes of isolates (both around 33%, Fig. 2 B and Supplementary Data ). On the annotation level, HMMs used on MAGs resulted in 84.76% hits in common with the respective reference isolate genome selected; 7.62% hits were present solely in the MAG dataset (false positives), and 7.62% hits were present in the complete genome dataset alone (false negatives). According to the selected dataset, KEMET HMM predictions therefore display a 91.75% precision and 91.75% sensitivity ( Supplementary Data ). Despite the addition of a limited number of protein sequences, the resulting models can thus be sensibly more accurate, leading to more precise inferences based on metabolic capabilities. For example, Selenomonas ruminantium MAG-derived GSMM (PATRIC genome id: 971.16) phenotype predictions were improved after KEMET usage. The original GSMM could not predict any known metabolic capability of S. ruminantium , while the modified GSMM could correctly reproduce metabolic exchanges involving cellobiose, salicin, mannitol, xylose, arabinose, fructose, maltose, lactose, and sucrose. In contrast, the GSMM based on the full genome annotation captured the correct exchanges for glycerol, cellobiose, salicin, mannitol, xylose, and arabinose. These results demonstrate that KEMET efficiently tackles the summarization of (meta)genomic potential in a user-friendly and scalable way. Other bioinformatics tools allow the evaluation of microbial genome annotation completeness (e.g. METABOLIC [19] ). However, to date and up to our knowledge, this is the only tool able to selectively fill the gaps in the annotation, and seamlessly add newly gathered information into GSMMs. At the moment, KEMET relies on KEGG given its structure allowing a systematic pathway completeness evaluation. Further development could include support towards other knowledgebases, such as MetaCyc [25] , to further expand the tool compatibility and predictive power. While other published programs, such as DRAM and Anvi’o [26] , [27] rely on specific KEGG releases, KEGG databases are constantly updated due to newly added sequences, or newly defined KO classifications. In contrast, KEMET allows users to update the downloaded KEGG GENES database through the KEGG API, in order to use the most up-to-date version of KEGG database without relying on fixed versions. The download of such a database represents the only limiting computational factor in KEMET ( Supplementary Data ), being a mandatory step to comply with the KEGG license. More efficient communication with KEGG servers could be obtained via license, while better solutions will be explored and implemented in future versions of KEMET. Nevertheless, this step is required only once at each database update, which can be decided by the user. Further, KEMET is based on HMMs given their broad applicability in the genomics and metagenomics fields. Other probabilistic graphical models, such as conditional random fields or Bayesian networks could be implemented in future versions of the software. Altogether, our experiments show that focusing on Module completeness down to single orthologs can aid in identifying missing annotations and enable their correction, not only supporting qualitative evaluation of microbial functions but also improving quantitative models of microbial metabolism. This enables a better mechanistic investigation of microbial ecological roles, allowing us to gather insights without relying necessarily on cultivation or in-depth characterization, which is impractical for most metagenomic studies."
} | 3,646 |
34001611 | PMC8166185 | pmc | 3,907 | {
"abstract": "Significance With simple DNA origami lever arms arranged in hinges and accordion structures, we amplify the nanometer displacements from DNA hairpin zippers to 4-μm motion, easily observable and quantified in real space and real time with conventional optical microscopy. Mechanically pulling a bead tethered on the accordion end, we measure high-energy recovery and retraction speeds up to 50 μm/s. On longer time scales, we have also opened and closed the hinges with light and heat. DNA nanotechnology, and particularly DNA origami, combined with colloids and emulsions can provide powerful architectures. The present study is a step toward activating such colloidal/cellular scale devices using DNA as a power source/fuel. We envision artificial active flagella, cilia, micropumps, and other cellular scale devices."
} | 204 |
23674611 | PMC3656441 | pmc | 3,908 | {
"abstract": "ABSTRACT Metabolic interactions with endosymbiotic photosynthetic dinoflagellate Symbiodinium spp. are fundamental to reef-building corals (Scleractinia) thriving in nutrient-poor tropical seas. Yet, detailed understanding at the single-cell level of nutrient assimilation, translocation, and utilization within this fundamental symbiosis is lacking. Using pulse-chase 15 N labeling and quantitative ion microprobe isotopic imaging (NanoSIMS; nanoscale secondary-ion mass spectrometry), we visualized these dynamic processes in tissues of the symbiotic coral Pocillopora damicornis at the subcellular level. Assimilation of ammonium, nitrate, and aspartic acid resulted in rapid incorporation of nitrogen into uric acid crystals (after ~45 min), forming temporary N storage sites within the dinoflagellate endosymbionts. Subsequent intracellular remobilization of this metabolite was accompanied by translocation of nitrogenous compounds to the coral host, starting at ~6 h. Within the coral tissue, nitrogen is utilized in specific cellular compartments in all four epithelia, including mucus chambers, Golgi bodies, and vesicles in calicoblastic cells. Our study shows how nitrogen-limited symbiotic corals take advantage of sudden changes in nitrogen availability; this opens new perspectives for functional studies of nutrient storage and remobilization in microbial symbioses in changing reef environments.",
"introduction": "Introduction Fundamental to the highly biodiverse reef ecosystems in (sub)tropical coastal waters is the endosymbiotic relationship between scleractinian corals (Cnidaria; Anthozoa) and autotrophic unicellular dinoflagellate protists of the genus Symbiodinium , commonly known as zooxanthellae ( 1 ). The disruption of this mutualistic association (coral bleaching) in response to, e.g., anthropogenic ocean warming is widely considered a key factor in the decline of coral reefs worldwide ( 2 ). In oligotrophic (nutrient-poor) waters, which often characterize the tropics, the dinoflagellate endosymbionts supply their animal host with photosynthetically fixed carbon compounds (photosynthates) that are essential to their respiration, growth, reproduction, and skeletal calcification ( 3 – 5 ). The dinoflagellates are also involved in the acquisition and retention of dissolved inorganic and organic nitrogen (N) from the seawater in the form of ammonium (NH 4 + ), nitrate (NO 3 − ), dissolved free amino acids (DFAAs), and urea, with a preference for ammonium ( 6 – 8 ), but the details of this process are much less clear. Although increasing loads of dissolved N in coastal waters (i.e., nutrification) is thought to potentially perturb the symbiotic relationship and threaten coral health ( 9 – 11 ), details of the coral response to a sudden natural or anthropogenic change in environmental N availability are unclear, in particular with regard to (i) the precise cellular and subcellular sites of N acquisition and storage, (ii) the nature and turnover of N storage metabolites, and (iii) the spatial and temporal pattern of host-symbiont nitrogenous nutrient exchange and subsequent metabolic utilization in the symbiotic system. Nanoscale secondary-ion mass spectrometry (NanoSIMS) isotopic imaging is a powerful analytical tool to simultaneously visualize and quantify in situ the incorporation and transfer of isotopically labeled metabolites in endosymbiotic organisms at subcellular-length scales ( 12 – 15 ). In this study, NanoSIMS analysis of coral tissue sections was combined with transmission electron microscopy (TEM) (the method is illustrated in Fig. S1 and Table S1 in the supplemental material) in order to investigate the dynamics of the assimilation and utilization of 15 N-labeled ammonium (NH 4 + ), nitrate (NO 3 − ), and DFAAs (here, aspartic acid) in the intact endosymbiosis between a reef-building coral and its dinoflagellates. Here, we define assimilation as the incorporation of N from inorganic and organic compounds initially dissolved in seawater into coral and dinoflagellate cell biomass. Pulse-chase labeling experiments with different N levels and multiple time scales (1- to 12-h pulse and 4-day chase) were performed at the Aquarium Tropical, Palais de la Porte Dorée (Paris, France), on coral nubbins of the very common and widely distributed Indo-Pacific reef-building coral species Pocillopora damicornis ( 16 ).",
"discussion": "DISCUSSION The ability of symbiotic reef corals to take up and assimilate various forms of inorganic (ammonium, nitrate) and organic (dissolved free amino acids, urea) N dissolved in the ambient seawater is well recognized, with ammonium representing the preferential source ( 6 – 8 ). However, the relative contribution to ammonium acquisition of each partner of the coral-dinoflagellate endosymbiosis has remained an open question ( 8 ). On one hand, early nutrient depletion experiments pointed to the dinoflagellates as the main assimilation site ( 17 , 20 ), a conclusion later supported by bulk-level tracer studies with [ 15 N]ammonium ( 21–24 ). The glutamine synthetase (GS)-glutamine oxoglutarate aminotransferase (GOGAT) enzymatic pathway was thus proposed as the main mechanism for ammonium assimilation by dinoflagellates, a view strongly supported by enzyme inhibitor experiments, with, e.g., the GS inhibitor methionine sulfoximine (MSX) ( 25 ) or the GOGAT inhibitor azaserine ( 26 ). On the other hand, detection within cnidarian animal tissue of both GS and glutamate dehydrogenase (GDH) indicates that the host potentially also has the enzymatic machinery for ammonium assimilation ( 26 – 28 ). Therefore, a host-mediated model of ammonium assimilation in which the animal host is the primary site for ammonium incorporation was proposed ( 29 ). A previous short-term pulse-chase (12 h) experiment with [ 15 N]ammonium, combining TEM and in situ quantitative NanoSIMS isotopic imaging of thick tissue sections (250 nm), indicated a primary role for dinoflagellates in ammonium assimilation compared to that of coral cells of the two epithelia in the oral tissue of the reef coral Acropora aspera ( 15 ). The present work confirms and expands this conclusion for the symbiotic reef coral P. damicornis , in which both dinoflagellate and coral cells, including all four coral epithelia (in oral and aboral tissue), are directly implicated in ammonium assimilation, with preferential incorporation by the dinoflagellate endosymbionts. Interestingly, ammonium assimilation by the symbiotic system appears less rapid in the present P. damicornis coral than in the previously studied A. aspera ( 15 ). Such kinetic differences may reflect variations of environmental parameters and/or species-specific effects and should be addressed in future experiments. Note that by comparing two different ammonium concentrations (20 µM versus 2 µM), our observations suggest a concentration-dependent mechanism for ammonium assimilation within both dinoflagellate and coral cells, a response previously identified only for the dinoflagellate endosymbionts of cnidarians ( 17 , 20 , 23 ). The apparent differences in the dynamics of ammonium incorporation between oral and aboral cellular layers of P. damicornis coral might be explained by various factors, including (i) different activities of ammonium-assimilatory enzymes among coral epithelia, (ii) different means of access of coral epithelia to [ 15 N]ammonium transported from seawater, i.e., by the oral tissue facing seawater versus the aboral tissue facing the skeleton, or (iii) the greater density of dinoflagellates in oral tissue, because these endosymbionts are thought to drive the uptake of ammonium from seawater ( 6 ). Additionally, we confirmed by NanoSIMS isotopic analyses that nitrate assimilation from seawater by the coral-dinoflagellate symbiosis takes place exclusively in the endosymbiotic cells, which possess nitrate- and nitrite reductase-assimilatory enzymes ( 18 , 19 ). Moreover, our NanoSIMS analyses indicate that aspartic acid dissolved in seawater is assimilated with similar efficiencies in the light by dinoflagellates and coral cells of all four epithelia, supporting previous observations from bulk-level tracer studies with 15 N-labeled DFAAs in various symbiotic cnidarians ( 7 , 22 ). Of interest, this conclusion strongly contrasts with early autoradiographic investigations performed on cnidarian tissue sections at the light microscopy level, proposing that incorporation of tritiated DFAAs occurred primarily in the oral epiderm rather than the oral gastroderm, with no incorporation of DFAAs into the dinoflagellates ( 30 , 31 ). These divergent results could be explained by the higher sensitivity and spatial resolution of NanoSIMS, which allows more precise, subcellular characterization of the sites for N assimilation. Aspartic acid is a major amino acid component of biocarbonate skeletal organic matrices ( 32 ). It is also involved in the purine pathway for uric acid metabolism ( 33 ). The observed contribution of dinoflagellate endosymbionts to its assimilation supports their potential role in providing precursors of skeletal organic matrix to fuel the biocalcification of their host. The coastal areas of the (sub)tropical oceans are characterized by large N concentration fluctuations due to, e.g., runoff from agricultural lands, sewage effluents, groundwater discharge, or fish excretion ( 10 , 11 , 34 ). A mechanism for temporary N storage in uric acid crystals in dinoflagellate cells has high nutritional importance for their coral host in such N-fluctuating environments; in addition, a central role has been demonstrated for endosymbiotic dinoflagellates in the rapid assimilation of dissolved N from ambient seawater. Macro- and microalgae are known to be capable of rapid assimilation and storage of N, which is then mobilized for growth during subsequent periods of N deficiency ( 35 – 37 ). Dinoflagellate crystalline deposits, previously assumed to be calcium oxalate ( 38 ), have recently been identified as uric acid (C 5 H 4 N 4 O 3 ) in the sea anemone Aiptasia sp. and were hypothesized to be subcellular sites for temporary N storage ( 39 ). Here, we demonstrate that in scleractinian corals, endosymbiotic dinoflagellates respond to a sudden increase in dissolved inorganic (ammonium, nitrate) and organic (aspartic acid) N in ambient seawater by storing N rapidly in intracellular uric acid crystals. Moreover, we demonstrate the rapid turnover of this metabolite, likely mobilized for maintaining the N status of the coral-dinoflagellate association under nutrient-limited environments ( 39 ). The formation and further remobilization of uric acid crystals are also reported (i) in other endosymbiotic associations involving marine animal hosts, such as in the tripartite endosymbiosis between the tunicate Molgula manhattensis , its apicomplexan Nephromyces protist (phylogenetically related to the dinoflagellates, in the Alveolates supertaxon), and their symbiotic intracellular bacteria ( 40 – 42 ), and (ii) in the endosymbiosis between the acoel flatworm Symsagittifera roscoffensis and the green microalgae Tetraselmis convolutae ( 43 ). However, in these associations, unlike with our results with symbiotic corals, it is the host partner that produces uric acid deposits as transitory N stores, which are later remobilized through the uricolytic activity of their endosymbionts. The dynamic pattern of uric acid production and remobilization that we have observed within dinoflagellate endosymbionts of scleractinian corals is likely to reflect the presence and activities of enzymes involved in de novo purine biosynthesis and further catabolism. Similar metabolic processes have been documented in the terrestrial endosymbiosis between leguminous plants and N 2 -fixing bacteria, which involves storage and subsequent conversion of uric acid into ureides ( 33 , 44 ). In the latter symbiotic system, bacteria convert atmospheric N 2 into ammonium, which is transferred to the plant host, where de novo purine biosynthesis and purine catabolism yield uric acid. Its subsequent metabolization into ureides (allantoin and allantoate) allocated to the entire host organism supports the plant N requirements. Our BLAST ( 45 ) analyses ( Table S4 ) of Symbiodinium expressed sequence tags (ESTs) ( http://medinalab.org/zoox/ ) indicate that transcripts encoding enzymes essential to purine synthesis and remobilization are present in the Symbiodinium transcriptome. This includes xanthine dehydrogenase ( 46 ), which catalyzes the conversion of hypoxanthine and xanthine to uric acid and uricase ( 19 ), which is involved in the subsequent oxidation of uric acid to the ureide allantoin. Our results therefore provide new evidence strengthening the hypothesis that activation of purine-related metabolic pathways is involved in controlling the N bio-economy of cnidarian endosymbioses under N-fluctuating marine environments ( 39 ). In reef corals, endosymbiotic dinoflagellate cells are thought to substantially contribute to the N requirements of their host by translocating N-bearing compounds in the form of amino acids ( 47 , 48 ) and/or larger N-containing glycoconjugates ( 49 , 50 ). However, direct evidence for this nutrient exchange, as well as its spatial pattern within the tissue layers and its precise timing, remained unclear ( 51 ). Here, combined TEM and NanoSIMS isotopic imaging of tissue sections provide direct in situ visualization and quantification of the dynamics of N fluxes in the symbiotic reef coral P. damicornis . Nitrogen translocation occurs first in the dinoflagellate-containing oral gastroderm about 6 h after exposure to a dissolved 15 N-labeled tracer. Subsequently, the three other cellular layers receive translocated N, indicating exchange between the different coral epithelia. Interestingly, the observed 6-h time lag is consistent with the hypothesis for delayed N release by the endosymbiotic dinoflagellates ( 52 ). Ultrastructural observations of the subcellular sites of coral host utilization of N acquired either directly by the host or translocated from the dinoflagellates suggest N utilization for important physiological functions of the coral, such as synthesis of mucus, proteins, and potential precursors of the skeletal organic matrix, known to be involved in coral biomineralization processes ( 53 ). In conclusion, by combining pulse-chase experiments using 15 N-enriched ammonium, nitrate, and aspartic acid with TEM ultrastructural observations and NanoSIMS isotopic imaging of thin coral tissue sections, we visualized and quantified in situ at the subcellular level the dynamics of dissolved N acquisition, storage, and utilization by the coral-dinoflagellate endosymbiosis. In particular, we provide experimental evidence for temporary N storage in dinoflagellate uric acid crystals in response to fluctuating environmental dissolved-nitrogen availability. In the context of marine environmental change due to both anthropogenic activities and natural fluctuations (e.g., ocean warming, acidification, pollution, and nutrification), this approach has the potential to provide new insights about how coral reef ecosystems will respond to such environmental perturbations."
} | 3,829 |
37999400 | PMC10672258 | pmc | 3,909 | {
"abstract": "Microalgae have gained attention as a promising source of chlorophylls and carotenoids in various industries. However, scaling up of conventional bubble columns presents challenges related to cell sedimentation and the presence of non-photosynthetic cells due to non-circulating zones and decreased light accessibility, respectively. Therefore, this study aimed to evaluate the newly developed continuously circulated bioreactor ROSEMAX at both laboratory and pilot scales, compared to a conventional bubble column. There was no significant difference in the biomass production and photosynthetic pigment content of Tetraselmis sp. cultivated at the laboratory scale ( p > 0.05). However, at the pilot scale, the biomass cultured in ROSEMAX showed significantly high biomass (1.69 ± 0.11 g/L, dry weight, DW), chlorophyll- a (14.60 ± 0.76 mg/g, DW), and total carotene (5.64 ± 0.81 mg/g, DW) concentrations compared to the conventional bubble column (1.17 ± 0.11 g/L, DW, 10.67 ± 0.72 mg/g, DW, 3.21 ± 0.56 mg/g, DW, respectively) ( p ≤ 0.05). Flow cytometric analyses confirmed that the proportion of Tetraselmis sp. live cells in the culture medium of ROSEMAX was 32.90% higher than that in the conventional bubble column, with a photosynthetic efficiency 1.14 times higher. These results support suggestions to use ROSEMAX as a bioreactor for industrial-scale applications.",
"conclusion": "5. Conclusions This study shows the potential of the ROSEMAX bioreactor as an ideal bioreactor for cultivation of Tetraselmis sp. biomass and photosynthetic pigment production. The cultivation of Tetraselmis sp. is essential to meet the consumer demands for bioactive compounds, and the pilot-scale ROSEMAX offers reliable product quality and quantity compared to conventional bubble columns. In pilot scale cultivations, the biomass concentration, chlorophyll- a content, and total carotene content of Tetraselmis sp. cultivated in ROSEMAX were significantly (1.44 times, 1.39 times, and 1.74 times, respectively) higher than those in conventional bubble columns ( p ≤ 0.05). A flow cytometry analysis showed that the live cell proportion in ROSEMAX was 84.36 ± 0.84% on the final day of cultivation, while that in conventional bubble columns was 51.46 ± 2.32%. Photosynthetic efficiency was 1.14 times higher in ROSEMAX. These results highlight ROSEMAX’s ability to prevent cell sedimentation and non-circulating zones during large-scale microalgae production. Notably, ROSEMAX’s thin and efficient design allows for effective light utilization, ensuring a consistent biomass concentration and photosynthetic pigment content, even in scaled-up cultivations.",
"introduction": "1. Introduction Currently, there is a growing interest among consumers to adopt healthy eating habits and prioritize food safety [ 1 ]. In response to this trend, food manufacturing companies are making efforts to develop and industrially produce compounds from natural sources to create nutritious and safe products [ 2 ]. Furthermore, there is a global trend to discourage the use of artificial synthetic colorants [ 3 ], which has led to a growing interest in biologically synthesized pigments that are considered nontoxic [ 4 ]. Natural pigments are gaining popularity because of their potential to meet the demands of consumers seeking healthy and safe food products [ 5 , 6 ]. By 2028, the European natural pigment market is expected to reach USD 2.18 billion, with a compound annual growth rate (CAGR) of 10.1% [ 7 ]. Generally, natural pigments are classified into several structural types, including tetrapyrroles, carotenoids, flavonoids, curcuminoids, betalains, and others [ 8 ]. Natural pigments are derived from biological sources, including animals, plants, microalgae, bacteria, and fungi [ 8 ]. Among the organisms, microalgae are considered to be promising biomass sources for natural pigments, because of their rapid growth, high photosynthetic efficiency, high productivity of metabolites, and low greenhouse gas emissions [ 9 , 10 ]. The main pigments produced in microalgae are tetrapyrroles and carotenoids [ 8 ]. Tetrapyrroles are a class of pigments composed of four pyrrole units, with chlorophyll being the most common representative. In particular, chlorophyll- a , which is abundant in green microalgae [ 11 ], has been suggested for medical purposes including wound healing, and for antimicrobial, anticancer, anti-mutagenic, antitumor, and anti-inflammatory properties [ 12 , 13 ]. Carotenoids are classified as non-oxygenated carotenes and xanthophylls. The types of carotenoids include alpha-carotene, beta-carotene, lutein, zeaxanthin, and astaxanthin [ 14 ]. These carotenoids, obtained from green microalgae, have demonstrated a wide range of health benefits including disease prevention, skin protection, and anti-aging, antioxidant, and antiviral effects [ 15 , 16 , 17 ]. These findings highlight the potential of microalgae biomasses and their derived photosynthetic natural pigments as promising candidates for use in functional foods and cosmetics [ 18 , 19 , 20 ]. In recent years, there has been growing interest in the efficacy of chlorophylls and carotenoids extracted from Tetraselmis sp. As a result, research on large-scale cultivation of Tetraselmis sp. biomasses for the production of these pigments has been reported [ 21 , 22 ]. These species are primarily cultivated to produce omega-3 fatty acid-rich aquafeeds [ 23 ]. Recent research has revealed the efficacy of Tetraselmis sp. biomasses across a wide range of applications, including antimicrobial activity, antioxidation, metal chelation, neuroprotection, and cell recovery [ 19 , 24 , 25 , 26 ]. For microalgal biomass production, the most widely used cultivation system in the industry is the open raceway pond (ORP) [ 27 ]. However, open-pond cultivation often results in severe biological contamination. In the worst-case scenario, the appearance of predators can lead to culture failure within a short time [ 21 ]. For this reason, only a few species capable of growing under highly alkaline and saline-selective environmental conditions, such as Arthrospira and Dunaliella , can overcome the contamination issues [ 28 ]. Therefore, several studies have suggested that photobioreactors may be more efficient for large-scale cultivation [ 28 ] and the development of innovative photobioreactors is encouraged for large-scale microalgae production. As the photobioreactors are scaled up in outdoor pilot-scale systems, where controlling the light can be challenging, the cell concentration tends to decrease, resulting in reduced light availability in the central region of the bioreactor. Recently, several photobioreactors have been proposed; however, most rely on artificial lighting to enhance light availability [ 29 ]. Adding artificial light can be costly and can produce heat, potentially leading to additional problems [ 29 ]. The reduction in light availability negatively affects the growth of microalgae [ 30 ]. Therefore, it is important to design a reactor with a thin structure to improve light availability. A bubble column bioreactor has been used as the conventional photobioreactor. Circulation is achieved using air bubbles (air injection) because stir-based mixing methods, which are generally used otherwise, are difficult to install in thin bioreactors [ 31 , 32 ]. The bubble column includes a rapid upward flow from the air injection points along the bioreactor axis and a downward flow near the upper wall, creating circulation throughout the entire culture medium [ 33 ]. Although this method is widely used in industry owing to its simplicity [ 34 ], relying only on bubbles can lead to the formation of partially non-circulating zones [ 35 , 36 , 37 ]. Due to these problems, closed photobioreactors encounter challenges in achieving the same level of productivity as in the laboratory during scaled-up processes [ 38 , 39 ]. Therefore, the development of a bioreactor that is thin—for light availability—has continuous medium circulation to prevent cell sedimentation, and uses simple air injection is necessary. The Jeju Bio Research Center of Korea Institute of Ocean Science and Technology (KIOST, Jeju Island, Republic of Korea) developed a novel, continuously circulating thin bioreactor (Korean Patent No. 1020180124656) for microalgal biomass production. This bioreactor, named Reproduction, Operation, Suspension, Easy, and Economized Max (ROSEMAX), was designed to prevent cell sedimentation by preventing non-circulating zones. Despite the fact that a previous study demonstrated that ROSEMAX could provide a stable and sufficient supply of Arthrospira ( Spirulina ) maxima biomass annually [ 40 ], more comprehensive evidence is needed to emphasize that ROSEMAX is a favorable system for microalgae cultivation. Therefore, this study aims to evaluate the newly developed ROSEMAX cultivation system at laboratory and pilot scales to overcome the problems of conventional bubble columns. ROSEMAX, which is thinly constructed and has a circular circulation form to reduce the light path and remove the cell sedimentation zones, will be evaluated for how advantageous it is for biomass, chlorophyll-a, and carotenoid production. To verify this, live cells in the culture medium will be selectively analyzed and photosynthetic efficiency measurements will be obtained. The results will serve as basic data for verifying the practicality of the ROSEMAX bioreactor and evaluating whether it is a viable cultivation system for scale-ups to an industrial-scale.",
"discussion": "3. Discussion Variations in productivity and biochemical composition between laboratory- and pilot-scale outdoor-cultivated microalgal biomasses pose challenges to scaling up [ 38 , 39 ]. In this study, we aimed to evaluate a newly developed ROSEMAX bioreactor for the practical cultivation of Tetraselmis sp. As shown in Table 1 , there was no significant difference in the growth of Tetraselmis sp. cultured in the conventional bubble column and ROSEMAX at the laboratory scale ( p > 0.05). However, at the pilot scale, the cell growth, and chlorophyll and carotenoid contents of Tetraselmis sp. cultured in ROSEMAX were significantly higher than those in the conventional bubble column ( p ≤ 0.05). Similar to our results, previous studies on Tetraselmis sp. cultivation also showed a higher content of photosynthetic pigment in PBRs with enhanced biomass production [ 41 ]. However, excessive light conditions of more than 1000 µmol photons/m 2 /s can cause photosynthetic inhibition in Tetraselmis sp., leading to negative cell growth and decreased chlorophylls and carotenoids [ 42 ]. In this study, the pilot-scale culture dependent on natural light was performed within a light intensity of 67.24–457.50 µmol photons/m 2 /s, and photosynthetic inhibition was not a concern. The chlorophyll- a (14.60 ± 0.76 mg/g, DW) and total carotene (5.64 ± 0.81 mg/g, DW) contents of Tetraselmis sp. produced in this study were comparable to those of the representative microalgae Spirulina platensis and Chlorella vulgaris , containing 10.6 and 6.11 mg/g, DW of chlorophyll- a and 2.4 and 4.7 mg/g, DW total carotene content, respectively [ 7 , 43 ]. Photosynthetic efficiency (F V /F M ) is positively correlated with photosynthetic pigments [ 44 ]. Therefore, in practice, there is a tendency to reduce the thickness of bioreactors to improve the photosynthetic efficiency of microalgae [ 45 ]. The chlorophyll fluorescence observed in the dark-adapted Tetraselmis sp. cells under a strong light pulse displayed an OJIP curve. Figure 3 shows the OJIP curves obtained from the minimum (F O ) to maximum fluorescence (F M ) [ 46 ]. The parameter F V /F M serves as a representative indicator of photosynthetic performance, specifically reflecting the ability of photosystem II to absorb light energy under dark-adapted conditions, which is essential for the photosynthetic process [ 47 , 48 ]. Light availability refers to the amount of light accessible to photosynthetic organisms in an environment, and F V /F M indicates the efficiency of light utilization in that environment. This study supports the idea that by increasing light accessibility in the central region of the photobioreactor, both biomass production and photosynthetic pigment content are likely to increase, as shown in our results ( Figure 1 and Figure 4 ). Flow cytometry can separate cells with and without pigments (cell debris) using the autofluorescence of the photosynthetic pigments [ 32 , 49 ]. As shown in Figure 1 , Tetraselmis sp. live cells were identified with high FL3-A signals and were distinguished from cell debris. Laboratory-scale cultivation conducted in an autoclaved medium with efficient fluid flow and an appropriate artificial light supply allowed both bioreactors to maintain a high proportion of live cells throughout the cultivation period. However, as shown in the cultivation results ( Figure 1 ), increasing the culture scale led to an increase in the proportion of cell debris in the conventional bubble columns. Although this phenomenon is common, ROSEMAX exhibited relatively low levels of cell debris during the culture period. Figure 2 shows the results of the flow cytometric analysis in the laboratory- and pilot-scale cultivations. In the pilot-scale cultivation, the ratio of live cells cultured in the conventional bubble column bioreactor decreased as the cultivation progressed from 95.06 ± 0.85% to 51.46 ± 2.32% on the final day. In contrast, ROSEMAX maintained a live cell proportion of over 84.36 ± 0.84% until the end of the cultivation. ROSEMAX was designed to overcome the disadvantages of the bubble column and airlift bioreactor. Airlift mixing involves the creation of a fluid flow by artificially separating the upward and downward flows in bubble columns [ 50 ]. To overcome the non-circulating zone of bubble columns during scale-up, transparent structures, such as baffles, are often added to bubble columns to physically enhance fluid flow [ 51 , 52 ]. However, installing baffles within a bioreactor can be costly and pose contamination issues, thus limiting its application in large-scale cultivation [ 53 , 54 ]. The structure of ROSEMAX allows for continuous fluid flow without a non-circulation zone and the need for attachments such as baffles. In the O-shaped ROSEMAX, injecting air at angles of 90° and 135° on one side creates an upward flow, whereas on the opposite side, between 180° and 360°, a downward flow naturally occurs along the curved shape of the reactor ( Figure 5 ). Previous studies also improved the circulation in bioreactors by modifying the reactor structure without baffles [ 55 ]. To our knowledge, no bioreactor with a horizontal rotating cylindrical shape similar to ROSEMAX has been used for continuous microalgae cultivation. However, a similar tank shape, known as a Kreisel tank, is commonly used in jellyfish cultivation. The Kreisel tank creates a continuous rotating water flow to prevent plankton sedimentation [ 56 ], supporting the structural advantages of ROSEMAX. Therefore, a high proportion of Tetraselmis sp. live cells was achieved in ROSEMAX because of structural air mixing, which minimized the cell debris. Photobioreactors are suggested to be designed with a thickness of 0.2 m or less, such as tubular and flat-panel photobioreactors, as this has been advantageously evaluated for large-scale production [ 57 , 58 ]. The pilot scale of ROSEMAX used in this study was in the form of a horizontally aligned cylindrical shape with a diameter of 1 m and a height of 0.2 m. This thickness is similar to that of general flat-panel and tubular reactors (0.2 m). Remarkably, despite ROSEMAX being scaled up approximately 36 times from its laboratory scale, there were no significant differences observed in biomass productivity and pigment content in the comparison between laboratory-scale and pilot-scale cultivations ( Table 1 ) ( p > 0.05). In contrast, the conventional bubble column bioreactor with a larger thickness (0.6 m) exhibited a substantial decrease in biomass concentration and photosynthetic pigment content. Additionally, the photosynthetic efficiency (F V /F M ) in ROSEMAX increased relative to that in the conventional bubble column bioreactor as cultivation progressed. These results indicate that the efficient design of ROSEMAX allows microalgae to effectively utilize light, even during scaled-up cultivation. Furthermore, thin photobioreactors such as ROSEMAX can increase the surface area-to-volume ratio through vertical alignment [ 59 , 60 , 61 ]. However, when several such bioreactors are installed, it is generally recommended to limit the bioreactor height to 4 m or less to avoid affecting the light availability near the photobioreactors [ 57 ]. Taking this into consideration, if ROSEMAX were scaled up from a 1 m diameter to a 4 m diameter, it would allow microalgae cultivation at an industrial scale of approximately 2500 L. In addition, if the width were increased by the depth (40 cm) recommended for ORP [ 21 ], the volume of the ROSEMAX bioreactor would be approximately 5000 L. Although it is not included in this study, a mixing efficiency or mass transfer analysis related to cell mixing, which can make an unexpected problem at a larger scale, is required to scale-up to the scale mentioned above [ 62 ]. The results of this study demonstrate that ROSEMAX can be applied to industrial scales. During scaled-up cultivation from the laboratory scale to the pilot scale, no significant decrease in biomass concentration or photosynthetic pigment content was observed ( p > 0.05). Following the pilot-scale cultivation, a noticeable difference in cell debris was observed between the biomass pellets cultivated in the conventional bubble column and ROSEMAX bioreactors. Therefore, we emphasize that the ROSEMAX bioreactor is suitable for reducing non-circulating zones in the culture medium and providing light accessibility, even in the central region of the bioreactor."
} | 4,522 |
20965727 | null | s2 | 3,910 | {
"abstract": "Cells constantly probe and respond to a myriad of cues that are present in their local surroundings. The effects of soluble cues are relatively straightforward to manipulate, yet teasing apart how cells transduce signals from the extracellular matrix and neighboring cells has proven to be challenging due to the spatially and mechanically complex adhesive interactions. Over the years, advances in the engineering of biocompatible materials have enabled innovative ways to study adhesion-mediated cell functions, and numerous insights have elucidated the significance of the cellular microenvironment. Here, we highlight some of the major approaches and discuss the potential for future advancement."
} | 175 |
40069804 | PMC11899718 | pmc | 3,913 | {
"abstract": "Background Methane emissions from livestock, particularly from dairy cattle, represent a significant source of greenhouse gas, contributing to the global climate crisis. Understanding the complex interactions within the rumen microbiota that influence methane emissions is crucial for developing effective mitigation strategies. Results This study employed Weighted Gene Co-expression Network Analysis to investigate the complex interactions within the rumen microbiota that influence methane emissions. By integrating extensive rumen microbiota sequencing data with precise methane emission measurements in 750 Holstein dairy cattle, our research identified distinct microbial communities and their associations with methane production. Key findings revealed that the blue module from network analysis was significantly correlated (0.45) with methane emissions. In this module, taxa included the genera Prevotella and Methanobrevibactor , along with species such as Prevotella brevis, Prevotella ruminicola, Prevotella baroniae, Prevotella bryantii, Lachnobacterium bovis, and Methanomassiliicoccus luminyensis are the key components to drive the complex networks. However, the absence of metagenomics sequencing is difficult to reveal the deeper taxa level and functional profiles. Conclusions The application of Weighted Gene Co-expression Network Analysis provided a comprehensive understanding of the microbiota-methane emission relationship, serving as an innovative approach for microbiota-phenotype association studies in cattle. Our findings underscore the importance of microbiota-trait and microbiota-microbiota associations related to methane emission in dairy cattle, contributing to a systematic understanding of methane production in cattle. This research offers key information on microbial management for mitigating environmental impact on the cattle population. Supplementary Information The online version contains supplementary material available at 10.1186/s42523-025-00386-z.",
"conclusion": "Conclusions This study employed WGCNA to investigate the interactions within the rumen microbiota and their associations with CH 4 emissions using a population-level of 750 Holstein cows. The MEblue was significantly correlated with CH 4 emission, revealing the critical roles of taxa such as Prevotella , Methanobrevibacter , and Methanomassiliicoccus luminyensis . These taxa underscore the complex interactions in carbohydrate fermentation and methanogenesis, key processes contributing to CH₄ production. Additionally, MEbrown was strongly associated with herd factors, revealing microbial networks influenced by farm management practices, diet composition, and feeding strategies. Functional predictions emphasized the complementary roles of bacteria and archaea in the rumen ecosystem, where bacteria provide substrates such as short-chain fatty acids (SCFAs) and hydrogen for methanogenic archaea, which are enriched in pathways linked to CH 4 production. The present study highlights the microbiota-trait and microbiota-microbiota interactions related to CH 4 emission in dairy cattle, contributing to a systematic understanding of CH 4 production in cattle and offering key information on microbial management for mitigating environmental impact in cattle population.",
"discussion": "Discussion Based on the WGCNA results, MEblue was the most significant module, correlating moderately with CH 4 emissions (r = 0.45, p = 7e-37) (Fig. 2 ). This module includes both methanogenic archaea ( Methanobrevibacter and Methanomassiliicoccus luminyensis ) and carbohydrate-fermenting bacteria ( Prevotella brevis, Prevotella ruminicol, Prevotella ruminicola, and Prevotella baroniae) (Fig. 3 ), forming a metabolic network that contribute to CH 4 emissions. Methanobrevibacter and Methanomassiliicoccus luminyensis are well-known for their roles in methanogenesis [ 32 ]. Both Methanobrevibacter and Methanomassiliicoccus luminyensis are methanogenic archaea that contribute to CH 4 production in rumen, utilizing hydrogen and CO 2 (hydrogenotrophy) or methylated compounds (methylotrophy) to produce CH 4 [ 33 , 34 ]. Notably, Methanomassiliicoccus luminyensis specially requires hydrogen as an electron donor, reducing methanol, methylamines into CH 4 [ 35 ]. The co-occurrence of Methanobrevibacter and Methanomassiliicoccus luminyensis within MEblue suggests functional interactions, where Methanobrevibacter helps maintain low hydrogen partial pressure in the rumen, indirectly supporting Methanomassiliicoccus by creating favorable conditions for its methylotrophic methanogenesis [ 36 ]. In addition to archaea, MEblue also includes bacteria from genus Prevotella , particularly networks among Prevotella brevis - Prevotella ruminicola - Prevotella baroniae , Prevotella brevis-Lachnobacterium bovis, Prevotella ruminicola - Prevotella bryantii, and Prevotella ruminicola - Prevotella brevis- Barnesiella viscericola , as well as Paraprevotella - xylaniphila within MEblue (Fig. 3 ). The interactions between Prevotella brevis , Prevotella ruminicola , and Prevotella baroniae suggest synergistic relationships that enhance polysaccharide breakdown and hydrogen production, both of which are crucial for CH 4 production [ 37 ]. These bacteria was highly abundant and co-clustered, emphasizing their role in carbohydrate fermentation, converting into short-chain fatty acids (SCFAs) [ 38 , 39 ]. Among SCFAs, acetate indirectly contributes to CH 4 production by methanogenic archaea. While acetate can serve as a substrate for methanogenesis in specific contexts, though its contribution to ruminal methane production is minimal due to the dominance of hydrogenotrophic and methylotrophic pathways and the rapid rumen passage rate, which limits acetogenic methanogens [ 39 ]. Furthermore, the cross-phylum microbial interaction ( Prevotella brevis - Lachnobacterium bovis ) is a suggests a synergistic metabolic network, where fermentative bacteria generate hydrogen as a byproduct, which is then utilized by methanogenic archaea for CH 4 production. Additionally, Lachnobacterium bovis , which was found to interact with Prevotella brevis , produces intermediates like lactate and acetate, further enhancing hydrogenotrophic pathways that contribute to CH 4 emissions [ 19 ]. Our results align with previous studies [ 17 , 38 ] but provide more detailed insights into microbial interactions at the species level. Furthermore, these findings reinforce that MEblue represents a functionally cohesive module linked to CH₄ emissions through hydrogen-mediated microbial interactions. While MEblue is significantly associated with CH 4 , other microbial modules also influence methane metabolism under different environmental conditions. For instance, MEbrown, another major module, contains similar fermentative bacteria ( Prevotella brevis and Prevotella ruminicola ), which are well-known contributors to carbohydrate fermentation and SCFAs production [ 39 ]. However, MEbrown networks exhibited greater diversity in bacterial taxa, including Flavonifractor plautii , Parabacteroides merdae , Barnesiella viscericola , and Alistipes putredinis (Supplementary Fig. 5). Unlike MEblue, which is moderately linked to CH 4 emissions, MEbrown is more influenced by herd effects (r = 0.43, p = 4e-34) (Fig. 3 ), suggesting that herd-specific factors such as feeding practices and diet composition [ 40 ]. Also, Flavonifractor plautii contributes to flavonoid degradation [ 41 ], influencing microbial dynamics and contributing to SCFAs production. PCA analysis (Supplementary Fig. 2) further revealed that the herd-specific factors significantly shaped microbial community structure, with lactation effects being less pronounced. This may be partly due to the fact that our study was conducted in commercial dairy cattle farms where cattle were reared with potentially different management conditions. Variations in feed composition, fiber content, and nutritional balance across farms likely contributed to the microbial diversity within MEbrown. This underscores the importance of dietary and environmental factors in modulating microbial networks and their roles in CH 4 emissions [ 40 ]. Understanding these environmental influences can help develop targeted intervention strategies to manipulate microbial communities for CH 4 mitigation. Since MEblue represents a microbial network closely associated with CH 4 production, dietary and management interventions targeting this module could be effective in reducing methane emissions. Recent studies reported that strategies such as dietary modifications, probiotics, and feeding additives can affect the CH 4 emissions by altering microbial community structure and metabolic pathways [ 42 – 44 ]. Probiotics can modulate gastrointestinal microbial. Their colonization in rumen improves feed efficiency, potentially reducing CH 4 emissions [ 23 ]. Similarly, dietary interventions, including high-lipid diets [ 45 ], nitrate supplementation [ 46 ], and plant secondary metabolites (e.g., tannins and saponins) [ 47 , 48 ], have been explored as effective CH 4 mitigation strategies. Lipid supplementation can suppress methanogens by reducing hydrogen availability, while tannins can directly inhibit methanogens activity [ 45 ]. Furthermore, dietary nitrate supplementation provides an alternative hydrogen sink, outcompeting methanogenesis and reducing CH 4 emissions [ 46 ]. Beyond diet, farm management practices such as precision feeding, controlled grazing, and strategic supplementation can also influence microbial communities. Precision feeding strategies that optimize fiber and protein balance can reduce hydrogen accumulation and CH 4 formation [ 49 ]. Future studies should investigate the long-term effects of such interventions on microbial networks and rumen functionality. Functional annotation of archaeal microbiota observed pathways associated with CH 4 production, which was enriched in “Methane metabolism”. This result provides an insight into the direct role of archaea in the rumen ecosystem. In contrast, bacterial ASVs exhibited different functional structures, including \"Transporters\", \"Ribosome\", and \"DNA repair and recombination proteins\". These differences reflect the complementary roles of bacteria and archaea in the rumen, where bacteria contribute to the breakdown and fermentation of complex carbohydrates, providing precursors like hydrogen and SCFAs for archaeal methanogens [ 17 , 34 , 39 ]. Interestingly, despite the variability in taxonomic composition, the rumen microbiota's functional pathways appear conserved. Both the general bacterial ASVs, the WGCNA MEblue or MEbrown ASVs, and their hub ASVs were enriched in KEGG level 3 pathways \"Transporters\", \"Ribosome\", and \"DNA repair and recombination proteins\", which belong to level 2 pathways \"Translators\", \"Membrane transport\", and “Replication and repair” (Fig. 4 , Supplementary Fig. 6). Our functional prediction results based on bacterial ASVs, WGCNA MEblue and MEbrown ASVs were similar. This similarity suggests functional redundancy among different bacterial taxa in the rumen, ensuring the stability and efficiency of microbial processes essential for host energy metabolism [ 16 , 27 , 50 , 51 ]. Functional redundancy is a well-documented phenomenon in microbial communities and is thought to arise from environmental selection for critical biochemical processes, as observed in a recent study on cross-biome microbial networks [ 52 ]. To elucidate the specific roles of these interactions in CH 4 production, future studies could employ metagenomics or metatranscriptomics to identify active metabolic pathways and their gene-level regulation within CH 4 emissions. Unlike SparCC [ 53 ], which focuses on pairwise correlations, WGCNA enables the identification of modules of highly correlated taxa or genes. This network-based approach provides deeper insights into the structure and potential roles of microbial or genes communities [ 17 , 54 ]. Through WGCNA, hub ASVs—such as those annotated to Prevotella brevis , Prevotella ruminicola , Prevotella bryantii , Methanobrevibacter and Methanomassiliicoccus luminyensi were identified as key taxa driving module dynamics and contributing to CH 4 -related metabolic pathways. Additionally, WGCNA excels in its ability to integrate multi-effects data, enabling the associations of microbial communities with external factors such as diet, management practices, or CH 4 emissions [ 29 ]. This comprehensive framework makes WGCNA a powerful tool for investigating complex relationships within microbial ecosystems and linking them to functional and environmental factors. Additionally, our study employed both MiSeq and HiSeq sequencing platforms, two of the leading choices for short-read sequencing. These platforms are considered leading choices for various genomic and microbiome studies due to their robust data output and high-quality sequencing capabilities [ 55 – 57 ]. However, despite their widespread use, the potential influence of platform-specific differences on data analysis remains unclear in the context of our study. A study reported that HiSeq performed longer read sequences and better assigned taxa compared to MiSeq in the context of oral microbiome samples [ 58 ]. Moreover, long-read platforms such as Oxford Nanopore [ 59 ] and PacBio [ 60 ], may provide complementary strengths to short-read technologies by generating longer read lengths that can span complex genomic regions, improve genome assemblies, and resolve ambiguities in repetitive sequences [ 61 ]. While PICRUSt2 enabled functional predictions, its reliance on human-derived databases poses limitations in accurately capturing the functional potential of rumen microbes [ 62 ]. To address this, COwPi [ 63 ]—a rumen microbiome-focused adaptation of PICRUSt [ 64 ]—offers a tailored database specific to the unique microbial communities and metabolic pathways of the rumen. This targeted approach reduces mismatches and improves the accuracy of functional predictions, particularly for KEGG pathway analysis [ 63 ]. However, it is important to note that the original PICRUSt has no longer developed, limiting its applicability to current datasets and novel discoveries in rumen microbiology and impeding the usage of COwPi. Despite these challenges, PICRUSt2 remains a practical and robust tool for predicting microbial functional genes, owing to its enhanced algorithm and extended database support. Incorporating rumen-specific features from CowPi into PICRUSt2 could further refine predictions, offering a comprehensive solution for rumen microbiome research. Despite providing valuable insights into microbial interactions related to CH 4 emissions, our study has several limitations. One key limitation is our reliance on inferred functional data from 16S rRNA gene amplicon sequencing. While predictive tools such as PICRUSt2 offer valuable insights, they lack the resolution of direct metagenomic or metatranscriptomic approaches, which could provide more accurate functional annotations. Integrating multi-omics data, including metabolomics and metaproteomic, would enhance our understanding of the active metabolic pathways driving CH 4 emissions. Another limitation is the absence of longitudinal measurements. Our study provides a snapshot of microbial interactions at a single time point, but microbiota composition and CH 4 emissions fluctuate over time due to factors such as diet changes, lactation stage, and seasonal variations. Longitudinal studies tracking microbial shifts across different feeding regimes and environmental conditions would be crucial to fully understanding the stability and adaptability of CH 4 -related microbial networks. Additionally, while herd-specific factors were considered, more detailed environmental metadata—such as individual feeding behaviors, rumination time, and precise nutrient intake—could further clarify the external influences on microbial community structures."
} | 4,027 |
40069804 | PMC11899718 | pmc | 3,913 | {
"abstract": "Background Methane emissions from livestock, particularly from dairy cattle, represent a significant source of greenhouse gas, contributing to the global climate crisis. Understanding the complex interactions within the rumen microbiota that influence methane emissions is crucial for developing effective mitigation strategies. Results This study employed Weighted Gene Co-expression Network Analysis to investigate the complex interactions within the rumen microbiota that influence methane emissions. By integrating extensive rumen microbiota sequencing data with precise methane emission measurements in 750 Holstein dairy cattle, our research identified distinct microbial communities and their associations with methane production. Key findings revealed that the blue module from network analysis was significantly correlated (0.45) with methane emissions. In this module, taxa included the genera Prevotella and Methanobrevibactor , along with species such as Prevotella brevis, Prevotella ruminicola, Prevotella baroniae, Prevotella bryantii, Lachnobacterium bovis, and Methanomassiliicoccus luminyensis are the key components to drive the complex networks. However, the absence of metagenomics sequencing is difficult to reveal the deeper taxa level and functional profiles. Conclusions The application of Weighted Gene Co-expression Network Analysis provided a comprehensive understanding of the microbiota-methane emission relationship, serving as an innovative approach for microbiota-phenotype association studies in cattle. Our findings underscore the importance of microbiota-trait and microbiota-microbiota associations related to methane emission in dairy cattle, contributing to a systematic understanding of methane production in cattle. This research offers key information on microbial management for mitigating environmental impact on the cattle population. Supplementary Information The online version contains supplementary material available at 10.1186/s42523-025-00386-z.",
"conclusion": "Conclusions This study employed WGCNA to investigate the interactions within the rumen microbiota and their associations with CH 4 emissions using a population-level of 750 Holstein cows. The MEblue was significantly correlated with CH 4 emission, revealing the critical roles of taxa such as Prevotella , Methanobrevibacter , and Methanomassiliicoccus luminyensis . These taxa underscore the complex interactions in carbohydrate fermentation and methanogenesis, key processes contributing to CH₄ production. Additionally, MEbrown was strongly associated with herd factors, revealing microbial networks influenced by farm management practices, diet composition, and feeding strategies. Functional predictions emphasized the complementary roles of bacteria and archaea in the rumen ecosystem, where bacteria provide substrates such as short-chain fatty acids (SCFAs) and hydrogen for methanogenic archaea, which are enriched in pathways linked to CH 4 production. The present study highlights the microbiota-trait and microbiota-microbiota interactions related to CH 4 emission in dairy cattle, contributing to a systematic understanding of CH 4 production in cattle and offering key information on microbial management for mitigating environmental impact in cattle population.",
"discussion": "Discussion Based on the WGCNA results, MEblue was the most significant module, correlating moderately with CH 4 emissions (r = 0.45, p = 7e-37) (Fig. 2 ). This module includes both methanogenic archaea ( Methanobrevibacter and Methanomassiliicoccus luminyensis ) and carbohydrate-fermenting bacteria ( Prevotella brevis, Prevotella ruminicol, Prevotella ruminicola, and Prevotella baroniae) (Fig. 3 ), forming a metabolic network that contribute to CH 4 emissions. Methanobrevibacter and Methanomassiliicoccus luminyensis are well-known for their roles in methanogenesis [ 32 ]. Both Methanobrevibacter and Methanomassiliicoccus luminyensis are methanogenic archaea that contribute to CH 4 production in rumen, utilizing hydrogen and CO 2 (hydrogenotrophy) or methylated compounds (methylotrophy) to produce CH 4 [ 33 , 34 ]. Notably, Methanomassiliicoccus luminyensis specially requires hydrogen as an electron donor, reducing methanol, methylamines into CH 4 [ 35 ]. The co-occurrence of Methanobrevibacter and Methanomassiliicoccus luminyensis within MEblue suggests functional interactions, where Methanobrevibacter helps maintain low hydrogen partial pressure in the rumen, indirectly supporting Methanomassiliicoccus by creating favorable conditions for its methylotrophic methanogenesis [ 36 ]. In addition to archaea, MEblue also includes bacteria from genus Prevotella , particularly networks among Prevotella brevis - Prevotella ruminicola - Prevotella baroniae , Prevotella brevis-Lachnobacterium bovis, Prevotella ruminicola - Prevotella bryantii, and Prevotella ruminicola - Prevotella brevis- Barnesiella viscericola , as well as Paraprevotella - xylaniphila within MEblue (Fig. 3 ). The interactions between Prevotella brevis , Prevotella ruminicola , and Prevotella baroniae suggest synergistic relationships that enhance polysaccharide breakdown and hydrogen production, both of which are crucial for CH 4 production [ 37 ]. These bacteria was highly abundant and co-clustered, emphasizing their role in carbohydrate fermentation, converting into short-chain fatty acids (SCFAs) [ 38 , 39 ]. Among SCFAs, acetate indirectly contributes to CH 4 production by methanogenic archaea. While acetate can serve as a substrate for methanogenesis in specific contexts, though its contribution to ruminal methane production is minimal due to the dominance of hydrogenotrophic and methylotrophic pathways and the rapid rumen passage rate, which limits acetogenic methanogens [ 39 ]. Furthermore, the cross-phylum microbial interaction ( Prevotella brevis - Lachnobacterium bovis ) is a suggests a synergistic metabolic network, where fermentative bacteria generate hydrogen as a byproduct, which is then utilized by methanogenic archaea for CH 4 production. Additionally, Lachnobacterium bovis , which was found to interact with Prevotella brevis , produces intermediates like lactate and acetate, further enhancing hydrogenotrophic pathways that contribute to CH 4 emissions [ 19 ]. Our results align with previous studies [ 17 , 38 ] but provide more detailed insights into microbial interactions at the species level. Furthermore, these findings reinforce that MEblue represents a functionally cohesive module linked to CH₄ emissions through hydrogen-mediated microbial interactions. While MEblue is significantly associated with CH 4 , other microbial modules also influence methane metabolism under different environmental conditions. For instance, MEbrown, another major module, contains similar fermentative bacteria ( Prevotella brevis and Prevotella ruminicola ), which are well-known contributors to carbohydrate fermentation and SCFAs production [ 39 ]. However, MEbrown networks exhibited greater diversity in bacterial taxa, including Flavonifractor plautii , Parabacteroides merdae , Barnesiella viscericola , and Alistipes putredinis (Supplementary Fig. 5). Unlike MEblue, which is moderately linked to CH 4 emissions, MEbrown is more influenced by herd effects (r = 0.43, p = 4e-34) (Fig. 3 ), suggesting that herd-specific factors such as feeding practices and diet composition [ 40 ]. Also, Flavonifractor plautii contributes to flavonoid degradation [ 41 ], influencing microbial dynamics and contributing to SCFAs production. PCA analysis (Supplementary Fig. 2) further revealed that the herd-specific factors significantly shaped microbial community structure, with lactation effects being less pronounced. This may be partly due to the fact that our study was conducted in commercial dairy cattle farms where cattle were reared with potentially different management conditions. Variations in feed composition, fiber content, and nutritional balance across farms likely contributed to the microbial diversity within MEbrown. This underscores the importance of dietary and environmental factors in modulating microbial networks and their roles in CH 4 emissions [ 40 ]. Understanding these environmental influences can help develop targeted intervention strategies to manipulate microbial communities for CH 4 mitigation. Since MEblue represents a microbial network closely associated with CH 4 production, dietary and management interventions targeting this module could be effective in reducing methane emissions. Recent studies reported that strategies such as dietary modifications, probiotics, and feeding additives can affect the CH 4 emissions by altering microbial community structure and metabolic pathways [ 42 – 44 ]. Probiotics can modulate gastrointestinal microbial. Their colonization in rumen improves feed efficiency, potentially reducing CH 4 emissions [ 23 ]. Similarly, dietary interventions, including high-lipid diets [ 45 ], nitrate supplementation [ 46 ], and plant secondary metabolites (e.g., tannins and saponins) [ 47 , 48 ], have been explored as effective CH 4 mitigation strategies. Lipid supplementation can suppress methanogens by reducing hydrogen availability, while tannins can directly inhibit methanogens activity [ 45 ]. Furthermore, dietary nitrate supplementation provides an alternative hydrogen sink, outcompeting methanogenesis and reducing CH 4 emissions [ 46 ]. Beyond diet, farm management practices such as precision feeding, controlled grazing, and strategic supplementation can also influence microbial communities. Precision feeding strategies that optimize fiber and protein balance can reduce hydrogen accumulation and CH 4 formation [ 49 ]. Future studies should investigate the long-term effects of such interventions on microbial networks and rumen functionality. Functional annotation of archaeal microbiota observed pathways associated with CH 4 production, which was enriched in “Methane metabolism”. This result provides an insight into the direct role of archaea in the rumen ecosystem. In contrast, bacterial ASVs exhibited different functional structures, including \"Transporters\", \"Ribosome\", and \"DNA repair and recombination proteins\". These differences reflect the complementary roles of bacteria and archaea in the rumen, where bacteria contribute to the breakdown and fermentation of complex carbohydrates, providing precursors like hydrogen and SCFAs for archaeal methanogens [ 17 , 34 , 39 ]. Interestingly, despite the variability in taxonomic composition, the rumen microbiota's functional pathways appear conserved. Both the general bacterial ASVs, the WGCNA MEblue or MEbrown ASVs, and their hub ASVs were enriched in KEGG level 3 pathways \"Transporters\", \"Ribosome\", and \"DNA repair and recombination proteins\", which belong to level 2 pathways \"Translators\", \"Membrane transport\", and “Replication and repair” (Fig. 4 , Supplementary Fig. 6). Our functional prediction results based on bacterial ASVs, WGCNA MEblue and MEbrown ASVs were similar. This similarity suggests functional redundancy among different bacterial taxa in the rumen, ensuring the stability and efficiency of microbial processes essential for host energy metabolism [ 16 , 27 , 50 , 51 ]. Functional redundancy is a well-documented phenomenon in microbial communities and is thought to arise from environmental selection for critical biochemical processes, as observed in a recent study on cross-biome microbial networks [ 52 ]. To elucidate the specific roles of these interactions in CH 4 production, future studies could employ metagenomics or metatranscriptomics to identify active metabolic pathways and their gene-level regulation within CH 4 emissions. Unlike SparCC [ 53 ], which focuses on pairwise correlations, WGCNA enables the identification of modules of highly correlated taxa or genes. This network-based approach provides deeper insights into the structure and potential roles of microbial or genes communities [ 17 , 54 ]. Through WGCNA, hub ASVs—such as those annotated to Prevotella brevis , Prevotella ruminicola , Prevotella bryantii , Methanobrevibacter and Methanomassiliicoccus luminyensi were identified as key taxa driving module dynamics and contributing to CH 4 -related metabolic pathways. Additionally, WGCNA excels in its ability to integrate multi-effects data, enabling the associations of microbial communities with external factors such as diet, management practices, or CH 4 emissions [ 29 ]. This comprehensive framework makes WGCNA a powerful tool for investigating complex relationships within microbial ecosystems and linking them to functional and environmental factors. Additionally, our study employed both MiSeq and HiSeq sequencing platforms, two of the leading choices for short-read sequencing. These platforms are considered leading choices for various genomic and microbiome studies due to their robust data output and high-quality sequencing capabilities [ 55 – 57 ]. However, despite their widespread use, the potential influence of platform-specific differences on data analysis remains unclear in the context of our study. A study reported that HiSeq performed longer read sequences and better assigned taxa compared to MiSeq in the context of oral microbiome samples [ 58 ]. Moreover, long-read platforms such as Oxford Nanopore [ 59 ] and PacBio [ 60 ], may provide complementary strengths to short-read technologies by generating longer read lengths that can span complex genomic regions, improve genome assemblies, and resolve ambiguities in repetitive sequences [ 61 ]. While PICRUSt2 enabled functional predictions, its reliance on human-derived databases poses limitations in accurately capturing the functional potential of rumen microbes [ 62 ]. To address this, COwPi [ 63 ]—a rumen microbiome-focused adaptation of PICRUSt [ 64 ]—offers a tailored database specific to the unique microbial communities and metabolic pathways of the rumen. This targeted approach reduces mismatches and improves the accuracy of functional predictions, particularly for KEGG pathway analysis [ 63 ]. However, it is important to note that the original PICRUSt has no longer developed, limiting its applicability to current datasets and novel discoveries in rumen microbiology and impeding the usage of COwPi. Despite these challenges, PICRUSt2 remains a practical and robust tool for predicting microbial functional genes, owing to its enhanced algorithm and extended database support. Incorporating rumen-specific features from CowPi into PICRUSt2 could further refine predictions, offering a comprehensive solution for rumen microbiome research. Despite providing valuable insights into microbial interactions related to CH 4 emissions, our study has several limitations. One key limitation is our reliance on inferred functional data from 16S rRNA gene amplicon sequencing. While predictive tools such as PICRUSt2 offer valuable insights, they lack the resolution of direct metagenomic or metatranscriptomic approaches, which could provide more accurate functional annotations. Integrating multi-omics data, including metabolomics and metaproteomic, would enhance our understanding of the active metabolic pathways driving CH 4 emissions. Another limitation is the absence of longitudinal measurements. Our study provides a snapshot of microbial interactions at a single time point, but microbiota composition and CH 4 emissions fluctuate over time due to factors such as diet changes, lactation stage, and seasonal variations. Longitudinal studies tracking microbial shifts across different feeding regimes and environmental conditions would be crucial to fully understanding the stability and adaptability of CH 4 -related microbial networks. Additionally, while herd-specific factors were considered, more detailed environmental metadata—such as individual feeding behaviors, rumination time, and precise nutrient intake—could further clarify the external influences on microbial community structures."
} | 4,027 |
40069804 | PMC11899718 | pmc | 3,914 | {
"abstract": "Background Methane emissions from livestock, particularly from dairy cattle, represent a significant source of greenhouse gas, contributing to the global climate crisis. Understanding the complex interactions within the rumen microbiota that influence methane emissions is crucial for developing effective mitigation strategies. Results This study employed Weighted Gene Co-expression Network Analysis to investigate the complex interactions within the rumen microbiota that influence methane emissions. By integrating extensive rumen microbiota sequencing data with precise methane emission measurements in 750 Holstein dairy cattle, our research identified distinct microbial communities and their associations with methane production. Key findings revealed that the blue module from network analysis was significantly correlated (0.45) with methane emissions. In this module, taxa included the genera Prevotella and Methanobrevibactor , along with species such as Prevotella brevis, Prevotella ruminicola, Prevotella baroniae, Prevotella bryantii, Lachnobacterium bovis, and Methanomassiliicoccus luminyensis are the key components to drive the complex networks. However, the absence of metagenomics sequencing is difficult to reveal the deeper taxa level and functional profiles. Conclusions The application of Weighted Gene Co-expression Network Analysis provided a comprehensive understanding of the microbiota-methane emission relationship, serving as an innovative approach for microbiota-phenotype association studies in cattle. Our findings underscore the importance of microbiota-trait and microbiota-microbiota associations related to methane emission in dairy cattle, contributing to a systematic understanding of methane production in cattle. This research offers key information on microbial management for mitigating environmental impact on the cattle population. Supplementary Information The online version contains supplementary material available at 10.1186/s42523-025-00386-z.",
"conclusion": "Conclusions This study employed WGCNA to investigate the interactions within the rumen microbiota and their associations with CH 4 emissions using a population-level of 750 Holstein cows. The MEblue was significantly correlated with CH 4 emission, revealing the critical roles of taxa such as Prevotella , Methanobrevibacter , and Methanomassiliicoccus luminyensis . These taxa underscore the complex interactions in carbohydrate fermentation and methanogenesis, key processes contributing to CH₄ production. Additionally, MEbrown was strongly associated with herd factors, revealing microbial networks influenced by farm management practices, diet composition, and feeding strategies. Functional predictions emphasized the complementary roles of bacteria and archaea in the rumen ecosystem, where bacteria provide substrates such as short-chain fatty acids (SCFAs) and hydrogen for methanogenic archaea, which are enriched in pathways linked to CH 4 production. The present study highlights the microbiota-trait and microbiota-microbiota interactions related to CH 4 emission in dairy cattle, contributing to a systematic understanding of CH 4 production in cattle and offering key information on microbial management for mitigating environmental impact in cattle population.",
"discussion": "Discussion Based on the WGCNA results, MEblue was the most significant module, correlating moderately with CH 4 emissions (r = 0.45, p = 7e-37) (Fig. 2 ). This module includes both methanogenic archaea ( Methanobrevibacter and Methanomassiliicoccus luminyensis ) and carbohydrate-fermenting bacteria ( Prevotella brevis, Prevotella ruminicol, Prevotella ruminicola, and Prevotella baroniae) (Fig. 3 ), forming a metabolic network that contribute to CH 4 emissions. Methanobrevibacter and Methanomassiliicoccus luminyensis are well-known for their roles in methanogenesis [ 32 ]. Both Methanobrevibacter and Methanomassiliicoccus luminyensis are methanogenic archaea that contribute to CH 4 production in rumen, utilizing hydrogen and CO 2 (hydrogenotrophy) or methylated compounds (methylotrophy) to produce CH 4 [ 33 , 34 ]. Notably, Methanomassiliicoccus luminyensis specially requires hydrogen as an electron donor, reducing methanol, methylamines into CH 4 [ 35 ]. The co-occurrence of Methanobrevibacter and Methanomassiliicoccus luminyensis within MEblue suggests functional interactions, where Methanobrevibacter helps maintain low hydrogen partial pressure in the rumen, indirectly supporting Methanomassiliicoccus by creating favorable conditions for its methylotrophic methanogenesis [ 36 ]. In addition to archaea, MEblue also includes bacteria from genus Prevotella , particularly networks among Prevotella brevis - Prevotella ruminicola - Prevotella baroniae , Prevotella brevis-Lachnobacterium bovis, Prevotella ruminicola - Prevotella bryantii, and Prevotella ruminicola - Prevotella brevis- Barnesiella viscericola , as well as Paraprevotella - xylaniphila within MEblue (Fig. 3 ). The interactions between Prevotella brevis , Prevotella ruminicola , and Prevotella baroniae suggest synergistic relationships that enhance polysaccharide breakdown and hydrogen production, both of which are crucial for CH 4 production [ 37 ]. These bacteria was highly abundant and co-clustered, emphasizing their role in carbohydrate fermentation, converting into short-chain fatty acids (SCFAs) [ 38 , 39 ]. Among SCFAs, acetate indirectly contributes to CH 4 production by methanogenic archaea. While acetate can serve as a substrate for methanogenesis in specific contexts, though its contribution to ruminal methane production is minimal due to the dominance of hydrogenotrophic and methylotrophic pathways and the rapid rumen passage rate, which limits acetogenic methanogens [ 39 ]. Furthermore, the cross-phylum microbial interaction ( Prevotella brevis - Lachnobacterium bovis ) is a suggests a synergistic metabolic network, where fermentative bacteria generate hydrogen as a byproduct, which is then utilized by methanogenic archaea for CH 4 production. Additionally, Lachnobacterium bovis , which was found to interact with Prevotella brevis , produces intermediates like lactate and acetate, further enhancing hydrogenotrophic pathways that contribute to CH 4 emissions [ 19 ]. Our results align with previous studies [ 17 , 38 ] but provide more detailed insights into microbial interactions at the species level. Furthermore, these findings reinforce that MEblue represents a functionally cohesive module linked to CH₄ emissions through hydrogen-mediated microbial interactions. While MEblue is significantly associated with CH 4 , other microbial modules also influence methane metabolism under different environmental conditions. For instance, MEbrown, another major module, contains similar fermentative bacteria ( Prevotella brevis and Prevotella ruminicola ), which are well-known contributors to carbohydrate fermentation and SCFAs production [ 39 ]. However, MEbrown networks exhibited greater diversity in bacterial taxa, including Flavonifractor plautii , Parabacteroides merdae , Barnesiella viscericola , and Alistipes putredinis (Supplementary Fig. 5). Unlike MEblue, which is moderately linked to CH 4 emissions, MEbrown is more influenced by herd effects (r = 0.43, p = 4e-34) (Fig. 3 ), suggesting that herd-specific factors such as feeding practices and diet composition [ 40 ]. Also, Flavonifractor plautii contributes to flavonoid degradation [ 41 ], influencing microbial dynamics and contributing to SCFAs production. PCA analysis (Supplementary Fig. 2) further revealed that the herd-specific factors significantly shaped microbial community structure, with lactation effects being less pronounced. This may be partly due to the fact that our study was conducted in commercial dairy cattle farms where cattle were reared with potentially different management conditions. Variations in feed composition, fiber content, and nutritional balance across farms likely contributed to the microbial diversity within MEbrown. This underscores the importance of dietary and environmental factors in modulating microbial networks and their roles in CH 4 emissions [ 40 ]. Understanding these environmental influences can help develop targeted intervention strategies to manipulate microbial communities for CH 4 mitigation. Since MEblue represents a microbial network closely associated with CH 4 production, dietary and management interventions targeting this module could be effective in reducing methane emissions. Recent studies reported that strategies such as dietary modifications, probiotics, and feeding additives can affect the CH 4 emissions by altering microbial community structure and metabolic pathways [ 42 – 44 ]. Probiotics can modulate gastrointestinal microbial. Their colonization in rumen improves feed efficiency, potentially reducing CH 4 emissions [ 23 ]. Similarly, dietary interventions, including high-lipid diets [ 45 ], nitrate supplementation [ 46 ], and plant secondary metabolites (e.g., tannins and saponins) [ 47 , 48 ], have been explored as effective CH 4 mitigation strategies. Lipid supplementation can suppress methanogens by reducing hydrogen availability, while tannins can directly inhibit methanogens activity [ 45 ]. Furthermore, dietary nitrate supplementation provides an alternative hydrogen sink, outcompeting methanogenesis and reducing CH 4 emissions [ 46 ]. Beyond diet, farm management practices such as precision feeding, controlled grazing, and strategic supplementation can also influence microbial communities. Precision feeding strategies that optimize fiber and protein balance can reduce hydrogen accumulation and CH 4 formation [ 49 ]. Future studies should investigate the long-term effects of such interventions on microbial networks and rumen functionality. Functional annotation of archaeal microbiota observed pathways associated with CH 4 production, which was enriched in “Methane metabolism”. This result provides an insight into the direct role of archaea in the rumen ecosystem. In contrast, bacterial ASVs exhibited different functional structures, including \"Transporters\", \"Ribosome\", and \"DNA repair and recombination proteins\". These differences reflect the complementary roles of bacteria and archaea in the rumen, where bacteria contribute to the breakdown and fermentation of complex carbohydrates, providing precursors like hydrogen and SCFAs for archaeal methanogens [ 17 , 34 , 39 ]. Interestingly, despite the variability in taxonomic composition, the rumen microbiota's functional pathways appear conserved. Both the general bacterial ASVs, the WGCNA MEblue or MEbrown ASVs, and their hub ASVs were enriched in KEGG level 3 pathways \"Transporters\", \"Ribosome\", and \"DNA repair and recombination proteins\", which belong to level 2 pathways \"Translators\", \"Membrane transport\", and “Replication and repair” (Fig. 4 , Supplementary Fig. 6). Our functional prediction results based on bacterial ASVs, WGCNA MEblue and MEbrown ASVs were similar. This similarity suggests functional redundancy among different bacterial taxa in the rumen, ensuring the stability and efficiency of microbial processes essential for host energy metabolism [ 16 , 27 , 50 , 51 ]. Functional redundancy is a well-documented phenomenon in microbial communities and is thought to arise from environmental selection for critical biochemical processes, as observed in a recent study on cross-biome microbial networks [ 52 ]. To elucidate the specific roles of these interactions in CH 4 production, future studies could employ metagenomics or metatranscriptomics to identify active metabolic pathways and their gene-level regulation within CH 4 emissions. Unlike SparCC [ 53 ], which focuses on pairwise correlations, WGCNA enables the identification of modules of highly correlated taxa or genes. This network-based approach provides deeper insights into the structure and potential roles of microbial or genes communities [ 17 , 54 ]. Through WGCNA, hub ASVs—such as those annotated to Prevotella brevis , Prevotella ruminicola , Prevotella bryantii , Methanobrevibacter and Methanomassiliicoccus luminyensi were identified as key taxa driving module dynamics and contributing to CH 4 -related metabolic pathways. Additionally, WGCNA excels in its ability to integrate multi-effects data, enabling the associations of microbial communities with external factors such as diet, management practices, or CH 4 emissions [ 29 ]. This comprehensive framework makes WGCNA a powerful tool for investigating complex relationships within microbial ecosystems and linking them to functional and environmental factors. Additionally, our study employed both MiSeq and HiSeq sequencing platforms, two of the leading choices for short-read sequencing. These platforms are considered leading choices for various genomic and microbiome studies due to their robust data output and high-quality sequencing capabilities [ 55 – 57 ]. However, despite their widespread use, the potential influence of platform-specific differences on data analysis remains unclear in the context of our study. A study reported that HiSeq performed longer read sequences and better assigned taxa compared to MiSeq in the context of oral microbiome samples [ 58 ]. Moreover, long-read platforms such as Oxford Nanopore [ 59 ] and PacBio [ 60 ], may provide complementary strengths to short-read technologies by generating longer read lengths that can span complex genomic regions, improve genome assemblies, and resolve ambiguities in repetitive sequences [ 61 ]. While PICRUSt2 enabled functional predictions, its reliance on human-derived databases poses limitations in accurately capturing the functional potential of rumen microbes [ 62 ]. To address this, COwPi [ 63 ]—a rumen microbiome-focused adaptation of PICRUSt [ 64 ]—offers a tailored database specific to the unique microbial communities and metabolic pathways of the rumen. This targeted approach reduces mismatches and improves the accuracy of functional predictions, particularly for KEGG pathway analysis [ 63 ]. However, it is important to note that the original PICRUSt has no longer developed, limiting its applicability to current datasets and novel discoveries in rumen microbiology and impeding the usage of COwPi. Despite these challenges, PICRUSt2 remains a practical and robust tool for predicting microbial functional genes, owing to its enhanced algorithm and extended database support. Incorporating rumen-specific features from CowPi into PICRUSt2 could further refine predictions, offering a comprehensive solution for rumen microbiome research. Despite providing valuable insights into microbial interactions related to CH 4 emissions, our study has several limitations. One key limitation is our reliance on inferred functional data from 16S rRNA gene amplicon sequencing. While predictive tools such as PICRUSt2 offer valuable insights, they lack the resolution of direct metagenomic or metatranscriptomic approaches, which could provide more accurate functional annotations. Integrating multi-omics data, including metabolomics and metaproteomic, would enhance our understanding of the active metabolic pathways driving CH 4 emissions. Another limitation is the absence of longitudinal measurements. Our study provides a snapshot of microbial interactions at a single time point, but microbiota composition and CH 4 emissions fluctuate over time due to factors such as diet changes, lactation stage, and seasonal variations. Longitudinal studies tracking microbial shifts across different feeding regimes and environmental conditions would be crucial to fully understanding the stability and adaptability of CH 4 -related microbial networks. Additionally, while herd-specific factors were considered, more detailed environmental metadata—such as individual feeding behaviors, rumination time, and precise nutrient intake—could further clarify the external influences on microbial community structures."
} | 4,027 |
40069804 | PMC11899718 | pmc | 3,914 | {
"abstract": "Background Methane emissions from livestock, particularly from dairy cattle, represent a significant source of greenhouse gas, contributing to the global climate crisis. Understanding the complex interactions within the rumen microbiota that influence methane emissions is crucial for developing effective mitigation strategies. Results This study employed Weighted Gene Co-expression Network Analysis to investigate the complex interactions within the rumen microbiota that influence methane emissions. By integrating extensive rumen microbiota sequencing data with precise methane emission measurements in 750 Holstein dairy cattle, our research identified distinct microbial communities and their associations with methane production. Key findings revealed that the blue module from network analysis was significantly correlated (0.45) with methane emissions. In this module, taxa included the genera Prevotella and Methanobrevibactor , along with species such as Prevotella brevis, Prevotella ruminicola, Prevotella baroniae, Prevotella bryantii, Lachnobacterium bovis, and Methanomassiliicoccus luminyensis are the key components to drive the complex networks. However, the absence of metagenomics sequencing is difficult to reveal the deeper taxa level and functional profiles. Conclusions The application of Weighted Gene Co-expression Network Analysis provided a comprehensive understanding of the microbiota-methane emission relationship, serving as an innovative approach for microbiota-phenotype association studies in cattle. Our findings underscore the importance of microbiota-trait and microbiota-microbiota associations related to methane emission in dairy cattle, contributing to a systematic understanding of methane production in cattle. This research offers key information on microbial management for mitigating environmental impact on the cattle population. Supplementary Information The online version contains supplementary material available at 10.1186/s42523-025-00386-z.",
"conclusion": "Conclusions This study employed WGCNA to investigate the interactions within the rumen microbiota and their associations with CH 4 emissions using a population-level of 750 Holstein cows. The MEblue was significantly correlated with CH 4 emission, revealing the critical roles of taxa such as Prevotella , Methanobrevibacter , and Methanomassiliicoccus luminyensis . These taxa underscore the complex interactions in carbohydrate fermentation and methanogenesis, key processes contributing to CH₄ production. Additionally, MEbrown was strongly associated with herd factors, revealing microbial networks influenced by farm management practices, diet composition, and feeding strategies. Functional predictions emphasized the complementary roles of bacteria and archaea in the rumen ecosystem, where bacteria provide substrates such as short-chain fatty acids (SCFAs) and hydrogen for methanogenic archaea, which are enriched in pathways linked to CH 4 production. The present study highlights the microbiota-trait and microbiota-microbiota interactions related to CH 4 emission in dairy cattle, contributing to a systematic understanding of CH 4 production in cattle and offering key information on microbial management for mitigating environmental impact in cattle population.",
"discussion": "Discussion Based on the WGCNA results, MEblue was the most significant module, correlating moderately with CH 4 emissions (r = 0.45, p = 7e-37) (Fig. 2 ). This module includes both methanogenic archaea ( Methanobrevibacter and Methanomassiliicoccus luminyensis ) and carbohydrate-fermenting bacteria ( Prevotella brevis, Prevotella ruminicol, Prevotella ruminicola, and Prevotella baroniae) (Fig. 3 ), forming a metabolic network that contribute to CH 4 emissions. Methanobrevibacter and Methanomassiliicoccus luminyensis are well-known for their roles in methanogenesis [ 32 ]. Both Methanobrevibacter and Methanomassiliicoccus luminyensis are methanogenic archaea that contribute to CH 4 production in rumen, utilizing hydrogen and CO 2 (hydrogenotrophy) or methylated compounds (methylotrophy) to produce CH 4 [ 33 , 34 ]. Notably, Methanomassiliicoccus luminyensis specially requires hydrogen as an electron donor, reducing methanol, methylamines into CH 4 [ 35 ]. The co-occurrence of Methanobrevibacter and Methanomassiliicoccus luminyensis within MEblue suggests functional interactions, where Methanobrevibacter helps maintain low hydrogen partial pressure in the rumen, indirectly supporting Methanomassiliicoccus by creating favorable conditions for its methylotrophic methanogenesis [ 36 ]. In addition to archaea, MEblue also includes bacteria from genus Prevotella , particularly networks among Prevotella brevis - Prevotella ruminicola - Prevotella baroniae , Prevotella brevis-Lachnobacterium bovis, Prevotella ruminicola - Prevotella bryantii, and Prevotella ruminicola - Prevotella brevis- Barnesiella viscericola , as well as Paraprevotella - xylaniphila within MEblue (Fig. 3 ). The interactions between Prevotella brevis , Prevotella ruminicola , and Prevotella baroniae suggest synergistic relationships that enhance polysaccharide breakdown and hydrogen production, both of which are crucial for CH 4 production [ 37 ]. These bacteria was highly abundant and co-clustered, emphasizing their role in carbohydrate fermentation, converting into short-chain fatty acids (SCFAs) [ 38 , 39 ]. Among SCFAs, acetate indirectly contributes to CH 4 production by methanogenic archaea. While acetate can serve as a substrate for methanogenesis in specific contexts, though its contribution to ruminal methane production is minimal due to the dominance of hydrogenotrophic and methylotrophic pathways and the rapid rumen passage rate, which limits acetogenic methanogens [ 39 ]. Furthermore, the cross-phylum microbial interaction ( Prevotella brevis - Lachnobacterium bovis ) is a suggests a synergistic metabolic network, where fermentative bacteria generate hydrogen as a byproduct, which is then utilized by methanogenic archaea for CH 4 production. Additionally, Lachnobacterium bovis , which was found to interact with Prevotella brevis , produces intermediates like lactate and acetate, further enhancing hydrogenotrophic pathways that contribute to CH 4 emissions [ 19 ]. Our results align with previous studies [ 17 , 38 ] but provide more detailed insights into microbial interactions at the species level. Furthermore, these findings reinforce that MEblue represents a functionally cohesive module linked to CH₄ emissions through hydrogen-mediated microbial interactions. While MEblue is significantly associated with CH 4 , other microbial modules also influence methane metabolism under different environmental conditions. For instance, MEbrown, another major module, contains similar fermentative bacteria ( Prevotella brevis and Prevotella ruminicola ), which are well-known contributors to carbohydrate fermentation and SCFAs production [ 39 ]. However, MEbrown networks exhibited greater diversity in bacterial taxa, including Flavonifractor plautii , Parabacteroides merdae , Barnesiella viscericola , and Alistipes putredinis (Supplementary Fig. 5). Unlike MEblue, which is moderately linked to CH 4 emissions, MEbrown is more influenced by herd effects (r = 0.43, p = 4e-34) (Fig. 3 ), suggesting that herd-specific factors such as feeding practices and diet composition [ 40 ]. Also, Flavonifractor plautii contributes to flavonoid degradation [ 41 ], influencing microbial dynamics and contributing to SCFAs production. PCA analysis (Supplementary Fig. 2) further revealed that the herd-specific factors significantly shaped microbial community structure, with lactation effects being less pronounced. This may be partly due to the fact that our study was conducted in commercial dairy cattle farms where cattle were reared with potentially different management conditions. Variations in feed composition, fiber content, and nutritional balance across farms likely contributed to the microbial diversity within MEbrown. This underscores the importance of dietary and environmental factors in modulating microbial networks and their roles in CH 4 emissions [ 40 ]. Understanding these environmental influences can help develop targeted intervention strategies to manipulate microbial communities for CH 4 mitigation. Since MEblue represents a microbial network closely associated with CH 4 production, dietary and management interventions targeting this module could be effective in reducing methane emissions. Recent studies reported that strategies such as dietary modifications, probiotics, and feeding additives can affect the CH 4 emissions by altering microbial community structure and metabolic pathways [ 42 – 44 ]. Probiotics can modulate gastrointestinal microbial. Their colonization in rumen improves feed efficiency, potentially reducing CH 4 emissions [ 23 ]. Similarly, dietary interventions, including high-lipid diets [ 45 ], nitrate supplementation [ 46 ], and plant secondary metabolites (e.g., tannins and saponins) [ 47 , 48 ], have been explored as effective CH 4 mitigation strategies. Lipid supplementation can suppress methanogens by reducing hydrogen availability, while tannins can directly inhibit methanogens activity [ 45 ]. Furthermore, dietary nitrate supplementation provides an alternative hydrogen sink, outcompeting methanogenesis and reducing CH 4 emissions [ 46 ]. Beyond diet, farm management practices such as precision feeding, controlled grazing, and strategic supplementation can also influence microbial communities. Precision feeding strategies that optimize fiber and protein balance can reduce hydrogen accumulation and CH 4 formation [ 49 ]. Future studies should investigate the long-term effects of such interventions on microbial networks and rumen functionality. Functional annotation of archaeal microbiota observed pathways associated with CH 4 production, which was enriched in “Methane metabolism”. This result provides an insight into the direct role of archaea in the rumen ecosystem. In contrast, bacterial ASVs exhibited different functional structures, including \"Transporters\", \"Ribosome\", and \"DNA repair and recombination proteins\". These differences reflect the complementary roles of bacteria and archaea in the rumen, where bacteria contribute to the breakdown and fermentation of complex carbohydrates, providing precursors like hydrogen and SCFAs for archaeal methanogens [ 17 , 34 , 39 ]. Interestingly, despite the variability in taxonomic composition, the rumen microbiota's functional pathways appear conserved. Both the general bacterial ASVs, the WGCNA MEblue or MEbrown ASVs, and their hub ASVs were enriched in KEGG level 3 pathways \"Transporters\", \"Ribosome\", and \"DNA repair and recombination proteins\", which belong to level 2 pathways \"Translators\", \"Membrane transport\", and “Replication and repair” (Fig. 4 , Supplementary Fig. 6). Our functional prediction results based on bacterial ASVs, WGCNA MEblue and MEbrown ASVs were similar. This similarity suggests functional redundancy among different bacterial taxa in the rumen, ensuring the stability and efficiency of microbial processes essential for host energy metabolism [ 16 , 27 , 50 , 51 ]. Functional redundancy is a well-documented phenomenon in microbial communities and is thought to arise from environmental selection for critical biochemical processes, as observed in a recent study on cross-biome microbial networks [ 52 ]. To elucidate the specific roles of these interactions in CH 4 production, future studies could employ metagenomics or metatranscriptomics to identify active metabolic pathways and their gene-level regulation within CH 4 emissions. Unlike SparCC [ 53 ], which focuses on pairwise correlations, WGCNA enables the identification of modules of highly correlated taxa or genes. This network-based approach provides deeper insights into the structure and potential roles of microbial or genes communities [ 17 , 54 ]. Through WGCNA, hub ASVs—such as those annotated to Prevotella brevis , Prevotella ruminicola , Prevotella bryantii , Methanobrevibacter and Methanomassiliicoccus luminyensi were identified as key taxa driving module dynamics and contributing to CH 4 -related metabolic pathways. Additionally, WGCNA excels in its ability to integrate multi-effects data, enabling the associations of microbial communities with external factors such as diet, management practices, or CH 4 emissions [ 29 ]. This comprehensive framework makes WGCNA a powerful tool for investigating complex relationships within microbial ecosystems and linking them to functional and environmental factors. Additionally, our study employed both MiSeq and HiSeq sequencing platforms, two of the leading choices for short-read sequencing. These platforms are considered leading choices for various genomic and microbiome studies due to their robust data output and high-quality sequencing capabilities [ 55 – 57 ]. However, despite their widespread use, the potential influence of platform-specific differences on data analysis remains unclear in the context of our study. A study reported that HiSeq performed longer read sequences and better assigned taxa compared to MiSeq in the context of oral microbiome samples [ 58 ]. Moreover, long-read platforms such as Oxford Nanopore [ 59 ] and PacBio [ 60 ], may provide complementary strengths to short-read technologies by generating longer read lengths that can span complex genomic regions, improve genome assemblies, and resolve ambiguities in repetitive sequences [ 61 ]. While PICRUSt2 enabled functional predictions, its reliance on human-derived databases poses limitations in accurately capturing the functional potential of rumen microbes [ 62 ]. To address this, COwPi [ 63 ]—a rumen microbiome-focused adaptation of PICRUSt [ 64 ]—offers a tailored database specific to the unique microbial communities and metabolic pathways of the rumen. This targeted approach reduces mismatches and improves the accuracy of functional predictions, particularly for KEGG pathway analysis [ 63 ]. However, it is important to note that the original PICRUSt has no longer developed, limiting its applicability to current datasets and novel discoveries in rumen microbiology and impeding the usage of COwPi. Despite these challenges, PICRUSt2 remains a practical and robust tool for predicting microbial functional genes, owing to its enhanced algorithm and extended database support. Incorporating rumen-specific features from CowPi into PICRUSt2 could further refine predictions, offering a comprehensive solution for rumen microbiome research. Despite providing valuable insights into microbial interactions related to CH 4 emissions, our study has several limitations. One key limitation is our reliance on inferred functional data from 16S rRNA gene amplicon sequencing. While predictive tools such as PICRUSt2 offer valuable insights, they lack the resolution of direct metagenomic or metatranscriptomic approaches, which could provide more accurate functional annotations. Integrating multi-omics data, including metabolomics and metaproteomic, would enhance our understanding of the active metabolic pathways driving CH 4 emissions. Another limitation is the absence of longitudinal measurements. Our study provides a snapshot of microbial interactions at a single time point, but microbiota composition and CH 4 emissions fluctuate over time due to factors such as diet changes, lactation stage, and seasonal variations. Longitudinal studies tracking microbial shifts across different feeding regimes and environmental conditions would be crucial to fully understanding the stability and adaptability of CH 4 -related microbial networks. Additionally, while herd-specific factors were considered, more detailed environmental metadata—such as individual feeding behaviors, rumination time, and precise nutrient intake—could further clarify the external influences on microbial community structures."
} | 4,027 |
28223930 | PMC5293783 | pmc | 3,916 | {
"abstract": "Modeling and simulating the neural structures which make up our central neural system is instrumental for deciphering the computational neural cues beneath. Higher levels of biological plausibility usually impose higher levels of complexity in mathematical modeling, from neural to behavioral levels. This paper focuses on overcoming the simulation problems (accuracy and performance) derived from using higher levels of mathematical complexity at a neural level. This study proposes different techniques for simulating neural models that hold incremental levels of mathematical complexity: leaky integrate-and-fire (LIF), adaptive exponential integrate-and-fire (AdEx), and Hodgkin-Huxley (HH) neural models (ranged from low to high neural complexity). The studied techniques are classified into two main families depending on how the neural-model dynamic evaluation is computed: the event-driven or the time-driven families. Whilst event-driven techniques pre-compile and store the neural dynamics within look-up tables, time-driven techniques compute the neural dynamics iteratively during the simulation time. We propose two modifications for the event-driven family: a look-up table recombination to better cope with the incremental neural complexity together with a better handling of the synchronous input activity. Regarding the time-driven family, we propose a modification in computing the neural dynamics: the bi-fixed-step integration method. This method automatically adjusts the simulation step size to better cope with the stiffness of the neural model dynamics running in CPU platforms. One version of this method is also implemented for hybrid CPU-GPU platforms. Finally, we analyze how the performance and accuracy of these modifications evolve with increasing levels of neural complexity. We also demonstrate how the proposed modifications which constitute the main contribution of this study systematically outperform the traditional event- and time-driven techniques under increasing levels of neural complexity.",
"conclusion": "Conclusions The way forward in computational neuroscience lies in the simulation of biologically plausible computational models of different nervous centers (cerebellum, inferior olive, cuneate nucleus, etc.) to better understand how the information is processed within these nervous centers. Computational neuroscience allows the study of these nervous center models without experimental restrictions using neural models that have been developed and validated according to experimental cellular data. These nervous center models can be simulated in different conditions and circumstances to give a consistent idea about how they may operate. In many cases, these models are becoming a fundamental tool in the neuroscience hypothesis-experimentation cycle. The computational models allow researchers to test their hypotheses in simulation. This fact leads to making a better hypothesis and better experiments designed with a greater probability of success. The road to model and simulate nervous centers has been progressively paved with increasing levels of mathematical complexity to include more and more biological features. However, this mathematical complexity comes at a computational cost (i.e., neural accuracy and computational performance). In this paper, we have proposed several new neural dynamic evaluation techniques to cope with the incremental mathematical complexity of well-known neural models (LIF, AdEx, and HH):\n The combined synchronous event-driven integration method combines the look-up tables to minimize the number of look-up table data queries needed to update the neural state variables during the simulation process. Additionally, this method also minimizes the look-up table data queries, making just one prediction about the emission of an output spike for each group of synchronous input spikes that arrive to each neuron. The bi-fixed-step integration method (optimized also in GPU) in which the neural dynamic equations that define the complex neural models (as HH) are accurately solved by switching between two time steps of different lengths during the simulation process. All these integration methods, with their own pros and cons, are meant to be used concurrently to increase the computational performance when simulating heterogeneous SNNs (such as those previously studied in Naveros et al., 2015 ). These heterogeneous SNNs consist of several layers with different neural properties, thus trying to mimic the neural heterogeneity found in different brain regions, such as the cerebellum (D'Angelo et al., 2016 ; Luque et al., 2016 ) or the cuneate nucleus (Bologna et al., 2011 ). The simulation platform used in this study integrates all these neural dynamic evaluation techniques in such a way that parts of the neural network (with low and sparse activity) can be simulated efficiently with event-driven methods (which have been optimized to more efficiently deal with relatively-complex neural models and synchronous activity) and parts of the neural networks (with higher activity in terms of number of spikes) can be simulated with time-driven methods (which have been optimized with bi-fixed-step integration methods and the capability of using highly parallel hardware, such as GPU engines). See Appendix B for a simulation accuracy study of neural networks with combined event- and time-driven methods. Choosing the most appropriate method or combination of methods for each neural center model to be simulated is a trade-off amongst three elements:\n The neural network architecture (number of neurons, neural model complexity, number of input and output synapses, mean firing rates, etc.). The hardware restrictions (number of CPU and GPU cores, RAM size). The simulation requirements and target (minimizing the execution time, maximizing accuracy, etc.). \n Finally, this study has been done using neural networks with medium-low connectivity ratios (from 10 to 1280 input synapses per neuron) oriented to fast simulations. However, the simulation performance results may change significantly when simulating neural networks with larger connectivity ratios (for example 10,000 input synapses per neuron). In this case the spike propagation task is usually more time consuming than the neural dynamics update task for time-driven methods. Nevertheless, as can be seen in Figure 8 , our synchronous event-driven method improves its performance in relation to the direct event-driven method when the number of synapses increases.",
"introduction": "Introduction Artificial neural networks (NNs) have been studied since the early 1940's (Mcculloch and Pitts, 1943 ). These NNs were born as mathematically tractable algorithms that attempted to abstract the learning mechanisms underlying our brain. The natural evolution of these NNs has lately resulted in diverse paradigms including Spiking Neural Networks (SNNs) (Ghosh-Dastidar and Adeli, 2009 ). These SNNs render a higher biological plausibility by bringing the concept of spike-timing into play. The idea behind the spike-timing concept is based on equipping the neural units (neurons) with the capability to emit spikes when their membrane potentials reach a specific dynamic range (firing regime). Leaky integrate-and-fire (LIF) models, for instance, emit spikes when their membrane potentials reach a specific firing threshold. When a spike is fired, it travels from the source neuron to the target neurons. The spike arrivals to the target neurons may increase or decrease their corresponding membrane potentials depending on their synaptic types and synaptic weights. The spike timing, that is, when a spike is either produced or received, constitutes the foundation for processing the neural information in SNNs and is fundamental to understand brain processing based on spike-timing codification. Spiking Neural Networks (SNNs) will be considered as highly parallelizable algorithms in which each neural-processing unit (neuron) sends and receives data (spikes) from other neurons. These SNNs are mainly defined by three key factors:\n The neural model that defines each neural-processing unit (neurons). The neural network topology, that is, how the neural-processing units (neurons) are interconnected. The learning mechanisms that drive adaptation within the SNN at both neural and network level. The parallelizable algorithmic nature of SNNs makes them perfect candidates for being implemented within a wide variety of specific hardware platforms, such as field programmable gate-array circuits (FPGAs) (Ros et al., 2006b ; Agis et al., 2007 ), very large-scale integration circuits (VLSI) (Pelayo et al., 1997 ; Schemmel et al., 2010 ) or specific purpose clusters, such as SpiNNaker (Furber et al., 2013 ) which are better suited for parallel processing. However, the wide-spread availability of general-purpose computers has drifted the SNN algorithmic development effort toward using hardware architectures better suited for sequential processing (Neumann, 1958 ). These general-purpose hardware architectures designed for sequential processing (also for parallel processing in the case of GPUs) do require tailor-made (customized) solutions that allow highly parallelizable SNN algorithms to run efficiently. Two main groups of techniques are traditionally used for simulating the neural units (neurons) of SNNs within general-purpose computers: event-driven and time-driven techniques (Brette et al., 2007 ). Whilst the first technique only computes the neural dynamics of a neuron when it is affected by a spiking-event (generation and propagation of neural activity), the second one iteratively updates the neural dynamics of all neurons in each simulation step time. Both groups have pros and cons (Brette et al., 2007 ) and the best choice depends on the SNN inner features. In this study, we have focused our efforts on developing tailor-made event-driven and time-driven solutions to overcome the architectural and processing computational problems derived from using a general-purpose computer for simulating SNNs. We have studied how the mathematical complexity of several neural models may affect the simulation accuracy and computational performance when different simulation techniques are used over a standard SNN configuration.",
"discussion": "Discussion Throughout this paper, different neural dynamic evaluation techniques are developed. Within the event-driven methods: the combined integration methods based on the combination of look-up tables and the synchronous integration methods based on the optimization of processing synchronous activities. These two integration method are clear improvements with respect to previously described event-driven neural dynamic evaluation techniques (Ros et al., 2006a ). As far as the time-driven methods are concerned, the bi-fixed-step integration methods and the CPU-GPU co-processing significantly increase the performance of time-driven neural dynamic evaluation techniques. The quality level of each proposed integration method is given in terms of neural accuracy and computational performance when simulating three neural models of incremental mathematical complexity (LIF, AdEx, and HH). These neural models are set up (Table 7 ) for reproducing similar activity patterns. All the simulation methods shall provide similar accuracy results to make them comparable. Fixed-step and bi-fixed-step time-driven integration methods for LIF and AdEx models are set up (Table 1 ) for obtaining similar accuracy results than event-driven methods (Figure 5 ). LIF and AdEx models are compiled in look-up tables of 249 and 712 MB, respectively (Figure 4 ). Table 7 Summary of parameters for LIF, AdEx and HH neural models . LIF AdEx HH C 0.19e–9 F C 110 pF C 120 pF E L −0.065 V E L −65 mV E L −65 mV g L 10e–9 S g L 10 nS g L 10 nS V T −0.050 V V T −50 mV V T −52 mV T ref 0.0025 s Δ T 2 mV g Na 20 nS E AMPA 0.0 V τ w 50 ms E Na 50 mV E GABA −0.080 V A 1 nS g Kd 6000 nS τ AMPA 0.005 s B 9 pA E K −90 mV τ GABA 0.010 s V r −80 mV E AMPA 0.0 mV E AMPA 0.0 mV E GABA −80 mV E GABA −80 mV τ AMPA 5 ms τ AMPA 5 ms τ GABA 10 ms τ GABA 10 ms The higher complexity of the HH model imposes a large storage memory capacity. An event-driven HH model with comparable accuracy levels to bi-fixed-step time-driven HH model would require up to 14 GB of storage memory capacity (estimation extrapolated from Figure 4A ). In this benchmark, the HH model has been compiled in look-up tables of 1195 MB that obtain larger accuracy errors results than the equivalent time-driven methods. Event-driven main functional aspects The main functional aspects in relation to the event-driven integration methods can be summarized as follows:\n The number of state variables defining a neural model represents, broadly speaking, the complexity of a neural model. When this number increases linearly, the memory requirements to allocate the pre-compiled look-up tables of the event-driven neural models increases geometrically. Thus, reducing the level of granularity of each dimension is the only way to reduce the total look-up table size, but this reduction directly affects the simulation accuracy (as shown in Figure 4A ). The more complex the neural models are or the smaller the look-up table sizes are, the higher van Rossum distance values (less accuracy) that are obtained. Boundaries in accuracy and memory capacity constrain the maximum neural complexity that these event-driven techniques can handle. The recombination of look-up tables improves the computational performance, maintaining the simulation accuracy. Actually, the combined event-driven integration methods slightly increase the computation time when the neural model complexity increases because the neural state update process of several variables using combined look-up tables is slightly more complex than the update of just one variable. Larger look-up table sizes cause higher rates of cache failures and, therefore, losses in computational performance (see Figure 4 ). This means that the computational performance is more impacted by the total look-up table size than by the mathematical complexity (the number of state variables) of the neural model, although both the mathematical complexity and the look-up table size are related. The computation mechanism used by synchronous methods to deal with synchronous activity significantly improves the computational performance. When a synchronous event-driven neuron receives input synapses coming from other synchronous event-driven neurons or time-driven neurons, the computational performance enhancement depends on either the synchronization period or the integration step size of the previous layers. The larger the synchronization period or the integration step size of the previous layers are, the more synchronous the activity that arrives to the synchronous model and the higher performance levels with respect to the direct non-synchronized integration methods (see Figures 5 – 8 ). Regarding the simulation accuracy, the look-up tables are precompiled maintaining a certain degree of precision. A lager synchronization period only generates a negligible error in the spike generation time which, in turn, causes small oscillations in the van Rossum distance measurements (Figure 5 ). Both neural dynamic evaluation techniques (the combination of look-up tables and synchronization of activity) are simultaneously applied by the combined synchronous event-driven method. This simulation technique outperforms the rest of event-driven techniques. This is more significant when the mathematical complexity of the neural models increases (see Figures 5 – 8 ). The main factor that finally constrains the computational performance of all these event-driven methods is the number of events that need to be processed. These events are mainly internal and propagated spikes (Ros et al., 2006a ) that linearly increase with the neural activity. Time-driven integration methods are preferred rather than event-driven integration methods for those neural networks with high levels of neural activity (see Figures 7 , 8 ). Conversely, there are particular cases in which the event-driven integration methods can be the best option. There are, actually, biologically realistic SNNs in which parts of their inner layers present a very low and sparse neural activity, such as the granular cells in the cerebellum (D'Angelo et al., 2016 ) or the mushroom bodies within the olfactory system in Drosophila (Serrano et al., 2013 ). The importance of these particular networks cannot be overlooked (i.e., just the granular cerebellar layer accounts for half of the neurons of the whole brain, its neurons receive between three and six input synapses with a low and very sparse activity, with most of them remaining silent and barely generating spikes). In these cases, event-driven integration methods perform better than time-driven integration methods. Time-driven main functional aspects The main functional aspects in relation to the time-driven integration methods can be summarized as follows:\n Hybrid CPU-GPU integration methods perform better than CPU methods. This is specifically relevant when the mathematical complexity of the neural models increases. GPU hardware architecture performs better computing parallel tasks than CPU architecture. The computation of the neural dynamics is a pure parallelizable task and consequently, GPU-friendly. In a hybrid CPU-GPU platform, the GPU only processes the neural dynamics, whilst the spike generation and propagation are processed in the CPU. When the mathematical complexity of the neural models increases, the workload assigned to the GPU increases, whilst the workload of the CPU remains equal. For this reason, CPU-GPU neural models perform better than purely CPU neural models, especially when the mathematical complexity of the neural models increases. This increase in performance is shown in Figures 5 – 8 . Bi-fixed-step integration methods outperform fixed-step integration methods for both CPU and GPU platforms when the mathematical complexity of the neural model increases (see Figures 5 – 8 ). Complex neural models usually demand small integration step sizes to better cope with the stiffness of their neural model equations during the spike shape generation. Figures 5E,F show how the maximum step size on a fixed-step integration method is constrained due to the differential equation stiffness (HH model). The adaptation mechanism used by the CPU bi-fixed-step integration methods improves the simulation performance by enlarging the simulation step size during those neural dynamic intervals out of the spike phase. The adaptation mechanism of the integration step size for GPU bi-fixed-step integration methods increases performance thanks to the minimization of the time spent in the synchronization and transfer of data between the CPU and GPU processors. Whilst CPU integration methods are better suited for small-medium groups of neurons (from one neuron to several thousands of neurons, depending on the mathematical complexity), the GPU integration methods are better suited for larger numbers of neurons (from thousands to millions of neurons). The computation time invested in the synchronization period and data transferences between CPU and GPU platforms dominates over the computation time invested in solving the neural dynamics when the number of neurons within the network is small (see Figure 6 ). In this case, the computational performance of the GPU integration methods reaches a plateau. The adaptation mechanism that the bi-fixed-step integration method uses in CPU may decrease the computational performance when the mean firing rate over the neural network is quite high. When the neural activity increases, the ratio of use between the local and global step also increases. The computational workload for the neural dynamic increases and the performance drops (see how the computation time increases in Figure 7 ). EDLUT hybrid architecture into perspective EDLUT is a simulator mainly oriented to efficiently simulate medium-scale neural networks (tens of thousands of neurons) pursuing real time simulations. EDLUT uses point neural models, such as LIF, AdEx or HH. EDLUT information transmission relies on spike timing rather than on the particular spike shape. What matters is when the spike is emitted rather than how the spike is generated. Neurons are just means to an end needed toward understanding the behavior of the neural network behind. The neural communication mechanisms are deployed at network level at very high simulation speeds on a single multicore computer, thus facilitating real time embodiment experiments (Carrillo et al., 2008 ; Luque et al., 2011a , b , 2014a , b , 2016 ; Garrido et al., 2013a ; Casellato et al., 2014 ; Antonietti et al., 2016 ). In these neurorobotic experimental set-ups the neural network and the body are coupled as a single entity. Conversely, NEURON (Hines and Carnevale, 1997 ) is mainly designed for the simulation of very complex and detailed neural models. What matters here is how the spike was generated rather than when it was emitted. Understanding neurons themselves is the goal. To be as biologically plausible as possible, NEURON is conceived to deal with high levels of mathematical complexity that usually require time-driven simulation methods (either fixed- or variable-step integration methods). The computational cost here highly depends on the mathematical complexity which makes the simulation of hundreds or tens of hundreds neurons conforming a network almost computationally intractable. Using NEURON for the benchmark analysis proposed here would be out of context. NEURON, lately, seems to be increasing its field of application toward medium- large-scale neural networks (see Lytton et al., 2016 ) that are comprised of highly simplified neural models (i.e., Izhikevich or the four dimensional HH models). Note that the time-driven simulation techniques here proposed may have a direct impact on NEURON if this tendency is finally consolidated. In contrast, BRIAN (Goodman and Romain, 2009 ) and NEST (Gewaltig and Diesmann, 2007 ) are simulators often considered to be playing in the same league as EDLUT. As is the case with EDLUT, Brian claims to be mainly oriented to efficiently simulate medium-scale neural networks (tens of thousands of neurons) while NEST is designed for very large-scale neural networks (up to 1.86 billion neurons connected by 11.1 trillion synapses on the Japanese K supercomputer; Kunkel et al., 2014 ). These simulators mainly implement point neuron models, although some models with few compartments can be simulated. Similarly, they consider neurons to be just means to an end. They use neurons to understand the behavior of the neural network behind. Both are natively implementing time-driven simulation methods in CPU and particularly BRIAN also implements a hybrid CPU-GPU co-processing scheme for time-driven models. Having said that, the conclusions and approaches proposed in the paper regarding time-driven methods would have a direct impact on Brian and a substantial impact on NEST since CPU-GPU co-processing is still missing. The other fundamental pillar of the methodology proposed here, the event-driven scheme, is not included in BRIAN but it does exist in NEST. Whilst the event-driven EDLUT framework (originally an event-driven scheme) was adapted to also perform time-driven neural simulations (Garrido et al., 2011 ), the time-driven NEST framework (originally a time-driven scheme) was adapted to also perform event-driven neural simulations (Morrison et al., 2007 ; Hanuschkin et al., 2010 ). Thus, both simulators can perform combined event- and time-driven simulations. In fact, NEST proposes an event-driven method that presents similarities to our synchronous event-driven method. Both event-driven methods minimize the number of spike predictions by processing all the synchronous input spikes conjointly and thus make only one prediction."
} | 6,024 |
21667086 | PMC3198195 | pmc | 3,919 | {
"abstract": "Recently discovered microorganisms affiliated to the bacterial phylum NC10, named “ Candidatus Methylomirabilis oxyfera”, perform nitrite-dependent anaerobic methane oxidation. These microorganisms could be important players in a novel way of anaerobic wastewater treatment where ammonium and residual dissolved methane might be removed at the expense of nitrate or nitrite. To find suitable inocula for reactor startup, ten selected wastewater treatment plants (WWTPs) located in The Netherlands were screened for the endogenous presence of M. oxyfera using molecular diagnostic methods. We could identify NC10 bacteria with 98% similarity to M. oxyfera in nine out of ten WWTPs tested. Sludge from one selected WWTP was used to start a new enrichment culture of NC10 bacteria. This enrichment was monitored using specific pmoA primers and M. oxyfera cells were visualized with fluorescence oligonucleotide probes. After 112 days, the enrichment consumed up to 0.4 mM NO 2 − per day. The results of this study show that appropriate sources of biomass, enrichment strategies, and diagnostic tools existed to start and monitor pilot scale tests for the implementation of nitrite-dependent methane oxidation in wastewater treatment at ambient temperature. Electronic supplementary material The online version of this article (doi:10.1007/s00253-011-3361-9) contains supplementary material, which is available to authorized users.",
"introduction": "Introduction Anaerobic nitrite-dependent methane oxidation is a recently discovered process performed by bacteria with doubling times of approximately 1–2 weeks (Raghoebarsing et al. 2006 ; Ettwig et al. 2008 ). The dominant bacteria present in the anaerobic enrichment cultures were members of the NC10 phylum (Ettwig et al. 2009 ; Hu et al. 2009 ). The genome of this dominant bacterium, named “ Candidatus Methylomirabilis oxyfera”, could be assembled from metagenomic data resulting in a 2.7-Mb circular single chromosome which contained genes of both anaerobic and, surprisingly, aerobic metabolic pathways (Ettwig et al. 2010 ). The genome harbored the complete aerobic pathway to oxidize methane, including the pmoCAB operon encoding the particulate methane monooxygenase (pMMO) complex for which recently PCR primers were developed (Luesken et al. 2011 ). Conversely, the denitrification pathway was not complete. Genes nosDFYZ encoding nitrous oxide reductase were missing from the genome and nosL appeared not to be expressed, indicating that another denitrifying pathway had to be operative (Wu et al. 2011 ). Dedicated stable isotope studies showed that this organism could make its own molecular oxygen from nitrite via nitric oxide (Ettwig et al. 2010 ). The produced oxygen was mainly used to oxidize methane in an anaerobic environment according to the expected stoichiometry: \\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}$$ 3\\;{\\text{C}}{{\\text{H}}_4} + 8\\;{\\text{N}}{{\\text{O}}_2}^{ - } + 8\\;{{\\text{H}}^{ + }} \\to 3\\;{\\text{C}}{{\\text{O}}_2} + 4\\;{{\\text{N}}_2} + 10\\;{{\\text{H}}_2}{\\text{O}} $$\\end{document} (Eq. 1) (Wu et al. 2011 ). Anaerobic wastewater treatment compared to conventional aerobic processes has advantages like a reduced production of sludge, a smaller footprint, and the production of biogas (methane) which can be used as an energy source (van Haandel and Lettinga 1994 ; Lema and Omil 2001 ; Aiyuk et al. 2006 ). One of the most established anaerobic techniques is the upflow anaerobic sludge blanket (UASB) first described by Lettinga et al. ( 1980 ). In the absence of oxygen, the microbial community, present in the UASB reactor, degrades organic matter eventually into the main products ammonium and methane (Toerien and Hattingh 1969 ). The produced ammonium should be removed in accordance with the stringent rules for nitrogen compounds in wastewater effluent ( http://ec.europa.eu/environment/water/water-urbanwaste/index_en.html ). Methane contributes to the greenhouse effect when released to the environment and should therefore be removed or used in an energy-efficient way (Cakir and Stenstrom 2005 ; Bogner et al. 2008 ). M. oxyfera -type bacteria could use methane to drive denitrification, circumventing the purchase of electron donors like methanol for nitrogen removal. To obtain sufficient oxidized nitrogen for methane oxidation by M. oxyfera -type bacteria, partial nitrification of ammonium could be used (van Dongen et al. 2001 ). This makes M. oxyfera -type bacteria important candidates for a novel way of sustainable anaerobic wastewater treatment. To date, enrichment cultures of M. oxyfera have been obtained from two different freshwater systems in The Netherlands (Raghoebarsing et al. 2006 ; Ettwig et al. 2009 ). In addition, enrichments of NC10 bacteria were obtained from a mixed sample of freshwater sediment, anaerobic sludge, and return activated sludge (Hu et al. 2009 ). In the present study, we detected M. oxyfera -type bacteria in nine out of ten screened wastewater treatment plants (WWTPs) located in The Netherlands using 16S rRNA gene analysis. One of these WWTPs was selected and biomass was used to start an enrichment culture of M. oxyfera . At the end of the experimental period, this enrichment was capable of nitrite-dependent methane oxidation with conversion rates of 0.3 nmol CH 4 min −1 mg protein −1 and 0.9 nmol NO 2 − min −1 mg protein −1 at ambient temperature.",
"discussion": "Discussion Ten WWTPs were screened with molecular tools to find suitable inocula to start pilot scale tests for application of nitrite-dependent anaerobic methane oxidation. In this study, NC10 specific primers targeting the 16S rRNA gene were used to identify M. oxyfera -type bacteria in the selected WWTPs. It appeared to be that a nested PCR approach was necessary to detect NC10 bacteria in the WWTPs sludge samples. Apparently, it was difficult to amplify the 16S rRNA gene from NC10 bacteria using a direct PCR, probably caused by low amounts of DNA from NC10 bacteria in the tested sludges. This is in contrast with previous studies using freshwater sediment samples, in which a direct PCR with 202F and 1492R resulted in strong PCR bands of the right size and subsequently obtained sequences belonged to the NC10 phylum (Ettwig et al. 2009 ). Alternatively, it could be that the WWTPs screened in this study harbored NC10 bacteria that have more mismatches with the primers used. These primers are based on a limited amount of NC10 sequences presently known, and therefore could miss members of this phylum which have a (slightly) different nucleotide composition at the primer positions. M. oxyfera -type bacteria were not detected in the WWTP in Varsseveld, which may be caused by severe mismatches of the primers or the prevailing conditions in this system. Varsseveld is the only plant in our selection that makes use of a membrane bio reactor (MBR). Before the biological treatment starts in this MBR, the wastewater is filtered thoroughly. Furthermore, to prevent the membranes from clogging, intensive aeration and periodical chemical cleaning are necessary. This might cause an environment not favorable for NC10 bacteria. Sludge from the treatment plant in Lieshout was selected to enrich a nitrite-dependent methane oxidizing culture. After 308 and 315 days of enrichment in the 3-l bioreactor, experiments to determine the conversion rates of methane and nitrite were performed. The observed stoichiometry for methane to nitrite was 3 CH 4 :9.8 NO 2 − (308 days) and 3 CH 4 :10.1 NO 2 − (315 days), which is comparable to the theoretical value 3 CH 4 :8 NO 2 − (Raghoebarsing et al. 2006 ; Ettwig et al. 2009 ). The slight deviation indicated that there was still additional denitrification. This process may have been using other electron donors than methane like organic compounds or ammonium. Similar observations were made in previous studies (Raghoebarsing et al. 2006 ; Hu et al. 2009 ). The measured conversion rates for methane to nitrite in this study are relatively low compared to previous experiments (Raghoebarsing et al. 2006 ; Ettwig et al. 2008 , 2009 ). It is possible that the protein content in the current enrichment was overestimated since organic compounds of activated sludge interfered with the BCA assay (Ras et al. 2008 ). Organic compounds might be present in the Lieshout enrichment, as a remainder of the UASB process. To detect M. oxyfera -type bacteria on a functional level, primers targeting the pmoA gene were used on samples originating from the WWTP Lieshout and the enrichment culture. The new pmoA primer A189_b (Luesken et al. 2011 ) was combined with the widely applied 682R primer (Holmes et al. 1995 ). Using this combination, pmoA sequences were found that did not cluster with M. oxyfera but clustered with M. capsulatus ( Gammaproteobacteria ) or C. polyspora ( Gammaproteobacteria ). This is in accordance with previous results where it was shown that M. oxyfera has some critical mismatches with, especially, the known pmoA reverse primers (Luesken et al. 2011 ). With a nested PCR approach, sequences clustering with the pmoA sequence present in the genome of M. oxyfera were retrieved in all tested samples. The sequences detected in the inoculum and in the WWTP 1 year after taking the inoculum were similar to each other and to the sequence of the M. oxyfera pmoA gene. This may indicate that there is a small but persistent population of NC10 bacteria present in this WWTP. Two different pmoA primer combinations were used in a direct PCR on samples of the enrichment originating from the WWTP Lieshout. After 332 days of enrichment, pmoA sequences could be retrieved. These results implied that when M. oxyfera -type bacteria were enriched, a direct PCR with specific pmoA primers can be used. Conclusively, we showed that nine out of ten selected WWTPs harbor small amounts of M. oxyfera bacteria clustering within the NC10 phylum, using molecular tools for screening and detection of these bacteria. In addition, an enrichment culture (60–70%) from wastewater sludge was obtained performing nitrite-dependent methane oxidation at ambient temperature. The M. oxyfera -type bacteria present in this enrichment were identified and monitored using specific pmoA primers. These data suggested that biomass from WWTP systems could be a potential source to start pilot scale reactors with very efficient biomass retention."
} | 2,675 |
29497408 | PMC5818412 | pmc | 3,921 | {
"abstract": "The roots of most terrestrial plants are colonized by mycorrhizal fungi. They play a key role in terrestrial environments influencing soil structure and ecosystem functionality. Around them a peculiar region, the mycorrhizosphere, develops. This is a very dynamic environment where plants, soil and microorganisms interact. Interest in this fascinating environment has increased over the years. For a long period the knowledge of the microbial populations in the rhizosphere has been limited, because they have always been studied by traditional culture-based techniques. These methods, which only allow the study of cultured microorganisms, do not allow the characterization of most organisms existing in nature. The introduction in the last few years of methodologies that are independent of culture techniques has bypassed this limitation. This together with the development of high-throughput molecular tools has given new insights into the biology, evolution, and biodiversity of mycorrhizal associations, as well as, the molecular dialog between plants and fungi. The genomes of many mycorrhizal fungal species have been sequenced so far allowing to better understanding the lifestyle of these fungi, their sexual reproduction modalities and metabolic functions. The possibility to detect the mycelium and the mycorrhizae of heterothallic fungi has also allowed to follow the spatial and temporal distributional patterns of strains of different mating types. On the other hand, the availability of the genome sequencing from several mycorrhizal fungi with a different lifestyle, or belonging to different groups, allowed to verify the common feature of the mycorrhizal symbiosis as well as the differences on how different mycorrhizal species interact and dialog with the plant. Here, we will consider the aspects described before, mainly focusing on ectomycorrhizal fungi and their interactions with plants and other soil microorganisms.",
"conclusion": "Conclusion and Perspectives In conclusion, genomic and transcriptomic sequencing projects starting with the first mycorrhizal genome sequencing (i.e., that of L. bicolor ) have allowed the identification of the common core of ECM symbiosis-related genes, as determinants of the symbiotic lifestyle, as well as the identification of species-specific traits. However, genome sequencing is only the first step to obtain information on how an organism interacts with the environment and with other organisms. The combination of functional, structural, cellular, and bioinformatics approaches is providing knowledge on the function of genes/proteins and permits to reconstruct the pathways of an organism in specific growth conditions, and in its natural environment. In fact, metagenomics, metatranscriptomics and metaproteomics studies are currently and fast providing a powerful mean for the analysis of environmental microorganisms without the need of culturing them. At the question: “can -omics provide insight into microbial ecology that cannot be achieved using traditional methods?”, Jansson and Prosser (2013) and Prosser (2015) reply that although -omics generate a large amount of ‘big data’ experiments that test hypotheses on microbe-environment associations may allow more direct identification and analysis of the ecological processes. The large volume of sequence data involved in the ECM symbiosis, provide a reference database for an estimation of the ECM fungal taxa number and their ecology. The next crucial research will be linking molecular and metabolic data to key processes such as the exchange of nutrients, the plant protection against stresses and diseases and the genes responsible of the symbiosis.",
"introduction": "Introduction The roots of most terrestrial plants are colonized by mycorrhizal fungi. They play a key role in terrestrial environments providing to plants an improvement in mineral nutrient uptake and earning in return carbon compounds ( Brundrett, 2009 ). Mycorrhizal interactions are usually classified on the basis of the features of the symbiotic interfaces and of the taxonomic identity of the plant and fungal symbionts ( Smith and Read, 2008 ). Among mycorrhizal symbioses (see van der Heijden et al., 2015 for a review), ectomycorrhizae are established by the mycelia of fungi almost exclusively belonging to the so called “higher fungi,\" i.e., Basidiomycetes and Ascomycetes, whose ecological strategies have been revisited by Tedersoo and Smith (2013) . Ectomycorrhizal (ECM) fungi are present all over the world, and their host plants include most angiosperm and gymnosperm trees, as well as shrubs ( Bonfante, 2010 ). Some ECM plants are economically important timber-producing tree species, while some ECM fungi are represented by the economically important truffles and porcini ( Mello et al., 2015a ). Between plant and soil there is a very specific environment, the ectomycorrhizosphere, in which diverse communities of microorganisms – fungi and bacteria – interact. It is known that ECM fungi have a key role in nitrogen cycling, particularly in boreal and temperate forests, and that they can help their host plants to tolerate abiotic stresses. ECM assemblages provide benefits for inorganic nitrogen uptake under environmental constraints, through stress activation of distinct ECM fungal taxa. This suggests that these taxa are functionally diverse and opens new opportunities to characterize the ECM fungal identities ( Pena and Polle, 2014 ). Furthermore, Gehring et al. (2017) demonstrated that tree genetics determines fungal partner communities that confer drought tolerance, highlighting the interlinked importance of the genetics of a tree and its microbiome. The development of an ECM symbiosis requires morphological changes in the two partners, to allow the formation of the symbiotic structures, through the regulation of several genes ( Martin et al., 2007 ; Kohler et al., 2015 ). From the first work in which cDNA arrays were used to study gene expression in the ECM symbiosis between Eucalyptus globulus and Pisolithus tinctorius ( Voiblet et al., 2001 ), important progress has been done in the comprehension of the mechanisms involved in the ectomycorrhiza development. Information on the functional diversity of the ECM interactions has been highlighted, leading to the discovery of many genes coding for plant/fungus symbiosis-regulated proteins. Among them, several mycorrhiza-induced small-secreted proteins (MiSSPs) that may act as effectors and are required for symbiosis establishment have been identified ( Plett et al., 2011 , 2014a ; Kohler et al., 2015 ; Martin et al., 2016 for a review). Additionally, Pellegrin et al. (2015) showed, through a bioinformatics pipeline, that the secretome of ECM fungi is enriched in SSPs in comparison to other species with a different life style. Shared- and lifestyle-specific SSPs have been identified in saprotrophic and ECM fungi, and the ECM-specific SSPs could be a signature of the ECM symbiosis lifestyle. This would suggest they have a role in a molecular dialog with host plants, leading to the formation of a functional ectomycorrhiza ( Pellegrin et al., 2015 ; Garcia and Ané, 2016 for a commentary). Despite similar anatomical patterns, the sequenced ECM genomes showed that differences are present in symbiosis regulated genes, revealing a diversity in the manner by which symbiotic fungi interact with their partners and suggesting the use of different molecular toolboxes to dialog with the host plant ( Martin et al., 2008 , 2010 ; Kohler et al., 2015 ; Peter et al., 2016 ). Remarkably, the role of MiSSPs (such as MiSSP7) to control host plant defense reactions has been elegantly demonstrated in Laccaria bicolor and Populus trichocarpa interaction ( Plett et al., 2011 , 2014b ), while such fungal effectors have not been found among the upregulated transcripts in Tuber melanosporum ECMs ( Martin et al., 2010 ), suggesting that different mechanisms may be involved in the development and maintaining of the ECM symbiosis. Transcript profiling of ECM roots from different plant/fungus interactions suggests that similar functional gene categories appear to be up-regulated, although these genes are not the same in the several ECM fungal species ( Kohler et al., 2015 ). The availability of more genome sequences from ECM fungi also confirms that they have a reduced set of genes encoding plant cell wall degrading enzymes (PCWDEs) ( Kohler et al., 2015 ), as already suggested from the genome sequence of the first two sequenced mycorrhizal fungi, i.e., L. bicolor and the black truffle T. melanosporum , respectively ( Martin et al., 2008 , 2010 ). In addition to genomic features and transcriptomic profiles, epigenetic variation is considered an important player in the evolution of biological diversity, and epigenetic regulatory systems have an important role in the response to environmental stimuli and stress factors ( Zhong, 2016 ). The availability of the genome sequences from several fungal species will allow the understanding on how DNA methylation regulatory components are evolved in ECM fungi, and the role of the epigenetic mechanisms to cope with different environmental conditions through modifications of gene expression mediated by DNA methylation and transposon activity profiles. Considering that DNA methylation in fungi lead to transposable elements (TEs) silencing, comparative methylome and transcriptome analyses have been performed in a TE-rich organism such as the ECM fungus T. melanosporum ( Chen et al., 2014 ), suggesting that a reversible methylation mechanism functions in truffles to cope with the multitude of TEs present in its genome. Information derived from these analyses, whether extended to individuals from different geographical areas, may also provide a new tool to explain intraspecific variability and adaptation to different environments and, in the case of truffles, commercially organoleptic properties (i.e., aroma). An ECM root is a complex organ, formed not only by two individuals, plant and fungus, but also by two fungal pseudotissues: the mantle (i.e., the sheat), which develops outside the root, and the Hartig net, which colonizes the apoplastic space between root cells ( Balestrini et al., 2012 ; Balestrini and Kottke, 2016 ). The two ECM compartments are thought to be functionally different. This has been first demonstrated by a study on Amanita muscaria ectomycorrhizae, where the mantle was manually dissected from the ectomycorrhizal root, revealing a differential expression for two fungal genes coding for a phenylalanine ammonium lyase ( AmPAL ) and a hexose transporter ( AmMst1 ) ( Nehls et al., 2001 ). While the first ( AmPAL ) was mainly expressed in the mantle, the expression of AmMst1 increased in the Hartig net. More recently, the combination of a laser microdissection (LMD) approach, which allows the collection of the two ECM fungal compartments, with microarray gene expression analysis, revealed a specificity in the transcript profiles, reflecting a functional specificity for these two ECM compartments, e.g., that the mantle is the responsible for the mineral elements (i.e., nitrogen) and water uptake from soil, while the expression of several transporters is enhanced in the Hartig net ( Hacquard et al., 2013 ). In the last years, different reviews have been focused on the molecular signals (mechanisms) underlying the ECM development and functioning ( Garcia et al., 2015 ; Martin et al., 2016 ). A role of flavonoids and hormones in the signaling pathway during the early stages of the ECM establishment has been proposed since several years ( Garcia et al., 2015 and references therein). More recently, two plant flavonoids have been suggested to trigger the expression of a fungal effector (MiSSP7, see below) in L. bicolor ( Plett et al., 2014a ). It has been also reported that accumulation at the root apex and redistribution of auxin, which is a hormone produced by both the symbiotic partners, may play a role to stimulate lateral short root development required for the ECM formation ( Felten et al., 2009 ). Martin et al. (2016) have recently speculated that secreted fungal MiSSPs may interact with auxin, gibberellin, and salicylate receptors to alter root development. Moreover, increased concentrations of ethylene and jasmonic acid repressed fungal colonization, with an impact on the development of the ECM roots ( Plett et al., 2014b ). However, the effective involvement of plant hormones in ECM establishment and maintenance has to be still fully elucidated ( Garcia et al., 2015 ). Here, some specific aspects related to the biology and ecology of the ECM fungi will be considered, starting with their in situ dynamics to the symbiotic interface creation, before and after their genome sequencing and the advent of the environmental genomics. Given that truffles are of high economic interest, crossing several research fields ranging from taxonomy to truffle cultivation, and are the first edible ECM fungi whose genome has been sequenced, extensive research has been focused on them in order to understand their life cycle and thus to increase their production. For this reason, particular attention will be given to some insights highlighted by the sequencing of the black truffle T. melanosporum genome."
} | 3,339 |
37293376 | PMC10246831 | pmc | 3,923 | {
"abstract": "Abstract Freezing phenomenon has troubled people for centuries, and efforts have been made to lower the liquid freezing temperature, raise the surface temperature, or mechanical deicing. Inspired by the elytra of beetle, we demonstrate a novel functional surface for directional penetration of liquid to reduce icing. The bionic functional surface is fabricated by projection microstereolithography (P µ SL) based three dimensional printing technique with the wettability on its two sides tailored by TiO 2 nanoparticle sizing agent. A water droplet penetrates from the hydrophobic side to the superhydrophilic side of such a bionic functional surface within 20 ms, but it is blocked in the opposite direction. Most significantly, the penetration time of a water droplet through such a bionic functional surface is much shorter than the freezing time on it, even though the temperature is as low as −90°C. This work opens a gate for the development of functional devices for liquid collection, condensation, especially for hyperantifogging/freezing.",
"conclusion": "Conclusion In conclusion, the near-to-zero Laplace force on the superhydrophilic side of a bionic functional surface (the other side is hydrophobic) with regularly arranged microholes enabled the ultrafast unidirectional penetration of a water droplet from its hydrophobic side to superhydrophilic side within 15 ms, which is shorter than the freezing time of a water droplet on such a bionic functional surface if the temperature is not lower than −90°C. Our bionic functional surface paves a new way for designing hyperantifreezing functional surfaces and liquid diode.",
"introduction": "Introduction In nature, water becomes ice at subzero temperature within 100 s ( 1 , 2 ), and such a phenomenon brings us a lot of inconvenience and disasters in the past thousands of years ( 3–5 ). After hundreds of years’ study, anti-icing can be achieved in passive and active ways ( 6 ) to inhibit ice nucleation, hinder ice crystal growth, and reduce ice adhesion, as well as water removal before icing ( 7 ). Typical passive anti-icing is ambiguous ( 8 ), including slippery surfaces with low energy of adhesion ( 9 , 10 ) and an ultralow ice nucleation temperature ( 11 ), as well as droplets bouncing away below the freezing point ( 12 ). Among them, slippery surfaces are achieved based on the microstructures or the surfaces themselves which contribute to the easy removal of liquid/ice, including superhydrophobic surfaces to reduce the adhesion of liquid ( 13–15 ), icephobic surfaces for easy removal of the ice ( 16 , 17 ), etc. Active anti-icing is induced by an external stimulus that induces magnetism ( 18 , 19 ), optothermal ( 20 , 21 ), and electric heating ( 22 , 23 ) that raise the temperature of these surfaces for preventing icing and melting the existing ice by using much energy. In nature, plants and animals have evolved over millions of years in a natural way that is not only faultlessly adapted to nature but also close to perfection. Scientists attempt to model the function of plants and animals in terms of characteristic and advancing designs, creative thinking and understanding of the principles of various physical processes that provide a bridge between biology and technology, which help us to solve technical problems ( 24 , 25 ). By reproducing the principles of biology with advancing means, people will not only break through existing bottlenecks but also perfectly solve the technical problems at the same time. Generally, bionic superhydrophobic surfaces (i.e. water contact angle (CA) >150° and sliding angle less than 10°) inspired by lotus leaf ( 26–28 ), butterfly wing ( 29 , 30 ), and antarctic scallop ( 31 ) are commonly used in those anti-icing phenomena ( 32 ). Inspired by the elytra of beetle, we herein propose a uniquely mimicked functional membrane with numerous microholes. The bionic porous membrane was manufactured by the projection microstereolithography (P µ SL) based three dimensional (3D) printing technique ( 33–36 ) precisely. One surface of the bionic porous membrane exhibited superhydrophilicity treated with TiO 2 nanoparticle sizing agent, while the other original surface was hydrophilic that induced the different Laplace pressures. Hence, the bionic porous membrane with asymmetric wettability enables the unique unidirectional microfluidic performance. Most significantly, liquid can penetrate the bionic porous membrane before icing that enables passive anti-icing even at −90°C. The proposed 3D printed bionic porous membrane promise applications in water resistible and breathable equipment ( 37 , 38 ), ant-icing devices ( 13 , 39 ), etc.",
"discussion": "Discussion Unidirectional fluidic performance for bionic functional surfaces The unidirectional microfluidic performance of our bionic functional surfaces is schematically shown in Fig. 2 A, during which the capillary force formed by the meniscus of liquid in the microholes within the bionic functional surface plays a crucial role. At the beginning, a deionized (DI) water droplet goes inside of the microholes because of the meniscus of liquid facing downward (Fig. 2 A-i). Then, the capillary force accelerates the DI water passing through the microholes (Figs. 2 B and C). However, the Laplace force dominates the further movement of the DI water upon reaching the upper openings of those microholes inside of the bionic functional surface (Online Supplementary Appendix, Fig. S8 ), \n (1) \n Δ p = 4 γ sin ( θ + α ) D \n where γ is the surface tension of the liquid, and D is the diameter of the microholes. θ is the static CA between a printed surface and a liquid droplet, and α is indicated in Fig. 2 B. In addition, the maximum Laplace pressure appears when θ reaches the advancing CA ( θ a ) of the bionic functional surface ( 41 ), \n (2) \n Δ p max = 4 γ sin θ a D \n Fig. 2. The ultrafast unidirectional penetration of water through a bionic functional surface. A) Schematic of unidirectional liquid penetration performance of a bionic functional surface. B) Capillary rise of water inside of a microhole. C) Capillary force inside a microhole. D) A droplet penetrates the bionic functional surface from the hydrophobic side to the superhydrophilic side. E) A water droplet is blocked on the superhydrophilic surface. F) Laplace pressure on the two sides of the bionic functional surface. G) The influence of membrane apertures on the penetration time through the bionic functional surface. The θ a of a superhydrophilic surface is around 14°, leading to a fact that the Laplace force is not large enough to decrease the velocity of the water to 0 before reaching its maximum value because of the inertia of the water (Fig. 2 A-ii and B). Then, the water will come out from the openings on the superhydrophilic surface (Fig. 2 D, Online Supplementary Appendix, Fig. S9 and Movie S1 ). In contrast, the DI water also comes into microholes from the superhydrophilic surface, reaching the bottom openings with a relatively high velocity. Similarly, the backward Laplace force decreases the velocity of the water gradually. Different from the situation on the superhydrophilic surface, the Laplace pressure is large enough to decrease the velocity of the water to 0 before the θ reaching the θ a (Fig. 2 A-ii). At last, the water will be forbidden to pass through the bionic functional surface when the water droplet is dropped on the superhydrophilic surface (Fig. 2 E, Online Supplementary Appendix, Fig. S10 and Movie S2 ). Moreover, the unidirectional penetration of water cannot be achieved with a superhydrophobic/superhydrophilic CA pair, a hydrophobic/superhydrophobic CA pair, or a hydrophobic/hydrophobic one (Online Supplementary Appendix, Figs. S11 and S12 and Movies S3–S5 ). The size of the droplet makes a difference in its CA (Online Supplementary Appendix, Fig. S13 ), and the CA reaches the maximum when the width of the microholes is 600 μm with a 2 μl droplet. Similarly, the size of the microholes also greatly influences the CA of the membrane, and the maximum CA appears when the width of microholes is 600 μm (Online Supplementary Appendix, Fig. S14 ). The theoretical Laplace pressure difference ( Δ p ) on the two sides of a bionic functional surface decreases with the increase of the microhole's size (Fig. 2 F), leading to a fact that the penetration time of a water droplet passing through a bionic functional surface decreases with the increase of the microhole's size (Fig. 2 G, Online Supplementary Appendix, Fig. S15 and Movie S6 ). In addition, the size of water droplets (Online Supplementary Appendix, Fig. S16 ), the length of the microholes covered with TiO 2 nanoparticles (Online Supplementary Appendix, Fig. S17 ), and the surface tension of the liquid (Online Supplementary Appendix, Figs. S18 and S19 and Movies S7 and S8 ) also make a big difference on the unidirectional microfluidic performance of our bionic functional surfaces. However, it should be pointed out that a bionic functional surface with large microholes might fail if ethyl alcohol is used as the working fluid because of its low adhesion force significantly affects the Laplace pressure at the openings of the microholes (Online Supplementary Appendix, Fig. S20 and Movie S9 ). Lattice Boltzmann method simulation of the unidirectional water transportation through bionic functional surfaces The unidirectional transport of water through bionic functional surfaces is found to be conditional by numerical calculations (Online Supplementary Appendix, Fig. S21 and Method). For the hydrophilic/superhydrophilic CA pair, successful unidirectional penetration of water droplets is found only when the bionic functional surface is placed upward, i.e. the water droplet is transported from the hydrophilic side to the superhydrophilic side (Fig. 3 A and B; Online Supplementary Movies S10 and S11 ). However, when the water droplet is placed on a bionic functional surface with superhydrophilic CA pair, it penetrates through the bionic functional surface from both directions (Fig. 3 C and D; Online Supplementary Appendix, Figs. S22 and S23 ). Fig. 3. Lattice Boltzmann method simulation of the unidirectional water transportation through a bionic functional surface. A) Dynamic process of a droplet permeating on hydrophilic side of a bionic functional surface with θ t = 60° (facing upward). B) Dynamic process of a droplet permeating on hydrophilic side of a bionic functional surface with θ t = 60° (facing downward). C) Dynamic process of a droplet permeating on superhydrophilic side of a bionic functional surface with θ t = 30° (facing upward). D) Dynamic process of a droplet permeating on superhydrophilic side of a bionic functional surface with θ t = 30° (facing downward). E) Effect of θ t on capillary pressure. F) The effect of h b/t on capillary pressure. G) The effect of microholes on capillary pressure. To look into the physics further, we propose a thermodynamic free energy model to calculate the capillary pressure along the penetration direction inside microholes on the bionic functional surface. For a bionic functional surface made of a single layer membrane with rectangular microholes and circular-shaped solid cross-section (Online Supplementary Appendix, Fig. S24 and Method), the local capillary force P c a p ( x ) = d E d V ( 42 ), which is the change in surface energy per volume, is calculated in permeation direction x with the consideration of meniscus shape. P cap ( x ) shares a similar trend of first increasing (due to flow channel geometry) and then decreasing (due to superhydrophilicity at the bottom surface) pattern with respect to x in all cases. The more hydrophilic θ t is, the larger range of x where P cap ( x ) is negative, indicating that it is thermodynamically preferred for the liquid–vapor interface to keep moving in x direction (Fig. 3 E). When θ t is hydrophobic (blue lines), P cap ( x ) = 0 occurs at much smaller x than hydrophilic θ t , which explains the no-penetration behaviors on such surfaces due to lack of penetration propulsion. The drop in P cap ( x ) is caused by the sudden change in intrinsic CA from θ t to θ b . With h b/t being smaller, the sudden decrease in P cap ( x ) occurs earlier (Fig. 3 F). In current settings, P cap ( x ) is always negative as long as h b/t < 70%, and its average value is decreasing with the decrease of h b/t , suggesting large penetration rate rising from great capillary pressure. Similarly, we find that the absolute value of capillary pressure is larger at smaller d a , showing a greater capillary wicking effect to propel liquid penetration (Fig. 3 G). However, it is worth noting that, ultrasmall d a will also bring large flow resistance (not incorporated in capillary pressure model), resulting in a nonmonotonic effect on penetration rate. The hyperantifreezing characteristics of our bionic functional surfaces The freezing time of a 2 μL DI water droplet on a printed flat surface made of the same material and roughness is tens of seconds (Fig. 4 A), which decreases with the decrease of the temperature (Fig. 4 B, and Online Supplementary Appendix, Fig. S25 ). In contrast, a 2 μL water droplet penetrates the bionic functional surface within 96 ms when the surrounding temperature is above −30°C (Figs. 4 A, C, and D; Online Supplementary Movies S12 and S13 ), and the water droplet is not frozen, while a water droplet on a similar membrane without CA difference is frozen around 18 s at −30°C (Online Supplementary Appendix, Fig. S26 ). The penetration speed of a water droplet decreases with the decrease of the surrounding temperatures (Fig. 4 C–G) because the surface tension of water decreases with the decrease of the surrounding temperatures ( 43 ), which reduces the Laplace force at the openings of the superhydrophilic surface. But the time of a water droplet passing through the bionic functional surface is still shorter than the freezing time of the water on it, leading to a fact that the water droplet will not be frozen on such a bionic functional surface even at −90°C (Figs. 4 C–E). In contrast, a small part of water is frozen on the hydrophobic side during the penetration process at −95°C ( Movie S14 ) because the water droplet starts to freeze at the interface of the water droplet and the bionic functional surface before passing through it, which slows down the penetration of the water passing through it. The quite high surface tension of a water droplet increases the penetration time longer than its freezing time on the hydrophobic side at −120°C, the water droplet is frozen to ice before penetrating the bionic functional surface (Fig. 4 G). Fig. 4. The hyperantifreezing performance of a bionic functional surface. A) The schematic diagram of liquid unidirectional penetration on the bionic functional surfaces at extremely low temperatures. B) The influence of temperature on freezing time of a water droplet on a printed flat surface. C–G) The penetration process of a water droplet passing through a bionic functional membrane at −10, −30, −60, −90, and −120°C, respectively."
} | 3,804 |
35495710 | PMC9039739 | pmc | 3,925 | {
"abstract": "The overlap of microbiology and electrochemistry provides plenty of opportunities for a deeper understanding of the redox biogeochemical cycle of natural-abundant elements (like iron, nitrogen, and sulfur) on Earth. The electroactive microorganisms (EAMs) mediate electron flows outward the cytomembrane via diverse pathways like multiheme cytochromes, bridging an electronic connection between abiotic and biotic reactions. On an environmental level, decades of research on EAMs and the derived subject termed “ electromicrobiology ” provide a rich collection of multidisciplinary knowledge and establish various bioelectrochemical designs for the development of environmental biotechnology. Recent advances suggest that EAMs actually make greater differences on a larger scale, and the metabolism of microbial community and ecological interactions between microbes play a great role in bioremediation processes. In this perspective, we propose the concept of microbial electron transfer network (METN) that demonstrates the “species-to-species” interactions further and discuss several key questions ranging from cellular modification to microbiome construction. Future research directions including metabolic flux regulation and microbes–materials interactions are also highlighted to advance understanding of METN for the development of next-generation environmental biotechnology.",
"introduction": "Introduction: Microbial Electron Transfer Network Near one hundred and a half years ago when Thomas Edison devoted himself to the improvement of light bulbs, he would probably never imagine that some of the bulbs could be powered by a bunch of tiny microbes named electroactive microorganisms (EAMs). After a rapid development in the past decade, scientists have contributed worthwhile endeavors to establish the scientific basis of potential microbial electrochemical technologies such as microbial fuel cells or microbial electrolytic cells to deal with various environmental issues ( Zhang and Angelidaki, 2014 ; Gude, 2016 ). Up to this day, the scope of “ electromicrobiology ” is far beyond microbial electron exchange with electrodes, but extended significantly to the microbe-mediated redox reactions between microbes, as well as between microbial cells and the ecosphere, driving both biotic and abiotic natural element cycles ( Nichols et al., 2015 ; Marzocchi et al., 2022 ). For instance, methanogens as ancient microbes were suggested to be responsible for annually producing one billion metric tons of methane in the global carbon cycle, taking micromolecular carbon chemicals (i.e., acetate) as terminal electron acceptors ( Prakash et al., 2019 ). Furthermore, some methanogens were recently evidenced to transfer electrons to exogenous Fe(III) for conserving energy, which was important in the early evolution of respiration ( Lueders and Friedrich, 2002 ; Seckbach, 2004 ; Prakash et al., 2019 ). Beyond that, there are other EAMs unintentionally influencing and shaping our blue planet Earth in totally different ways like mineral diffusion in soil or sediments ( Lovley et al., 2011 ). In particular, the recent findings on cable bacteria and filamentous bacteria have revolutionarily expanded the wide spectrum of bio-electron flow from micrometer to centimeter scale, cooperatively connecting assorted biochemical reactions [e.g., sulfate oxidation, reduction of dissolved oxygen, and Fe(II)/Fe(III) transformation] from anoxic to oxic conditions in sediments ( Liu et al., 2021a ; Yang et al., 2021 ). On this topic, a microbial electron transfer network (METN), which is a three-dimensional collection of extracellular electron transfer (EET) behaviors among aggregative microflora of the phylogenetically diverse EAMs and even non-electroactive ones, is certainly evident and ubiquitous in both natural and engineering environments. In the viewpoint of environmental bioremediation, it continues to make us wonder if it is possible that the development of biotreatments could be guided by the theories of METN, which had been less focused on previously. We hope that this idea of METN could help the scientific community bring a deeper understanding to answer the following two fundamental questions: (1) How does one modify and optimize functional microbiome including EAMs and non-electroactive ones in a real environment? (2) To what extent could this artificial modification be regulated for the environmental biotechnology development?",
"discussion": "Discussion Why Is Microbial Electron Transfer Network Important? On a cross-sectoral scale, environmentalists may be more interested in how the METN works in some biological treatment processes, e.g., membrane bioreactor and anaerobic digestion, especially for granular sludge-based structures in which microbiome shares micro-niche intimately and intensive mass-transfer flow both individually and collectively. Taking anaerobic digestion as an example, besides either H 2 or acetate as electron donors for methanogenesis processes, direct interspecies electron transfer (DIET) is recognized as a highly efficient and stable process connecting both respiratory and fermentative bacteria/archaea for bioconversion from (macro-molecular) organics to methane ( Lovley and Holmes, 2021 ). Speaking of which, the fast development and broad application scenarios of genomics could provide plenty of correlative information from other transboundary research. For instance, a two-species microbial coculture was established for value-added chemical evolution driven by bio- and light energy ( Huang et al., 2022 ). In this case, Rhodopseudomonas palustris harvested and transferred solar energy into bioenergy (bio-electrons) while the other, Methanosarcina barkeri , conducted CO 2 -to-CH 4 conversion powered by the bio-electron flow. The key question was how those bio-electrons passed through cytomembranes. Revealed by metatranscriptomic analyses, both multihaem cytochrome c and nanofilaments (direct contact) and electron shuttles (indirect connect) wired two species to construct a biological hybrid system ( Huang et al., 2022 ). Coincidentally, a model of DIET between acetate-consuming bacteria and methanogens was recently established via genome-centric metatranscriptomics analysis; either electrically conductive pili (e-pili) and cytochromes or artificial materials (hydrochar) were evident as available electric conduits for DIET ( Shi et al., 2021 ). Those intriguing findings, of course, would be helpful and constructive for environmentalists to optimize related environmental biotechnology. Now, back to the question of whether METN is important for future environmental biotechnology. More than 400 scientific papers per year have been published within the broad scope of ‘‘ biological treatment ’’ and ‘‘ electron transfer ’’ in the past 5 years (based on the web search results 1 ). The answer is obvious. New scientific discoveries and insights have expanded upon new electroactive species and novel electric bridges; however, a coherent and comprehensive picture or framework of METN on how it works in biological applications and its associated environmental implications is still not available. It would be of great significance to formulate this concept to guide the construction and operation of biological treatments from lab-scale to real application, in a multidiscipline view. In What Area Could Microbial Electron Transfer Networks Be Improved? As a manifestation of functional microflora behaviors, cell performance in METN is paid more attention prior to METN improvement in the eyes of environmentalists. For example, it is known that the electric conduits of EAM are arrayed disjunctively on the cell surface ( Lovley and Holmes, 2021 ). This fact, however, makes the bioelectric connection of e-pili or other EET-related proteins with extracellular electron barriers/terminals become random behaviors that require close contact and more active interfacial area. In this sense, METN is fundamentally an issue of mathematic probability. The fast-growing literature has suggested “top-down” strategies. Theoretically, the inactive electric conduits on a single-EAM cell could be wired up by covering abiotic conductive electron collectors (e.g., FeS or polydopamine), achieving record-high EET efficiency ( Yu et al., 2020 ). In this way, it is an encouraging story in which EET had been evolved from natural “dot-to-dot” contact to artificial “cell sphere-to-cell sphere” connection. Recently, this story is enriched by another article published in Science . The Shewanella sp., a famous EAM model strain, was in vivo embedded with silver nanoparticles for excellent fuel-utilization efficiency in microbial fuel cells ( Cao et al., 2021 ). Although they are followed by universal controversies, the presented findings strongly imply that the cytomembrane-level deficits are responsible for the sluggish electron transfer efficiency. However, the cytomembrane-level modification is still not “three-dimensional” enough before its competing mechanisms on “species-to-species” connections in METN uncovered from mysteries. To this end, some redox-active substances (e.g., elemental sulfur) are suggested to be useful for mediating fundamental interactions on the species level ( Zhang et al., 2021 ). Those redox-active substances (or generally called electron shuttles) could act as driving forces for indirect interspecies electron transfer (IIET). IIET was also intensively researched in fields like iron cycle in sedimentary environments, anaerobic digestions, and microbial electrosynthesis ( Weber et al., 2006 ; Rabaey and Rozendal, 2010 ; Shi et al., 2016 ). It could be a critical framework of METN that makes the “microcosm-to-microcosm” communication possible. However, unlike DIET, progress on IIET has been slow to sufficient, considering that the vast candidates for electron shuttles ranged from artificial additives (e.g., H 2 , biochar, and flavins) to microbial secreta (e.g., soluble c -type cytochromes) ( Liu et al., 2020 ; Wu et al., 2020 ; Zavarzina et al., 2020 ). Overall, such “top-down” strategies mainly focus on modifying and optimizing the natural-given properties of EAM, whereas how to piece together those nature-given properties is still a fundamental question awaiting an answer. Though great strides have been made, an ideal niche for METN remains difficult to maintain in real biotreatments. It was found that the interspecific competition even within the same genus (i.e., Geobacter spp.) would largely alter the electron transfer networks in complicated microbial consortia ( Yan et al., 2021 ). Here, we also propose bottom-up strategies on research of the eco-niches and microbial interaction of METN in artificial and engineering biosystems for fundamental and practical interests. For example, the synthetic microbiome, a rationally programmed microbial consortia with engineering strategies (i.e., quantification, standardization, and modularization) into the assembly of functional microbiome, opens a modular toolbox for scientists to break the limitation of natural evolution of METN ( Lawson et al., 2019 ; Jaiswal and Shukla, 2020 ). In other words, we can now create a synthetic METN microcosm with known microbial consortia ( Figure 1 ). Recently, a synthetic METN microcosm within a three-species microbial consortium (engineered Escherichia coli , Bacillus subtilis , and Shewanella oneidensis ) was constructed following a “division-of-labor” principle, resulting in better bioenergy generation during which the production of electron donors/shuttles and bioenergy recovery were separately allocated in the three-species microbial consortium ( Liu et al., 2017 ). The principle was generally exerted and tested through a cross-feeding strategy in most research. For example, it was found that chromate [Cr(VI)] could be reduced in an anaerobic digestion sludge coupled with elemental sulfur [S(0)] or zerovalent iron [Fe(0)] as the electron donor ( Shi et al., 2019 ). This process was mediated by a typical cross-feeding strategy. The volatile fatty acids produced by S(0)- or Fe(0)-oxidizing bacteria (like Thiobacillus spp. and Ferrovibrio spp., respectively) could be used to further metabolize the chromate-reducing bacteria (like Geobacter spp. or Desulfovibrio spp.). This finding was quite important as it provided a potential microbial consortia design for Cr(VI) removal in groundwater and other water streams where proper organic electron donors are insufficient and any treatments that potentially bring secondary pollution are strictly forbidden. Though promising, the scientific community may expect a database-like toolbox recording interspecific synergy or even competition to better advise the research and application attempts of the synthetic METN microcosm. Ignoring immense technology transfer issues, it is not surprising to expect that the METN effectiveness could be largely improved and dynamically controlled by means of constructing synthetic microbial consortia with designed objectives. FIGURE 1 METN development for next-generation environmental biotechnology. Future Research Directions Though successful examples have been introduced above, the metabolic pathway segregation for establishing rational “division of labor” should be noted for further verification and implementations. Taking METN as a whole bio-unit in biological treatments, how to regulate metabolic flux (mass, energy, and information flows) in METN to avoid unwanted selection bias is a huge challenge. To answer this, cytocompatible EET circuits and eco-compatible strains should be carefully constructed, selected, and assembled in the METN microcosm. Though most proposals of “cytocompatible EET” establishment are based on the modifications of electroactive species, the unique EET capability of EAMs may place themselves as potential chassis cells for “eco-compatible strains” construction to deeply modify the metabolic flux (especially energy flow) in such biotechnology (synthetic biology). In this case, biosafety should be carefully noted in the research to avoid the diffusion of modified genes into the natural gene pool. Nevertheless, intensive “Design–Build–Test–Learn cycles” should be both essentially and iteratively conducted for objective-driven and precise optimization. Strategies including competition-related elements (e.g., toxin secretion systems) and cross-feeding can be applied to maintain the balance of this synthetic METN microcosm. Though it seems impossible to universally know the specific functions of each microorganism, piecing together clues behind the mechanisms of METN would be a potential shortcut. Speaking of which, quorum sensing (QS) could be probably taken into consideration on the microbial compatibility control. QS is basically an interspecies communication process during microbial aggregation, biofilm formation, and granulation, induced by a series of QS signals like homoserine lactones and autoinducing peptides ( Maddela et al., 2019 ). Of most interest, the signals could increase the concentration and redox activities of extracellular polymeric substances from electroactive biofilm ( Chen et al., 2017 ). Here, we appeal for more efforts on microbial interactions (synergy, mutualism, competition, etc.) of METN, especially on connections between key functional METN microbiomes and non-electroactive species. Microbial electron transfer network is clearly not an exclusive concept of EAMs or other microorganisms; more innovative research on the microbes–materials interactions should be conducted both technically and economically to increase the knowledge base and the competitiveness of related environmental biotechnologies. In particular, questions on how to expand the influence range of METN and to what extent could METN be domesticated still perplex environmental researchers. Energy taxis, as a key branch of chemotaxis, was recently proposed to control the transport and motility of S. oneidensis MR-1 in porous media ( Liu et al., 2021b ). It is important as the migration of those functional microbes could hopefully be controlled toward (micro)pollutant sources along gradient redox-active material surfaces. Such effectiveness provides a complementary balance strategy between two major contradictions: excessive population growth and biomass running off. Either overpopulation or flushing loss of biomass is disfavored in environmental biotechnology since it would cause severe sludge accumulation and inexorable crash in efficiency, resulting in added complications on reactor operation. Thus, harnessing energy taxis to different redox materials could be effective for METN regulation, but relevant research is still in its infancy. Biochar is also an excellent candidate for METN regulation as its sources are earth-abundant, and importantly, it is highly redox-active with sufficient micro- and macropores ( Mohanty and Boehm, 2014 ; Zhou et al., 2019 ). Attempts have been made to substantially expand the electronic reach of METN by wheat straw-derived biochar for bioremediation of pentachlorophenol-contaminated soils ( Cai et al., 2020 ). On the other hand, once it involves usage of materials, the operation cost should be calculated and reported ( Zhang and Angelidaki, 2016 ). We are expecting more voices and more research activities on this topic for further advances made in the near future."
} | 4,363 |
30841503 | PMC6429283 | pmc | 3,927 | {
"abstract": "Tartaric acid is an important chiral chemical building block with broad industrial and scientific applications. The enantioselective synthesis of l (+)- and d (−)-tartaric acids has been successfully achieved using bacteria presenting cis -epoxysuccinate hydrolase (CESH) activity, while the catalytic mechanisms of CESHs were not elucidated clearly until very recently. As biocatalysts, CESHs are unique epoxide hydrolases because their substrate is a small, mirror-symmetric, highly hydrophilic molecule, and their products show very high enantiomeric purity with nearly 100% enantiomeric excess. In this paper, we review over forty years of the history, process and mechanism studies of CESHs as well as our perspective on the future research and applications of CESH in enantiomeric tartaric acid production.",
"introduction": "1. Introduction Tartaric acid (TA) is a well-known organic acid that naturally occurs in many kinds of fruit, most notably in grapes. The chemical chirality of TA was first discovered by Jean-Baptiste Biot in 1832 [ 1 ]. The naturally-occurring form of the acid is l (+)-TA, while d (−)-TA rarely exists in natural sources [ 2 , 3 ]. l (+)-TA is widely used in the food, wine, pharmaceutical, chemical, and polyester industries. d (−)-TA is also important in pharmaceutical manufacturing [ 4 , 5 , 6 ]. Both are well-known chiral chemical building blocks with broad industrial and scientific applications [ 7 , 8 ]. In enantioselective chemical synthesis, TA serves not only as a resolving agent or chiral auxiliary in the synthesis of bioactive molecules, but also a source of new asymmetric organocatalysts [ 7 , 8 , 9 ]. Traditionally, l (+)-TA is obtained as a solid by-product during wine fermentation, and this kind of production method is strongly influenced by the growth of grapes and the climatic conditions. Chemical synthesis of l (+)-TA with maleic acid is also possible but this gives a much less soluble racemic product (DL-form) which is not suitable for inclusion in foods because d (−)-TA in the product is considered to be harmful to human health. Commercial application of the chemical method is limited by both the product form and the high production cost [ 10 ]. Currently, microbial methods are considered to be much simpler and more economical for the production of l (+)-TA and d (−)-TA. Epoxide hydrolases (EHs, EC 3.3.2.3) are biocatalysts that are ubiquitous in Nature. They can hydrolyze racemic epoxides to their corresponding optically active epoxides and pure vicinal diols, which are versatile intermediates for chiral pharmaceutical synthesis. In general, this enzymatic process occurs under mild conditions without the need for any cofactors, prosthetic groups, or metal ions [ 11 ]. The high transformation rate and enantioselectivity of epoxide hydrolases have gained them increasing attention in recent years, and they have found more and more applications in the organic chemical industry [ 11 ]. Epoxide hydrolases are found in a variety of sources, such as plants, insects, mammals, and microbes [ 12 , 13 , 14 ]. Mammalian epoxide hydrolases have been the subject of many studies because of their key role in xenobiotic detoxification in the liver, but their use as biocatalysts has been hindered by their limited availability [ 15 ]. Lately, bacterial epoxide hydrolases have been increasingly recognized as highly versatile biocatalysts owing to their abundance, high efficiency, and environmental friendliness [ 16 ]. cis -Epoxysuccinic acid hydrolases (CESHs) are epoxide hydrolase members that catalyze the asymmetric hydrolysis of cis -epoxysuccinate (CES) to form an enantiomeric tartrate [ 17 , 18 , 19 ]. Bacteria presenting CESH activity were discovered in the 1970s, and the synthesis of l (+)-TA was the first application of an epoxide hydrolase [ 20 ]. Since then, a large number of bacteria with CESH activity have been discovered and successfully applied for industrial TA production. The sequences and mechanism of CESHs have been partly elucidated since 2000, when it was revealed that CESH[L] and CESH[D], which produce l (+)-TA and d (−)-TA, respectively, are completely different proteins in terms of both sequence and structure. Therefore, CESHs are interesting EHs not only for TA production, but also for enantiomer biosynthesis in general. In this review, we summarize the body of literature on CESHs including both process optimization for industrial application and mechanism studies to understand how their regio- and stereoselectivity makes them efficient biocatalysts. We also provide our perspective on the use of CESHs in future research and applications."
} | 1,163 |
23083487 | PMC3534718 | pmc | 3,930 | {
"abstract": "Background Synthetic biology allows the development of new biochemical pathways for the production of chemicals from renewable sources. One major challenge is the identification of suitable microorganisms to hold these pathways with sufficient robustness and high yield. In this work we analyzed the genome of the propionic acid producer Actinobacteria Propionibacterium acidipropionici (ATCC 4875). Results The assembled P. acidipropionici genome has 3,656,170 base pairs (bp) with 68.8% G + C content and a low-copy plasmid of 6,868 bp. We identified 3,336 protein coding genes, approximately 1000 more than P. freudenreichii and P. acnes , with an increase in the number of genes putatively involved in maintenance of genome integrity, as well as the presence of an invertase and genes putatively involved in carbon catabolite repression. In addition, we made an experimental confirmation of the ability of P. acidipropionici to fix CO 2 , but no phosphoenolpyruvate carboxylase coding gene was found in the genome. Instead, we identified the pyruvate carboxylase gene and confirmed the presence of the corresponding enzyme in proteome analysis as a potential candidate for this activity. Similarly, the phosphate acetyltransferase and acetate kinase genes, which are considered responsible for acetate formation, were not present in the genome. In P. acidipropionici , a similar function seems to be performed by an ADP forming acetate-CoA ligase gene and its corresponding enzyme was confirmed in the proteome analysis. Conclusions Our data shows that P. acidipropionici has several of the desired features that are required to become a platform for the production of chemical commodities: multiple pathways for efficient feedstock utilization, ability to fix CO 2 , robustness, and efficient production of propionic acid, a potential precursor for valuable 3-carbon compounds.",
"conclusion": "Conclusions The genome-scale information that we have obtained combined with already developed physiological, genetic and metabolic approaches will increase our knowledge of P. acidipropionici and may help establish strategies that will increase and optimize its use in industrial processes. The experimental results and genome annotations have revealed some P. acidipropionici features that may explain the plasticity and hardiness of this specie. P. acidipropionici has about 1,000 genes more than P. acnes and P. freudenreichii , suggesting less specialization and a more flexible metabolism. Genes for the defense of genome integrity by the CRISPR/Cas system, polyphosphate accumulation, and the large amount of rRNA and carbohydrate ABC transporters that were identified might also render some competitive advantages and allow the bacterium to adapt in different environments. In addition, the high acid tolerance, wide range of substrates metabolized, hardiness and microaerophilic life style suggest that P. acidipropionici is a microorganism that has a potential use in the industrial fermentation of commodities, such as propionic acid. Moreover the antifungal property of propionic acid will help avoid the contamination of the process by indigenous yeast, the main source of contamination in sugarcane juice\n[ 38 ]. Propionic acid fermentation from sugarcane juice, therefore, could be performed without strict microbial control and, consequently, in a cost-effective manner. Worldwide interest in chemical compounds produced by biological reactors has increased considerably, especially in the petrochemical industry that has as a main objective, the low cost production of compounds that could be fixed in high durable materials. In this context, propionic acid fermentation by P. acidipropionici shows a great potential to satisfy the world demand for this commodity.",
"discussion": "Results and discussion General genome features The genome of P. acidipropionici comprises a circular chromosome of 3,656,170 base pairs (bp) with 68.8% GC content and a low-copy plasmid of 6,868 bp with 65.4% GC content. The chromosome contains 3,336 protein coding sequences (CDSs) with an average length of 967.5 bp, 53 tRNAs and four 16S-23S-5S rRNA operons, accounting for 88.8% of genomic DNA. Putative functions were assigned to 2,285 (68.5%) of the CDSs, while 556 (16.7%) were classified as conserved hypothetical and 495 (14.8%) had no significant similarity with data in the public databases (e-value >1E-10) (Table \n 1 ). Table 1 General features of P. acidipropionici genome Chromosome pRGO1 Length (bp) ~3,656,170 6,868 Copy number 1 ~7.4 Coding content (%) 88.8 50.5 G + C content of total genome (%) 68.8 65.4 G + C content of coding regions (%) 69.0 70.8 G + C content of non-coding regions (%) 66.9 59.7 rRNA 4 X (16S-23S-5S) 0 Pseudogenes 32 0 Genes 3389 8 tRNAs 53 0 Protein Coding (CDS) 3336 8 conserved with assigned function 2285 (68.5%) 5 conserved with unknown function 556 (16.7%) 0 Nonconserved 495 (14.8%) 3 Average CDS length (bp) 967.5 519.4 Preliminary proteomic analysis of P. acidipropionici growing on different carbon sources allowed the identification of 649 (19.5%) of the CDSs (Methods; Additional file\n 1 ). The assembled plasmid matched exactly the previously sequenced pRGO1\n[ 13 ] from P. acidipropionici and pLME106 from P. jensenii [ 14 ]. The number of plasmid copies per cell was estimated to be 7.4 based upon relative read coverage. We identified one new CDS [Genbank:AB007909.1; 1292–1552] in the plasmid coding for an 87 amino acid peptide similar to an InterPro family of proteins putatively involved in plasmid stabilization [InterPro: IPR007712]. This peptide sequence may be useful for the development of P. acidipropionici vectors, a fundamental molecular biology tool for the biotechnological use of any organism. Comparative analysis The genome of P. acidipropionici , a species that can live in several environments such as soil, rumen and cheese, was compared with the genomes of three closely related species, but with different habitats and ecology: (i) P. acnes [GenBank: NC_006085.1], a major inhabitant of human skin and considered to be an opportunistic pathogen that has been associated with acne vulgaris\n[ 15 ]; (ii) P. freudenreichii subs. shermanii CIRM-BIA1 T [GenBank: FN806773.1] which has known use in cheese manufacture and, more recently, as a probiotic\n[ 16 ]; and (iii) Microlunatus phosphovorus [GenBank: NC_015635.1], a species that belongs to the same family ( Propionibacteriaceae ), which is found in soil and has been isolated from activated sludge by its ability to accumulate polyphosphate and polyhydroxyalkanoates (PHA)\n[ 17 ]. The general features of these genomes are summarized in Table \n 2 . Table 2 General features of four species used in comparative analysis Organism Genome size (Mb) %GC No. of Proteins No. of rRNA operons No. of tRNAs Propionibacterium acidipropionici ATCC4875 3.6 68.8 3336 4 53 Propionibacterium acnes KPA171202 2.6 60.0 2297 3 45 Propionibacterium freudenreichii CIRM-BIA1 2.6 67.3 2375 2 45 Microlunatus phophovorus NM-1 5.7 67.3 5338 1 46 Clustering of all proteins encoded by these four bacteria resulted in 6469 families. From the total proteins encoded by P. acidipropionici , 1009 were clustered into 296 families containing 2 to 54 members. The comparative clustering of proteins from P. acidipropionici, P. acnes and P. freudenreichii is summarized in Figure\n 2A , and a four-set Venn diagram including M. phosphovorus is presented in an additional figure (Figure S1 in Additional file\n 2 ). P. acidipropionici shows great expansion in families shared by the three Propionibacteria, most notably in transport proteins, two-component regulatory systems, transcriptional regulators and proteins with oxidoreductase activity (Table \n 3 ). Families of transport proteins were compared in detail using the Transporter Classification (TC) system. A total of 469 transport proteins were annotated in P. acidipropionici genome and almost half (46.9%) of these proteins were classified as members of the ATP-binding cassette family (ABC). This proportion is greater than in any of the other three bacteria species used in the comparison (Table \n 4 ). The complete comparison of transport proteins in these four bacteria species is presented in an additional file (Additional file\n 3 ). Figure 2 Comparative protein clustering and GO annotation. A . Venn diagram showing the number of common and unique protein clusters for three Propionibacteria. The values in parentheses represent the number of protein clusters. The values in square brackets represent the number of single proteins (proteins not in clusters). The number of proteins clustered in each group are also indicated, not enclosed and color-coded by organism: values in green represent P. acidipropionici proteins, values in blue represent P. freudenreichii proteins and values in red represent P. acnes proteins. B . Pie chart depicting the result of GO annotation of proteins unique to P. acidipropionici. C . Multi-level pie chart detailing GO annotation of proteins unique to P. acidipropionici . Inner circle represent level 3 terms. Outer circle represent lower level (more specific) terms. Table 3 A selection of paralogous families of P. acidipropionici and close related genomes Cluster functional annotation Number of proteins found in: P. acidipropionici P. acnes P. freudenreichii M. phosphovorus ABC transporters 54 45 34 59 ABC transporters 20 11 1 13 ABC transporters 18 9 1 12 ABC transporters 9 8 4 19 ABC transporters 8 2 4 3 ABC transporters 7 3 2 6 ABC transporters 7 3 2 6 ABC transporters 5 1 2 1 ABC transporters 4 2 1 1 MFS transporters 10 3 10 21 Proton antiporters 3 1 1 2 Anaerobic C4-dicarboxylate transporters 0 2 2 0 Two-component system regulators 24 13 14 48 Two-component system kinases 8 3 4 11 Transcriptional regulators 27 10 4 22 Transcriptional regulators 9 5 5 12 Transcriptional regulators 5 2 1 1 MarR family transcriptional regulators 7 4 4 13 ArsR family transcriptional regulators 3 1 3 20 Oxidoreductases 23 9 8 29 Oxidoreductases 12 8 10 36 Oxidoreductases 12 4 4 21 Oxidoreductases 11 5 7 21 Dehydrogenases 8 2 4 15 Catalases 1 1 1 0 Catalases 1 0 0 1 Glycerol-3-phosphate dehydrogenases 2 1 1 0 Glutathione S-transferases 3 0 1 1 Aminotransferases 8 3 5 8 Aminotransferases 4 1 1 1 Amidotransferases 3 1 1 1 Glycosyl transferases 6 4 3 1 Glycosyl transferases 3 1 0 0 Sugar kinases 4 2 2 0 Polyphospate kinases 1 1 1 1 Polyphospate kinases 2 1 1 1 3 Polyphosphate-dependent glucokinases 1 1 1 2 Lipase esterases 3 1 0 0 Alpha-galactosidase 3 0 0 1 Putative sucrose-6-phosphate hydrolases (invertases) 1 0 0 2 Plasmid maintenance system antidote proteins 3 0 0 0 DNA-binding proteins 2 0 0 11 HNH endonucleases 6 0 0 0 HNH endonucleases 2 0 0 0 Hypothetical proteins 5 1 1 0 Hypothetical proteins 5 0 0 0 Hypothetical proteins 4 3 0 2 Hypothetical proteins 3 0 0 1 Methylases 0 0 2 2 S-layer proteins 0 0 2 0 cAMP factors 0 5 0 0 Adhesion proteins 0 4 0 0 Magnesium chelatase 0 2 0 1 Alpha-L-fucosidases 0 2 0 1 Sialic acid transporters 0 2 0 0 Endoglycoceramidases 0 2 0 0 Adhesion proteins 0 2 0 0 Lysophospholipases 0 2 0 0 Triacylglycerol lipase precursors 0 2 0 0 Table 4 A selection of transporter families of P. acidipropionici and close related genomes TC system Description Number of proteins found in: P. acidipropionici P. acnes P. freudenreichii M. phosphovorus 1.C.70 CAMP Factor 0 5 0 0 2.A.1 Major Facilitator Superfamily 38 21 36 68 2.A.1.1 Sugar Porter (SP) 8 5 5 8 2.A.1.3 Drug:H + Antiporter-2 (14 Spanner) (DHA2) 10 3 12 21 2.A.1.6 Metabolite:H + Symporter (MHS) 7 2 4 10 2.A.13 C4-Dicarboxylate Uptake 0 2 2 0 2.A.2 Glycoside-Pentoside-Hexuronide :Cation Symporter 4 1 1 1 2.A.3 Amino Acid-Polyamine-Organocation 12 12 10 6 2.A.47 Divalent Anion:Na+ 0 0 2 1 2.A.66 Multidrug/Oligosaccharidyl-lipid/Polysaccharide Flippase 5 1 1 2 3.A.1 ATP-binding Cassette (ABC) 220 137 105 219 3.A.1.1 Carbohydrate Uptake Transporter-1 (CUT1) 57 29 7 40 3.A.1.2 Carbohydrate Uptake Transporter-2 (CUT2) 17 7 4 9 3.A.1.3 Polar Amino Acid Uptake Transporter (PAAT) 19 5 8 11 3.A.1.4 Hydrophobic Amino Acid Uptake Transporter (HAAT) 5 0 5 5 3.A.1.5 Peptide/Opine/Nickel Uptake Transporter (PepT) 28 14 9 29 3.A.1.12 Quaternary Amine Uptake Transporter (QAT) 13 6 7 6 3.A.11 Bacterial Competence-related DNA Transformation Transporter 0 2 2 2 3.B.1 Na + −transporting Carboxylic Acid Decarboxylase 7 3 4 4 4.A.1 PTS Glucose-Glucoside 1 8 1 0 4.A.2 PTS Fructose-Mannitol 2 4 0 0 4.A.3 PTS Lactose-N.N′-Diacetylchitobiose-β-glucoside 0 3 0 0 4.A.4 PTS Glucitol 5 4 0 0 4.A.5 PTS Galactitol 4 2 0 0 4.A.6 PTS Mannose-Fructose-Sorbose 0 0 1 0 4.A.7 PTS L-Ascorbate 0 2 0 0 5.A.3 Prokaryotic Molybdopterin-containing Oxidoreductase 3 6 2 7 8.A.9 rBAT Transport Accessory Protein 4 2 2 6 9.A.10 Iron/Lead Transporter 5 0 2 5 9.A.40 HlyC/CorC Putative Transporters 4 2 2 2 The assignment of Gene Ontology terms to proteins unique to P. acidipropionici is depicted in Figures\n 2B and\n 2C . Two proteins that distinguish the central fermentative pathway of P. acidipropionici from that of P. acnes and P. freudenreichii have ligase activity and are discussed in the section “Propionic acid fermentation”. Proteins with sequence-specific DNA binding and methyltransferase activities could be involved in DNA repair, gene expression and chromosome replication. These proteins could also compose a restriction modification system serving as defense against foreign DNA. Therefore, the study of these proteins that are unique to P. acidipropionici may be important in developing genetic tools suitable for this strain. The number of common genes is consistent with the previously reported phylogeny\n[ 18 ]. P. acidipropionici is more closely related to P. acnes and colinearity of genes was observed most of the time (Figure\n 3 ) . However, P. acnes has approximately 1,000 genes fewer than P. acidipropionici , a fact that could reflect the specialization of this species as an opportunistic pathogen of human skin. It is generally considered that specialization leads to gene loss\n[ 19 ], a process that can be very fast in bacteria. For example, in Lactobacillus the transition of strains to nutritionally rich environments leads to metabolic simplification and the loss of several genes\n[ 20 ]. Although P. freudenreichii has all the enzymes needed for the de novo biosynthesis of amino acids and vitamins\n[ 16 ], the metabolic simplification premise may indicate that the reduced number of genes found in P. freudenreichii in comparison to P. acidipropionici is suggestive of the beginning of a specialization process as a consequence of the continuous use of this species in cheese ripening. This reasoning is consistent with the analysis of the M. phosphovorus genome, a soil bacterium, with 5,338 annotated genes. Most of the gene clusters that were conserved in the three Propionibacterium are also present in M. phosphovorus (895 out the 1,026) (Figure\n 2A ; Figure S1 in Additional file\n 2 ). P. acidipropionici , with 3,336 genes, would seem to hold an intermediary position. The 32 pseudogenes identified in P. acidipropionici are disrupted by a single frameshift or point mutation, suggesting that these events are recent and that some genome reduction is underway. Figure 3 Syntenic dot plot. Dot plot alignment between the chromosome of P. acidipropionici (vertical) against the chromosomes of M. phosphovorus , P. freudenreichii and P. acnes (horizontal) at protein level. The dnaA gene is located at the beginning of all four sequences. The horizontal transfer of DNA fragments might supply the recipient microorganism with the necessary genetic resources to be able to adapt to new environments; for example, an antibiotic resistance gene and/or a gene that encodes a peptide in a biodegradative pathway\n[ 21 ]. Furthermore, Jain et al. [ 22 ] described the horizontal transfer of genes as a mechanism to spread genetic diversity across species and showed that horizontal gene transfer occurs between organisms that share similar factors like G/C content, genome size, oxygen tolerance and carbon utilization. Predicted genomics islands account for 3.8% (126) of total P. acidipropionici genes. This number is similar to predictions in P. acnes and M. phosphovorus genomes (3.0% and 4.8%, respectively) but lower than the prediction in the P. freudenreichii genome (6.7%). Defense of genome integrity Analysis of the P. acidipropionici genome revealed some defense mechanisms that may allow it to withstand viral and nucleic acid invasion; these include restriction enzymes and the Clustered Regularly Interspaced Short Palindromic Repeats (CRISPRs). CRISPRs can provide the cell with an acquired resistance against bacteriophages and conjugative plasmids, possibly acting as a RNA interference-like mechanism. The spacers between direct repeats are derived from invader sequences and determine the specificity of the system. In addition, the mechanism is composed by CRISPR-related sequences (Cas), which are proteins encoded in the vicinity of CRISPR loci\n[ 23 ]. Some of these proteins show similarity to helicases and repair proteins. Seven CRISPR-associated proteins were annotated in the P. acidipropionici genome (Cas1, Cas2, Cas3, Cas5, Cse1, Cse3, Cse4), while only two were identified in the P. freudenreichii and M. phosphovorus genomes. Moreover, using the CRISPR finder tool (crispr.u-psud.fr/Server/CRISPRfinder.php,\n[ 24 ]), three CRISPR loci were annotated in P. acidipropionici genome. The CRISPR1 locus contains 1248 bp and harbors 20 spacer sequences, the CRISPR2 locus contains 2043 bp and harbors 33 spacer sequences and the CRISPR3 locus contains 3689 bp and harbors 60 spacer sequences. The length of direct repeat sequence in all CRISPR loci is 29 bp. The spacers do not show strong similarity to phage and bacteria sequences available in databases from the National Center for Biotechnology Information (NCBI). Since Cas genes and CRISPRs are known to have undergone extensive horizontal transfer and the identified proteins do not share high sequence similarity (were not clustered together), these regions could have been acquired horizontally from different sources, explaining this inequality. No CRISPR-related proteins were found in the P. acnes genome. The most common cause of slow or incomplete bacterial fermentations in the dairy industry is bacteriophage infections that lead to substantial economic loss\n[ 25 ]. To deal with phage diversity, lactic acid bacteria have developed several systems to withstand the infections; one of these is the CRISPR/Cas system. CRISPR/Cas loci have been identified in many lactic acid bacteria, such as Streptococcus thermophilus , Lactobacillus casei , Lactobacillus delbrueckii , Lactobacillus helveticus and Lactobacillus rhamnosus [ 26 ]. The CRISPR/Cas system was found in the genome of P. acidipropionici , and could thus be interpreted as another feature of this bacterium that could render it robustness against phage infections in an industrial setting for the production of chemical commodities. Furthermore, the genomic data for the CRISPR/Cas system found in P. acidipropionici will provide information about the defense mechanism of this bacterium and allow the development of studies to improve its resistance to phage infections. Lifestyle and environmental adaptation Many of the annotated proteins families in P. acidipropionici have the potential to render this species flexibility to adapt to different environments (Tables\n 3 and\n 4 ). Among these protein families are the carbohydrate ABC transporters, transcriptional regulators, two-component response regulators, and oxidoreductases; all of these families had greater numbers of members in P. acidipropionici when compared with P.acnes or P. freudenreichii . Moreover, there are more rRNA operons and tRNA genes in P. acidipropionici compared with the other three species (Table \n 2 ). This feature could potentially confer competitive advantages such as a faster growth rate and quick response to environmental changes\n[ 27 ]. M. phosphovorus contains the same set of 45 tRNAs as P. acnes and P. freudenreichii , plus a selenocysteine tRNA that could be present solely to support formate dehydrogenase expression and is unique to M. phosphovorus among the four species that were compared. Other protein families that appear expanded in P. acidipropionici in comparison to the other three species are glutathione-S-transferases (GSTs) and HNH endonucleases. Compared with the single copy of a glutathione S-transferases (GST) coding gene found in P. freudenreichii genome, three copies of the gene were found in the P. acidipropionici genome (Table \n 3 ). In bacteria, this enzyme has been reported to be involved on growth on recalcitrant chemicals and in the degradation of aromatic compounds\n[ 28 ]. HNH endonucleases coding genes were absent from the genome of P. freudenreichii , P. acnes e M. phosphovorus , while eight copies grouped into two clusters were found in P. acidipropionici (Table \n 3 ). This enzyme is a type of homing endonuclease that could be potentially involved in DNA rearrangements or could act as bacteriocin\n[ 29 , 30 ]. Taken together, the notable expansion of these two gene families may confer to P. acidipropionici robustness in its capacity for growth under adverse conditions and competitiveness against other bacterial species. However, further studies are needed to establish the function of these enzymes in the lifestyle of P. acidipropionici . On the other hand, previously reported probiotic related genes were found only in the P. freudenreichii genome (gluconate kinase, S-layer proteins, cell-wall peptidase NlpC/p60, sortase and microcin resistance)\n[ 16 ], thus agreeing with the possible adaptation of the species to the nutritional environment of the gut. Similarly, some protein families that were identified only in P. acnes are probably related to pathogenicity and host specialization. Some examples of these families include cAMP factor, endoglycoceramidases, sialic acid transporter, sialidases, and lysophospholipase. Energy reserve and stress resistance Polyphosphates (polyPs) are linear polymers composed of orthophosphate residues. PolyPs are not only used as energy reserves, but also have been related to stress resistance and adaptation to extreme environments\n[ 31 ]. P. freudenreichii ST33 has been reported to accumulate up to 3% of cell dry weight of polyP\n[ 32 ] and M. phosphovorus NM-1 has been reported to accumulate up to 11.8%\n[ 33 ]. Falentin et al. [ 16 ] identified the genes related to polyphosphate metabolism in P. freudenreichii CIRM-BIA1; all of these genes were present in the four genomes that were compared in the present study. Similar numbers of the Nudix hydrolases were present in all four organisms, suggesting that they had similar levels of metabolic complexity and adaptability\n[ 34 ]. In addition to the one copy of polyphosphate kinase (PPK) per genome that we found, PPK2 genes were also identified in the four genomes. PPK is an important enzyme in bacterial polyP synthesis, transferring reversibly the terminal phosphate of ATP to polyP. PPK2 differs from PPK by its use of either GTP or ATP, its preference for Mn 2+ over Mg 2+ and for being stimulated by polyP\n[ 34 ]. P. acidipropionici (PACID_16380), P. acnes (GenBank: NC_006085.1; locus: PPA1186) and P. freudenreichii (GenBank: FN806773.1; locus PFREUD_12510; annotated as hypothetical) have one PPK2 copy each. M. phosphovorus contains three PPK2 copies (Genbank: NC_015635.1; loci: MLP_05750, MLP_23310 and MLP_50300), which may play an important role in the large polyP accumulation in this organism. Metabolic reconstruction The predicted metabolic map of P. acidipropionici contained 221 pathways and 1,207 enzymatic reactions. The gene annotations and metabolic maps revealed that the complete pathways of glycolysis, gluconeogenesis, pentose-phosphate and all de novo amino acid biosynthesis were represented in the P. acidipropionici genome. The TCA cycle was also complete and genes corresponding to aerobic and anaerobic respiration pathways were detected. The biosynthetic pathways for most vitamins were found; however, pathways for biotin and pantothenic acid synthesis were notably absent, as has been reported for other species from this genus\n[ 16 ]. Figure\n 4 summarizes the basic metabolic pathways that were present in P. acidipropionici . Figure 4 Overview of P. acidipropionici metabolism. Reactions of fermentative pathway are in green while reactions of respiratory pathway are in blue. Reactions described in literature but absent in the genome are in gray dotted lines. Putative reactions for CO 2 fixation are in orange. The total number of transporters in each major category are shown. Substrate utilization It has been reported that P. acidipropionici can use a wide variety of substrates for the heterofermentative production of propionic acid\n[ 7 ]. In agreement with this report, the genome analysis of P. acidipropionici has revealed a notable set of transporters (Table \n 4 ) and enzymes that could be related to uptake and degradation of numerous substrates, putatively including glucose, fructose, sucrose, lactose, xylose, threalose, mannose, chitobiose and arabinose. The genome comparison showed that two families within ABC superfamily concerned exclusively with carbohydrate uptake are present in greater number in the genome of P. acidipropionici. The carbohydrate uptake transporter-1 (CUT1) and −2 (CUT2) families exhibit specificity for oligosaccharides and monosaccharides, respectively. There are two principal amino acid uptake families in the ABC superfamily. The family specific for polar amino acids uptake (PAAT) is more numerous in the P. acidipropionici genome. The family concerned with hydrophobic amino acids uptake (HAAT) had the same number of genes identified in all bacteria compared, except in P. acnes, where no members were identified\n[ 35 ]. The propanediol utilization operon of P. freudenreichii, probably acquired through horizontal transfer\n[ 16 ], is not present in P. acidipropionici , P. acnes or M. phosphovorus . P. acidipropionici possesses one invertase coding gene (PACID_33010) and could thus potentially use sucrose as a carbon source, while M. phosphovorus has two copies of putative invertase coding genes (MLP_06620 and MLP_06630). P. acnes and P. freudenreichii lack such a gene, and for that reason, the growth of these species in sucrose containing feedstocks like molasses needs pretreatment, with consequent cost increase\n[ 36 ]. This feature enables the development of P. acidipropionici fermentations at low cost from renewable feedstocks, such as sugarcane juice readily available in Brazil. Experimental batch fermentations using P. acidipropionici with sugarcane juice showed a significant yield and productivity of propionic acid, although significant amounts of acetic and succinic acids still emerged as by-products (Figure\n 5 ). In addition, the ability of this bacterium to metabolize the C5 sugar xylose has also been previously reported\n[ 37 ], which makes this species a strong candidate for the metabolization of second generation feedstocks derived from biomass. Industrial ethanol fermentations are commonly contaminated by indigenous yeasts, which many times are less productive than the commercial yeasts used in ethanol fermentation. This causes considerable yield losses and in some cases indigenous yeasts are even able to replace commercial yeasts\n[ 38 ]. Preliminary experiments showed that the production of propionic acid inhibited the growth of indigenous yeasts in sugarcane juice (data not shown). Hence, industrial fermentation to produce propionic acid with P. acidipropionici using sugarcane juice could be performed without strict microbial control, potentially reducing production costs. Figure 5 Preliminary results of P. acidipropionici experimental batch fermentation with sugarcane juice. Biomass A600 (blue diamond); sucrose (orange circle); glucose (red square); fructose (gray line); propionic acid (blue asterisk); acetic acid (purple cross); succinic acid (green line). In summary, we report here the genetic basis for the ability of P. acidipropionici to metabolize many carbon sources, confirming previous studies of carbon source utilization by this species. These carbon sources include the readily available sucrose from sugar cane juice\n[ 7 ] and the C5 sugar xylose that is abundant in second generation feedstocks derived from biomass\n[ 37 ]. These results coupled to the preliminary findings suggesting that the propionic acid produced by propionibacteria may help control possible contamination by indiginous yeast species in an industrial setting emphasizes the potential of P. acidipropionici as a biological reactor of industrial interest. Catabolic repression and phosphotransferase system Given the wide range of substrates that P. acidipropionici is able to metabolize, we performed a preliminary study to verify the presence of carbon catabolite repression. P. acidipropionici was grown in various carbon sources (glucose, fructose, sucrose, lactate and glycerol) with or without addition of the hexose analogous 2-deoxy-glucose (2-DG). 2-deoxyglucose is taken up by cell and phosphorylated into 2-DG 6-phosphate. This molecule cannot be metabolized; however it is able to trigger the glucose repression phenomenon\n[ 39 ]. Therefore, the presence of 2-DG in the cultures led to a clear decrease in the P. acidipropionici growth rate, suggesting the presence of a possible mechanism of carbon catabolite repression by glucose (Figure S2 in Additional file\n 2 ). The analysis of variance test (ANOVA) indicated that there was a significant difference between the growth rate of P. acidipropionici grown in the media containing glucose and in the media containing glucose with 2-DG (ANOVA factorial: F 7.32 = 8.346; p <0.0001); fructose and fructose with 2-DG (ANOVA factorial: F 7.32 = 22.363; p < 0.0001); sucrose and sucrose with 2-DG (ANOVA factorial: F 7.32 = 35.70; p < 0.0001); lactate and lactate with 2-DG (ANOVA factorial: F 7.32 = 24.53; p < 0.0001) and glycerol and glycerol with 2-DG (ANOVA factorial: F 7.32 = 11.46; p < 0.0001). The molecular basis of this phenomenon is generally related to a multiprotein phosphorelay system, the phosphotransferase system (PTS). At least three enzymes, enzyme I (EI), histidine protein (HPr) and enzyme II (EII), that are responsible for carbohydrate transporting, phosphorylation and triggering the catabolic repression respectively, are present in the PTS\n[ 40 ]. Twenty-three genes coding for general proteins of the PTS were identified in the P. acidipropionici genome (Additional file\n 4 ). Moreover, two transcriptional factors probably involved in this system (CRP – cyclic AMP receptor factor and ccpA – catabolite control protein A) and 1 PTS-regulatory domains (PRDs), which could act as antitermination proteins or as transcriptional activators, were found in the genome of P. acidipropionici . In addition, carbon catabolite repression can regulate the expression of virulence factors in many pathogenic bacteria\n[ 40 ], and this could explain the great number of PTS related genes (31 genes) found in the P. acnes genome. On the other hand, P. freudenreichii has only 5 genes related to the PTS. These observations could suggest that P. acidipropionici has the potential ability for selective carbon source utilization and genome plasticity to adapt to different environments, but further studies would be necessary in order to confirm this hypothesis. Aerobic metabolism P. acidipropionici is a facultative anaerobic microorganism. In practical terms, the cells need anaerobic conditions to grow on solid media; however, they can grow in liquid media with oxygen added (data not shown)\n[ 41 ]. Inspection of its genome showed that all the genes related to aerobic respiration and oxidative stress are present. P. acidipropionici had two putative catalase genes while P. acnes and P. freudenreichii contained only one copy. Thus, the susceptibility of P. acidipropionici to oxygen could be related to deficient expression of the oxidative defense genes or to a redox imbalance. These preliminary data are encouraging in suggesting oxygen resistance in this species. However, further studies on the oxygen resistance of P. acidipropionici will be needed in order to establish non-strict anaerobic conditions for growth in an industrial setting, where the maintenance of a strict anaerobic atmosphere would be too costly. Electron transport chain The electron transport chains of prokaryotes vary widely, having multiple terminal oxidases. This variation allows bacteria to adapt their respiratory systems to different environmental growth conditions\n[ 42 ] by choosing the composition of enzymes that will achieve: (i) the highest possible coupling efficiency (H + /e - ratio); (ii) the rapid removal of excess reducing equivalents such as NADH/NADPH; and (iii) the regulation of intracellular oxygen concentration\n[ 43 ]. Cytochrome bd (PACID_05290 and PACID_05280) is present in all four organisms that were compared. In E. coli , this oxidase has a high affinity for oxygen but it does not pump protons being preferred in low-oxygen conditions because the respiratory chain is then still able to provide energy, even in low amounts\n[ 44 ]. Although the cytochrome bd complex does not pump protons in E.coli , there is a net proton transfer (1 H + /e - ) by this oxidase due to quinol oxidation and oxygen reduction on the periplasmic and the cytoplasmic sides of membrane, respectively\n[ 44 - 46 ]. On the other hand, cytochrome c oxidase (CcO) reduces molecular oxygen to water coupled with the pumping of four protons across the membrane, generating more energy\n[ 47 ]. The four subunits of CcO (PACID_12210-12230 and PACID_12290), the cytochrome C reductase (PACID_12260, PACID_12280 and PACID_12290) and cytochrome c biogenesis and assembly proteins were found in P. acidipropionici, P. acnes and M. phosphovorus, suggesting a greater flexibility of these strains to adapt to different oxygen conditions in comparison to P. freudenreichii, which lacks these genes. However, P. acidipropionici showed a frameshift in CcO subunit I (Figure S3 in Additional file\n 2 ). Because the CcO subunit I is a strongly conserved protein and the P. acidipropionici version have lost most of its domain identity and its copper-binding site, this oxidase is probably not functional. The alignment of the 454 and Illumina reads against the frameshift region can be viewed in the additional figure S4 (Additional file\n 2 ). A recent transcriptome work using P. acnes [ 48 ] detected expression of the whole respiratory chain under anaerobic growth conditions, including NADH dehydrogenase/complex I (NDH-1) and succinate dehydrogenase/fumarate reductase (SdhABC). Wackwitz and colleagues\n[ 49 ] showed that NDH-1 is stimulated under fumarate-dependent respiration in E. coli . Brzuszkiewicz and colleagues\n[ 48 ] hypothesized a scenario of P. acnes metabolism with ATP generation via F o F 1 ATP synthase and proton translocation by NDH-1 and SdhABC, with the former feeding reducing equivalents to the respiratory chain, and the latter using fumarate as terminal electron acceptor. These proteins were also detected in our proteome analysis of anaerobic fermentation samples, indicating a similar behavior in P. acidipropionici, capable to use different electron acceptors (NDH-1: PACID_26280-26410; F o F 1 ATP synthase: PACID_19340-19410; SdhABC: PACID_21310-21330). P. acidipropionici has two SdhABC/FrdABC gene clusters, SdhABC (PACID_21310-21330) and SdhA2B2C2 (PACID_18400-18420). Only SdhABC was detected in the anaerobic fermentation samples and hence could be acting as fumarate reductase, while SdhA2B2C2 could be used in the oxidation of succinate to fumarate. The membrane anchor subunits SdhC and SdhC2 have five transmembrane helices each and predicted molecular weight of 25 and 28 kDa respectively, characteristic for Sdh/Frd proteins lacking Subunit D. When the cluster contains a SdhD/FrdD subunit, the SdhC/FrdC proteins are distinctly smaller (13 to 18 kDa) and contain only three transmembrane helices each\n[ 50 ]. Given its importance in propionibacteria metabolism, a better understanding of electron transport chain can help design engineered strains with fine-tuned redox balance. Nitrate and sulfate reduction Kaspar (1982)\n[ 51 ] demonstrated that strains of P. acidipropionici , P. freudenreichii, P. jensenii, P. shermanii and P. thoenii can reduce nitrate to nitrite and further to nitrous oxide. This ability was later shown to vary from strain to strain, being strongly affected by environmental factors\n[ 52 ]. P. acidipropionici ATCC 4875 seems to be able to use nitrate as an electron acceptor, since all the genes of the respiratory nitrate reduction to nitrous oxide were all found: respiratory nitrate reductase (EC:1.7.99.4; PACID_02700-02730), nitrite reductase (cytochrome) (EC:1.7.2.1; PACID_02330) and nitric-oxide reductase (cytochrome c) (EC:1.7.2.5; PACID_31710 and PACID_27170). These genes are also present in P. acnes and M. phosphovorus , while P. freudenreichii subsp. shermanii cannot reduce nitrate due to lack of a functional nitrate reductase\n[ 16 ]. Also, P. acidipropionici could reduce the nitrite to ammonium through the enzyme nitrite/sulfite reductase (EC:1.7.7.1; PACID_27260) and incorporate the ammonium formed into amino acids. This gene is also present in P. freudenreichii and M. phosphovorus , but not in P. acnes. Additionally, genes coding for an assimilatory nitrite reductase (EC:1.7.1.4; PACID_33080 and PACID_33090) are found in P. acidipropionici and M. phosphovorus . However, the enzyme of the former is nonfunctional due to a frameshift in the large subunit. The sulfur assimilation pathway also varies among propionibacteria, with species utilizing from the most oxidized sources (sulfate) to the most reduced ones (sulfide)\n[ 53 ]. An adenylylsulfate kinase (EC:2.7.1.25) was only found in M. phosphovorus. The coding genes for nitrite/sulfite reductase (EC:1.7.7.1; PACID_27260), sulfate adenylyltransferase (EC:2.7.7.4; PACID_01720 and PACID_01730) and phosphoadenylyl-sulfate reductase (EC:1.8.4.8; PACID_01710) coding genes were found in the genomes of P. acidipropionici, P. freudenreichii and M. phosphovorus, but not in P. acnes. However, it was not possible to identify the complete pathway of sulfate reduction in P. acidipropionici, thus requiring more studies. P. freudenreichii and P. acnes have genes coding for all three subunits of anaerobic dimethyl sulfoxide (DMSO) reductase, allowing these bacteria to use DMSO and other highly oxidized substrates as electron acceptors. Vitamin B12 biosynthesis Cobalamin (vitamin B12) is one of the most structurally complex nonpolymeric biomolecule described and is an essential cofactor for several important enzymes like methylmalonyl-CoA mutase\n[ 48 ]. The complexity of its synthesis makes a chemical production too challenging and expensive. Hence, the industrial production of this vitamin is currently performed by biosynthetic fermentation processes using two organisms: Propionibacterium freudenreichii (P. shermanii) and Pseudomonas denitrificans . Since some Propionibacterium species do not produce either endo- or exotoxins, they are preferred for the production of food additives or medicines. Pseudomonas produce Vitamin B12 by the aerobic pathway, whereas Propionibacteria use the anaerobic one, which differs from the first in that the cobalt ion is inserted much earlier, at the precorrin-2 intermediate. The oxidation step also differs as it cannot use O 2 and generates a 6- rather than a 5-membered lactone\n[ 54 ]. M. phosphovorus uses the aerobic pathway and the characteristic oxygen-requiring C-20 hydroxylase CobG and the distinct cobalt insertion complex (cobNST) were identified [Genbank: AP012204.1]. The organization of B12 biosynthesis genes of P. acidipropionici is similar to that of P. acnes , with genes grouped into two clusters (Figure S5 in Additional file\n 2 ), while P. freudenreichii has its genes grouped into four gene clusters. The small cluster (PACID_28120-28220) harbors genes responsible for providing aminolaevulinic acid and converting it to uroporphyrinogen III, as well as the interconversion of uroporphyrinogen III into haem. The large cluster (PACID_08770-09000) harbors genes responsible for cobalt transport and adenosylcobalamin synthesis from uroporphyrinogen III. Some genes coding for transport proteins are only present in the cluster from P. acnes. Fused genes also differ between the three genomes. While cbiEGH appears in the three Propionibacteria, cobT, cobU and bluB are organized differently. P. acnes large cluster harbors cobT and cobU genes fused into a single protein (cobTU), while P. acidipropionici large cluster contains two separate adjacent genes (cobT:PACID_08960 and cobU:PACID_08970). P. freudenreichii have cobU as a separate gene inside one cluster, and cobT fused to the bluB gene and located elsewhere in its genome. This fused gene, cobT/bluB, is also present in the P. acnes genome, but a bluB homolog was not found in P. acidipropionici genome. BluB, first identified in Rhodobacter capsulatus, was shown to be required for the conversion of cobinamide to cobalamin in various organisms, being responsible for the oxygen-dependent synthesis of dimethylbenzimidazole from reduced flavin mononucleotide (FMNH 2 )\n[ 55 ]. Since P. acidipropionici does not possess this gene, another unidentified enzyme may catalyze this reaction. Propionic acid fermentation As expected, the dicarboxylic acid pathway, known as Wood Werkman cycle, was identified in the P. acidipropionici genome (Figure\n 4 ). Many bacteria of different genera as Rhodospirillum , Mycobacterium , Rhizobium , Micrococcus and Propionibacterium have the mechanisms required for propionic acid fermentation metabolism\n[ 56 ]. Nonetheless, it is only in Propionibacterium that this is the main pathway for energy generation, resulting in high propionic acid production. Propionic acid production is a cyclic process, in which propionate formation is related to the oxidation of pyruvate to acetate and to reduction of fumarate to succinate. We identified thirteen genes in the P. acidipropionici genome that could encode the enzymes involved in propionate production from pyruvate: methylmalonyl-CoA carboxyltransferase (EC:2.1.3.1; PACID_07970-08000), malate dehydrogenase (EC:1.1.1.37; PACID_24560), fumarate hydratase (EC:4.2.1.2; PACID_33440), succinate dehydrogenase (EC:1.3.99.1; PACID_21310-21330), propionyl-CoA:succinate CoA-transferase (EC:2.8.3.-; PACID_06950), methylmalonyl-CoA epimerase (EC:5.1.99.1; PACID_19260) and methylmalonyl-CoA mutase (EC: 5.4.99.2; PACID_11200 and PACID_11210). Importantly, two main divergences from the literature were found. The phosphoenolpyruvate carboxytransphosphorylase (PEPC) that would lead to the formation of oxalacetate from CO 2 and phosphoenolpyruvate is not present in P. acidipropionici, P. acnes , P. freudenreichii or M. phosphovorus genomes. In addition, the acetate dissimilation pathway differs from previous reports\n[ 7 , 57 ]. P. acidipropionici does not contain the phosphate acetyltransferase (PTA) and acetate kinase (ACK) genes; therefore, the two step PTA-ACK pathway cannot be the one that is used for acetate assimilation or dissimilation as was thought previously. Instead, this bacteria seems to dissimilate acetate using an ADP-forming acetyl-CoA synthetase (ADP-ACS, EC:6.2.1.13 PACID_02150), an enzyme that is distinct from the broadly distributed AMP-ACS and which is instead related to the ADP-forming succinyl-CoA synthetase complex (SCSC)\n[ 58 ]. Unlike PTA and ACK, ADP-ACS converts acetyl-CoA, inorganic phosphate and ADP into acetate, ATP and CoA in one step\n[ 59 ]. The protein encoded by ADP-ACS gene was found in all proteome samples of P. acidipropionici , where acetate production was also observed. The PTA and ACK genes were identified in the genomes of P. freudenreichii and M. phosphovorus but not in the genomes of P. acidipropionici and P. acnes . An ADP-ACS gene was only identified in P. acidipropionici and related genes from the SCSC (α and β subunits) were also found in all the compared organisms except P. freudenreichii . The acetate assimilation is probably performed by AMP-forming ACS enzymes (6.2.1.1; PACID_13940 and PACID_13950). Although reversible in vitro, the reaction carried out by these enzymes is irreversible in vivo because of the presence of intracellular pyrophosphatases. In E. coli, this high affinity pathway scavenges small amounts of environmental acetate, while PTA-ACK pathway works only with large concentrations of the substrate\n[ 60 ]. In this way, P. acidipropionici resembles some halophilic archaea which do not have the PTA-ACK pathway and use AMP-ACS and ADP-ACS enzymes to assimilate and dissimilate acetate, respectively. n-Propanol production In addition to propionic, succinic and acetic acid, it has been reported that n-propanol might be produced under specific conditions; for example, in the fermentation of P. acidipropionici using glycerol as the carbon source\n[ 61 ]. In this case, n-propanol production by P. acidipropionici could have resulted by a necessity to balance the intracellular redox potential, since glycerol is more reduced when compared to glucose\n[ 62 ]. The genome analysis identified a possible metabolic pathway for n-propanol production: propionyl-CoA could be converted to propionaldehyde and then to n-propanol by oxidation-reduction reactions catalyzed by the aldehyde and alcohol dehydrogenase enzymes. Genes coding for aldehyde dehydrogenases that could act on propionyl-CoA are PACID_33630, PACID_02980 and PACID_24580. The genome of P. acidipropionici contains many alcohol dehydrogenase enzymes, and one or more could be responsible for producing propanol, including PACID_02970, PACID_04160, PACID_05040, PACID_05370, PACID_05800, PACID_06750, PACID_07130, PACID_08230, PACID_11040, PACID_11390, PACID_14120, PACID_16510, PACID_30120, PACID_30260, PACID_30590, PACID_30730, PACID_30890, PACID_31640, PACID_32980, PACID_33650 and PACID_33970. Indeed, to change the fermentation pattern of P. acidipropionici towards higher yields of reduced compounds like propionic acid and n-propanol, it would be necessary to modify the NAD(P)H:NAD(P) cofactor ratio. The influence of the NAD(P)H:NAD(P) ratio in metabolic pathways has been demonstrated, for example, by using substrates with different oxidation states\n[ 61 ] or by supplementing anaerobic growth with electron acceptors, such as nitrate and fumarate\n[ 62 ]. Some studies have also report that the cofactor ratio can be varied through NADH regeneration by the NAD-dependent formate dehydrogenase found in some species of yeast and bacteria. This enzyme catalyzes the oxidation of formate to CO 2 and the reduction of NAD to NADH\n[ 63 ]. Moreover, the intracellular redox potential could be modified in the presence of a low-potential electron mediator, which transfers electrons between a working electrode and a bacterial cell. Fermentation studies carried out with different strains of Propionibacterium in a three-electrode amperometric culture system showed that high yields of the reduced compounds could be produced from oxidized substrates, such as glucose and lactate\n[ 64 , 65 ]. n-propanol production was also observed in bioelectrical reactors with P. acidipropionici ATCC4875 fermenting sucrose in different mediator concentrations\n[ 66 ]. CO 2 fixation Heterotrophic CO 2 fixation was reported for the first time in P. pentosaceum , later renamed as P. acidipropionici [ 67 ]. Wood and Leaver\n[ 68 ] tested the fermentation of 3, 4, 5 and 6 carbon compounds by this bacterium and found that the best CO 2 assimilation occurred when cells were grown on a medium containing glycerol, yeast extract, phosphate and vitamin B. In the same study, they also reported the inhibition of CO 2 fixation by NaF (Sodium Fluoride). In 1961, Siu et al. [ 69 ] reported that the PEPC enzyme obtained from sonic extracts of P. shermanii was able to catalyze the formation of oxaloacetate from CO 2 and phosphoenolpyruvate. PEPC is widespread in plants, algae, cyanobacteria, bacteria and protozoa and has a highly conserved domain; the bacterial proteins have approximately 870 residues\n[ 70 ]. However, similarity searches did not identify a homolog of PEPC in the genomes of P. acidipropionici P. acnes , P. freudenreichii and M. phosphovorus . To investigate the CO 2 assimilation by P. acidipropionici we performed fermentation assays using glycerol as the substrate in the presence of 13 CO 2 . The NMR results (Figure\n 6 ) revealed an increase in the carbonyl group peak of propionic acid in relation to the control, indicating that a certain amount of isotope 13 C was fixed by the CO 2 assimilation pathway. Hence, we concluded that P. acidipropionici can indeed fix CO 2 ; however, the enzyme responsible for this reaction may not be PEPC. An inspection of the P. acidipropionici genome led to the identification of a gene encoding pyruvate carboxylase (EC 6.4.1.1, PACID_00400), an enzyme that catalyzes the following reaction: (none) Pyruvate + ATP + HCO 3 − + H + → oxalacetate + ADP + phosphate Figure 6 P. acidipropionici culture under CO 2 atmosphere. A , Culture under 13 CO2 (labeled) and B , Culture with CO 2 (unlabeled). A pyruvate carboxylase was also identified in the M. phosphovorus genome (MLP_46180) but not in the P. acnes and P. freudenreichii genomes. CO 2 assimilation by P. acidipropionici occurs mainly when glycerol is used as the carbon source\n[ 68 ]. Therefore, this pathway could be related to the necessity to balance the intracellular redox potential when the glycerol reaches a high reduction potential. Moreover, in the presence of glucose as carbon source the gene for carbon dioxide fixation could be repressed due to a carbon catabolite repression mechanism\n[ 71 ]. The proteome experiments led to the identification of this protein in all samples obtained from P. acidipropionici cells growing in glycerol, but not from cells grown in glucose (data not shown). Although this result supports our hypothesis, further experiments should be performed to elucidate the CO 2 assimilatory pathway and its regulation. The pyruvate carboxylase is member of the biotin-dependent carboxylase family. Members of this family are post-transcriptionally biotinylated by a biotin protein ligase, usually present at one copy in prokaryotes and able to biotinylate different carboxylases. Prokaryotes have either one monofunctional protein that only performs the biotinylation or a bifunctional protein containing an N-terminus regulatory domain that participates in the transcriptional control of genes involved in biotin biosynthesis\n[ 72 ]. The four bacteria compared seem to have one copy each of the monofunctional protein, since the typical N-terminal biotin operon repressor domain [TIGRFAMs: TIGR00122] was not found in these proteins. The heterotrophic CO 2 fixation emphasizes the potential use of P. acidipropionici in industrial applications, once it enables the industrial propionic acid production from low feedstock amounts and fermentation released CO 2 ."
} | 12,994 |
26867858 | null | s2 | 3,931 | {
"abstract": "Photosynthesis in C3 plants is limited by features of the carbon-fixing enzyme Rubisco, which exhibits a low turnover rate and can react with O2 instead of CO2 , leading to photorespiration. In cyanobacteria, bacterial microcompartments, known as carboxysomes, improve the efficiency of photosynthesis by concentrating CO2 near the enzyme Rubisco. Cyanobacterial Rubisco enzymes are faster than those of C3 plants, though they have lower specificity toward CO2 than the land plant enzyme. Replacement of land plant Rubisco by faster bacterial variants with lower CO2 specificity will improve photosynthesis only if a microcompartment capable of concentrating CO2 can also be installed into the chloroplast. We review current information about cyanobacterial microcompartments and carbon-concentrating mechanisms, plant transformation strategies, replacement of Rubisco in a model C3 plant with cyanobacterial Rubisco and progress toward synthesizing a carboxysome in chloroplasts."
} | 245 |
39893176 | PMC11787297 | pmc | 3,935 | {
"abstract": "Lignocellulose, an abundant renewable resource, presents a promising alternative for sustainable energy and industrial applications. However, large-scale adoption of lignocellulosic feedstocks faces considerable obstacles, including scalability, bioprocessing efficiency, and resilience to climate change. This Review examines current efforts and future opportunities for leveraging lignocellulosic feedstocks in bio-based energy and products, with a focus on enhancing conversion efficiency and scalability. It also explores emerging biotechnologies such as CRISPR-based genome editing informed by machine learning, aimed at improving feedstock traits and reducing the environmental impact of fossil fuel dependence.",
"introduction": "Introduction The reliance on fossil fuels has become increasingly unsustainable in light of climate change. Fossil fuels, such as coal, oil, and natural gas, have been the primary drivers of environmental pollution and anthropogenic greenhouse gas (GHG) emissions worldwide 1 . Unprecedented levels of atmospheric carbon dioxide (CO 2 ) are leading to a rise in global average temperatures 2 , which in turn, has triggered a cascade of environmental challenges, including more frequent and severe weather events, rising sea levels, disruptions to ecosystems and biodiversity, and have materialized into significant threats to the well-being of humanity 3 . Consequently, interest in renewable natural resources and alternative energy solutions has surged, driving increased efforts into climate change mitigation. This shift is endorsed by international policies and regulatory frameworks, including the United Nations Paris Agreement, which set a target to keep global warming at well below 2 °C, with an ideal ceiling of 1.5 °C relative to pre-industrial levels 2 , 4 . Many nations have also committed to achieving net zero GHG emissions by 2050 2 , 5 . Renewable energy sources, such as wind, solar, and hydroelectric power, offer cleaner, sustainable alternatives to fossil fuels by converting natural energy into electricity without emitting GHG during operation. These technologies are vital for reducing carbon emissions, but their widespread adoption faces obstacles, including geographical constraints, intermittency, and technological readiness to up-scale 6 , 7 . Moreover, they cannot fully replace fossil fuels in all sectors, such as petrochemical production for industries traditionally reliant on fossil fuels, including maritime transport and aviation 7 . According to the U.S. Department of Energy’s (DOE) 2023 Billion-Ton Report, despite significant advances in battery technology for electrification, biomass feedstocks will continue to be critical for renewable energy production and bioproducts, particularly for biofuels and industrial applications that cannot easily transition to electrification 8 . One promising source of biomass energy is lignocellulosics derived from bioenergy crops and woody plants. Lignocellulose is the most abundant form of biomass on Earth 9 , accounting for approximately 57% of the planet’s biogenic carbon 10 . Innovation toward deploying biomass energy has been projected to displace 30% of fossil fuel consumption in the near future with biofuels, biochemicals and biomaterials 7 . Unlike fossil fuels, which release CO 2 into the atmosphere from permanently sequestered subterrestrial carbon sinks, sustainably managed lignocellulose biomass can be carbon-neutral while alleviating food-security concerns of starch-based bioenergy by not competing with agricultural land used for food production 11 . Despite its tremendous potential, the development and utilization of lignocellulosic feedstocks have not yet met industrial expectations. Bottlenecks include production scalability 12 , long production cycles and timelines, bioprocessing efficiency, and cost-effectiveness 13 , 14 , as well as concerns over the adaptation and resilience of feedstocks to climate change. Biotechnological advances to modifying lignocellulosic feedstocks offer enormous potential to enhance biomass yield, improve bioenergy traits, reduce agricultural inputs, and increase resilience to climate change and pathogens. Combined with innovations in bioprocessing methods and feedstock pretreatments, this integrative approach creates transformative opportunities for scaling up bio-based applications to meet the global demands for sustainable energy and materials. This holistic approach not only addresses critical environmental challenges but also broadens the scope of feedstock utilization into emerging and diverse sectors 15 . Here, we review current efforts and future perspectives in harnessing lignocellulosic feedstocks for bio-based energy and products, with a focus on reducing environmental impacts of fossil fuel dependence. Advances in feedstock applicability and utilization across various bioprocessing and feedstock engineering strategies to enhance conversion efficiency and scalability are described. Additionally, we explore the potential of emerging biotechnologies to accelerate trait modification and expanding feedstock capabilities in response to the rapidly evolving challenges of climate change, at a time when CRISPR-based genome editing technologies are informed by democratized artificial intelligence (AI) resources and sophisticated machine learning (ML) models that enable next-generation mutagenesis and breeding in lignocellulose feedstocks. Existing and future lignocellulosic-based applications Lignocellulosic feedstocks have a long and impactful history in human civilization and remain deeply embedded in modern society. Traditionally, their use has centered on applications such as heat and energy production, lumber for construction and furniture, and pulping of cellulosic fibers. However, recent advances in biomaterials science have opened the door to a wide array of emerging applications. Lignocellulosic biomass predominantly consists of the thickened secondary cell walls of vascular plants, which are mainly composed of cellulose and hemicelluloses, impregnated with lignin 16 , 17 . This section provides an overview of both current and emerging applications for the individual components of lignocellulose (Fig. 1 ). Fig. 1 Pathway to large-scale bio-based production. a The choice of biomass (e.g., energy crops or woody plants) depends on factors like availability and recalcitrance level, which in turn influence the required bioprocessing methods, treatment approaches, and trait improvements; b Overview of the processing and treatment of lignocellulosic biomass; c Optimizing feedstock composition for efficient conversion, improving plant growth, and increasing resilience to biotic and abiotic stresses are critical for ensuring a stable and sustainable biomass supply; d Large-scale production of bio-based applications can be achieved by integrating optimal feedstock selection, processing methods, treatments, and trait improvements, contributing to sustainable solutions in the face of climate change. Cellulose based applications Cellulose is the most abundant biopolymer on earth. It is primarily used in a fibrous form for paper-based products and textiles, extracted through pulping. Fiber characteristics vary widely across biomass sources, providing flexibility for different applications. Softwood fibers, known for their length and coarseness, are ideal for packaging, sanitary, and hygiene products due to their bulk and absorbency. In contrast, hardwood fibers are preferred for soft tissues 18 . Highly pure cellulose fibers ( > 90%), known as dissolving pulp, are used in clothing (textiles) and are also processed into cellulose derivatives. These derivatives have broad applications, including the production of bioplastics for food and medical industries 19 . The exploitation of cellulose at the nano- to micro-scale has started a green revolution, with the successful implementation of large-scale manufacturing of cellulose nanomaterials. Renewable nanomaterials 20 represent the next generation of nanotechnology, outperforming their synthetic analogs owing to their sustainability features. Nanocelluloses not only retain the inherent properties of cellulose but also exhibit key characteristics of nanomaterials, such as a large surface area, versatile reactive sites, and stable scaffolding for hosting nanoparticles 21 . Applications of nanocelluloses are being explored across various fields, including energy storage, photovoltaic devices, thin-film transistors, medical applications like biomedical scaffolds and drug delivery, functional barrier packaging, and mechanical reinforcement in composites or concrete, among others 22 , 23 . Recently, top-down methods of wood engineering leveraging cellulose fibrils while maintaining wood anisotropic structure has gained increasing interest, such as densified, cooling, thermal insulating, and transparent wood 24 , 25 . These advanced wood materials are part of the broader push towards sustainable and energy-efficient building solutions, combining the natural benefits of wood with enhanced physical and chemical properties that could replace traditional materials in various applications. Finally, the conversion of cellulose into biofuels and renewable chemicals has gained worldwide attention due to climate change and the projected 6% increase in total energy demand by 2030, as estimated by the International Energy Agency 26 . The significance of cellulose-based applications, derived from biomass through sustainable technologies, in addressing social and environmental challenges cannot be overstated, as they offer solutions for reducing dependence on fossil fuels, minimizing waste, and promoting resource efficiency while contributing to the development of a circular economy and mitigating climate change impacts. Hemicellulose based application Unlike cellulose, hemicelluloses are amorphous and heterogenous in chemical composition 27 . In the pulping processes, the majority of hemicelluloses are degraded and burnt to produce energy. Meanwhile, in biorefinery, hemicelluloses are often partially extracted in pretreatments to ensure efficient enzymatic digestion of cellulose, and the residual part is converted to pentoses and hexoses by hemicellulases 28 . The monomeric sugars from hemicelluloses can be further processed into bio-ethanol by fermentation or catalytically converted into platform chemicals, such as sorbitol, xylitol, and 2,3-Butanediol 28 . Hemicelluloses have been shown to improve the mechanical strength of paper when used as a wet-end additive 29 , enhance food properties 28 , and serve as an alternative source for pharmaceutical and healthcare products 30 . Other emerging hemicellulose-based applications include films for packaging, absorbent material, aerogels, raw materials for carbon quantum dots and porous carbons 27 , 28 , 31 . Lignin based application Remarkably, lignin represents the relatively highest carbon content ( ~ 60%) of the lignocellulosic biomass 32 due to its polymer size and naturally branched structure. However, it remains largely underutilized. To date, lignin is principally burned for heat and energy production (e.g., in pulp and paper mills) 33 . Research efforts have been directed towards valorization of the lignin stream. However, the complexity and heterogeneity of the lignin polymer, along with its recalcitrance to depolymerization, pose considerable challenges to lignin valorization for producing biochemicals and bioproducts. Promising conversion methods, including hydrogenolysis, acidolysis, and depolymerization, have been exploited to convert lignin into value-added products 34 , 35 . The use of specific solvents, organic compounds (e.g., formaldehydes) and metal formates have been investigated to promote lignin depolymerization through C-O bond cleavages, ring opening, and subsequent oxidation, meanwhile preventing the recondensation reactions of the lignin fragments. Among the possible lignin derivatives that can be produced, vanillin has been the major phenol aldehyde, industrially manufactured from the guaiacyl and syringyl units of lignin 36 . While chemical approaches have been the primary means to lignin valorization, the use of bacteria such as Rhodococcus 37 and Pseudomonas 38 species have also shown promise in the production of biopolymers (e.g., polyhydroxyalkanoates, PHAs) or polymer precursor molecules (e.g., nylon 39 ) from depolymerized lignin. Recently, studies are investigating the potential of the filamentous white-rot fungi for lignin valorization 40 . Processing and treatments Conversion of lignocellulosic feedstocks into biofuels and bioproducts involves a series of successive, often integrated steps, from feedstock selection, preparation and pre-processing to conversion and valorization through principally, thermochemical and biological means. Advances in technologies and sustainable practices continue to drive efficiency and productivity improvements of these processing and treatment steps, while expanding the range of economically viable biomass-derived, value-added products. Yet, the conversion of lignocellulosic feedstocks into biofuels and bioproducts still faces major challenges that curb enthusiasm for industrial deployment, given the inherent difficulty of breaking down this complex and recalcitrant polymer. The development of biorefineries 12 , 41 , or integrated systems that combine farming and conversion processes to produce bioenergy and biomaterials, has been proposed as a sustainable industrial alternative to fossil-based refineries. These biorefineries aim to enhance scalable production across industrial sectors, address climate change, reduce dependence on fossil fuels, and stimulate job creation and development in rural areas. The conversion of biomass into fuels, chemicals, and energy occurs through various processes, including liquefaction, fractionation, hydrolysis, pyrolysis, gasification, catalysis, and fermentation. While there have been several examples of lab-scale and industrial-scale implementation, expanding research and development at scale is needed to overcome biomass conversion bottlenecks and techno-economic challenges, while minimizing environmental impact 12 . Feedstocks is the starting point of a biorefinery Starch and lignocellulosic biomass in all its forms remain a large reserve of fixed carbon that could be used to displace petroleum as a source of fuels and chemical products. The source, availability, type and recalcitrance level of the biomass dictate the selection, runnability and productivity of the main converting operations 42 (Fig. 1a ). First-generation biorefineries 43 rely on food-grade feedstocks for biofuel production, which has led to growing socio-economic and environmental concerns. Also, their excessive use of nitrogen fertilizers can result in the emission of nitrous oxide (N₂O), a GHG more potent than CO₂ 34 . Alternatively, lignocellulosic feedstocks, such as wood and energy crops, have driven the development of second-generation biorefineries 44 . These feedstocks could lower GHG emissions, may be cultivated on marginal lands that are unsuitable for food crops, while also minimizing the strain on scarce water resources. However, the type of lignocellulosic biomass and its intrinsic properties significantly impacts the cost, productivity, and efficiency of the conversion process 44 . Studies show that feedstock supply logistics and handling can account for nearly 50% of the total cost of biofuels and bioproducts 45 . Pretreatment strategies of lignocellulosic biomass The pretreatment step is an essential part of the biomass conversion process 34 , 42 (Fig. 1b ). It aims to breakdown the recalcitrant structure of the lignocellulosic biomass to enzymatic or microbial deconstruction and allows for separation and further treatment of each of the biomass components (i.e., cellulose, hemicelluloses and lignin). This step is also key to maximize transportation efficiency through biomass size reduction, drying, and compaction. Ideally, the pretreatment process should result in maximum recovery with minimum degradation of the biomass components, minimum capital and operating costs, and should be applicable to a variety of lignocellulosic materials. However, most conventional pretreatment techniques are limited by their high energy demand and high cost and can form inhibitory –potentially toxic– intermediates and by-products, resulting in low productivity and the need for additional costly steps 42 . Pretreatment methods can be categorized into physical, chemical, physicochemical, and biological methods 34 , 42 , 46 . Physical pretreatment uses mechanical forces like friction and shear to reduce biomass size, increase surface area, and improve reaction rates, such as for saccharification. When combined with fractionation methods (e.g., sieving), particle size control can be achieved. However, these methods require significant mechanical and electrical energy, making them costly, time-consuming, and eventually damaging the lignocellulose. Although large-scale equipment is available, particularly in the pulp and paper, and polymer industries (e.g., refiners, ball-milling, extrusion), these processes still face efficiency challenges. Chemical pretreatments require less energy but involve chemicals and digesters, which can raise costs and environmental concerns. Originally established in the pulp and paper industry, processes like Kraft, sulfite, organosolv or dissolving pulping remove controlled amount of lignin for further biomass conversion 47 . While effective for paper, packaging, and cellulose products, these methods are less suitable for biofuels. Acid pretreatment removes hemicellulose and produces inhibitors, while alkali pretreatment can dissolve too much biomass. Residual lignin also limits sugar yield, and alternative methods like organosolv and ionic liquids are costly and not yet scalable due to limited solvent/chemical recovery systems. Physicochemical pretreatments commonly employ high pressure and temperature to alter biomass structure by disrupting intra- and inter-molecular linkages. These pretreatments often require specialized reactors or costly setups. Steam explosion works well on hardwood and agricultural residues, but liquid hot water pretreatment is ineffective at removing lignin from hardwood and softwood. Wet oxidation effectively removes lignin but requires oxygen or air and may use catalysts, further increasing costs 42 . Alternatively, biological pretreatments offer sustainable, cost-effective alternatives, operating under mild conditions without chemicals or high energy input. Specialized enzymes and fungi can facilitate lignin degradation and partially remove hemicelluloses without producing toxic byproducts, eliminating the need for a neutralization step before saccharification. This process can be integrated with enzymatic hydrolysis for sugar conversion, promoting a safer, more sustainable biorefinery. However, improving reaction rates, titer, and sugar yields, along with large bioreactors and continuous monitoring, is needed for large-scale use 42 . Combining different pretreatment methods has been explored to improve sugar yields and conversion efficiency. While this synergy accelerates biomass breakdown and enhances conversion rates, they still results in high operational costs. Moreover, since conversion efficiency depends on biomass type and structure, finding a universal pretreatment method for all lignocellulosic materials remains elusive. Pretreatment of biomass is estimated to contribute approximately 40% of the total production costs in biomass conversion, with these costs depending on the specific pretreatment method used. The remaining 60% includes costs for enzymatic hydrolysis (the saccharification step), which alone can account for nearly 25% of the overall production costs due to enzyme prices, along with additional costs for energy, chemicals, and reactor operation 41 , 42 . Substantial research efforts are needed to optimize the pretreatment step, ensuring higher yield and productivity at lower costs. Over the last few decades, bioconversion of lignocellulosic feedstocks has primarily focused on the carbohydrate components, resulting in the development of methods to remove lignin, the primary recalcitrance to biomass utilization 33 . However, for specific bioproducts like cellulose nanomaterials, the presence of lignin in the biomass can, instead, show benefits for endowing the carbohydrates with additional performance such as hydrophobicity and resistances against biological degradation 48 , 49 . It is essential to select pretreatment methods that align with the desired properties of specific biofuels and bioproducts. Saccharification of lignocellulosic biomass Biomass pretreatment is usually followed by enzymatic hydrolysis that converts the extracted carbohydrates into soluble sugars (Fig. 1b ). Enzymatic hydrolysis has proven to be quite effective and cost-efficient with reported conversion rates that can exceed 80% 41 . This step utilizes a cocktail of enzymes that selectively hydrolyze either hemicelluloses or cellulose. The high cost of enzymes, especially for large-volume sugar productions, remains a barrier for industrial scalability of high-efficiency biorefineries 50 . While some enzymatic cocktails can be an economically viable solution for ethanol production, the enzymatic hydrolysis step only converts the hexose (C6) sugars, leaving behind the pentose (C5) sugars along with proteins and other carbon sources as residues. Given the growing need for a more sustainable carbon solution, the complete utilization of both hexoses and pentoses would significantly improve the overall biorefinery carbon efficiency. Genetic engineering possibilities, for instance, are being investigated to improve the fermentation performance of microorganisms 51 , 52 , while other novel contributions involve the development of statistical tools and computational models to design, predict, and validate the performance and stability of enzyme cocktails 33 . From an economic perspective, combining enzyme hydrolysis with pretreatment or fermentation could create a path towards improving the process economics of other fermentation products (e.g., butanol), since such a procedure has been successful with ethanol production 50 . Feedstock utilization challenges for industrial use Cutting-edge technologies, in bioprocessing and feedstock genetic modification, have the potential to overcome key challenges related to the economic viability and environmental impacts of lignocellulosic feedstock, enabling their industrial implementation for commercial applications. According to the Billion-Ton report by the U.S. Department of Energy, the future supply of the lignocellulosic biomass for the U.S. could achieve 174 million dry metric tons per year for bioenergy use 53 , highlighting the significant potential for large-scale deployment. Life-cycle analysis, encompassing steps from raw material extraction to end-of-life 54 , 55 , has demonstrated reduced carbon footprints and environmental impacts compared to traditional industrial methods in several bio-based applications, including biofuels 56 , hydrogen 57 , renewable natural gas 58 , sustainable aviation fuel 59 , 60 , electricity 58 , biogas 61 , nanocelluloses 62 , platform chemicals 63 , and structural materials 64 . Additionally, techno-economic assessments looking at economic indicators like minimum selling price, internal rate of return, and net present value are key to overcome the issues in a sustainable manner for large-scale practices. The expansion of the circular bioeconomy enhances environmental sustainability and economic welfare, besides providing social benefits for future generations 65 . Among the ongoing challenges is the need for effective detoxification methods to neutralize fermentation-inhibitory compounds generated during biomass pretreatment 66 . These methods are crucial for enhancing conversion rates and efficiency, such as achieving higher biofuel yields. However, these benefits are accompanied by additional costs. The recovery of enzymes and chemicals/solvents is another aspect to address for decreasing the environmental footprint of a biorefinery and improve its economics 67 . The use of green solvents and chemicals (i.e., ionic liquids), development of solvent recovery systems, implementation of on-site enzyme production technologies and integrated conversion steps are all strategies being investigated to minimize environmental impacts and improve process economics 12 . In addition, cost-effective strategies for management and exploitation of each of the generated residual streams into viable, value-added products are critical for commercial viability. A plethora of other novel value-added products from lignocellulosic feedstocks are within reach. For instance, biochar, a carbonizing material with a high carbon content, can be derived from the thermo-chemical decomposition of lignocellulosic materials by pyrolysis or gasification in the absence of oxygen 68 . Biochar can be used as a soil supplement, a water treatment or for CO 2 capture and sequestration 68 . Co-production of renewable nanomaterials, biochemicals and biofuels is a promising way to offset some of the processing costs for biofuel and biochemical production by generating new revenues through market diversification and waste residues reduction 69 . Finally, an important driver for the industrial development of lignocellulosic bioenergy and bioproduct industries is the development of strategic policies and regulatory frameworks. Over the years, many countries have initiated national policies and mandates to reduce their dependence on conventional sources of energy and combat climate change. The United States biofuels initiative was kicked off in 2001 with the authorization of blending E10 and E15 (i.e., 10% and 15% ethanol, respectively) with gasoline for all cars 70 . The US Energy Independence and Security Act (EISA) set up a target production of 136 billion liters of biofuels by 2022 to meet 18% of the nation’s transportation fuel demands 71 . Similar initiatives have been implemented by the European Union, China, India, and others with the intend of achieving percent biofuel blending mandates and strengthening the global development of biorefineries 68 . Modifying feedstocks to optimize biomass conversion efficiency Alongside advancements in bioprocessing technologies, optimizing the composition of lignocellulosic feedstocks for efficient conversion is essential for ensuring a sustainable feedstock supply, particularly in the context of climate change (Fig. 1c ). Lignin represents a significant impediment in the processing of lignocellulosic biomass for various downstream applications 16 , 17 , 72 . Lignin is formed by free radical polymerization of three major monolignol precursors, p -coumaryl, coniferyl, and sinapyl alcohols, which form the p -hydroxyphenyl, guaiacyl, and syringyl (in angiosperms) units in lignin. The monolignols are biosynthesized from phenylalanine through a series of enzymatic reactions in a metabolic grid consisting of at least 24 metabolites and 11 enzyme families. The pathway is regulated in a multilayer manner, encompassing transcriptional, post-transcriptional, and post-translational modifications, protein-protein interactions, and metabolic regulations 72 . A hierarchical transcription regulatory network, involving transcription factors like MYBs and NACs, controls secondary cell wall formation in plants by coordinating the spatial-temporal expression of lignin biosynthetic genes 73 – 76 . Non-coding RNAs (ncRNAs) like miRNAs and lncRNAs 77 , 78 , along with alternative splicing mechanisms 77 , further regulate lignin biosynthesis by targeting TFs and enzyme-encoding genes involved in the pathway. Protein-protein interactions 79 and post-translational modifications 80 , such as phosphorylation and ubiquitination, modulate the metabolic fluxes, and alter the activity or stability of enzymes, resulting in changes in lignification. Understanding how these multilayer regulations respond to developmental and environmental stimuli is crucial for engineering plants with modified lignocellulosic properties, which could improve biomass utilization for industrial applications. Given the importance of lignin for biomass conversion efficiency, its modification has been a major focus for lignocellulosic feedstock engineering 81 . Downregulation of monolignol biosynthesis genes can lead to significant changes in lignin content, subunit composition, and lignin–carbohydrates linkages. These alterations have been demonstrated to enhance biomass deconstruction, saccharification, and fermentation efficiencies in numerous energy crops and woody species such as Populus spp ., Pinus spp ., Panicum virgatum , Miscanthus sinensis , and Mendicago sativa 82 . In P. virgatum and M. sinensis , the downregulation of aldehyde O-methyltransferase ( AldOMT ) led to a reduction in lignin content by up to 15% and 63%, respectively 83 , 84 . Low lignin transgenic P. virgatum showed up to 38% increase in saccharification efficiency and 42% increase in ethanol yield per unit of biomass 83 . In P. trichocarpa , Wang et al. generated ~2000 engineered P. trichocarpa trees, downregulating 21 monolignol genes individually, in pairs, or entire gene families. Some transgenic lines showed substantial changes in wood properties, leading to improvements in glucose and xylose release by up to 351% and 828%, respectively, compared to wild-type trees 17 . These transgenic poplars also increased biomass solubilization via microbial digestion by Caldicellulosiruptor bescii , from 20% in wild-type trees to 79% in transgenic trees, resulting in a 7.6-fold improvement in bioethanol production 52 . Modifications to lignin composition and its interactions with other cell wall components can lead to architectural changes in the plant cell walls. The heterologous expression of feruloyl–coenzyme A monolignol transferase ( FMT ) and 3-dehydroshikimate dehydratase ( QsuB ) in P. alba × grandidentata resulted in the incorporation of non-canonical conjugate subunits into lignin 85 , 86 . These subunits introduce readily cleavable ester bonds in the lignin backbone and facilitate the deconstruction of feedstocks and subsequent release of sugars 85 , 86 . The overexpression of the callose synthase gene in P. tremula × tremuloides induces callose deposition in the secondary cell walls, which is not typically observed in most plant species. Callose deposition modulated cell wall porosity, water, and lignin contents, and increased the lignin–cellulose distance. These modifications ultimately resulted in a substantial reduction in biomass recalcitrance, evidenced by improvements in enzymatic hydrolysis 87 . Research on modifying cell wall polysaccharides has been less extensive than lignin. Perturbations to cellulose and hemicellulose biosynthesis have been achieved by manipulating secondary cell wall TFs, cellulose synthases (CesA), sucrose synthases (SuSy), and glycosyltransferases 88 , 89 . The downregulation of glycosyltransferases from the GT8 , GT43 , and GT47 families decreases xylan content, the predominant hemicellulose in angiosperms, resulting in reduced feedstock recalcitrance in woody species 90 – 92 . In P. virgatum and P. deltoides , the downregulation of Galacturonosyltransferase 4 ( GAUT4 ), a key enzyme in pectin biosynthesis, reduced cell wall pectin content and the molecular weight of hemicelluloses, while altering lignin composition, lignin inter-unit linkages, and lignin-polysaccharide cross-linkages. These alterations in the plant cell wall likely synergistically contribute to the reduced feedstock recalcitrance 92 , 93 . Advances in feedstock modifications for biofuels and bioproducts Despite notable advances in the genetic improvement of bioenergy and woody plants, challenges remain in developing lignocellulosic feedstocks with optimal bioprocessing, growth, and resilience to climate change. Altering the expression of cell wall-related genes often impairs growth and reduces biomass yield. Moreover, RNA interference (RNAi) and artificial microRNAs (amiRNAs) techniques to gene perturbation can lead to unpredictable outcomes, such as off-target effects or variability in transgene expression across different individuals and environmental conditions. These challenges contribute to inconsistent feedstock performance, restricting the practical application of trait improvements. Additionally, most studies have concentrated on individual traits (e.g., cell wall composition), with less attention given to the understanding the broader impacts of how bioenergy trait modifications on plant development and resilience. Climate change is exposing plants to more frequent and severe abiotic stresses, such as prolonged droughts and nutrient limitations 94 . Rising temperatures, increased CO₂ levels, and extended droughts also have been shown to increase the spread of pathogens and the vulnerability of crops and trees to diseases 95 . These environmental stresses ultimately lead to reduced growth and higher plant mortality and are projected to reduce plant productivity and decrease the global planted area by 2–9% 96 . Furthermore, agricultural expansion and urbanization are forcing non-food crops and tree plantations into less fertile, non-arable lands, limiting water and nutrient availability 97 , 98 . Given the clear interplay between biomass recalcitrance and feedstock scalability, fine-tuning characteristics of resilience, growth, and bioprocessing are essential for practical applications (Fig. 1c ). Expanding lignocellulosic trait enhancement by genome editing CRISPR-Cas genome editing technology has emerged as a powerful tool for inducing precise and heritable mutagenesis in lignocellulosic feedstocks, allowing for predictable modification of traits for bio-based utility. CRISPR-modified lignocellulosic feedstocks with loss-of-function mutations of single genes were successfully developed in energy crops and woody plants, leading to enhancements in feedstock composition 72 , 99 , bioprocessing 72 , 100 , engineered wood 101 , growth 102 , plant architecture 103 and resilience to climate stress 104 . However, the potential of CRISPR extends far beyond canonical single-gene loss-of-function mutations. Recent advancements in CRISPR now allow for more complex and powerful editing modalities, encompassing multiple-gene mutations, whole plants or tissue-specific alterations, the transcriptional regulation of thousands of genes, and even editing entire chromosomes. This versatility expands the scope of modifications possible for enhancing feedstocks, particularly in response to challenges posed by climate change (Fig. 2 ). Fig. 2 Advanced CRISPR-based genome editing technologies for expanding the range of possible lignocellulose feedstock modifications. a CRISPR-mediated gene knockout at the whole-plant, tissue-specific, or cell-type-specific levels. b Gene expression modulation achieved through transcriptional regulators fused to a catalytically dead Cas9 (dCas9) or by editing gene regulatory sequences. c CRISPR-mediated chromosome rearrangements to modulate genetic linkages for the inheritance of desired traits. CDS Coding sequence, UTR Untranslated region, TSS Transcription start site. Targeted editing and spatial regulation in feedstock modifications Tissue-specific CRISPR editing has been demonstrated in A. thaliana , Lycopersicon esculentum , and Gossypium hirsutum using tissue-specific promoters to drive Cas9 expression, enabling editing of specific cell types and tissues (e.g., stomatal lineage, fibers, root-caps, lateral roots and fruits) 105 – 108 . While this approach has not yet been demonstrated in bioenergy crops and woody feedstocks, it holds promise for enhancing woody tissues while preserving other cell types to preserve plant growth and development. Recent studies have demonstrated spatial modulation of lignin biosynthesis. Target suppression of LTF1 , a transcriptional repressor of lignin biosynthesis, was achieved in a fiber-specific manner. Concurrent loss-of-function editing of Cinnamoyl-CoA reductase 1 ( CCR1 ) while overexpressing a modified version of the same gene in vessel cells, resulted in significant improvements in wood bioprocessing and saccharification efficiency 16 , 109 . These findings highlight the potential for spatial regulation of lignin biosynthesis, avoiding disruptions to vessel lignification that could lead to vessel collapse and preventing plant developmental problems. Targeted editing of regulatory sequences Recent efforts have increasingly explored the mutagenesis of non-coding sequences to fine-tune gene expression patterns. Manipulating cis-regulatory elements within promoters or 5’UTR regions can yield diverse quantitative changes across traits of interest, from photosynthetic efficiency to disease resistance. Introducing indels and sequence inversions into the 5’ untranslated region (UTR) of the photosystem II (PSII) subunit S ( OsPSBS1 ) in transgene-free O. sativa CRISPR lines produced photoprotective phenotypes 110 , opening the possibility of applying similar strategies across plant species to optimize photosynthetic efficiency and minimize energy loss during light transitions. Significant losses in CO 2 fixation during transitions from full light to shaded conditions are attributed to energy dissipation as heat by nonphotochemical quenching (NPQ) 111 , 112 . Overexpression of NPQ-related genes has been shown to enhance photosynthetic efficiency in transgenic Nicotiana tabacum and Glycine max , resulting in a 15% biomass increases and a 33% grain yield increase during light transition in greenhouse and field conditions 111 , 112 . However, unlike CRISPR-edited plants, this approach relied on transgene expression, which is susceptible to gene silencing or expression instability. By editing the regulatory sequences (e.g., suppressors) of other NPQ genes in combination with PSII , and extending this strategy to lignocellulosic feedstocks (NPQ genes are present in all plants 110 ), could offer promising opportunities for improving yield and resilience under field environments. Similarly, gene expression modulation through promoter editing has shown great potential in enhancing disease resistance in plants. Xanthomonas spp . use transcription activator-like effectors (TALEs) to transactivate susceptible host genes, promoting pathogen growth and symptom development 113 – 115 . Editing the promoter of the susceptibility gene CsLOB1 by introducing indels to disrupt the bacterial effector binding elements has enhanced resistance to canker disease in CRISPR-edited lines of Citrus sinensis and Citrus x paradisi 113 , 114 . Alternatively, introducing TAL effector binding elements from susceptibility genes into the promoter of dysfunctional resistance genes has enabled inducible TALE-dependent resistance in O. sativa 115 . Critical cis-elements in the promoters of pathogen-responsive genes have been identified in Populus ssp 116 . By editing these cis -elements, introducing new copies, or disrupting transcriptional regulator sites, gene expression patterns can be modulated to enhance disease resistance in lignocellulosic feedstocks, offering another promising avenue for improving resilience. Building on the strategies of editing regulatory sequences for gene expression modulation, recent innovations have expanded the potential of CRISPR-based systems to achieve more sophisticated modifications. One such approach involves fusing Cas9 or gRNA with transcriptional regulators (either activators or repressors) to simultaneously edit the genome and modulate gene expression 117 . The development of a CRISPR-Combo system exemplifies this dual functionality by enabling the concurrent activation of endogenous genes and targeted gene mutation 118 . This system was used to demonstrate enhanced plant regeneration and propagation of P. alba × tremula and O. sativa by activating morphogenic genes WUS and WOX11 while editing a lignin biosynthesis gene. The multiplex functionality of CRISPR-Combo is a step towards more complex metabolic engineering and sophisticated modulation of signaling pathways that require simultaneous gene editing and activation. Altogether, the integration of these techniques provides a versatile toolkit for optimizing plant traits through precise gene expression control. Modulating genetic linkages for the inheritance of desired traits Chromosome rearrangement is a promising approach for the development of feedstocks with novel traits. Plant attributes, including resilience to pathogens, adaptation to climate conditions, biomass yield, and quality, have been associated with single or multiple quantitative trait loci (QTLs) in different species, including woody plants like Populus ssp . and Salix ssp . and energy crops such as M. sinensis and P. virgatum 119 , 120 . The core genes associated with those traits are not always fully characterized, posing significant challenges for canonical approaches to target gene manipulation, especially at non-model plants like lignocellulosic feedstocks. Modulating the genetic linkages between a gene or QTL and other QTLs is crucial for ensuring the inheritance of desired traits across generations. This is particularly relevant for woody species with long generation cycles, where this approach could support effective breeding programs by enabling the development of strategic progeny genotypes that achieve consistent phenotypic outcomes more rapidly. Chromosomal inversions, permutations, and translocations can disrupt or stabilize genetic linkage groups through physical proximity in specific chromosome regions 121 . In A. thaliana , the utilization of frequent-cutter restriction enzymes induced multiple DNA double-strand breaks across the genome, resulting in chromosome rearrangements (inversions and translocations) rates up to 17.4% and 66.6% in diploid and tetraploid plants, respectively. Chromosome rearrangements have been used to improve biomass yield and tolerance to high-salinity 122 , 123 . Once QTLs for traits of interest have been identified, we can now inform the manipulation of the plant genome precisely using CRISPR-Cas to induce chromosome rearrangement at the desired locations. The first CRISPR-Cas-mediated inversions were recently achieved in plants and involved heritable inversions of 75.5-Mb and >17-Mb-long chromosome fragments in Zea mays and A. thaliana 124 , 125 , leading to suppressed meiotic recombination across nearly the entire chromosome in A. thaliana 125 . Similarly, targeted induction of heritable reciprocal chromosomal translocations was achieved in A. thaliana between chromosomes 1 and 2, and between 1 and 5, involving fragments of approximately 0.5 Mb and 1 Mb-long, respectively 126 . These pioneering studies offer optimism that comparable achievements may be replicated in energy crops and woody plants 121 , 124 – 126 , potentially eliminating or stabilizing linkage drags to enhance growth, yield, resilience, and bioprocessing efficiency. However, challenges may arise for lignocellulosic species without adequate genotypic and phenotypic data, and more studies will be necessary to identify relevant QTLs. Revealing complex genetic traits using CRISPR library and ML High-throughput forward genetic screens are powerful tools for deciphering novel genes and functions without prior knowledge. Even in P. trichocarpa , the first woody species with a sequenced genome, the number of annotated genes validated by experimental data remains limited. The functions of many genes are still speculative, often inferred through comparative genomics with distant model plants like Arabidopsis , underscoring gaps in functional genomic studies related to economically and ecologically significant traits in lignocellulosic feedstocks. Large-scale CRISPR gRNA library screening has been used for loss-of-function screens on a genome-wide scale to establish causal links between genotypes and phenotypes (Fig. 3a ). Despite their potential, it has been underutilized in plants, limited to a few food crops and model species 127 – 132 . Remarkably, a library of gRNAs targeting 34,234 genes (83% of all genes) in O. sativa was generated, resulting in the production of 91,004 loss-of-function mutants 127 . O. sativa and Z. mays lines generated from CRISPR libraries exhibited phenotypic changes across several traits, including growth, plant architecture, leaf morphology, and disease susceptibility 127 , 128 , 132 . Since the regulation of complex plant traits are typically multigenic, a multiplex CRISPR library can generate mutants that harbor concurrent edits in multiple genes, uncovering phenotypes that are obscured by genetic redundancy or epistasis. In A. thaliana , a CRISPR library of 59,129 gRNAs targeting concurrently two to ten genes within a gene family were designed for over 16,152 genes. This library enabled the identification of novel mutant phenotypes masked by genetic redundancy 131 . These CRISPR screens were based on loss-of-function mutagenesis. However, it would be very useful to apply CRISPR-based gain-of-function screens for trait engineering, using CRISPR activation systems such as CRISPR-Act3.0 133 . High-throughput CRISPR library of lignocellulosic feedstocks could significantly advance breeding programs by increasing genetic diversity and enhance bioenergy traits. Notably, further research is critical for expanding the availability of genome sequences for more species, maximizing the potential of new genome editing technologies in lignocellulosic feedstocks. Fig. 3 Identification of new genes and traits of interest for feedstock modifications. High-throughput CRISPR gRNA libraries targeting multiple genes, individually or in combination ( a ), along with the utilization of ML-based models ( b ), enable the identification of key genes and regulatory networks associated with desirable traits. ML is a powerful tool to rapidly uncover the non-linear relationship between genotypes and phenotypes and enable the identification of key genes and genetic networks involved in complex traits, when conventional genetic approaches fail to capture these regulatory complexities (Fig. 3b ). In P. trichocarpa , a predictive model based on ML for lignin biosynthesis was used to explore 69,123 multigenic editing strategies to identify optimal gene targeting combinations for improving the bioprocessing characteristics of wood for kraft pulping. CRISPR-edited trees showed significant alterations in wood properties, resulting in more efficient fiber pulping, and could increase productivity on an industrial scale, while reducing CO 2 emissions from pulp mills, thereby supporting a more sustainable and efficient fiber bioeconomy 72 . Feedstocks with properties tailored for applications other than pulping, such as bio-based fuels and products can be generated by integrating advanced computational tools with CRISPR-based applications to address environmental, social, and industrial needs 134 . A new generation of predictive models will likely be crucial for enhancing predictability and advancing feedstock improvements by integrating multi-omics data with powerful ML-based algorithms while accounting for the complexity, plasticity, and dynamic spatial-temporal responses of plants to environmental stimuli 135 . Challenges and frontiers in lignocellulose utilization CRISPR-Cas genome editing offers significant potential for identifying and targeting genes linked to valuable traits in lignocellulosic feedstocks. However, it faces notable challenges, primarily due to the variable feasibility of genetic transformation across different lignocellulosic species and genotypes. The most common method for transforming woody plants and energy crops is Agrobacterium tumefaciens -mediated transformation and particle bombardment 120 , 136 . Despite its effectiveness, A. tumefaciens infection is restricted to certain host species and genotypes 136 . Particle bombardment, though less genotype-specific, often introduces high copy numbers of DNA, and is overall inefficient. Another widely used method is PEG-mediated transformation of protoplasts (plant cells lacking cell walls) for the transient delivery of CRISPR-derived plasmids and ribonucleoprotein complexes (RNPs). With the plant cell walls removed, a major barrier to transformation is eliminated, allowing for high transformation rates and enabling precise, targeted changes introduced by CRISPR at desired locations. However, it requires well-developed protoplast regeneration systems. Regeneration of transformed plant cells to entire plants through tissue culture techniques remains a major obstacle, as many species and genotypes show limited response to conventional hormone-based regeneration systems. This limitation hampers the broader application of techniques such as genome editing, particularly in diverse plant groups like lignocellulosics 137 . Recently, various approaches have been explored to overcome the genotype dependence and low transformation rates associated with the limited regeneration capacity of some plants, alongside public initiatives from the U.S. Department of Energy (DOE) to define transformation and editing needs and barriers focused on bioenergy crops 120 . The utilization of less genotype-dependent explant tissues (e.g, meristematic tissues), screening and engineering of Agrobacterium strains and other non-pathogenic bacteria, exogenous application of phytohormones to pretreat explants before transformation, and ectopic expression of genes encoding morphogenic transcription factors (e.g., Bbm and Wus2 ) were demonstrated to increase the substantially the transformation rates of recalcitrant genotypes and species 138 . Concomitantly, tissue culture-free transformation methods, such as electroporation, cell-penetrating peptides, carbon nanotubes, viral vectors, and inoculation Agrobacterium strains suspensions onto the wound sites also holds promise for delivering cargos directly into plant tissues without the need for in vitro plant regeneration, though they are not yet universally effective. Enhancing transformation efficiencies in plant species now extends beyond traditional screening of culture media compositions and in vitro growth conditions 138 . Optimizing existing methods and developing new transformation techniques 138 , combined with insights into how differentiated cells reprogram to achieve pluripotency or totipotency for regeneration 139 , are essential for making genome editing technologies accessible across diverse plant species, and to provide sustainable lignocellulosic feedstock in a timely manner. While efforts toward biofuel innovations continue to advance, exciting new frontiers are emerging, including renewable nanomaterials, advanced wood products, bioplastics, and platform chemicals. These innovations broaden the versatility of lignocellulosic materials, offering sustainable alternatives across various industries. The future utility of lignocellulosic biomass hinges not only on improving existing applications but also expanding into new, high-value sectors, paving the path for a circular bioeconomy. New bioprocessing technologies are vital for industrial-scale implementation of lignocellulosic feedstocks. Improving pretreatment methods, enzyme recovery, and integrated biorefineries are essential to overcoming the economic and environmental challenges associated with utilizing lignocellulosic biomass. Life-cycle analyses demonstrated that these innovations could significantly reduce the carbon footprint of bio-based products, from sustainable aviation fuel to advanced nanomaterials. However, persistent challenges remain, including the removal of undesirable inhibitors, recovering valuable byproducts, and enhancing process efficiency to ensure the economic viability of biorefineries. Additionally, governmental policy and regulatory frameworks play a pivotal role in driving the large-scale adoption of these sustainable technologies. Without strong political endorsement and public support, scaling these innovations will remain constrained. Modifications to cell wall composition have significantly improved bioprocessing efficiency in energy crops and woody plants. However, these feedstocks have primarily been tailored for biofuels and platform chemical production, with a relatively limited focus on developing sustainable biomaterials for broader applications, including medical, electronic, and structural materials. Recently, CRISPR-mediated mutagenesis tailoring wood composition by reducing lignin content has enabled the production of densified woody materials without the need for chemical delignification, a step typically required in traditional wood engineering processes. Engineered wood is increasingly being explored as a sustainable alternative to conventional structural materials like steel, cement, glass, and plastic. These materials, produced through a combination of genetic and wood engineering, have the potential to lower costs and support the development of a CO₂-negative bioeconomy by providing renewable, environmentally friendly alternatives to traditional materials 101 . In addition, more work is needed to ensure feedstocks thrive under changing climate conditions and elevated environmental stresses. As drought, nutrient deficiency, and pathogen infections intensify, new traits must be explored to enhance feedstock resilience. CRISPR-Cas technology offers promising tools for accelerating plant breeding by enabling precise mutagenesis of the genome. However, the potential of CRISPR-Cas remains to be broadly democratized across lignocellulosic feedstocks. Beyond canonical applications, CRISPR-Cas can be expanded to edit regulatory sequences, perform tissue-specific mutations, use chromosome editing to induce the inheritance of desired traits, or conduct library screening for gene discovery. Furthermore, CRISPR-derived precise genome editing techniques like base editors, prime editors, and homology direct repair will augment our capabilities to introduce desirable gain-of-function traits 140 , which remain challenging to achieve with traditional CRISPR-Cas tools in plants. Additionally, beyond editing the genomic sequence, there are also opportunities to manipulate the transcriptome and epigenome by exploring the sophisticated CRISPR toolbox encompassing Cas proteins fused with diverse effectors such as transcriptional activators and repressors, as well as methylases and acetylases 117 . Similarly, integrating ML into breeding programs can accelerate the discovery of new genes and genetic networks, complementing these advanced CRISPR-Cas techniques. Looking ahead, interdisciplinary research, technological innovation, and policy support will be critical in overcoming these obstacles and fully unlocking the potential of lignocellulosic biomass. Collaboration between academia, industry, and policymakers is essential to creating an environment that fosters the growth of this sector, at speed and at scale. As the world strives to meet climate goals and transition to more sustainable energy and materials sources, lignocellulosic feedstocks offer a promising and perhaps the only path for reducing reliance on fossil fuels and advancing a sustainable bioeconomy."
} | 13,953 |
25162943 | PMC4145238 | pmc | 3,937 | {
"abstract": "Background Styrene is an important building-block petrochemical and monomer used to\nproduce numerous plastics. Whereas styrene bioproduction by Escherichia coli was previously reported, the\nlong-term potential of this approach will ultimately rely on the use of hosts\nwith improved industrial phenotypes, such as the yeast Saccharomyces cerevisiae . Results Classical metabolic evolution was first applied to isolate a mutant capable of\nphenylalanine over-production to 357 mg/L. Transcription analysis revealed\nup-regulation of several phenylalanine biosynthesis pathway genes including\n ARO3 , encoding the bottleneck enzyme DAHP\nsynthase. To catalyze the first pathway step, phenylalanine ammonia lyase\nencoded by PAL2 from A. thaliana was constitutively expressed from a high copy\nplasmid. The final pathway step, phenylacrylate decarboxylase, was catalyzed by\nthe native FDC1 . Expression of FDC1 was naturally induced by trans -cinnamate, the pathway intermediate and its\nsubstrate, at levels sufficient for ensuring flux through the pathway. Deletion\nof ARO10 to eliminate the competing Ehrlich\npathway and expression of a feedback-resistant DAHP synthase encoded by\n ARO4 K229L preserved and promoted the endogenous availability precursor\nphenylalanine, leading to improved pathway flux and styrene production. These\nsystematic improvements allowed styrene titers to ultimately reach 29 mg/L at a\nglucose yield of 1.44 mg/g, a 60% improvement over the initial strain. Conclusions The potential of S. cerevisiae as a host\nfor renewable styrene production has been demonstrated. Significant strain\nimprovements, however, will ultimately be needed to achieve economical\nproduction levels. Electronic supplementary material The online version of this article (doi:10.1186/s12934-014-0123-2) contains supplementary material, which is available to authorized\nusers.",
"conclusion": "Conclusions By coupling the classical approach of metabolic evolution with systematic strain\nand pathway engineering, styrene bioproduction directly from glucose by S. cerevisiae has been demonstrated for the first time.\nWhile providing an important demonstration of concept, future strain engineering\nefforts will be required to ultimately achieve economical production levels.",
"discussion": "Results and discussion Evolving phenylalanine over-production by S. cerevisiae As phenylalanine serves as the immediate endogenous precursor to the styrene\npathway, its over-production by S. cerevisiae \nis an essential pre-requisite to styrene biosynthesis. Thus, to develop a\nphenylalanine over-production phenotype in S.\ncerevisiae , a classic antimetabolite selection strategy was first\nemployed [ 29 ]. In this case,\n m -fluoro-DL-phenylalanine was chosen to\nprovide the necessary selection pressure whereas exposure to the chemomutagen\nEMS increased mutation rates and frequency [ 30 ]. In the first round of selection, a total of only two\nmutants were isolated when using either 18 mg/L (strain 18A) or 22 mg/L (strain\n22A) m -fluoro-DL-phenylalanine. As seen in\nFigure 2 , said mutants were\nsubsequently characterized in shake flask cultures. Since phenylalanine is not\nexported from S. cerevisiae , the established\npractice of correlating enhanced flux through the pathway with net extracellular\naccumulation of 2-phenylethanol and 2-phenylacetate (both of which are naturally\nand readily produced as degradation products of phenylpyruvate, the precursor to\nphenylalanine; Figure 1 ) was employed\n[ 24 ]. In S. cerevisiae , it has previously been shown that\nall available phenylalanine is efficiently shuttled through one or both of these\npathways [ 31 ], with the relative\ndistribution of products being dictated by the cell’s redox state (specifically,\nthe relative intracellular ratio of NAD + to NADH) [ 32 ]. For example, in glucose grown cultures\nwith limited aeration (as would be expected in the sealed shake flasks used in\nthis study), 9:1 mixtures of 2-phenylethanol:2-phenylacetate are typically\nobserved [ 32 ]. Throughout this\nstudy, the mixture of products obtained was similarly consistent (an average of\n89% 2-phenylethanol; Figure 2 ). Relative\nto BY4741 (the parent strain and control), strains 18A and 22A showed 3.3- and\n6.4-fold improvements in net production of 2-phenylethanol and 2-phenylacetate\n(0.63 ± 0.05 and 1.13 ± 0.10 mM), respectively. Figure 2 \n Evolution of phenylalanine overproducing\nmutants of \n S. cerevisiae \n . Mutants were evolved through\nthe use of EMS mutagenesis and high-throughput selection on\nsolid agar plates using m -fluoro-DL-phenylalanine as anti-metabolite.\n2-Phenylethanol (gray) and 2-phenylacetate (hashed) production\nby isolated mutants was determined after 48 h of growth in SD\nmedia by measuring the concentration of 2-phenylethanol and\n2-phenylacetate in the supernatant. Error bars reported at one\nstandard deviation from triplicate experiments. To further deregulate and enhance phenylalanine biosynthesis, a second round\nof mutagenesis and selection was performed, in this case using the isolated\nmutants 18A and 22A as parents. The selection pressure was accordingly elevated\nby increasing the content of m -fluoro-DL-phenylalanine in SD agar plates to 25 mg/L, 50 mg/L, or\n75 mg/L. In contrast to the first round, in this case numerous colonies (i.e.,\n> 50) were obtained at all three selection pressures. Thus, to screen just\nfor the best performers, only the fastest growing mutants were selected (i.e.,\nthose whose colony forming units were largest after overnight incubation). A\ntotal of 18 additional mutants were chosen and subsequently characterized, as\nabove. As seen in Figure 2 , the top\nperforming mutant, 22A75D, produced a combined total of about 2.17 ± 0.22 mM\n2-phenylethanol and 2-phenylacetate in 48 h, representing ~3.3- and ~21-fold\nimprovements over 22A and BY4741, respectively. Investigating the evolved phenotypes A series of characterizations were next performed on strains 22A and 22A75D\n(with BY4741 as control) to begin to understand the underlying factors\nresponsible for imparting the evolved phenotypes. The native aromatic amino acid\nbiosynthesis pathways of S. cerevisiae are\nshown in Figure 1 , where it can be seen\nthat two known control points are principally responsible regulating metabolite\nflux. The first occurs at DAHP synthase (for which S.\ncerevisiae possesses two isoenzymes), which is allosterically\nfeedback inhibited by either phenylalanine (ARO3) or tyrosine (ARO4)\n[ 33 - 35 ]. The second, meanwhile, occurs at\nchorismate mutase (ARO7), which converts chorismate to prephenate, the precursor\nto both phenylalanine and tyrosine. Transcription of ARO7 is repressed in the\npresence of as little as 0.5 mM tyrosine but remains, however, insensitive to\nphenylalanine [ 33 , 36 , 37 ]. Here, overcoming feedback repression of ARO3 thus\nconstitutes a key priority. However, whereas relief from tyrosine repression of\nARO4 has been reported to result from a single mutation (K229L) [ 24 ], a phenylalanine feedback resistant\nmutant of ARO3 remains unreported to date. Sequences of several key genes in the phenylalanine biosynthesis pathway\n( ARO3 , ARO4 , ARO7 , ARO8, and PHA2 )\nwere first determined for all three strains (including coding regions as well as\n500 bp upstream of each start codon). Interestingly, however, mutations were not\nobserved in the sequence of any investigated gene, including with respect to\nboth its coding and upstream non-coding regions. Transcription levels of all\ngenes in the phenylalanine biosynthesis pathway (namely ARO1 , ARO2 , ARO3 , ARO4 ,\n ARO7 , ARO8 , ARO9 , and PHA2 ; see Figure 1 ) were next examined in the mutants 22A and 22A75D and\nquantified relative to that of the wild-type control (BY4741). The results are\ncompared in Figure 3 , wherein it can be\nseen that, in strain 22A75D, up-regulation of ARO8 was found to be most significant (a 9.3-fold increase),\nfollowed by ARO1 (6.8-fold), ARO2 (5.8-fold), and ARO3 (4.5-fold). Note that similar but less significant\ndifferences were also observed in strain 22A. Furthermore, only modest increases\nin ARO4 and ARO7 expression were observed in 22A75D (about 2.7- and\n1.8-fold, respectively), with no significant changes occurring in 22A for either\ngene. Figure 3 \n Transcriptional analysis of top\nphenylalanine overproducing \n S. cerevisiae \n mutants. Relative transcript\nlevels of the top first (22A) and second (22A75D) round evolved\nyeast mutants, normalized to the parent (BY4741). Measured genes\nincluded ARO1 (black),\n ARO2 (right diagonal),\n ARO3 (dark gray),\n ARO4 (horizontal),\n ARO7 (light gray),\n ARO8 (left diagonal),\n ARO9 (no fill), and\n PHA2 (hashed). Error bars\nreported at one standard deviation from triplicate\nexperiments. The collective findings point to the prospects of several interesting\nmechanisms in the mutant strains. For example, with no change to its sequence,\nthe evolved phenotype clearly did not arise as a result of relieving allosteric\ninhibition at the known bottleneck enzyme, ARO3. However, as a significant\nincrease in its expression was observed in both mutants, this could suggest that\nup-regulation of wild type ARO3 occurred as\nan alternative strategy. That is, despite the fact that the wild type enzyme\npossesses lower specific activity in the presence of phenylalanine, with more\ncopies net DAHP synthase activity may have been sufficiently high so as to\neffectively overcome the flux bottleneck. Simultaneous up-regulation of\n ARO1 and ARO2 (the two subsequent steps in the pathway), meanwhile, may\nhave aided in this process by ensuring that produced DAHP was then promptly\nassimilated further along the pathway, thereby maintaining a maximum driving\nforce. Up-regulated expression of ARO1 has\nbeen successfully employed in yeast as a rational approach for enhancing the\nbiosynthesis of cis,cis- muconic acid – a\nproduct derived from the shikimate pathway intermediate 3-hydroshikimate\n[ 38 ]. Meanwhile, up-regulation\nof ARO8 was evolved perhaps as a mechanism to\ncompete with the native activity of ARO10 ,\nensuring that metabolite flux was efficiently routed towards phenylalanine\nbiosynthesis rather than through the degradative Ehrlich pathway [ 32 , 39 ]. This prospect is further supported by the fact that, in\ncontrast to ARO8 , no appreciable change was\nobserved with respect to the expression of ARO9 , which functions primarily in the reverse direction (i.e.,\nfor phenylalanine assimilation from the culture medium) [ 40 ]. With no change to either the coding or upstream non-coding regions for any of\nthe four up-regulated genes ( ARO1 , ARO2 , ARO3 , and\n ARO8 ) another factor must be responsible\nfor this observed result. Increased copy number through genomic amplifications\nis one possibility, however, a more efficient mechanism may have involved the\nmutation of one or more transcription factors controlling their expression.\n GCN4 encodes one such major transcription\nfactor [ 41 ], however, further\nsequencing revealed no changes there either. To identify other prospective\ntranscription factors involved, the promoter regions of the four up-regulated\ngenes were further investigated by aligning the 1000 bp sequences prior to each\nstart codon. Whereas a possible consensus sequence of 5’-AACATC-3’ was located\nat positions −292, −307, −289, and −290 for ARO1 , ARO2 , ARO3 , and ARO8 ,\nrespectively, said sequence does not match binding site of any known\ntranscription factor. Among all known transcriptional regulators, eleven are\nshared between the four up-regulated genes [ 42 ] (note, an annotated list is provided in Additional file\n 1 : Table S3). The ability to\ndetermine which if any of these regulators are responsible for the evolved\nchanges will only be possible through the collective analysis of their gene\nsequences, or better, to ensure full elucidation of all changes in the mutants,\nthrough whole-genome sequencing. However, given that the achievable titers are\nstill quite modest (a total of only 2.17 ± 0.22 mM 2-phenylethanol and\n2-phenylacetate), such an undertaking was deemed as unwarranted at this time.\nFor now, and for the purpose of this study, these efforts were successful in\ndeveloping a host strain to serve as a test platform for engineering styrene\nbiosynthesis from glucose. Investigating native FDC1 activity and factors influencing its\nexpression Although it had previously been shown that, when cultured in the presence of\nexogenous trans -cinnamate, S. cerevisiae is capable of catalyzing its\ndecarboxylation to styrene [ 11 ],\nseveral factors related to the native function and expression of FDC1 remained initially unclear and deserving of\nfurther investigation. Most importantly, it was wholly unknown as to if, when,\nand how the native expression of FDC1 would\nbe induced in the context of the styrene pathway and under the culture\nconditions of interest. BY4741 was initially cultured in SD minimal media\nsupplemented with potential inducers of interest. In addition to trans -cinnamate and phenylalanine (the pathway\nintermediate and precursor, respectively), p- coumarate and ferulate were also screened as positive controls\n(note, both are structural homologs of trans -cinnamate and known inducers of trans -cinnamate decarboxylase activity [ 11 ]). As seen in Table 1 , in vitro\ntrans- cinnamate decarboxylase activity was detected in the\nlysates of cells cultured in the presence of each of trans- cinnamate, p- coumarate,\nand ferulate (with the former serving as the strongest inducer), but not with\nphenylalanine or in the control. With respect to the styrene pathway, this\nimplies that native FDC1 expression will be\nwholly contingent upon the heterologous expression of PAL2 to provide trans- cinnamate as inducer (a realization that could be of benefit\nwith respect to minimizing overall metabolic burden). Table 1 \n Assaying the in vitro decarboxylase\nactivity of FDC1 against a pool of structurally-related,\nphenylacrylic acid substrates \n \n Compound \n \n Induced activity \n \n mU mg \n −1 \n total protein \n \n trans -cinnamate + 0.46 ± 0.02 \n p -coumarate + 0.39 ± 0.02 ferulate + 0.21 ± 0.03 phenylalanine - N.D. none (control) - N.D. Positive, ‘+’; Negative, ‘-‘; Not Detected,\n‘N.D.’. Evaluating the styrene pathway via the exogenous addition of\nphenylalanine Preliminary studies were next performed to begin probing the functionality of\nthe styrene pathway in wild type S.\ncerevisiae . Strains BY4741-PAL and BY4741 ΔFDC1 -PAL were first cultured in SD-Leu minimal media\nsupplemented with 200 mg/L (1.21 mM) phenylalanine while monitoring the\nextracellular accumulation of trans -cinnamate, styrene, and 2-phenylethanol, 2-phenylacetate, and\nconcomitant depletion of phenylalanine. As seen in Figure 4 , while only 37% of supplied the phenylalanine\nwas consumed by BY4741-PAL after 24 h, styrene and 2-phenylethanol constituted\nthe major end-products, reaching titers of up to 20 ± 1 and 43 ± 1 mg/L\n(0.19 ± 0.01 and 0.35 ± 0.01 mM), respectively, with trans -cinnamate being undetected. In contrast, 2-phenylethanol\nwas produced to a final titer of 98 ± 3 mg/L (0.80 ± 0.02 mM) by BY4741 ΔFDC1 -PAL with styrene being undetected throughout.\nIn addition, in this case the extracellular accumulation of trans -cinnamate was also observed, reaching final\nconcentration of 26 ± 3 mg/L (0.18 ± 0.02 mM) by 24 h. Only trace levels of\n2-phenylacetate were observed throughout. Figure 4 \n Assessing the trans-membrane export\nof \n trans \n -cinnamate. Depletion of\nexogenous phenylalanine (black; initially 200 mg/L) by growing\ncultures of wild-type S.\ncerevisiae BY4741-PAL and BY4741 ΔFDC1 -PAL and the resultant\nproduction of trans -cinnamate\n(diagonal), styrene (horizontal), and 2-phenylethanol (gray)\nafter 24 h. Error bars reported at one standard deviation from\ntriplicate experiments. These results demonstrate several key points. First, styrene can be produced\nfrom phenylalanine by S. cerevisiae via the\nheterologous expression of PAL2 and native\nexpression of FDC1 . Second, both styrene and\n trans -cinnamate are naturally excreted\nfrom S. cerevisiae , at least to a certain\ndegree. Third, since trans -cinnamate did not\naccumulate in BY4741-PAL cultures, this implies that the net activity afforded\nby native FDC1 expression was sufficiently\nhigh so as to preclude the creation of a flux bottleneck at the final pathway\nstep (at least with respect to the specific PAL2 expression level examined). And, lastly, that under the\nconditions examined, synthesis of byproduct 2-phenylethanol significantly\ncompetes with the styrene pathway for precursor availability, even when\n PAL2 is constitutively expressed on a\nhigh copy number plasmid. Styrene production from glucose Based on the above findings, a series of strains were next constructed and\nevaluated with respect to their styrene production potential from glucose, the\nresults of which are compared in Figure 5 . Although only as much as ~5 mg/L (0.05 mM) styrene was\ndetected in BY4741-PAL cultures, 22A75D-PAL accumulated up to 18 ± 2 mg/L\n(0.17 ± 0.02 mM) styrene in 48 h. In the latter culture, however, more\nsignificant accumulation of byproduct 2-phenylethanol was also observed,\nreaching up to 54 ± 5 mg/L (0.44 ± 0.04 mM; again, 2-phenylacetate did not\naccumulate above trace levels). As this again suggested that the native Ehrlich\npathway was competitively inhibiting the styrene pathway with respect to\nprecursor availability, the effect of deleting ARO10 – which converts phenylpyruvate to phenylacetaldehyde\n[ 32 , 43 ] – on styrene production was explored. As\nseen in Figure 5 , deletion of ARO10 to preserve phenylpyruvate availability\nimproved styrene production by 22A75D10-PAL by ~28%, reaching up to 23 ± 2 mg/L\n(0.22 ± 0.02 mM). Lastly, to further improve styrene production, a feedback\nresistant mutant of ARO4 – namely ARO4 K229L – was introduced into 22A75D10-PAL. Although ARO4 encodes a tyrosine-sensitive DAHP synthase, in\nprevious works ARO4 K229L over-expression in S.\ncerevisiae was shown to increase flux through the shikimic acid\npathway by as much as 4.5-fold [ 16 , 24 ]. Here,\nexpressing ARO4 K229L , 22A75D104-PAL displayed an additional 25% increase in styrene\ntiter, reaching up to 29 ± 2 mg/L (0.28 ± 0.02 mM) at a glucose yield of about\n1.44 ± 0.11 mg/g (0.0025 ± 0.0002 mol/mol; or just 0.6% of theoretical).\nMeanwhile, trans -cinnamate was not detected\nin the culture media of any styrene producing strain at any time (data not\nshown). Figure 5 \n Styrene biosynthesis from glucose by\nengineered \n S. cerevisiae \n strains. Styrene (gray) and\n2-phenylethanol (lined) production by strains BY4741-PAL,\n22A75D-PAL, 22A75D10-PAL, and 22A75D104-PAL after 48 h in SD-Leu\nminimal media in sealed shake flask cultures. Error bars\nreported at one standard deviation from triplicate\nexperiments. As a volatile product, it was also noted that styrene significantly\naccumulated within the headspace of each sealed flask. We have previously\nconfirmed that said vapor–liquid partitioning behaves according to Henry’s Law,\nwith a dimensionless Henry’s Law coefficient of 0.113 at 32°C [ 10 ]. For the conditions examined here, this\nmeant that an additional ~45% styrene was produced and accumulated in each case.\nWith this in mind, the maximum volumetric styrene production achieved here would\nbe more accurately represented as 42 ± 3 mg/L (0.40 ± 0.03 mM) with a glucose\nyield of 2.09 ± 0.16 mg/g (0.0036 ± 0.0003 mol/mol; 0.9% of theoretical). The\nvolatile nature of styrene may also prove useful as a product recovery strategy\nin the future [ 44 ]. Even at this adjusted output, however, achievable styrene production remains\nonly 18% of the net production of 2-phenylethanol and 2-phenylacetate\ndemonstrated by 22A75D (2.17 ± 0.22 mM) and is 65% of the net production of\nstyrene and 2-phenylethanol by 22A75D-PAL (0.61 ± 0.02 mM). This suggests that\nmultiple limiting factors may have arisen during strain construction. Although\nit was not anticipated to be problematic at such low aqueous titers, to ensure\nthat end-product inhibition was not the central productivity-limiting factor, a\ncursory evaluation of styrene toxicity was performed. S.\ncerevisiae growth and viability was found to be only minimally\ndisrupted in the presence of styrene at up to at least 200 mg/L (1.92 mM; data\nnot shown), suggesting that styrene toxicity was not a critical barrier at this\npoint. Over-expression of PAL2 appears to\nhave had the greatest negative impact, reducing net aromatic production by 72%.\nThis was likely due to the metabolic burden imposed by its expression from a\nhigh-copy plasmid. While decreasing PAL2 \nexpression will likely reduce burden, as trans -cinnamate was never detected in styrene producing\ncultures, this already points to the fact that PAL2 activity was rate-limiting\nin the styrene pathway. Thus, future and careful optimization of PAL2 expression and/or the identification of other\nPAL homologs displaying greater inherent activity in S.\ncerevisiae will be key to achieving further improvements in\nstyrene production. Lastly, although seemingly low, it should be appreciated that level of styrene\nproduction demonstrated here agrees well with that of prior reports by others\nwhom have engineered S. cerevisiae to produce\naromatic chemicals (for which Curran et al. \npreviously provided a comprehensive examination). Notable examples include\n p- hydroxycinnamate [ 45 ], p- aminobenzoic acid [ 46 ], p- hydroxybenzoic acid\n[ 46 ], and vanillin\n[ 20 ], which have been produced\nto maximal titers of up to 33.3, 34.3, 89.8, and 45 mg/L, respectively, at\nyields of 1.7, 2.3, 6.0, and 2.3 mg/g. Meanwhile, when compared with the\nbaseline for styrene production established using an E.\ncoli platform [ 10 ],\nthe achievable titers and yields demonstrated with S.\ncerevisiae currently lag by about 9- and 21-fold, respectively.\nTo achieve higher styrene titers with S.\ncerevisiae , further de-regulation of metabolite flux through its\nphenylalanine biosynthesis pathway will ultimately be required."
} | 5,536 |
28916814 | PMC5601950 | pmc | 3,938 | {
"abstract": "Brown midrib (bmr) mutants in sorghum ( Sorghum bicolor (L.) Moench) and several other C4 grasses are associated with reduced lignin concentration, altered lignin composition and improved cell wall digestibility, which are desirable properties in biomass development for the emerging lignocellulosic biofuel industry. Studying bmr mutants has considerably expanded our understanding of the molecular basis underlying lignin biosynthesis and perturbation in grasses. In this study, we performed quantitative trait locus (QTL) analysis, identified and cloned a novel cinnamyl alcohol dehydrogenase allele ( SbCAD2 ) that has an 8-bp deletion in its 5′-untranslated region (UTR), conferring the spontaneous brown midrib trait and lignin reduction in the sorghum germplasm line PI 595743. Complementation test and gene expression analysis revealed that this non-coding region alteration is associated with the significantly reduced expression of the SbCAD2 in PI 595743 throughout its growth stages. Moreover, a promoter-GUS fusion study with transgenic Arabidopsis thaliana plants found that SbCAD2 promoter is functionally conserved, driving a specific expression pattern in lignifying vascular tissues. Taken together, our results revealed the genetic basis of bmr occurrence in this spontaneous sorghum mutant and suggested the regulatory region of the SbCAD2 can be a target site for optimizing lignin modification in sorghum and other bioenergy crops.",
"introduction": "Introduction Sorghum is one of the premier biomass feedstocks for biofuel production because of its high biomass yield, outstanding drought tolerance and efficient nutrient usage 1 . In the renewable bioenergy industry, overcoming the intrinsic recalcitrance of biomass is critical for the development of lignocellulosic biofuel as a cost-effective alternative to fossil fuels 2 . Biomass recalcitrance is mainly due to the complex configuration of lignin polymer chains and the intertwined network of lignin and polysaccharides in plant cell walls 3 . In plants, the highly interwoven cell wall matrix rigidified by lignin is crucial for their structural integrity, water conduction and pathogen resistance 4 , 5 . Hence, as a major impediment to plant biomass utilization and an essential component of the secondary cell wall, lignin has become a main research object in plant biology and its biosynthetic pathway has been extensively studied in the past several decades. In general, cell wall lignin is oxidatively polymerized primarily from three types of alcohols (monolignols), which are synthesized in cytosol from phenylalanine through successive deamination, reduction, hydroxylation and methylation steps 5 . Knowledge from the well-delineated monolignol biosynthetic pathway has led to the successful genetic manipulation of lignin content and composition in bioenergy grasses 6 , 7 . Recently, in order to minimize developmental defects in lignin-modified plants, lignin modification strategy has evolved from simple perturbations of the general monolignol biosynthetic pathways to coordinated orchestrations of the lignin synthesis, deposition and integration into the cell wall network 8 . Therefore, a profound knowledge and advanced understanding of the underlying regulatory mechanisms of lignin production would benefit the long-term goal of manipulating cell wall structure in biomass feedstocks based on our needs. Brown midrib (bmr) is a visual trait first documented in maize over eighty years ago and later mutagenesis-induced in sorghum and pearl millet 9 – 11 . In grasses, this trait is characterized by the reddish-brown pigment accumulated in leaf midribs and stems during the vegetative stage. In earlier years, this distinct pigmentation was found to be associated with higher ruminant digestion rate and thus became a desired feature in livestock forage development 12 , 13 . Later, the property of reduced lignin content in bmr mutants was appreciated by plant geneticists and utilized for identifying candidate genes for manipulating the lignin production in grasses 14 . A number of mutagenized bmr mutants of sorghum have been characterized and the corresponding genes involved in specific enzymatic steps of monolignol biosynthetic pathway have been identified in recent years 15 – 18 . For instance, bmr6 mutant, originally generated from a diethyl sulfate mutagenesis sorghum population, was found to be associated with reduced cinnamyl alcohol dehydrogenase (CAD) activity 19 . CAD is a specialized enzyme of the alcohol dehydrogenase family involved in the conversion of the cinnamaldehydes into alcohols, the monomeric precursors of lignin. Downregulation of CAD genes generates atypical lignin polymers with the incorporation of phenolic aldehydes and becomes a promising strategy to improve cell wall digestibility 20 – 22 . Later on, a nonsense mutation in a sorghum CAD gene ( SbCAD2 ) was identified to be responsible for the bmr6 phenotype. Specifically, the nonsense mutation truncates the coding region of SbCAD2 prior to the conserved NADPH-binding and C-terminal domains and consequently, abolishes the function of the encoded protein 16 , 17 . Fiber digestion tests and field evaluation performed on sorghum varieties carrying the mutagenized bmr mutations found that they generally show higher digestibility and inferior agronomic performance relative to their wild-type counterparts, even though the effects of mutations are not uniformly expressed across different genetic backgrounds and are influenced by environmental conditions 14 , 23 . Stacking different bmr mutations in sorghum has an additive effect on lignin content and cellulose-to-ethanol conversion 24 . Therefore, there are some recent efforts focusing on characterization of novel bmr sorghum lines derived from EMS (ethyl methanesulfonate)-mutagenized populations with the aim to incorporate new sources into the forage and biomass feedstock improvement 25 , 26 . In summary, most sorghum bmr mutants studied so far are from mutagen-induced populations, and no spontaneous bmr phenotypes in sorghum have been characterized in detail at the molecular level. In this study, we focused on PI 595743, a sorghum germplasm line showing a distinct bmr trait without any observable growth and developmental defects. Phenotypic evaluation confirmed the lignin deficiency in its brown vascular tissues. We next used QTL mapping, DNA sequencing and complementation test to identify that an 8-bp deletion in the 5′ UTR of SbCAD2 is responsible for this novel spontaneous bmr phenotype. This deletion results in the down-regulation of SbCAD2 expression in PI 595743 in a mechanism that is different from that underlying the reduced expression of SbCAD2 in the bmr6 lines. Furthermore, we found that the function of SbCAD2 promoter is conserved in both sorghum and Arabidopsis and this deletion region is implicated in the transcriptional regulation of SbCAD2 expression in lignifying tissue. We expect that this alternative allele of SbCAD2 gene will expand the repertoire of genetic resources for biomass improvement in sorghum, and more importantly, provide new insights into the complex underlying regulatory mechanism of the lignin biosynthetic pathway in grasses.",
"discussion": "Discussion Developing brown midrib mutant lines and identifying the genetic basis of lignin deficiency in these mutants are of increased interest due to their potential to reduce biomass recalcitrance. There are more than ten enzymatic steps involved in the monolignol biosynthetic pathway, leading to the synthesis of monolignol precursors for lignin polymerization 5 , 29 . Theoretically speaking, any perturbation in one of these steps could affect the lignin production. To date, only a few steps of the monolignol pathway were found to be directly involved in the exhibition of bmr in sorghum 15 , 16 , 30 and most sorghum bmr mutants developed from mutagenesis are allelic 25 , 30 . Hence, our main objective was to characterize a naturally-occurring bmr mutant in sorghum, with an attempt to isolate novel bmr loci for further study. Interestingly, the causal gene underlying the spontaneous bmr mutant in this study was identified to be SbCAD2 , the same gene responsible for the bmr6 phenotype. This high occurrence of allelic bmr mutants could be simply ascribed to the formation of brown color itself. In other words, disruption of a particular step in the monolignol biosynthetic pathway may not be sufficient to induce the generation of a specific bmr phenotype. While the exact cause of brown pigmentation in lignified tissue is elusive, the bmr trait in mutants or transgenic lines with impaired CAD activity has been attributed to the incorporation of cinnamyl aldehydes into lignin in place of cinnamyl alcohols 14 . This fact is consistent with the phloroglucinol staining result we observed in PI 595743. Another possible explanation of the discrepancy between bmr occurrence and lignin disruption is that modifying some enzymatic steps in the monolignol biosynthetic pathway could result in relatively more severe side-effects on plant growth and development. This hypothesis is supported by several previous studies on lignin-modified mutants. For example, Arabidopsis plants defective in C3H ( p -coumarate 3-hydroxylase), the second enzyme in the monolignol pathway, exhibit severe dwarfism and sterility 31 . Similar deleterious defects were also observed in the diethyl sulfate (DES)-induced bmr lines of sorghum, precluding the subsequent genetic study on those mutants 9 . Due to the importance and plasticity of the last step of monolignol pathway, the impacts of CAD disruption on cell wall properties have been intensively studied in several plant species. CAD-deficient mutants 32 , 33 and transgenic lines with down-regulated CAD activity 34 , 35 were generally associated with reduced lignin content and/or altered lignin composition change, even though there were some reported inconsistencies partly due to functional redundancy of CAD paralogs or incomplete disruption of CAD expression. In sorghum, the CAD-deficient bmr6 mutant showed reductions in lignin content, altered lignin composition and an improved cell wall digestibility 13 , 24 . Despite the substantial impact on lignin biosynthesis, agronomic evaluation found that the bmr6 mutant was not associated with significant negative impacts on plant fitness in comparison with its near-isogenic counterpart 25 . In the current work, a spontaneous sorghum bmr mutant, PI 595743, showed a modest decrease in lignin content and dramatic change of lignin composition in certain lignified cell types, in agreement with the aforementioned fact that monolignol biosynthesis is highly plastic as to allow CAD-deficient plants to form lignin polymers directly from cinnamyl aldehydes. The accumulation of aldehydes and reduction of the S lignin subunit in PI 595743 were revealed by phloroglucinol staining and the Maule staining, respectively. These histochemical results provide the first clue that the CAD enzymatic activity is likely affected in PI 595743. Gene expression assay further confirmed the down-regulation of SbCAD2 gene during the development of PI 595743 plants. The significant reduction of CAD activity due to a nonsense mutation was reported in bmr6 17 . Based on the allelic variation, it may prove of interest to examine to what extent the CAD activity is impaired in PI 595743. Furthermore, comparative evaluation of different CAD mutants under the same genetic background, for example, within the near-isogenic lines would be a good starting point to study whether genetic variants in the SbCAD2 gene have different effects on lignin composition, biomass digestibility and the overall agronomic performance. Previous research on the bmr6 mutant revealed that a nonsense mutation-mediated mRNA decay mechanism was very likely involved in the downregulation of SbCAD2 in its vascular tissues 17 . In the present work, based on collective evidences from histochemical analysis, QTL mapping, association study and complementation test, an 8-bp deletion in the 5′ UTR of SbCAD2 was linked to the dark-brown pigmentation and the significant reduction of SbCAD2 expression in the spontaneous mutant PI 595743. Thus, an obvious question from these results is: how does this deletion in an untranslated region affect the expression of SbCAD2 and lead to a mutated phenotype? As a fundamental structural and regulatory region of eukaryotic genes, 5′ UTR mainly plays a role in the regulation of mRNA translation, providing a novel layer of coordinated control of gene expression 36 , 37 . Parts of the 5′ UTR may also contain regulatory elements and may be part of the promoter. 5′ UTR-mediated regulation of transcript abundance and translational efficiency has been reported in plant development and stress responses 38 – 40 . In the case of PI 595743, the 8-bp deletion within the 5′ UTR could affect its qualitative features, such as length, GC content and potential for secondary structure formation, associated with the translational regulation. There is also a possibility that this deletion is involved in the transcriptional regulation of SbCAD2 by interfering with potential binding sites, an interaction mechanism totally different from the mRNA decay mechanism underlying the reduced expression of SbCAD2 in the bmr6 mutant. Further studies on heterologous expression of a reporter gene fused with this mutated 5′ UTR or promoter deletion analysis are needed to dissect the association between this specific deletion and lignin pathway perturbation. The active expression of GUS gene in the lignifying tissue of Arabidopsis plants driven by the promoter of the sorghum CAD2 gene reflects the conservation of mechanism that underlies the transcriptional regulation of monolignol biosynthetic pathway in higher plants. The conservative property of lignin biosynthesis in plants has been investigated and discussed previously from different perspectives 41 – 43 . Evolutionarily speaking, the signals controlling vascular expression of lignin biosynthetic genes are more likely to be highly conserved in higher plants since lignin biosynthesis is an early adaptation feature for vascular plants to survive in the terrestrial environment 41 , 44 . In addition, the maintenance of conserved motifs in many lignin biosynthetic genes suggested a common catalytic mechanism might be shared during the process of lignification in plants 45 . Lignin biosynthesis is a highly coordinated process that involves the expression of multiple monolignol genes and the coordinated regulation of transcription factors 44 , 46 . Knowledge gained from this study could open an avenue for us to decipher the regulatory aspects of the SbCAD2 gene and reveal the occurrence of bmr phenotype and lignin production in grasses at the transcription and post-transcription level."
} | 3,738 |
24046771 | PMC3763484 | pmc | 3,939 | {
"abstract": "Plants interact with a variety of other community members that have the potential to indirectly influence each other through a shared host plant. Arbuscular mycorrhizal fungi (AMF) are generally considered plant mutualists because of their generally positive effects on plant nutrient status and growth. AMF may also have important indirect effects on plants by altering interactions with other community members. By influencing plant traits, AMF can modify aboveground interactions with both mutualists, such as pollinators, and antagonists, such as herbivores. Because herbivory and pollination can dramatically influence plant fitness, comprehensive assessment of plant–AMF interactions should include these indirect effects. To determine how AMF affect plant–insect interactions, we grew Cucumis sativus (Cucurbitaceae) under five AMF inoculum treatments and control. We measured plant growth, floral production, flower size, and foliar nutrient content of half the plants, and transferred the other half to a field setting to measure pollinator and herbivore preference of wild insects. Mycorrhizal treatment had no effect on plant biomass or floral traits but significantly affected leaf nutrients, pollinator behavior, and herbivore attack. Although total pollinator visitation did not vary with AMF treatment, pollinators exhibited taxon-specific responses, with honey bees, bumble bees, and Lepidoptera all responding differently to AMF treatments. Flower number and size were unaffected by treatments, suggesting that differences in pollinator preference were driven by other floral traits. Mycorrhizae influenced leaf K and Na, but these differences in leaf nutrients did not correspond to variation in herbivore attack. Overall, we found that AMF indirectly influence both antagonistic and mutualistic insects, but impacts depend on the identity of both the fungal partner and the interacting insect, underscoring the context-dependency of plant–AMF interactions.",
"conclusion": "CONCLUSION The outcomes of plant–AMF interactions have historically focused on the direct effects of the fungi on plants, such as plant growth or nutrient content. However, plant growth and fitness are also influenced by community members, whose interactions may be modified by AMF-driven changes in plant traits. Here we show that colonization by different AMF species has consequences for pollinator visitation and herbivory in an agroecosystem, but these effects vary with both AMF and insect identity. For AMF–plant–pollinator interactions, future work should focus on the multiple floral traits that can mediate pollination, including how different AMF species (or AMF communities from different ecological contexts) influence floral traits like visual cues, nectar production and composition, and floral scent. A more detailed understanding of AMF effects on these flower traits will allow better predictions of pollinator responses based on the floral signals used by different pollinator taxa. Similarly, understanding AMF effects on herbivory will require experiments that measure plant nutrients and chemical defenses in a field setting or controlled, but ecologically realistic, laboratory conditions. Our results demonstrate the potential for above- and belowground communities to interact in complex ways via species-specific responses of insects and their effects on plant fitness. Thus, even in relatively simple agroecosystems, diversity may provide an important buffer maintaining critical species interactions.",
"introduction": "INTRODUCTION Plants interact with a variety of organisms both above and below the soil surface. Belowground interactions between plants and other organisms influence, and are influenced by, interactions aboveground ( Bardgett and Wardle, 2003 ; Wardle et al., 2004 ; van der Putten et al., 2009 ). Among the most abundant and widespread soil microbes are arbuscular mycorrhizal fungi (AMF), members of the phylum Glomeromycota that form associations with plant roots and exchange nutrients, such as phosphorus and nitrogen, for plant-derived carbon ( Smith and Read, 2008 ). These globally important fungi interact with 60–80% of terrestrial plant species ( Smith and Smith, 2011 ) and generally confer growth and fitness benefits ( Smith and Read, 2008 ). The impacts of AMF on host plants are usually evaluated based on these direct effects alone. However, the direct effects of AMF on plants may also alter plant traits that mediate interactions between plants and insects, such as pollinators or herbivores, with important consequences for plant fitness ( Wolfe et al., 2005 ; Koricheva et al., 2009 ; Vannette and Rasmann, 2012 ). Colonization by AMF can affect floral traits such as flower number ( Schenck and Smith, 1982 ; Lu and Koide, 1994 ; Gange et al., 2005 ) and size ( Gange and Smith, 2005 ; Kiers et al., 2010 ; Varga and Kytöviita, 2010 ), as well as nectar and floral volatile characteristics ( Gange et al., 2005 ; Kiers et al., 2010 ; Becklin et al., 2011 ). Although the number of studies measuring AMF effects on pollinator visitation to plants in the field is very limited, they have demonstrated that AMF can influence pollinator behavior. Wolfe et al. (2005) and Gange and Smith (2005) both found increased visitation to plants inoculated with AMF compared to non-mycorrhizal plants, and the latter study found that this pattern was consistent among both hymenopteran and dipteran pollinators. In another experiment, the preferences of these two taxonomic groups differed depending on the AMF species used to inoculate the plant ( Varga and Kytöviita, 2010 ). Thus, the direction and magnitude of AMF impacts on plant–pollinator interactions likely depend on both the pollinator and the AMF species colonizing the plant ( Gehring and Bennett, 2009 ). Insect herbivory is also frequently influenced by AMF colonization ( Koricheva et al., 2009 ), and these effects may be due to mycorrhizal effects on plant biomass, nutrient content, or defenses ( Bennett et al., 2006 ). For example, increased nutrient content of mycorrhizal plants may increase their quality as a food source for herbivores, but the resources made available by this interaction may also be allocated toward defense against herbivores ( Vannette and Hunter, 2011 ). Additionally, AMF may also play an important role in induced resistance of plants against insects by priming the jasmonic acid-dependent responses of plants to herbivory ( Pozo and Azcón-Aguilar, 2007 ; Koricheva et al., 2009 ; Jung et al., 2012 ). Regardless of the underlying mechanisms, AMF can indirectly affect plant fitness through changes in herbivory. The effects of AMF on pollination or herbivory are likely to differ among AMF species or strains. For example, both constitutive and induced levels of defensive chemicals in Plantago lanceolata varied among plants inoculated with three different AMF species ( Bennett et al., 2009 ). In a recent study, performance of herbivores feeding on Fragaria vesca varied when plants were inoculated with different strains of the AMF Rhizophagus irregularis ( Roger et al., 2013 ). These results underscore the importance of examining multiple species in AMF–plant–insect interactions to understand the variation in indirect mycorrhizal effects ( Gehring and Bennett, 2009 ). There are additional challenges to studying insect responses to mycorrhizal variation in a realistic field setting. For example, most field studies of these interactions have manipulated AMF by applying fungicide to plots and observing insect responses ( Koricheva et al., 2009 ). Although this is an effective method of eliminating AMF from experimental plots, there may be unintended effects by altering non-mycorrhizal fungi and other soil organisms. In this study we tested the hypothesis that plant–AMF interactions belowground influence aboveground interactions between plants, herbivores and pollinators. We used an inoculation experiment to manipulate multiple species/strains of AMF in the roots of Cucumis sativus (cucumber, Cucurbitaceae). We transferred plants to an agricultural field setting, measured traits that may affect plant reproduction directly and indirectly, and determined pollinator and herbivore preferences. Although we made no specific predictions about the impacts of each inoculum on plant–insect interactions, we hypothesized that AMF-free plants would have reduced pollinator visitation and based on previous similar studies ( Gange and Smith, 2005 ; Wolfe et al., 2005 ). Given the role of AMF in induced defenses of C. sativus ( Barber, 2013 ), we also expected greater herbivory on these non-mycorrhizal plants.",
"discussion": "DISCUSSION Species interactions can be very context-dependent, and outcomes will vary depending on various biotic and abiotic factors. This is evident in natural systems as well as in agroecosystems ( Tscharntke et al., 2008 ). In agricultural fields, mycorrhizal and other symbioses may modify a range of plant traits that alter the nature or frequency of plant–insect interactions important for plant reproduction. This can happen via direct effects on plant growth, nutrition, and other traits. Here, we found that associations between plants and AMF influenced aboveground plant interactions with both pollinators and herbivores and that these effects differed among both AMF and insect species, highlighting the context-dependent nature of these interactions. Colonization varied significantly among AMF treatments, with the highest colonization (fungal structures present in >34% of root length on average) by the mix treatment containing three AMF species. This level is greater than the average colonization by any of the component species alone and greater than the sum of these single species, suggesting there may be synergistic interactions that benefit the fungi when multiple species are present. The lowest colonization, other than the AMF-free control, was by R. irregularis. This was much lower than the commercial inoculum, also a strain of R. irregularis, illustrating the wide variation in colonization potential possible among even taxa categorized as conspecific ( Stockinger et al., 2009 ; Roger et al., 2013 ). POLLINATION Inoculation by different AMF species influenced the behavior of several taxonomic groups of insect pollinators. Honey bees, bumble bees, and Lepidoptera behavior varied with inoculation treatment, and the patterns of visitation and flower probing differed among these groups. Differences in visitation to plants by bumble bees was driven by apparent greater preference for plants inoculated with R. irregularis, although this contrast was marginally significant after adjustment. Similarly, there was a trend toward greater preference by Lepidoptera for plants inoculated with G. clarum . While the decision to begin foraging on a plant (i.e., a plant visit) may be determined by long- and short-range cues, the proportion of flowers probed may be a better indicator of pollinator assessment of floral quality ( Mitchell and Waser, 1992 ). Pollinators are expected to probe a greater number of flowers on a high-reward plant. Conversely, a visit to a less-rewarding plant may be terminated before all flowers have been visited. We found that the proportion of flowers probed by both honey bees and Lepidoptera varied with AMF inoculation treatment. Honey bee flower probing rates were significantly reduced on plants that had been inoculated with single species of AMF compared to non-mycorrhizal controls; this is surprising given that AMF usually increase floral reward and pollinator preference ( Gange and Smith, 2005 ; Gange et al., 2005 ; Aguilar-Chama and Guevara, 2012 ). Lepidoptera flower probes were significantly greater for mixture plants than single species inocula plants combined, suggesting that the three AMF had interactive effects on floral traits that increased Lepidoptera preference. Previous studies have found that the effects of AMF on pollinator behavior differ among plant species and pollinator taxa. Gange and Smith (2005) compared pollinator visitation to mycorrhizal and non-mycorrhizal individuals of three plant species. Although AMF generally increased visitation the effect differed with specific combinations of pollinator taxa and plant species, with increased Hymenoptera visitation to two species and Diptera visitation to the other species. Similarly, we show that visitation or flower probes varied by taxa (bumble bees, honey bees, and Lepidoptera) and inoculum type. This suggests that AMF species or strains alter plant traits in different ways, and that these pollinator taxa differ in their response to these traits. In a study that manipulated two AMF species, including R. irregularis , pollinator visitation increased with both AMF, although only a subset of the pollinator community was examined ( Wolfe et al., 2005 ). Of the few prior experiments on AMF effects on pollinator behavior, only one ( Varga and Kytöviita, 2010 ) both manipulated AMF species identity and examined multiple pollinator taxa, as we did here. Interestingly, they found reduced Syrphidae visitation to female Geranium sylvaticum (Geraniaceae) when plants were inoculated with one species of AMF, compared to control and the other AMF species. They also showed reduced visitation by small Hymenoptera to plants inoculated with the other species. Syrphidae and small Hymenoptera (here, Halictidae) were also common visitors in our experiment, but we found no effects of AMF inocula on visitation or flowers probed. Taken together, these results indicate that different AMF species likely have distinct effects on floral traits and that pollinators have taxa-specific responses to these trait changes. Pollinators responded to AMF treatments, despite the lack of treatment effects on the floral traits we measured. Gange and Smith (2005) attributed increased pollinator visitation on mycorrhizal plants to greater inflorescence number or size for two Asteraceae species. However, our AMF treatments did not affect flower number or size, which contrasts with many studies that find increased flower production in plants associating with AMF ( Bryla and Koide, 1990 ; Perner et al., 2007 ; Varga and Kytöviita, 2010 ). In previous work R. irregularis increased male flower production and flower diameter in C. sativus, although the effect on flower size was eliminated by addition of methyl jasmonate ( Kiers et al., 2010 ). Herbivore attack triggers jasmonic acid responses in plants ( Farmer et al., 2003 ), so insect herbivory on our plants in the field may have erased any positive effects of AMF on floral traits. However, herbivore attack was not significantly correlated with total flower production or male flower size (data not shown). Nectar production and composition and floral volatiles can also have profound effects on pollinator behavior ( Schemske and Bradshaw, 1999 ; Dudareva and Pichersky, 2006 ; Adler, 2007 ), but were not measured in this study. Effects of AMF on nectar quantity and quality vary among plant species ( Gange and Smith, 2005 ; Becklin et al., 2011 ), and AMF use of plant photosynthates may reduce plant carbohydrate availability for nectar ( Laird and Addicott, 2007 ). Nectar production in C. sativus, like flower size, was also reduced by methyl jasmonate application ( Kiers et al., 2010 ), so herbivory may interact with mycorrhizal status to affect nectar. Floral scent from volatile production affects pollinator attraction, and experimental elimination of soil fungal communities altered volatile production in Polemonium viscosum ( Becklin et al., 2011 ). However, inoculation with commercial AMF or farm AMF communities had no effect on C. sativus volatiles compared with non-mycorrhizal controls (Barber et al. unpublished data), suggesting that this trait may not explain the indirect effects of AMF on pollinators observed here. HERBIVORY Mycorrhizal treatment significantly affected the probability of herbivore damage to leaves, with probability of attack varying from 0.3 in plants inoculated with G. clarum to nearly 0.6 in plants with a mixture of AMF species. Control plant herbivory was intermediate, so individual treatments did not differ significantly from control ( Figure 4 ). Inoculation affected leaf nutrient content, but surprisingly not P or N, the nutrients that are most frequently studied in plant–AMF research. Rather, commercial AMF inoculum significantly increased leaf K and Na content relative to non-mycorrhizal plants, although the increase in K was modest. Recent work has emphasized the potential importance of less-studied elements that exist in both organic molecules and ionic forms, but are essential to herbivores ( Behmer and Joern, 2012 ; Joern et al., 2012 ). Sodium can be limiting for insect herbivores because it occurs in low concentration in plant tissues ( Kaspari et al., 2008 ; Behmer and Joern, 2012 ; Chavarria Pizarro et al., 2012 ), and potassium was also identified as a predictor of insect herbivore abundance ( Joern et al., 2012 ). If AMF alter plant concentrations of these elements (in organic or inorganic forms) that are important to insect nutrition, it may provide an additional mechanism of indirect mycorrhizal effects on insect herbivore preference and performance. Future work could address whether the magnitudes of these differences in elemental content (20–50 ppm Na, 1–2 ppt K) are sufficient to influence insect herbivore preference or performance. FIGURE 4 Effects of AMF inoculation treatments on probability of herbivore attack. Values are fitted means ± 1 SE, transformed from logits to probability for ease of interpretation. Treatment effects on herbivory could also be caused by AMF influences on plant defenses. Colonization of plant roots by AMF is thought to induce both local and systemic responses that allow the plant to respond more rapidly or efficiently to attack by herbivores or pathogens ( Jung et al., 2012 ). Mycorrhization increased induced defenses against a generalist herbivore ( Spodoptera exigua ) in C. sativus , with herbivores consuming more leaf tissue on induced mycorrhizal plants without increasing their biomass ( Barber, 2013 ). Given this finding, we would expect reduced herbivory on inoculated plants compared to control plants in this experiment, but instead we found lower herbivory on control plants ( Figure 4 ). This may in part be explained by the dominant wild herbivore in this agroecosystem, Acalymma vittatum (striped cucumber beetle), a specialist that responds positively to cucurbitacins, the primary defensive chemicals in Cucumis and its relatives ( Metcalf et al., 1980 ). The role of AMF in inducing plant defenses may be more important for generalist herbivores than specialists. This hypothesis is supported by a meta-analysis of AMF–herbivore experiments that found more negative effects of mycorrhizae on generalist chewing herbivores than on specialists ( Koricheva et al., 2009 )."
} | 4,760 |
28362721 | PMC5520029 | pmc | 3,940 | {
"abstract": "The Zetaproteobacteria are ubiquitous in marine environments, yet this class of Proteobacteria is only represented by a few closely-related cultured isolates. In high-iron environments, such as diffuse hydrothermal vents, the Zetaproteobacteria are important members of the community driving its structure. Biogeography of Zetaproteobacteria has shown two ubiquitous operational taxonomic units (OTUs), yet much is unknown about their genomic diversity. Genome-resolved metagenomics allows for the specific binning of microbial genomes based on genomic signatures present in composite metagenome assemblies. This resulted in the recovery of 93 genome bins, of which 34 were classified as Zetaproteobacteria. Form II ribulose 1,5-bisphosphate carboxylase genes were recovered from nearly all the Zetaproteobacteria genome bins. In addition, the Zetaproteobacteria genome bins contain genes for uptake and utilization of bioavailable nitrogen, detoxification of arsenic, and a terminal electron acceptor adapted for low oxygen concentration. Our results also support the hypothesis of a Cyc2-like protein as the site for iron oxidation, now detected across a majority of the Zetaproteobacteria genome bins. Whole genome comparisons showed a high genomic diversity across the Zetaproteobacteria OTUs and genome bins that were previously unidentified by SSU rRNA gene analysis. A single lineage of cosmopolitan Zetaproteobacteria (zOTU 2) was found to be monophyletic, based on cluster analysis of average nucleotide identity and average amino acid identity comparisons. From these data, we can begin to pinpoint genomic adaptations of the more ecologically ubiquitous Zetaproteobacteria, and further understand their environmental constraints and metabolic potential.",
"conclusion": "Conclusions Coverage analysis of our composite metagenome indicates that carbon is fixed primarily by Zetaproteobacteria containing Form II RubisCO. Through an assessment of the diversity of Form II RubisCO genes and the abundance of cbb 3 -type cytochrome c oxidase genes, many Zetaproteobacteria show an adaptation to life at very low oxygen levels in conjunction with high-ferrous iron and dissolved CO 2 levels. Denitrification is less common, and our data indicates that bioavailable nitrogen is primarily metabolized. Nearly all of the Zetaproteobacteria genome bins contain genes for the detoxification of arsenate as well as representatives from each of the other classes that were detected in these microbial mat communities. This shows that metagenomics analyses can also provide insights into geochemical conditions. The lineage represented by zOTU 2 is monophyletic suggesting an ancestral bottleneck during its more recent evolutionary history. This zOTU is also the most prevalent of our Zetaproteobacteria genome bins, indicating it is the most ecologically successful manifestation of both sheath and stalk morphology. Through the use of genome-resolved metagenomics, we have better constrained patterns observed in metabolic potential and divergence across many of the Zetaproteobacteria growing within microbial mats at Lō’ihi Seamount.",
"introduction": "Introduction Microbes are everywhere, and in many ecosystems they are the key drivers of biogeochemical cycles. Iron is the most abundant element in the earth and only microbes are able to utilize it as an energy source. Mineralogical evidence of iron- oxidizers has been found, dating to 1.89 Ga, making iron oxidation a very ancient metabolism ( Planavsky et al. , 2009 ). Early Earth hosted a ferruginous ocean where iron oxidation may have been the dominant metabolism ( Ilbert and Bonnefoy, 2013 ; Guilbaud et al. , 2015 ). Microbial iron oxidizers are found suspended in the water column ( Field et al. , 2016 ), but extensive microbial growth by iron oxidation is limited to areas of high ferrous iron and low oxygen concentrations, such as hydrothermal vents ( Emerson and Moyer, 2010 ; Scott et al. , 2015 ). Reduced iron released by hydrothermal vent systems fuels primary production by lithoautotrophic microbes, which in turn support additional trophic levels making hydrothermal vent systems some of the most biologically active regions of the deep-sea ( Sievert and Vetriani, 2012 ). It is estimated that 3 × 10 11 mol of Fe(II) is released each year through hydrothermal venting in Earth’s oceans ( Holland, 2006 ), and is transported in the water column thousands of kilometers away from the source ( Resing et al. , 2015 ), where it can be utilized by phototrophs in the upper ocean; however, iron is still a limiting factor for phototrophs in the upper ocean ( Raven et al. , 1999 ). The abiotic oxidation of Fe(II) by O 2 is rapid in fully aerated seawater ( Konhauser et al. , 2005 ; Druschel et al. , 2008 ). Therefore, from a microbe’s perspective, Fe(II) is potentially a vast food source, yet it is as ephemeral as it is abundant and bioavailable. Microbial iron oxidation has been recognized in freshwater systems since the 1890s, whereas microbial iron oxidation in marine systems is just beginning to be recognized ( Emerson et al. , 2013 ; Fleming et al. , 2013 ). The isolates of the newest class of Proteobacteria, the Zetaproteobacteria, are described as neutrophilic marine iron-oxidizers ( Emerson et al. , 2007 ). Zetaproteobacteria have been identified throughout the Pacific and Atlantic Oceans at hydrothermal vent habitats and estuaries ( McAllister et al. , 2011 ; Scott et al. , 2015 ; Field et al. , 2016 ). At sites where the predominant vent effluent is high in ferrous iron, Zetaproteobacteria are the dominant microbial mat community members, with the classes of the Gamma-, Delta- and Epsilon-proteobacteria as well as Nitrospira consistently detected in these habitats ( Moyer et al. , 1995 ; Rassa et al. , 2009 ; Fleming et al. , 2013 ). Several Zetaproteobacteria operational taxonomic units have been identified and two are globally ubiquitous in iron-driven microbial mat communities ( McAllister et al. , 2011 ). Zetaproteobacteria are considered ecosystem engineers due to their foundational role in the formation of the microbial mat architecture. This architecture is comprised of exopolysaccharide structures, including twisted helical stalks or tubular sheaths as observed by microscopic analysis of cultures and microbial mats ( Chan et al. , 2011 ; Fleming et al. , 2013 ; Chan et al. , 2016b ). Through the production of stalks or sheaths, the Zetaproteobacteria can alter their physical and chemical environment ( Chan et al. , 2016a ). Furthermore, Zetaproteobacteria are lithoautotrophs and the primary producers in iron-dominated hydrothermal vent systems ( Singer et al. , 2011 ; Field et al. , 2015 ). Previous molecular analysis of microbial mats at Lō’ihi Seamount showed that Zetaproteobacteria correlate with the abundance of key functional genes, but that functional gene abundance did not vary across varying mat morphologies; furthermore, vent chemistry was found to be associated with the observed mat morphologies ( Jesser et al. , 2015 ), suggesting unrealized genomic diversity within the Zetaproteobacteria. Zetaproteobacteria were first described at Lō’ihi Seamount, which is located 35 km south-east of the big island of Hawai’i and hosts a plethora of dynamic hydrothermal vents ( Moyer et al. , 1995 ). In 1996, a major eruption formed Pele’s Pit, a 300 m wide caldera near the summit, with several active hydrothermal venting sites ( Figure 1 ). Before the 1996 eruption, Lō’ihi was dominated by low-temperature diffuse-flow hydrothermal vents emitting fluids up to ~70 °C and elevated levels of Fe(II), CO 2 , CH 4 and NH 4 + ( Sedwick et al. , 1992 ; Wheat et al. , 2000 ) and has now returned to these pre-eruption conditions ( Glazer and Rouxel, 2009 ). Biogeographic patterns for marine microbes remain poorly understood in terms of distribution scale and evolutionary divergence rates. To address this, we sequenced six distinct microbial mat communities collected from Lō’ihi Seamount. From this, a shotgun metagenomics approach was used, where we were able to construct a composite assembly for genome binning. We used differential coverage analysis to reconstruct site-specific community composition and compared this to the community structure as determined by taxa specific Quantitative PCR (qPCR) analyses. Here we present genome-resolved metagenomics to further explore patterns of biodiversity and adaptation of Zetaproteobacteria populations, including two ecologically significant Zetaproteobacteria OTUs (zOTUs).",
"discussion": "Results and Discussion Site description and community structure Microbial mats vary in and around Pele’s Pit in gross morphology and color, from white-yellow to burnt orange ( Figures 2a–f ). In addition to variation in color, the mats had variable textures that were assigned to three specific mat morphological groups associated with variable fluid flow regimes. These were described as curds in the presence of direct flow ( Figures 2a and b ), veils associated with diffuse flow ( Figures 2c and d ), and streamers also found in direct flow ( Figures 2e and f ). Pohaku is the only sample site located on the outside of Pele’s Pit on the southern rim of this caldera ( Figure 1 ), and has been characterized as highest in reduced iron, at nearly 1 m M ( Glazer and Rouxel, 2009 ). Microscopic analysis of the curd-type mat shows the predominance of helical stalks ( Chan et al. , 2016b ), whereas analysis of veil-type mats showed a prevalence of the sheathed morphology ( Fleming et al. , 2013 ). A comprehensive community fingerprint analysis by T-RFLP of 25 mat communities from seven vent sites showed three distinct groups, which corresponded to the gross mat morphology of curds, veils and streamers ( Figure 3 ); however, these groups did not correlate with location or site temperature. All of the microbial mats were collected with a Biomat Syringe sampler, allowing for precision sampling of the topmost active layer of the mat. The morphology of Group I mats are characterized as light yellow to light orange curds, Group II are yellow veiled-type mats and Group III are comprised of white to dark orange streamers attached to the vent orifice. Group I mats had the greatest abundance (56.3%±15.5%) of Zetaproteobacteria within the bacterial community, whereas Group II had significantly less Zetaproteobacteria (20.5%±2.7%) and Group III had the lowest (17.7%±13.7%) as determined by qPCR. SSU rRNA gene clone libraries were constructed from representative mat communities in an attempt to identify the microbial community members driving the T-RFLP clustering. Group I mats had a lower bacterial diversity compared to the other two groups, and exhibited high levels of zOTUs 1 and 2. Group II mats contained a higher abundance of zOTUs 4, 6 and 10 along with Gammaproteobacteria. Group III mats were dominated by sulfur- and hydrogen-metabolizing Epsilonproteobacteria, with a smaller contribution from zOTUs. These results highlight a clear difference between Fe-rich (Groups I and II) and S-rich (Group III) habitats, and between direct flow (Groups I and III) and indirect/diffuse-flow (Group II) environments ( Figures 2 and 3 ; Supplementary Figures 1 and 2 ). Assembly and annotation Six samples, two representatives from each morphotype group, were chosen for metagenomic sequencing. The resulting composite assembly had 162 376 contigs comprised of 289 114 522 bases with an overall GC% of 51.1% and an n50 of 3483. This composite assembly was separated into genome bins based on coverage and tetranucleotide frequencies of the scaffolds with MaxBin 2.0. These bins contained 77.9% of the total composite metagenome bases and 37.4% of the scaffolds. There were no sequences in multiple bins. Genome binning of the composite metagenome resulted in 93 total bins. These genome bins were assessed for completeness and taxonomic classification using CheckM ( Parks et al. , 2015 ). Two of the bins were identified as Archaea, which is consistent with previous analysis showing Archaea were either below the detection limit or less than 5% of the community, and generally derived from deep-sea archaeoplankton retention in the mats ( Moyer et al. , 1998 ; Rassa et al. , 2009 ). Nine genomes were unresolved to the class level; however, one of these bins contained a full-length SSU rRNA gene identified as a Deferribacteres (LoihiBin_014). Unclassified bins were removed from further analysis. The most numerous genome bins identified belonged to the Zetaproteobacteria ( Table 1 ). Overall, the bins had an average n50 of 12 376 in an average of 707 scaffolds. The genome bins range in completeness from 6.55 to 100%, with an average of 70.8% (±30.8%). Contamination ranged from 0.0 to 83%, with an average of 12.6% (±15.0%). On average, the Zetaproteobacteria genome bins were 62.9% (±34.1%) complete, with an average contamination level of 11.6% (±13.4%). T-RFLP and qPCR results both indicate that Group I mat communities were less diverse than Group II or Group III. This is again corroborated by coverage analysis of the genome bins, where communities from Group I have the lowest diversity and Group III had the highest diversity. Zetaproteobacteria genome bins were still present, though as minor community members, in the representative Group III communities ( Figure 4 ). Group III also had higher coverage estimates within the Nitrospira, Gamma-, Epsilon- and Alphaproteobacteria. Zetaproteobacteria genome bins had the highest coverage in the Group I mat communities. The bacterial taxa distribution observed in the clone libraries is consistent with that estimated by genome binning from metagenomics ( Figure 4 ; Supplementary Figure 1 ). Carbon utilization All isolates of Zetaproteobacteria grow via lithoautotrophy and encode for the RubisCO protein for carbon fixation from CO 2 . Mariprofundus ferrooxydans PV-1, M. ferrooxydans JV-1, M. ferrooxydans M34 and Mariprofundus DIS-1 encode for both Form I and Form II large subunit RubisCO gene, whereas Zetaproteobacterium TAG-1 and Mariprofundus sp. EFK-M39 only encodes a Form II RubisCO ( Field et al. , 2015 ). In total, 87 genes were identified as the large subunit of RubisCO. Of these, 67 were Form II and 11 were Form I. Of the Form I genes, only one was binned into a Zetaproteobacteria genome bin (ZetaBin022). The majority of the RubisCO Form II genes belonged to Zetaproteobacteria and 28 of the Zetaproteobacteria genome bins encoded a Form II gene, including the bin with the Form I gene (ZetaBin022). The Gammaproteobacteria had the second highest abundance of RubisCO genes, with four Form I and sixteen Form II genes detected. Twenty-three of the RubisCO genes were not placed into genome bins, but nine of these had the highest similarity to Zetaproteobacteria genes and six were most similar to Gammaproteobacteria ( Supplementary Table 5 ). In comparison, only seven ATP citrate lyase (encoded by aclB ) genes were identified. This is a key gene in the reductive tricarboxylic acid cycle and is found in autotrophic Epsilonproteobacteria and Aquifacales ( Hügler and Sievert, 2011 ). Five of the seven aclB genes were binned into Epsilonproteobacteria genome bins ( Supplementary Table 6 ). The closest taxonomic hits were to Sulfurovum sp AR, Sulfurimonas autotrophica OK10 and Nitratiruptor sp SB155-2. Two of these organisms, S. autotrophica OK10 and Nitratiruptor sp. SB155-2, were isolated from Iheya North hydrothermal field sediments and chimneys, respectively ( Sikorski et al. , 2010 ; Inoue et al. , 2016 ). Sulfurovum sp. AR was isolated from deep marine sediments collected near Svalbard, within the Arctic Circle ( Park et al. , 2012 ). There was a high diversity of Form II RubisCO proteins recovered from Zetaproteobacteria genome bins and unbinned proteins identified as Zetaproteobacteria by IMG ( Figure 5 ; Supplementary Table 5 ). Many of these RubisCO proteins were most similar to RubisCO proteins from the Zetaproteobacteria SAGs belonging to zOTU 2. This zOTU was one of the two considered as cosmopolitan because it is found throughout the Pacific Ocean ( McAllister et al. , 2011 ). Targeted qPCR on RubisCO Form II ( cbbM ) showed high abundance of the gene correlated strongly with a high abundance of Zetaproteobacteria ( Jesser et al. , 2015 ). The abundance of Form II RubisCO genes in comparison to Form I is indicative of adaptations to high CO 2 and very low O 2 environments ( Hernandez et al. , 1996 ; Tabita et al. , 2008 ). The prevalence of Form II RubisCO in the genome bins of the Zetaproteobacteria ( Table 2 ) shows an adaptation to growth in very low O 2 environments similar to what is found in and around Pele’s Pit ( Glazer and Rouxel, 2009 ). Zetaproteobacteria SAGs showed a similar pattern, in that Form I RubisCO was undetected ( Field et al. , 2015 ). Only a single Zetaproteobacteria genome bin contained both forms of RubisCO, suggesting that genotypes containing only Form II are the most prevalent. Nitrogen cycling Biological nitrogen fixation is a key process in any ecosystem. The gene nifH encodes the nitrogenase reductase subunit, and is commonly used to track abundance and diversity among nitrogen-fixing organisms ( Gaby and Buckley, 2012 ). Of the Zetaproteobacteria, Mariprofundus sp. EKF-M39, DIS-1 and M. ferrooxydans M34 encode a nifH gene, and qPCR estimates showed very low occurrence of nifH in microbial mat communities from Lō’ihi Seamount ( Jesser et al. , 2015 ). Consistent with this notion, only eleven nifH genes were identified and only two of these were within Zetaproteobacteria genome bins (ZetaBin035 & ZetaBin089). ZetaBin089 also encodes for nifD and nifK , the nitrogenase alpha and beta subunits, respectively. These genes are encoded on the same contig and are syntenous with the other identified nifH -containing Zetaproteobacteria isolates ( Supplementary Figure 3 ). ZetaBin035 is lacking the alpha and beta subunits, but encodes the dinitrogenase iron-molybdenum cofactor, which is involved in the synthesis of the iron-molybdenum cofactor that binds the active site of the nitrogenase enzyme. Based on these annotations, it appears that these two Zetaproteobacteria bins (ZetaBin035 and ZetaBin089) are potentially capable of nitrogen fixation. Diverse nifH genes have been identified at Axial Seamount, located along the Juan de Fuca Ridge ( Mehta et al. , 2003 ) and interestingly, ammonium has been detected to similar levels as found at Lō’ihi microbial mats, where nifH genes were either below detection or at very low abundance ( Jesser et al. , 2015 ). Ammonium transport proteins ( amt ) were found in 26 of the Zetaproteobacteria genome bins, including the genome bins that encode a nifH ( Table 2 ). Use of nitrate and/or nitrite as a nitrogen source appears to be the most common across the Zetaproteobacteria genome bins. The majority of the Zetaproteobacteria genome bins contained genes for nitrate reduction ( nasAB ) and/or nitrite reductase ( nirBD ) for the assimilation of nitrogen. The dissimilatory nitrate reductase ( napAB ) and nitrite reductase ( nirK/nirS ) were also identified in 18 of the Zetaproteobacteria genome bins ( Table 2 ) showing that denitrification is also possible. The prevalence of the ammonium transport proteins, presence of assimilatory nitrogen pathways, and the low recovery of nifH suggest that Zetaproteobacteria rely more on the presence of bioavailable nitrogen compounds accessed from the environment, rather than by dinitrogen fixation. Arsenic cycling Arsenic has been found at hydrothermal vents and the arsenic detoxification gene, arsenate reductase (encoded by arsC ) has been identified in abundance in microbial mats from Lō’ihi hydrothermal habitats ( Jesser et al. , 2015 ). ArsC reduces arsenate to arsenite, which can then be exported from the cell via an arsenite specific transporter. In the composite assembly, there were 195 identified arsC genes in 67 of the genome bins representing every taxonomic class. The majority of the binned arsC genes were contained within either the Zetaproteobacteria or Gammaproteobacteria genome bins, with 71 and 28 gene copies, respectively. Of the identified arsC genes that were unbinned, taxonomic placement by IMG shows these genes to again be similar to genes from Zetaproteobacteria and Gammaproteobacteria. Arsenite transport proteins were identified in 23 of the Zetaproteobacteria genome bins. All of the Zetaproteobacteria genome bins with an arsenite transport protein contained an arsenate reductase as well. At Tutum Bay, a shallow water hydrothermal vent system, ~1.5 kg of arsenic per day is released into the environment ( Meyer-Dombard et al. , 2013 ). This system also releases reduced iron, and Zetaproteobacteria were shown to heavily colonize slides incubated in situ . Although arsenic geochemistry has yet to be recorded at Lō’ihi vents, the abundance of arsenic-related genes found in our composite assembly suggests that arsenate is abundant in this environment. However, to show this, further geochemical analysis targeting arsenic redox states at Lō’ihi would be required. Electron transport chain Zetaproteobacteria SAGs and isolate genomes encode for a cbb 3 -type cytochrome c oxidase ( Field et al. , 2015 ; Fullerton et al. , 2015 ). M. ferrooxydans PV-1, encodes for subunits I–III ( ccoNOP ) and appears to be lacking subunit IV ( ccoQ ) according to Singer et al. (2011) . Only the CcoNO subunits were identified in the proteomic profile of M. ferrooxydans PV-1 ( Barco et al. , 2015 ). Eleven of the 34 Zetaproteobacteria bins encode all four subunits of the cbb 3 -type cytochrome c oxidase. Mariprofundus sp. EKF-M39, DIS-1, M. ferrooxydans JV-1 and six of the Zetaproteobacteria SAGs encode all four subunits of the cbb 3 -type cytochrome c oxidase. Nine of the Zetaproteobacteria genome bins encode for subunits I–III and appear to lack subunit IV. The ccoQ gene product is a membrane-spanning protein of unclear function; ccoN gene encodes for the catalytic subunit and ccoO, a monoheme c-type cytochrome. Only the ccoNO subunits are common to all gene clusters across multiple bacterial phyla ( Ducluzeau et al. , 2008 ). The cbb 3 -type cytochrome c oxidase has a high affinity for O 2 and is predominately used under microaerophilic conditions and may also be used to prevent O 2 poisoning ( Sievert et al. , 2008 ; Jewell et al. , 2016 ). The aa 3 -type cytochrome c oxidase is encoded by coxABC where expression is repressed in facultative anaerobes under low oxygen conditions ( Pitcher and Watmough, 2004 ). Ten of the Zetaproteobacteria genome bins contain the coxA gene (aa 3 -type cytochrome c oxidase), and all but one of these genome bins encodes for the ccoNOP (cbb 3 -type cytochrome c oxidase) as well. This suggests that like other facultative anaerobes and microaerophiles, Zetaproteobacteria are able to modulate their electron transport chain to account for variable oxygen conditions. Only one of the 24 Zetaproteobacteria SAGs encodes both types of the cytochrome c oxidases ( Field et al. , 2015 ). There is no direct evidence that Zetaproteobacteria can grow anaerobically using nitrate as the terminal electron acceptor; however, a number of other iron-oxidizing Proteobacteria can grow anaerobically this way ( Hedrich et al. , 2011 ; Beller et al. , 2013 ). In the Zetaproteobacteria genome bins there was one bin, ZetaBin084, which encoded the respiratory nitrate reductase, NarG. This genome bin also encodes the cbb 3 and aa 3 type cytochrome c oxidases, that is, both the ccoNO and coxA genes. Iron oxidation is hypothesized to occur on the outer membrane and is coupled to cytoplasmic and membrane-bound electron transfer proteins ( Hedrich et al. , 2011 ; Ilbert and Bonnefoy, 2013 ). From M. ferrooxydans PV-1 genome analysis, a molybdopterin oxidoreductase (Mob, SPV1_03948) was hypothesized to be important in Fe(II) oxidation ( Singer et al. , 2011 ), and showed synteny with two contigs contained in a fosmid library generated from a suction-sample collected from Hiolo South ( Singer et al. , 2013 ). This protein was also identified in the top 25 most abundant proteins of M. ferrooxydans PV-1; however, its function in iron oxidation is questionable due to high similarity to proteins found in non-iron oxidizers ( Barco et al. , 2015 ). In the Zetaproteobacteria genome bins, similar proteins were detected and annotated by IMG as different molybdopterin-containing oxidoreductases (for example, nitrate reductase NapA; Supplementary Table 4 ). Proteomic analysis of M. ferrooxydans PV-1 revealed a membrane bound cytochrome that was highly expressed and distantly related to cytochrome c 2 of Acidothiobacillus ferrooxydans ( Barco et al. , 2015 ; White et al. , 2016 ). It has been proposed that this protein, referred to as Cyc2 PV-1 , is the site of electron transfer from iron to a cytoplasmic cytochrome (Cyc1 PV-1 ), which was also identified as high-abundant by proteomic analysis. From Cyc1 PV-1 , electrons are hypothesized to be shuttled into a membrane-bound electron transport chain, terminating with the cbb 3 -type cytochrome c oxidase. Using the amino acid sequence of Cyc1 PV-1 and Cyc2 PV-1 to search the composite metagenome, 24 and 41 gene copies, respectively, were identified within the Zetaproteobacteria genome bins ( Table 2 ). The open reading frames most similar to Cyc1 PV-1 were annotated as cytochrome c553, whereas the Cyc2 PV-1 genes were annotated as hypothetical proteins by IMG ( Supplementary Tables 2 and 3 ). The identification of Cyc1 PV-1 and Cyc2 PV-1 in our Zetaproteobacteria genome bins, supports the hypothesis that a Cyc2-like protein is the site of iron oxidation, as opposed to the alternative hypothesis using the Mob protein ( Hedrich et al. , 2011 ; Singer et al. , 2011 ; Ilbert and Bonnefoy, 2013 ; Barco et al. , 2015 ). These Cyc2-like proteins were identified in every zOTU detected, indicating their ubiquity across the Zetaproteobacteria, including within the ecologically significant taxa ( Table 2 ). Whole genome comparisons In this composite metagenome study, there were 249 total SSU rRNA genes recovered. Of these, 41 were contained within Zetaproteobacteria genome bins as determined by CheckM and 37 SSU rRNA genes were identified as Zetaproteobacteria by the RDP classifier ( Wang et al. , 2007 ; Parks et al. , 2015 ). Previous studies on Zetaproteobacteria SSU rRNA diversity identified two operational taxonomic units that were ubiquitous across the Pacific Ocean, referred to as zOTUs 1 and 2 ( McAllister et al. , 2011 ). Genomes were compared at the nucleotide level to assess genomic diversity across the Zetaproteobacteria genome bins as compared to isolate genomes and SAGs ( Figure 6 ) by ANI. Hierarchical clustering of the genomes based on ANI showed that genome bins most similar to zOTUs 1 and 2 are the most highly represented, with 10 and 13 out of the 34 Zetaproteobacteria genome bins, respectively. Based on Form II RubisCO phylogeny, these zOTUs constitute a single lineage that diverged more recently than any of those that occurred in other lineages ( Figure 5 ). Both these zOTUs were also found to be the most abundant phylotypes detected in microbial mats from Lō’ihi hydrothermal habitats by SAGs and SSU clone library analyses ( McAllister et al. , 2011 ; Field et al. , 2015 ). Based on the cluster analysis of ANI comparisons from our Zetaproteobacteria genome bins, this study has shown that zOTU 2 represents a monophyletic cluster and is distinct from all the other zOTU clusters ( Figure 6 ), and based on estimated genome size hints, that genome streamlining may be occurring within this group. This zOTU was also the first to be identified from any hydrothermal system ( Moyer et al. , 1995 ). Our whole genome cluster analysis also showed that zOTUs 1, 4, 6 and 10 have much greater genomic dissimilarity (that is, diversity) than what would be expected based on SSU rRNA identity alone. The distribution of Zetaproteobacteria genome bins across the three different groups of mat communities shows that zOTU 2 is the most abundant in both Group I and Group II (that is, both curds and veils) type mats based on gross morphology, representing twisted-stalks and sheaths, respectively. The Group III mats (streamers), which have a low Zetaproteobacterial abundance relative to the other members of the community, included zOTU 11 as the most highly represented within this mat-type ( Supplementary Figure 4 ). Using this hierarchical cluster analysis approach, patterns of metabolic potential across zOTUs can also be realized. The only two bins with a nifH gene (ZetaBin035 and ZetaBin089) were also most closely related to isolates that are able to fix nitrogen. All cultured isolates remain within the same tight cluster, including the type strain M. ferrooxydans PV-1, possibly indicating a narrow range of selection pressure resulting from our present culturing techniques. Furthermore, there were few Zetaproteobacteria genome bins with similarity to any cultured isolates, suggesting environmental parameters are poorly mimicked in the lab. In general, the RubisCO protein relationships and genome relationships identified by ANI were conserved (that is, similar). None of the genome bins within zOTU 2 contained genes for the aa 3 -type cytochrome c oxidase, further supporting adaptation to the low O 2 levels found at Lō’ihi hydrothermal habitats."
} | 7,423 |
38814057 | PMC11218620 | pmc | 3,941 | {
"abstract": "ABSTRACT The oxidation of sulfide-bearing mine tailings catalyzed by acidophilic iron and sulfur-oxidizing bacteria releases toxic metals and other contaminants into soil and groundwater as acid mine drainage. Understanding the environmental variables that control the community structure and metabolic activity of microbes indigenous to tailings (especially the abiotic stressors of low pH and high dissolved metal content) is crucial to developing sustainable bioremediation strategies. We determined the microbial community composition along two continuous vertical gradients of Cu/Ni mine tailings at each of two tailings impoundments near Sudbury, Ontario. 16S rRNA amplicon data showed high variability in community diversity and composition between locations, as well as at different depths within each location. A temporal comparison for one tailings location showed low fluctuation in microbial communities across 2 years. Differences in community composition correlated most strongly with pore-water pH, Eh, alkalinity, salinity, and the concentration of several dissolved metals (including iron, but not copper or nickel). The relative abundances of individual genera differed in their degrees of correlation with geochemical factors. Several abundant lineages present at these locations have not previously been associated with mine tailings environments, including novel species predicted to be involved in iron and sulfur cycling. IMPORTANCE Mine tailings represent a significant threat to North American freshwater, with legacy tailings areas generating acid mine drainage (AMD) that contaminates rivers, lakes, and aquifers. Microbial activity accelerates AMD formation through oxidative metabolic processes but may also ameliorate acidic tailings by promoting secondary mineral precipitation and immobilizing dissolved metals. Tailings exhibit high geochemical variation within and between mine sites and may harbor many novel extremophiles adapted to high concentrations of toxic metals. Characterizing the unique microbiomes associated with tailing environments is key to identifying consortia that may be used as the foundation for innovative mine-waste bioremediation strategies. We provide an in-depth analysis of microbial diversity at four copper/nickel mine tailings impoundments, describe how communities (and individual lineages) differ based on geochemical gradients, predict organisms involved in AMD transformations, and identify taxonomically novel groups present that have not previously been observed in mine tailings.",
"conclusion": "Conclusions Mine tailings present a challenge in waste management practices. AMD environments harbor taxonomically and metabolically diverse microorganisms, which have evolved a wide variety of adaptations to thrive in highly heterogeneous contaminated environments. We utilized a 16S rRNA amplicon sequencing approach to characterize microbial communities at narrow (~10 cm) depth intervals at four locations at two tailings sites, providing a high-resolution profile of community changes across tailings gradients of varying geochemical compositions. Our research shows that microbial communities are highly diverse between and within each sampling location. Our findings suggest that future sampling efforts that capture a wide range of tailings locations and depths could contribute to the isolation of a broad range of bioremediation-relevant species. We identified that overall community composition is correlated with pH, Eh, alkalinity, salinity, and some metal ions, which surprisingly did not include the major constituents Cu and Ni. Other variables that did not correlate with overall diversity were DOC, F, SO 4 , NO 3 , Al, As, Co, Li, Mg, Pb, S, Si, Sr, and Zn. The abundance of most individual lineages was also closely associated with different geochemical factors. However, due to the covarying nature of many environmental variables, it is important to frame predictions about which factors microorganisms are directly responding to, as well as the trajectory of contaminant cycling within tailings, around what is currently known about the physiology and ecology of each individual organism (or its closest known relatives). Our results showed that amplicon sequencing data can be used to identify specific microenvironments of tailings harboring novel or currently uncultured species, although this approach is limited in its ability to predict the physiological and metabolic properties of these uncharacterized organisms. Many ASVs present in Strathcona Mill and Nickel Rim tailings could not be classified at the species level, including potential iron/sulfur oxidizers and reducers, though this result is likely influenced by the fact that most samples were circumneutral and low in metal concentration. Functional profiling of tailings known to contain these novel lineages using metagenomics and metaproteomics will be an important future direction to identify genes associated with metal (Fe, Cu, Ni) resistance and transformation. By informing on geochemical variables required to provide niche-specific conditions for growth in the lab our work also provides an insight into improved approaches for establishing enrichment cultures that target these species.",
"introduction": "INTRODUCTION Mine tailings are the fine-grained residues left as waste as part of the extraction process of valuable metals from ores. Sulfide-rich tailings (generated from selective flotation of sulfide minerals) stored in subaerial impoundments are a major source of groundwater contamination due to release of acid mine drainage (AMD). When sulfidic minerals (e.g., pyrite, FeS 2 ) are exposed to oxygen and water, the oxidation of ferrous iron (Fe II ) and sulfur generates high amounts of Fe II SO 4 and H + . The Fe II release by sulfide mineral oxidation may be further oxidized to Fe III , which contributes to the oxidation and dissolution of additional pyrite and other metal-bearing minerals. The overall effect of these reactions includes the formation of acidic porewater with high concentrations of heavy metals. Release of AMD to natural environments and freshwater is detrimental to natural ecosystems and human health ( 1 ). Microbial activity can accelerate the rate of AMD formation by hundreds to millions of times compared to abiotic oxidation alone ( 2 ). Acidophilic, chemolithotrophic bacteria such as Acidithiobacillus ferrooxidans ( 3 , 4 ) obtain energy through the oxidation of iron and sulfur compounds, releasing additional H + , SO 4 , and regenerating the Fe III ion. While these organisms are detrimental in the context of tailings waste management, they have also been extensively studied for industrial bioleaching applications to extract valuable metals from tailings ( 4 , 5 ). Microbial consortia isolated from AMD environments also include acidophilic sulfate reducers that precipitate metal ions from solution as metal sulfides ( 6 ), and species that immobilize metal ions through bioaccumulation or biosorption ( 7 – 10 ); all of these processes serve as potential remediation approaches to remove metals from tailings porewater and mitigate the negative environmental and human health impacts of mine wastes. Physical remediation strategies may also be employed to reduce acidity and metal leaching of tailings, such as by the application of cover layers (e.g., fine-grained silt and clay or organic carbon) that limit oxygen and water ingress into the tailings and facilitate the growth of vegetation ( 11 – 14 ). Vegetation (phytoremediation) and organic matter in organic covers also increase microbial community diversity of the tailings by promoting the establishment of organoheterotrophs and nitrogen-fixing rhizobacteria (increasing plant growth, nutrient turnover, and ecosystem productivity) ( 15 , 16 ). Many studies have described microbial genera common across varied AMD environments, such as the acidophilic iron and/or sulfur oxidizing Acidithiobacillus , Leptospirillum , and Sulfobacillus spp ( 17 , 18 ). However, mine tailings can be diverse in composition depending on the metals targeted for extraction, resulting in substantial differences in the extent of oxidation and acidification and concentrations of contaminants. As a result, the microbial community compositions found within mine tailings are also highly heterogeneous. pH is an important driving factor in the relative abundance of several lineages, with Proteobacteria generally dominant in slightly to moderately acidic tailings (pH 4–7) and Euryarchaeota (especially Ferroplasma spp.) more abundant in highly acidic tailings (pH <3) ( 19 , 20 ). The concentrations of metal ions (including Fe, Cu, Ni, and trace elements such as Cd, Hg, and U) select for organisms with resistance mechanisms against these contaminants ( 21 – 24 ). Mine tailings also exhibit vertical stratification, with distinct zones of oxidation, neutralization, and unaltered tailings, heterogeneous physical characteristics (e.g. , particle size) and temporal variability in moisture content; all of which influence the microbial guilds dominating each microenvironment ( 25 ). A wide variety of microbial guilds are applicable to AMD bioremediation (e.g., bioreactors with engineered sulfate-reducing consortia, in situ remediation via iron-oxidizing consortia in aerobic wetlands) ( 26 ) but making improvements in bioreactor/ in situ remediation efficiency and cost-effectiveness depends on understanding the appropriate isolates/consortia to leverage given the specific geochemistry of the target mine tailings. Tailings are highly selective environments that may be host to unexplored phylogenetic radiations of bacteria and archaea that have evolved unique adaptations to survive in heavy metal contaminated areas. Therefore, tailings provide an excellent starting point to identify AMD-tolerant microbes that can be used to improve the efficiency of engineered bioreactors and in situ remediation strategies. Recent advances in high-throughput sequencing methods such as 16S rRNA amplicon sequencing have enabled the high-resolution examination of the taxonomic diversity of AMD microbiomes ( 27 ). In this study, we compared microbial community diversity across vertical and horizontal transects of tailings associated with two Ni/Cu mining sites in the Sudbury Basin region of Ontario, Canada. 16S rRNA amplicon sequencing was used to characterize community composition across 3–5 m depth transects at four tailings locations. A temporal comparison was made for one tailings location to assess community stability over time (2 years). Our objectives were to identify the main geochemical factors influencing microbial diversity and variability between locations, predict the contribution of microbial guilds to iron/sulfur cycling within tailings, as well as to explore microbial novelty within the context of mine tailings environments.",
"discussion": "RESULTS AND DISCUSSION Sampling site and tailings core characteristics The Sudbury Basin is a major geological structure in the Canadian Shield. Formed by a meteorite impact ( 28 ), it consists of mainly Fe-Ni-Cu sulfide deposits (e.g. , pyrrhotite, chalcopyrite, and pentlandite), which may be enriched in precious metals ( 29 ). Tailings stored at two mines owned and operated by Glencore Sudbury Integrated Nickel Operations (Sudbury INO) were sampled for characterization of microbial community diversity and function: the Strathcona Waste Water Treatment System (SWWTS), which received tailings from 1968 to 2012 ( 30 , 31 ) and Nickel Rim North tailings area, which received tailings from 1953 to 1958 ( 32 ). Operations at Sudbury INO are focused primarily on extracting and processing nickel and copper, with cobalt and precious metals as by-products ( 33 ). At Strathcona, two locations, Moose Lake 25 (ML25) and ML34, were chosen to assess spatial (vertical) variation. The sulfide-rich (9–18 wt. % S) ( 34 ), coarse-grained tailings at ML25, which have been exposed to atmospheric weathering since deposition decades ago, are extensively oxidized. The porewater at ML25 is low pH (pH <5) and contains high concentrations of contaminants ( 35 ) ( Fig. 1b ; Data File S1). In contrast, the high sulfide tailings at ML34 are overlain by a 2 m layer of desulfurized tailings (DST) cover of <1.4 wt. % S mainly as pyrrhotite, and a 50 cm organic carbon cover composed of a mixture of a mixed municipal compost amended with biosolid fertilizer. The cover materials are not potentially acid generating, and the fine-grained DST cover maintains a high moisture content, limiting O 2(g) diffusion. The pH is circumneutral and concentrations of dissolved metals are lower than at ML25 ( 12 , 35 ) ( Fig. 1b ; Data File S1). Fig 1 Distribution of bacterial and archaeal genera along tailings depth gradients and pore-water chemistry analyses. The heatmap ( a ) indicates the percent relative abundance of genera that were present at an abundance of at least 15% in at least one sample, across all depth intervals at each sampling location. The phylum and genus are labeled (if the genus was unknown, the family is shown instead). Relative abundances were log transformed, with 0% abundance values set to −2.999. Genera with predicted iron/sulfur cycling capabilities are indicated in color. The corresponding values for pH, Eh, and contaminant concentrations (Cu, Ni, Fe, SO 4 ) are plotted in ( b ). For contaminant plots, the coordinates (depth in meters, concentration in mg/L) of the peak concentration value are indicated directly on the plot. At Nickel Rim North (NR), locations NR18 and NR3 were selected for comparison. The unoxidized tailings contain on the order of 3 wt. % S mainly as pyrrhotite ( 36 ). Contaminant profiles are similar at these two locations ( Fig. 1b ; Data File S1), although NR3 has shallow water-table elevation and the shallow tailings have a high moisture content, leading to a shallower depth of oxidation and lower concentrations of dissolved metals than at NR18 ( 36 ). NR18 also has a shallow (< 10 cm) organic soil, deposited in the early 1990s ( 36 ). In 2021, tailings core samples extracted from each location were collected in lengths of aluminum pipe (with inner diameters measuring 5.08 and 7.62 cm) driven into the tailings using the technique of Starr and Ingleton ( 37 ). The core samples were subsectioned in lengths of 10 cm (except for the ends of each core, which were 10–17 cm), generating a total of 112 samples (Table S1). A core collected, frozen, and processed in the same way from NR18 in 2019 was used for comparison to 2021 samples (Table S1). 16S rRNA amplicon sequencing statistics Out of a total of 7,610,282 raw reads (ranging from 161 to 217,042 per sample), 6,100,701 were kept after filtering, denoising, merging, and detection of chimeric reads. From these, 11,116 amplicon sequence variants (ASVs) were identified. The median ASV frequency per sample was 50,912, though the range varied from 141 to 199,300, reflecting the variability in community composition within samples. The median frequency of each ASV across all samples was 29, with a maximum frequency of 546,045. ASVs with a total frequency of 100 or less were generally confined to one or two samples, while higher abundance ASVs were present across multiple cores and/or multiple depth ranges. Community composition Multiple changes in the dominant phyla occur along the depth gradients for each sampled core ( Fig. 2 ). In general, upper layers are mainly comprised of Proteobacteria and Acidobacteriota, which is consistent with most studies on mine tailing communities ( 19 , 20 , 24 ). Desulfobacterota and Firmicutes are abundant in deeper tailings; members of these phyla include anaerobic sulfate-reducing bacteria commonly found in anoxic layers of tailings ( 38 – 41 ). Actinobacteriota were highly abundant in ML25 tailings below 2 m in depth ( Fig. 2 ), but not at the other locations. While members of the Actinobacteriota phylum display a diverse range of metabolism, the genera that are commonly found in mine tailings (especially metal tolerant Arthrobacter spp.) ( 42 – 44 ) are usually obligate aerobes associated with the rhizosphere of metal hyper-accumulating plants. Therefore, we would expect to see them in the surface layers of vegetated tailings (e.g., ML34) as opposed to the deeper, ML25 tailings. The Actinobacteriota ASVs in ML25 appear to belong to novel clades not previously found in mine tailings; their classification is discussed in the next section. Fig 2 Phylum-level microbial community composition of mine tailings. Bars representing samples are grouped by location (ML25, ML34, NR18, and NR3) and depth. Bars depict relative frequency of ASVs corresponding to all phyla present at a frequency of ≥5% in at least one sample. The bar widths (y-axis) are not scaled based on the depth range covered. The WPS-2 phylum is also novel in terms of its presence in mine tailings and seems to be unique to Nickel Rim samples (particularly at NR18 at depths between ~1 and 3 m) ( Fig. 2 ). WPS-2 (or Ca. Eremiobacterota) is a recently described phylum, consisting of a wide range of metabolically diverse bacteria found in many types of terrestrial environments ( 45 ). Metagenomes isolated from Antarctic desert soil suggest some WPS-2 members are acidophilic ( 45 ). Another study found that the WPS-2 was abundant (with ASV relative frequencies of up to 24%) in the unvegetated soils of extinct iron-sulfur springs in British Columbia ( 46 ). Given that sampling occurred during near-freezing temperature periods and the acidic, iron/sulfur contaminated nature of AMD, it is not surprising that the WPS-2 could thrive in tailings environments. However, Ji et al. ( 45 ) and Sheremet et al. ( 46 ) found no evidence of microaerophilic or anaerobic metabolic capacity from genomic data in their studies, meaning the WPS-2 ASVs present at Nickel Rim likely represent organisms with novel metabolic capacity within this understudied lineage. Community change along geochemical gradients A more detailed analysis of community composition changes (in terms of genus-level relative abundance) across tailings samples is shown in Fig. 1a . The distribution of the most abundant genera (≥ 15% in at least one sample) was highly variable. Most populations did not show consistent abundance across samples and tended to be confined to and/or enriched in specific depth ranges. To track how variation in geochemistry along cores impacted the distribution of the lineages highlighted in Fig. 1a , pH/Eh and the concentration of the most environmentally relevant contaminants (Fe, Ni, Cu, SO 4 ) were linked to genera abundances ( Fig. 1b ). The pH of ML25, NR18, and NR3 ranged from moderately to slightly acidic (~3–6) wherein pH increased with depth, while the pH of ML34 was slightly above seven and more stable across all depths. The general trend of peak Fe, Ni, and Cu concentrations was similar between locations: the highest Cu concentrations were close to the surface, followed by nickel, and finally iron. This observation is consistent with previous hydrogeochemical assessments of Nickel Rim ( 32 ). These differences are due to differences in the susceptibility of pyrrhotite (Fe (1-x) S), pentlandite [(Fe,Ni) 9 S 8 ], and chalcopyrite (CuFeS 2 ) to oxidation, and to the formation of secondary Fe oxyhydroxide phases (e.g., goethite; αFeOOH) and the secondary Cu sulfide mineral covellite (CuS) ( 32 ). Optical examination of mineral thin sections from Nickel Rim and similar tailings impoundments have shown that pyrrhotite is the most susceptible (least resistant) to oxidation, followed by pentlandite, chalcopyrite and pyrite ( 47 ). Other factors influencing the localization of dissolved metals include differences in metal sorption capabilities to soil/clay particles (competitive sorption of Cu ions is greater than Ni ions, reducing transport to lower depths) ( 48 , 49 ) as well as pH-dependent precipitation of Fe (oxy)hydroxides (which can sorb Ni) and covellite (CuS) ( 36 ). Sulfate concentrations generally followed Fe, as both are released during the dissolution of sulfide minerals (see Fig. S1 for a summary of correlations between geochemical variables). Correlations between abundance data ( Fig. 1a ) and geochemistry (variables from Fig. 1 , plus Ni, Cu, and SO 4 ) identified several significant connections ( Fig. 3 ). We included Cu and Ni in our analyses despite their concentrations not having a significant impact on overall community diversity because they are the main contaminants released from these tailings, and increased tolerance to these metals would be an asset for bioremediation-relevant organisms. Dissolved Cu concentrations were moderately correlated with the abundance of almost half of the dominant genera (21 of 44 genera, 48%). Ni showed comparatively fewer significant correlations ( Fig. 3 ). However, it was difficult to determine which specific factors were directly interacting with microorganisms, due to the interconnected nature of environmental variables (Fig. S1). Correlation matrices were also individually calculated for each sampling location (Fig. S2), which differed from the pooled data set (most notably, ML25 had much stronger R values due to a wider range of contaminant concentrations, contributing to stronger selective forces). Fig 3 Correlation matrix of genus-level relative abundance and geochemical factors. Spearman’s rank correlation coefficients were calculated between the relative abundance of the abundant genera ( Fig. 3 ) and select geochemical factors (Dep = depth, EC = electrical conductivity, Alk = Alkalinity), across all sampling locations. Significant correlations ( P < 0.05, Bonferroni corrected) are shown. For location-specific correlations, see Fig. S1. Diverse consortia associated with iron and sulfur cycling were identified in each core, so their abundance and localization were specifically examined ( Fig. 1 and 3 ). Iron/sulfur-oxidizing genera common to AMD were typically confined to the upper 2 m of tailings. Acidithiobacillus ASVs (the majority of which were classified as A. ferrooxidans ) were the most abundant predicted iron-oxidizer in the subsurface (0.5–1.5 m). The surrounding environment is likely microaerophilic at this depth, which are optimal growth conditions for A. ferrooxidans ( 50 ). Pore gas oxygen data were limited to near the tailings surface due to the low gas-filled pore space present in deeper tailings with a high water content and was shown to decrease from atmospheric levels at ground surface to <10% over 1 m in ML25, ML34, and NR18 (Fig. S3). It is worth noting that A. ferrooxidans also exhibits considerable metabolic diversity: not only is this species able to oxidize both ferrous iron and reduced sulfur species in oxic environments, but it can also use ferric iron as an electron acceptor in anoxic conditions. Iron reduction is likely the predominant metabolism supporting A. ferrooxidans populations under anoxic conditions below depths of 2 m. The abundance of Acidithiobacillus ASVs in ML34 despite pH >7 conditions is unusual, but could be associated with the period of exposure to atmospheric oxygen and sulfide oxidation prior to the installation of the DST and organic carbon cover layers. The persistence of Acidithiobacillus ASVs could be leveraged for bioremediation approaches, such as in aerobic wetlands that use neutrophilic iron oxidizers to precipitate ferric iron minerals (e.g., ferrihydrite and goethite) ( 51 ). Existing iron-oxidizers currently applied in aerobic wetlands include Gallionella and Leptothrix spp., which are not native to AMD ( 26 ). Acidithiobacillus species/strains that have adapted to survive in neutral or alkaline environments may be more efficient at immobilizing iron by oxidation while also being tolerant to higher concentrations of copper, nickel, and other toxic metals in tailings. Leptospirillum spp. (mostly L. ferrooxidans ), exclusive iron-oxidizers, follow a similar distribution but are much less abundant. L. ferrooxidans prefers lower pH ranges (0.5–0.7) than A. ferrooxidans (1–3) in extreme AMD environments ( 52 ), but was co-localized in moderately acidic environments such as ML25. A. ferrooxidans abundance did not significantly correlate with any factor in the combined data set ( Fig. 3 ), but did correlate with a few factors such as pH and Eh within each sampling location (Fig. S2). L. ferrooxidans abundance was moderately-to-strongly correlated with several environmental factors in both the combined data set and in ML25. Based on these observations, A. ferrooxidans could be described as more of a “generalist” species that is able to inhabit many types of mine tailings (see earlier discussion on their metabolic diversity), although they will still localize to a niche within each tailings environment based on preferences in pH, oxygen availability, etc. In contrast, L. ferrooxidans is a “specialist” that is adapted to survive at high concentrations of metals, but is outcompeted in mine tailings with low metal contamination. L. ferrooxidans abundance correlates strongest with copper, R = 0.64, although previous studies have shown that their higher affinity for ferrous iron and tolerance to higher concentrations of ferric iron is what drives niche partitioning between A. ferrooxidans and L. ferrooxidans ( 51 ). It is also important to note that bioleaching with mixed cultures containing both Acidithiobacillus spp. and Leptospirillium spp. is more effective than pure cultures ( 53 – 56 ). Therefore, oxidation and AMD formation within ML25 tailings are expected to be higher due to the combined activity of both organisms (as evidenced by the lower pH and higher dissolved metal content) ( Fig. 1b ), even if their total abundance is similar to single populations of A. ferrooxidans at other locations. Genera predicted to exclusively be sulfur oxidizers included Sulfurifustis , Thiobacillus , Halothiobacillus , Sulfuritalea , Sulfuriferula, Sulfuricella , and Desulfovibrio . The Sulfurifustis ASV most abundant in the surface of ML25 and NR18 tailings appears to be a novel species tolerant of acidic environments, as previously isolated Sulfurifustis species are neutrophilic ( 57 ). The abundance correlation for ASVs in the Sulfurifustis genus with geochemical factors resembles that of Leptospirillum rather than other sulfur oxidizers, suggesting that members of the Sulfurifustis are also specialized for resistance to high metal concentrations ( Fig. 3 ). Thiobacillus and Halothiobacillus ASVs were also not classified to the species level and are predicted to be novel neutrophilic sulfur oxidizers ( 58 ) most abundant in ML34; Thiobacillus abundances in particular exhibit a strong positive correlation with pH (R = 0.71) and alkalinity (R = 0.77). The localization of Desulfovibrio (and Sulfuricella within NR18) ASVs to the deeper levels of the tailings is in line with their expected use of sulfur oxidation pathways coupled to NO 3 reduction rather than O 2 (in the case of Desulfurivibrio , it is expected to use the dissimilatory sulfate reduction pathway in the reverse direction), potentially allowing these populations to co-exist with anaerobic sulfate reducers by metabolizing the sulfides generated by the sulfate reducers ( 59 , 60 ). The distribution of potential iron reducers was more varied than iron oxidizers, and included Geobacter , Acinetobacter , Thermincola, Metallibacterium, and Pseudomonas populations ( 61 – 63 ). Geobacter and Thermincola generally localize to greater depths in NR18/NR3, where dissolved iron is high. At pH >2.5, ferric iron precipitates as secondary oxyhydroxide minerals ( 51 ), and reductive dissolution is problematic in these conditions, as it re-mobilizes ferrous iron. The Pseudomonas and Acinetobacter do not display a consistent localization pattern, but as these genera are highly diverse and ubiquitous in various environments, predictions about iron-reducing capabilities should only be done on the species level, which was not possible as the most abundant ASVs were only classified at the genus level. The Metallibacterium is a recently described genus, with one isolate ( M. scheffeleri ) from an acidic biofilm demonstrating iron reduction ( 63 ), although Bartsch et al. ( 64 ) could not identify any iron-reducing genes through genomic, transcriptomic, and proteomic analyses. Unlike the other potential iron reducers, Metallibacterium ASVs co-localized with iron oxidizers ( L. ferrooxidans in ML25, and A. ferrooxidans in ML34/NR18); as they are facultative anaerobes and only reduce iron under anoxic conditions ( 63 ), it is unlikely that they are actively reducing iron at these depths. However, as Metallibacterium have shown the potential to alkalinize their surroundings through the release of ammonium ( 64 ), they are a promising candidate for AMD bioremediation and will be a target for investigation in future multi-omic studies. Sulfate reducers including members of the Firmicutes ( Desulfosporosinus , Ca . Desulforudis, Desulfotomaculum ) and Desulfobacterota (unclassified Desulfuromonadaceae) were generally dominant at depths > 2 m, which is below the depth of oxygen ingress. As expected, the sulfate reducers were generally positively correlated with pH (most are inhibited below pH 5.5) ( 65 ) and negatively correlated with Eh. They also show moderate negative correlations with dissolved copper concentration, as they are expected to stimulate secondary covellite (CuS) precipitation via the production of sulfide. Relative abundances were not consistent between locations, which could reflect preferences of individual taxa to specific environmental conditions, or legacy impacts from the initial subsurface community present in the ore from which the tailings were generated. The Desulfotomaculum , almost exclusively localized to ML25, is a genus mostly associated with deep subsurface environments ( 66 ). Ca . Desulforudis (localized to NR18) is also typically found in very deep environments, the most notable being Ca . Desulforudis audaxviator, which was the sole organism present 2.8 km below the surface in a gold mine ( 67 ). Desulfosporosinus are common in natural soil/sediments as well as mine tailings in cold climates ( 68 ) and were more evenly distributed in Strathcona Mill tailings compared to Nickel Rim. Desulfuromonadaceae abundance followed a similar trend. Finally, the Sva0485 clade ( Ca . Acidulodesulfobacterales) have recently been described as dominant members of sulfate-reducing consortia in AMD and ferruginous lakes ( 69 ). While the characterized members of the Sva0485 lineage possess genes for dissimilatory sulfate reduction, they are facultative anaerobes and also encode genes for sulfur oxidation, iron cycling, and methanogenesis ( 69 , 70 ). This varied metabolic potential could explain why their depth range was not consistent between tailings locations here, and why their abundance was not significantly correlated to geochemistry ( Fig. 3 ), although individual locations showed variable correlations between Sva0458 and salinity (Fig. S2). Biological sulfate reduction is leveraged in a variety of AMD bioremediation strategies, including anaerobic (or compost) bioreactors, sulfidogenic bioreactors, and permeable reactive barriers ( 26 ). All these approaches rely on biogenic sulfide production, typically in the form of H 2 S, which precipitates various metal-sulfide minerals ( 26 , 51 ). The archaeon E-plasma (and other unclassified Thermoplasmataceae ASVs) abundant at ~1 m depth at ML25 and NR18 have been identified in various AMD and non-AMD environments ( 17 , 71 ). E-plasma were found to dominate microbial AMD communities within the Parys Mountain mine (UK), characterized by low-to-moderate temperatures and extremely low pH ( 72 ), suggesting that they are adapted to colder temperatures (unlike most members of the order Thermoplasmatales, which are mesophilic to moderately thermophilic) ( 71 ). The metabolic potential of E-plasma and other related “alphabet-plasma” lineages belonging to the Thermoplasmataceae is currently unknown, as isolation attempts have not yet been successful. However, genomic data suggests that they are heterotrophic scavengers ( 72 ). Many of the most abundant lineages are taxonomically novel and/or previously not identified within mine tailings ( Fig. 1a ). The Actinobacteriota (most abundant in ML25 but present at all locations) ASVs mostly belong to the order Gaiellales (common in deep sea sediments) ( 73 ), WCHB1-81 (microcystin-contaminated lakes) ( 74 ), and OPB41 (subsurface environments) ( 75 ). Other ASVs that were unclassified at the genus/species level include the Acidobacteriaceae (associated with mine tailings but highly diverse ( 76 ), as shown by the fact that the two subgroups identified in this study exhibit very different localization patterns), as well as the Gemmatimonadaceae, Pirellulaceae, Comamonadaceae, Chitinophagaceae, and Caulobacteraceae, which have not previously been associated with mine tailings. The high amount of unresolved microbial diversity from 16S rRNA gene sequencing data implicates that the Strathcona Mill and Nickel Rim mine tailings host currently uncharacterized microorganisms that could participate in the biogeochemical processes governing the release, mobility, and/or attenuation of contaminants associated with AMD. Nonetheless, it is important to note that most samples analyzed in this study were low in metal concentration and circumneutral pH, and therefore not representative of microbial communities that could be leveraged for the remediation of highly acidic tailings. Future multi-omic and culture-based approaches are required to elucidate the full range of microbial activities present at these sites. The ultimate goal of these combined approaches is to contribute to the development of in situ bioremediation strategies for AMD, or to limit the extent of contaminant release that is accelerated by microbes ( 77 ). Alpha diversity Plots comparing Faith’s phylogenetic distances, Shannon indices, and Observed ASVs of samples (grouped by location) showed consistent trends across the four cores ( Fig. 4 ). Samples from ML25 consistently scored low on α-diversity metrics across all depths, although the specific taxonomic composition of samples was highly variable when comparing upper and lower depths (discussed in the Community Composition section). The α-diversity metrics of NR3 samples were overall similar to ML25, although the deeper fractions of the core tend to have higher diversity ( Fig. 4 ). It is possible that the deposition of NR3 tailings on top of native vegetation in the 1950s, the presence of nearby plants, and periodic flooding events ( 36 ) account for the higher diversity at greater tailings depths, supported by the increased DOC concentrations in these samples (Data File S1). Fig 4 α-diversity metrics for all sampling locations and depths. Faith’s phylogenetic distances ( a ), Shannon diversity indices ( b ), and observed ASVs ( c ) were calculated with QIIME2 using a rarefied sampling depth of 12,000 reads. Individual samples overlaid on box plots are colored according to the vertical depth below ground level. Significant differences (Kruskal-Wallis, P < 0.05) between locations are designated by different letters above each boxplot. ML34 and NR18 showed significantly higher α-diversity metrics compared to ML25/NR3, which we attribute to the cover layers and vegetation present at these locations ( 20 , 78 ), and is especially evident in the fact that ML34 diversity was highest in the samples closest to the surface. A similar trend was seen for NR18, but due to low read numbers in the deeper NR18 sections, most samples below 2 m were excluded after rarefying to a minimum sampling depth of 12,000, hindering a full comparison. Beta diversity Samples did not appear to discretely cluster by location in any of the β-diversity ordinations ( Fig. 5a ; Fig. S4a and S5a), indicating that the variability between samples at the same location was greater than the overall differences between locations. Fig 5 Weighted unifrac PCoA plot for ( a ) all sampling locations combined, ( b ) ML25, ( c ) ML34, ( d ) NR18, and ( e ) NR3. Samples in plots b-e are colored by depth. The weighted unifrac principal coordinate analysis ( Fig. 5a ) was the most effective in mapping distances between samples, with the two principal axes explaining 91.7% of variation in the data. ML25 samples showed particularly high separation along the primary axis, which may reflect the much broader range of contaminant concentrations along the core, such as dissolved Fe, Ni, Cu, and SO 4 (Data File S1). Analysis of individual locations showed that weighted unifrac PCoA was also best for mapping distances ( Fig. 5b through e ), with the principal axes explaining 66.1 – 96.4% of the variation. There was some separation of samples along one or both axes based on depth, which was more apparent in the Bray-Curtis and unweighted unifrac PCoA plots (Fig. S4b through e and S5b through e), suggesting the presence of environmental gradients along the depth of the core that influence community composition. NMDS analysis ( Fig. 6 ) also did not show clustering of samples by location, though there was some separation between ML and NR samples along the secondary axis. Out of all geochemical variables mapped to the samples (Data File S1), the ones significantly correlating with community composition included pH, Eh, alkalinity, and salinity (inferred by electrical conductivity (EC), and dissolved Na, Cl, and K). It was expected that major community shifts occur across these environmental gradients, as microorganisms segregate into niches based on tolerance to selective forces (e.g., stress from low pH or high salinity) ( 79 – 81 ) and metabolic potential (influenced by redox gradients and oxygen availability) ( 82 – 84 ). Pore-gas oxygen measurements were taken at ML25, ML34, and NR18 tailings, but were not included in the statistical analyses due to limited data points (Fig. S3). However, the iron-dominated redox potential (Eh) of tailings environments can be used as a proxy to infer the metabolic capacity of samples (Data File S1); with aerobic respiration only supported at highly positive Eh values (above 300–500 mV) ( 85 ), although this threshold increases with decreasing pH. Fig 6 NMDS biplot of Bray-Curtis distances and environmental variables. Samples are colored by location. Environmental variables that significantly correlated with community dissimilarity ( P < 0.05, Bonferroni corrected) were included. For a plot showing all variables, see Fig. S6. Diversity also correlated with the concentrations of Fe, Mn, B, Ba, Ca, Tl, and V. Ferrous and ferric iron availability is known to correlate with the abundance of iron oxidizing and reducing bacteria, such as members of the Actinobacteriota and Gammaproteobacteria ( 21 ). Many iron oxidizing/reducing organisms can also oxidize/reduce manganese ( 86 , 87 ). As for the other elements, correlation between microbial diversity and concentration is more likely due to the co-variance between ion concentration and another geochemical factor, such as pH, rather than direct interactions between the elements and microorganisms (Fig. S1). For example, calcium (i.e., Ca in Fig. 6 ) concentration was strongly correlated with alkalinity (measured in mg/L CaCO 3 ) (R = 0.79). However, it is possible that selection for resistance to Tl and V toxicity may influence community composition ( 81 , 88 ), as these elements are highly toxic at trace concentrations. In particular, vanadium concentrations in several samples from ML25 and ML34 far exceed the recommended safe limit for drinking water (0.05 mg/L) ( 89 ) so selection for vanadium resistance mechanisms such as the reduction of V V to V IV would be relevant in these communities. Interestingly, β-diversity did not significantly correlate with Ni or Cu concentrations, despite these elements being the major porewater contaminants present in these tailings besides Fe (with upper concentration limits of 1,166 mg/L for Cu and 561.6 mg/L for Ni) (Data File S1). A possible explanation could be that Ni/Cu tolerance is widely distributed across many microbial taxa commonly found in mine tailings, and thus the overall variance in community composition is determined by other selective forces. Ni and Cu are both micronutrients required for the function of metalloenzymes and most prokaryotes encode homeostatic mechanisms to regulate intracellular concentrations ( 90 ). Nickel efflux pumps such as RcnA have been identified in organisms isolated from metal contaminated environments (e.g., Cupriavidus metallidurans strain CH34), as well as in non-Ni-resistant E. coli and H. pylori , with homologs found in other proteobacteria, cyanobacteria, and archaea ( 91 ). Similarly, extremophiles used in industrial biomining such as A. ferrooxidans and Sulfolobus metallicus share the same ATPases and transporters for copper export (CopA and CusA) with other bacteria, archaea, and eukaryotes ( 92 – 94 ). The presence of these pathways in Nickel Rim/Strathcona Mill tailings could be confirmed with multi-omic sequencing. Temporal variation at NR18 Changes in genus-level abundance of ASVs sequenced from NR18 tailings in 2019 compared to 2021 samplings show that the abundance of most genera remained relatively stable over the 2-year period ( Fig. 7 ). Both Acidithiobacillu s and Pseudomonas ASVs show an overall trend of decreased abundance at various depths between 0.8 and 2.6 m. We note that prominent increases and decreases in abundance in adjacent core sections is more likely due to imperfect depth mapping between the 2 years’ cores rather than large changes in community composition at NR18 as a whole. The WPS-2 relative abundance increased at multiple depth ranges between 1 and 2 m. Many of the NR18 vadose zone pore-water samplers were dry in 2021, and core moisture content is likely to influence WPS-2 abundance, as they are known to be common in dry, bare soil environments ( 46 ). Drier tailings could also explain the decrease in A. ferrooxidans , which are not desiccation tolerant ( 95 ). The taxonomic representation of sulfate-reducing bacteria also changes slightly ( Ca . Desulforudis appears to be superseded by Desulfosporosinus between 3 and 3.5 m), which may be a response to changes in an environmental factor such as pH or alkalinity (moderately positively correlated with Desulfosporosinus abundance but not Ca . Desulforudis) ( Fig. 3 ). Water samples were not collected in 2019, so we were unable to further investigate this hypothesis. Fig 7 Change in relative abundance of microbial genera across varying depths at the NR18 tailings dump, 2019–2021. The genera listed in Fig. 1 and 3 were compared to NR18 samples taken in 2019, those present in both data sets were included in this figure. The percent relative abundance of the 2019 samples was subtracted from the NR18 2021 samples. Samples were mapped to each other based on the proximity of the midpoint depths of each core section. Conclusions Mine tailings present a challenge in waste management practices. AMD environments harbor taxonomically and metabolically diverse microorganisms, which have evolved a wide variety of adaptations to thrive in highly heterogeneous contaminated environments. We utilized a 16S rRNA amplicon sequencing approach to characterize microbial communities at narrow (~10 cm) depth intervals at four locations at two tailings sites, providing a high-resolution profile of community changes across tailings gradients of varying geochemical compositions. Our research shows that microbial communities are highly diverse between and within each sampling location. Our findings suggest that future sampling efforts that capture a wide range of tailings locations and depths could contribute to the isolation of a broad range of bioremediation-relevant species. We identified that overall community composition is correlated with pH, Eh, alkalinity, salinity, and some metal ions, which surprisingly did not include the major constituents Cu and Ni. Other variables that did not correlate with overall diversity were DOC, F, SO 4 , NO 3 , Al, As, Co, Li, Mg, Pb, S, Si, Sr, and Zn. The abundance of most individual lineages was also closely associated with different geochemical factors. However, due to the covarying nature of many environmental variables, it is important to frame predictions about which factors microorganisms are directly responding to, as well as the trajectory of contaminant cycling within tailings, around what is currently known about the physiology and ecology of each individual organism (or its closest known relatives). Our results showed that amplicon sequencing data can be used to identify specific microenvironments of tailings harboring novel or currently uncultured species, although this approach is limited in its ability to predict the physiological and metabolic properties of these uncharacterized organisms. Many ASVs present in Strathcona Mill and Nickel Rim tailings could not be classified at the species level, including potential iron/sulfur oxidizers and reducers, though this result is likely influenced by the fact that most samples were circumneutral and low in metal concentration. Functional profiling of tailings known to contain these novel lineages using metagenomics and metaproteomics will be an important future direction to identify genes associated with metal (Fe, Cu, Ni) resistance and transformation. By informing on geochemical variables required to provide niche-specific conditions for growth in the lab our work also provides an insight into improved approaches for establishing enrichment cultures that target these species."
} | 11,526 |
27287321 | PMC4968545 | pmc | 3,943 | {
"abstract": "ABSTRACT Subseafloor sediment hosts a large, taxonomically rich, and metabolically diverse microbial ecosystem. However, the factors that control microbial diversity in subseafloor sediment have rarely been explored. Here, we show that bacterial richness varies with organic degradation rate and sediment age. At three open-ocean sites (in the Bering Sea and equatorial Pacific) and one continental margin site (Indian Ocean), richness decreases exponentially with increasing sediment depth. The rate of decrease in richness with increasing depth varies from site to site. The vertical succession of predominant terminal electron acceptors correlates with abundance-weighted community composition but does not drive the vertical decrease in richness. Vertical patterns of richness at the open-ocean sites closely match organic degradation rates; both properties are highest near the seafloor and decline together as sediment depth increases. This relationship suggests that (i) total catabolic activity and/or electron donor diversity exerts a primary influence on bacterial richness in marine sediment and (ii) many bacterial taxa that are poorly adapted for subseafloor sedimentary conditions are degraded in the geologically young sediment, where respiration rates are high. Richness consistently takes a few hundred thousand years to decline from near-seafloor values to much lower values in deep anoxic subseafloor sediment, regardless of sedimentation rate, predominant terminal electron acceptor, or oceanographic context. IMPORTANCE Subseafloor sediment provides a wonderful opportunity to investigate the drivers of microbial diversity in communities that may have been isolated for millions of years. Our paper shows the impact of in situ conditions on bacterial community structure in subseafloor sediment. Specifically, it shows that bacterial richness in subseafloor sediment declines exponentially with sediment age, and in parallel with organic-fueled oxidation rate. This result suggests that subseafloor diversity ultimately depends on electron donor diversity and/or total community respiration. This work studied how and why biological richness changes over time in the extraordinary ecosystem of subseafloor sediment.",
"introduction": "INTRODUCTION Subseafloor sediment contains a diverse microbial ecosystem ( 1 – 3 ), with a total cell abundance comparable to that in terrestrial soil and in the world ocean ( 4 ). Subseafloor sedimentary communities push the boundaries of life as we know it; per-cell rates of respiration are often orders of magnitude lower than those in the surface world ( 5 , 6 ), biomass turnover can take hundreds to thousands of years ( 7 , 8 ), cell abundance can be as low as 10 cells per cm 3 ( 9 ), and microbes in deep subseafloor sediment may be isolated from the surface world for millions of years (Ma) to tens of Ma. Subseafloor sediment, therefore, provides an unprecedented opportunity to investigate drivers of microbial diversity on a time scale of thousands to millions of years. In the broadest context, distributions of microbial diversity result from combined effects of speciation, selection, dispersal, and ecological drift ( 10 , 11 ). However, subseafloor conditions may severely impact the relative influence of these processes. For example, exceedingly low per-cell energy fluxes may place very high selection pressure on subseafloor populations, severely limit active dispersal ( 6 ) and cell abundance, and cause mean generation times to greatly exceed the already-long few-hundred-year to few-thousand-year time scale of biomass turnover ( 7 ) in subseafloor sediment ( 12 ). Generation times of hundreds to millions of years may in turn greatly lower the rates of speciation. To document microbial diversity and its potential drivers in subseafloor sediment, we extracted and sequenced PCR amplicons for the V4 to V6 hypervariable region of the bacterial 16S rRNA gene from the sediment of four distinct locations: the Bering Sea (Integrated Ocean Drilling Program [IODP] expedition 323 site U1343) ( 13 ), the eastern equatorial Pacific ( Knorr expedition 195-3 site EQP1), the central equatorial Pacific Ocean ( Knorr 195-3 site EQP8), and the Bay of Bengal continental margin (Indian National Gas Hydrate Program [NGHP] site NGHP-1-14) ( 14 ) ( Fig. 1 ). FIG 1 Sampling locations. (Map created with Generic Mapping Tools.)",
"discussion": "DISCUSSION The relationships between abundance-weighted community composition and redox zones ( Fig. 2 ) indicate that some taxa are influenced by the predominant terminal electron-accepting activity. In contrast, the lack of clear correspondence between bacterial richness and redox zonation suggests that the predominant terminal electron-accepting pathway does not exert primary control on OTU richness of subseafloor bacterial communities. Possible explanations of this lack of relationship between OTU richness and predominant terminal electron acceptors include the following: (i) most OTUs may represent taxa that are not involved in terminal electron acceptance and that operate similarly in successive redox zones (for example, fermentative taxa may be active in all of the anoxic zones), (ii) taxa directly involved in terminal electron acceptance may be capable of processing multiple kinds of electron acceptors, and (iii) terminal electron acceptance may not be limited to the predominant pathway, with terminal electron-accepting activity predominant in one redox zone also existing in other zones (for example, iron reducers may be present and active where sulfate reduction, methanogenesis, or sulfide oxidation predominate [ 51 , 52 , 53 ]). The exponential decline in bacterial richness from seafloor to greater sediment depth is consistent with recent comparisons of bacterial OTUs in the ocean to OTUs in marine sediment ( 43 ). Based on the relative abundance of 16S V6 tags in the water column, shallow sediment (0 to 0.1 mbsf), and subseafloor sediment, these studies show that (i) marine sedimentary bacteria are dispersed via the ocean, and (ii) subseafloor sedimentary lineages are selected from the community present in shallow sediment ( 43 ). The close match between the exponential decline in bacterial richness and the depth distribution of organic degradation rates at our open-ocean sites indicates that vertical variation in richness is closely tied to organic-fueled community activity. The pattern of organic degradation exponentially declining from seafloor to greater sediment depth was first observed decades ago. It is often explained with a “multi-G model,” in which organic matter is assumed to be composed of diverse organic compounds with different levels of reactivity ( 44 ). In such models, the most labile or biologically reactive organic substrates are respired at much higher rates than the least-labile substrates, leading the net rates of organic-fueled respiration to decrease exponentially with increasing sediment depth ( 44 , 45 ). A single EQP1 sample at 7.21 mbsf constitutes the only exception to the close correspondence between exponentially declining richness and exponentially declining organic degradation at these sites. This relatively OTU-rich sample contains an unusually high concentration of organic matter relative to adjacent sediment; its exceptionality is consistent with previous research that showed that discrete horizons of organic-rich sediment may sustain locally high respiration for millions of years ( 46 ). Near continental shelves (such as the Indian Margin) and in upwelling regions (such as the Bering Sea and the equatorial Pacific), organic matter is the primary electron donor for subseafloor sedimentary communities ( 2 ). Consequently, the close correspondence between vertical patterns of richness and vertical patterns of organic degradation suggests that selection for organismal properties related to either total catabolic activity or electron donor diversity exerts the primary influence on bacterial OTU richness in anoxic subseafloor sediment. This correspondence also indicates that many bacterial taxa that are poorly adapted for subseafloor sedimentary conditions are degraded in the geologically young sediment where respiration rates are high. This result sets a clear boundary for understanding bacterial OTU richness in anoxic subseafloor sediment. It also provides a potential basis for ultimately integrating OTU richness with other key properties that appear to be broadly related to total catabolic activity in subseafloor sedimentary communities, such as cell ( 4 ) or viral particle abundance ( 47 ) and activity ( 48 ). However, the exact traits that preferentially aid survival as catabolic activity and/or electron donor diversity decline remain to be determined; candidate traits include specialization to metabolize recalcitrant organic substrates, specific energy-conserving properties, such as membrane permeability ( 6 ), use of sodium ions for energy storage ( 6 ), spore formation ( 6 ), prophage modulation of metabolic activity ( 49 ), and/or a wide range of other properties ( 12 )."
} | 2,274 |
35869258 | PMC9307576 | pmc | 3,945 | {
"abstract": "Cyanobacteria of the genus Synechococcus play a key role as primary producers and drivers of the global carbon cycle in temperate and tropical oceans. Synechococcus use phycobilisomes as photosynthetic light-harvesting antennas. These contain phycoerythrin, a pigment-protein complex specialized for absorption of blue light, which penetrates deep into open ocean water. As light declines with depth, Synechococcus photo-acclimate by increasing both the density of photosynthetic membranes and the size of the phycobilisomes. This is achieved with the addition of phycoerythrin units, as demonstrated in laboratory studies. In this study, we probed Synechococcus populations in an oligotrophic water column habitat at increasing depths. We observed morphological changes and indications for an increase in phycobilin content with increasing depth, in summer stratified Synechococcus populations. Such an increase in antenna size is expected to come at the expense of decreased energy transfer efficiency through the antenna, since energy has a longer distance to travel. However, using fluorescence lifetime depth profile measurement approach, which is applied here for the first time, we found that light-harvesting quantum efficiency increased with depth in stratified water column. Calculated phycobilisome fluorescence quantum yields were 3.5% at 70 m and 0.7% at 130 m. Under these conditions, where heat dissipation is expected to be constant, lower fluorescence yields correspond to higher photochemical yields. During winter-mixing conditions, Synechococcus present an intermediate state of light harvesting, suggesting an acclimation of cells to the average light regime through the mixing depth (quantum yield of ~2%). Given this photo-acclimation strategy, the primary productivity attributed to marine Synechococcus should be reconsidered.",
"introduction": "Introduction Marine photosynthesis by single-celled microorganisms accounts for nearly 50% of global primary productivity 1 . Numerically, the vast majority of primary producers in the oceans are cyanobacteria, the only extant prokaryotic group of oxygenic photoautotrophs. Among these, the two cyanobacterial genera— Prochlorococcus and Synechococcus —are responsible for a significant fraction of primary production, mainly in open ocean waters in subtropical and tropical settings 2 – 4 . The basic photosynthetic apparatus in all cyanobacteria consists of two photochemical reaction centers: Photosystem I and Photosystem II. Most cyanobacteria, including Synechococcus that are the focus of our study, possess a supramolecular light-harvesting antenna coupled mainly to PSII, the Phycobilisome (PBS). Prochlorococcus however, use membrane internal light-harvesting systems 5 . In Synechococcus , the PBS contains proteins that bind phycoerythrin chromophores (PE) absorbing blue light (peak at 497 nm), the wavelength that best penetrates seawater 6 , 7 . Owing to this adaptation, this genus specializes in light harvesting in the deeper ocean 5 . Light regimes through the water column can change dramatically in space and time. Its intensity attenuates exponentially with depth, and its spectrum is narrowed to blue wavelengths. Moreover, the conditions in an open ocean water column vary seasonally 8 . Generally, during summer periods, as the surface warms up and temperature declines monotonically with depth, the water is stratified, and vertical movements of plankton are restrained. Under these conditions, cells inhabiting different water layers acclimate to the available light regime. However, during winter, cooling of the surface drives vertical mixing of the water column. This in turn requires phytoplankton to entrain to a light regime which exposes them to changes on an hour-to-day time scales 9 . Photosynthetic cells deploy acclimation mechanisms to cope with light regime changes, which impacts photosynthetic performance and thus productivity 10 – 12 . Among the phytoplankton, cyanobacterial Synechococcus species are known to exhibit extensive photo-acclimation capacities 13 – 15 . Known acclimation strategies to low light conditions include increasing both the number and the size of photosynthetic units 16 , a term defining the number of antennae chromophores coupled to a photosystem reaction center 17 . Synechococcus cells under low light will contain a higher number of thylakoid membranes per cell, higher chlorophyll content, and larger phycobilisomes with additional PE units 18 . The plasticity of the Synechococcus is enabled by the position of the PBS antenna in the inter-thylakoid space. However, at the same time, the intermediate chromophore coupling regime determines energy transfer efficiencies that are considered lower than those of thylakoid membrane internal antenna complexes 19 . Recently, we showed that, in response to low light, the Synechococcus WH8102 strain can improve its phycobilisomes’ light-harvesting efficiency 20 . From a physical point of view, this discovery is surprising, since, with the larger antenna, the absorption cross-section increases but requires the excitation energy to travel longer distances. In land plants, the longer energy migration path was shown to decrease energy transfer rates 21 , as expected according to Forster Resonance Energy transfer calculations 22 . However, in Synechococcus WH8102 the energy transfer rate through the antenna to the reaction centers increased when grown under lower light. We demonstrated that this is achieved by enhanced coupling between chromophores in the phycobilisome 20 . When light is absorbed in a photosynthetic light-harvesting complex (PBS in the case of Synechococcus ), the energy has to migrate through the antenna and reach a reaction center, where photochemical energy conversion takes place. There are three competing pathways that light energy can follow: (i) dissipation through heat; (ii) emission as fluorescence (iii) photochemical reactions 23 – 26 . In the upper water layers, light intensities are high and excess light can be extremely harmful to the cell, due to the generation of reactive oxygen species (ROS) 27 . Photosynthetic organisms use a variety of mechanisms to dissipate excess energy, collectively called non-photochemical quenching (NPQ) mechanisms 28 , 29 . NPQ levels may vary significantly and therefore influence the heat dissipation rate in surface waters. However, when examining photosynthetic organisms in deeper layers under lower irradiance, heat dissipation is expected to be minimal and constant 30 . In this scenario, changes in the quantum yield of photochemistry ( \\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}$${\\Phi }_{{\\it{p}}}$$\\end{document} Φ p ) are inversely related to the quantum yield of fluorescence ( \\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}$${\\Phi }_{{\\it{f}}}$$\\end{document} Φ f ). Comparing the quantum yields of the different processes can be achieved by using fluorescence lifetime measurements. This is a standard method for estimating light-harvesting efficiency in laboratory studies 31 . Using time-correlated single-photon counting (TCSPC) technique to measure fluorescence lifetime in the picosecond time domain, we can quantitatively relate the fluorescence lifetime to the absolute quantum yield of fluorescence 13 , 14 , 30 (Eq. 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}$${\\Phi }_{{\\it{f}}}=\\frac{\\tau }{{\\tau }_{{{{{{\\rm{n}}}}}}}}$$\\end{document} Φ f = τ τ n , where \\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}$${\\tau }_{{{{{{\\rm{n}}}}}}}$$\\end{document} τ n is the intrinsic or natural lifetime of a phycobilisome complex. Here we examine how phycobilisome light-harvesting efficiencies correlate with depth in native Synechococcus populations. We sampled seawater along a depth gradient during different seasons in the Gulf of Aqaba (GoA), using high-resolution fluorescence lifetime and flow cytometry measurements to specifically capture the energy transfer properties of Synechococcus PBS. Our study site, the GoA, is located in the northeastern-most section of the Red Sea. During summer (April–September), the oceanographic conditions are markedly stratified and oligotrophic, resembling an open ocean gyre ecosystem 32 . However, during winter (October–March), surface cooling progressively drives convective mixing of the water column, reaching hundreds of meters in depth, depending on how cold the winter is 9 , 33 , 34 . In turn, deep mixing leads to the homogenization of plankton across the mixing depth and the entrainment of ample nutrients to the upper photic layer. This results in the formation of major spring blooms, that are uncommon in (sub)tropical oligotrophic ecosystems 35 , 36 . Synechococcus are numerically a major component of microbial plankton in the Gulf, both during the spring bloom and stratification periods, where higher densities can be found along the deep chlorophyll maximum (DCM) around 80–100 m depth 37 – 39 . Therefore, this is an attractive study site to examine photo-acclimation dynamics of Synechococcus populations in situ in the natural environment.",
"discussion": "Discussion Our results show that the fluorescence lifetime of native Synechococcus phycobilisomes varies with depth, and follows a clear trend, which correlates to the conditions in the water column. In a stratified water column, when Synechococcus remain at a certain depth for sufficient time to acclimate, the physiological state of the cells at each depth is determined by the ambient light radiation. As depth increases, their size and cellular complexity increase in response to light limiting conditions; the number of photosynthetic units and pigment content increases; yet their fluorescence lifetime becomes shorter. Hence, during stratification, the quantum efficiency of light-harvesting increases as light availability decreases. During mixing, cells are continuously exposed to varying light regimes, requiring them to optimize their photosynthetic machinery to the average available light intensity perceived. Therefore, sampling during mixing served as a natural control experiment for our stratified water column results. Indeed, lifetime was found to be uniform across the mixing depth. Beyond contrasting the results obtained in stratified water, these results show how Synechococcus phycobilisome systems cope with the challenges imposed by mixing. Since light intensities change on a short (hours to days) time scale, they cannot optimize to a specific light regime and adopt a likely average state that can serve light harvesting across the mixed layer. This is an “intermediate state,” which resembles the state optimized for a depth of ~70 m during the stratified season. A comparison of photo-acclimation mechanisms between natural populations of Synechococcus from the stratified water column, in this study, and a previous study done on light acclimated marine Synechococcus strains grown in culture, show similarities. In both cases, PBS fluorescence lifetime was shorter when light intensity was low. Under comparable light intensities, natural populations showed longer fluorescence lifetimes (Fig. 3a, b ). For example, based on CTD data, in the August dataset (during stratification), under light intensities of 137 µmol photons m −2 s −1 and 6 µmol photons m −2 s −1 , fluorescence lifetimes values were 0.25 and 0.17 ns, accordingly. Under similar illumination conditions in the lab, PBS fluorescence lifetimes were 0.15 (at 150 µmol photons m −2 s −1 ) and 0.1 (at 10 µmol photons m −2 s −1 ) 20 . In both cases, Synechococcus acclimated to lower light intensities exhibited shorter lifetimes. The difference in values may be attributed to the diverse population of Synechococcus strains in the GoA 39 , compared to the axenic WH8102 strain grown in culture. In addition, changes in the spectral composition of single-cell fluorescence measured by flow cytometry reported here are comparable to those reported by Six and coworkers 44 . The photo-acclimation response of the PBS of natural Synechococcus populations from the mixed water column are therefore in contrast to both natural stratified Synechococcus and laboratory light acclimated Synechococcus strains. Going beyond marine Synechococcus , studies of freshwater Synechocystis further demonstrate the plasticity of PBS fluorescence properties and indicate lifetimes in the range observed here 45 – 47 . Additional photo-acclimation responses of Synechococcus cells to the increasing depth, are the increase in cell size and in chlorophyll and phycobilin content. Note that the emission intensity of a pigment or chromophore cannot be used to directly quantify its content and to determine the photosynthetic unit size. Yet, previous studies showed that the phycobilisome size increases with increasing depth and decreasing light intensity 44 , 48 . Forward and side scatter of Synechococcus showed a depth profile similar to fluorescence lifetime, with a distinct difference between cells residing in the shallow layers and the deeper layers during stratification, and uniform properties across the water column during mixing (Fig. 2 and Supplementary Fig. 5 ). Coordinated dynamics of cell size, photosynthetic unit content, and PBS fluorescence lifetime were previously shown in a laboratory study 20 . Their co-occurrence in our stratified field samples supports the interpretation of the fluorescence lifetime measurements. During mixing fluorescence lifetime had an “intermediate state”, and concurrently forward scatter and side scatter values were low—similar to cells in the shallower layer during stratification. The ability to manipulate energy transfer efficiency in the phycobilisome is not trivial. Previously, it was known that in response to low light, organisms increase their light-harvesting antennae size 21 , 49 , 50 , therefore increasing the absorption cross-section (absorbing light from a larger surface). However, in a larger antenna, excitation energy must travel a longer distance to reach the reaction centers in the photosystems 21 . In PBS, where pigment-pigment distances are larger than in plants, the effect of a bigger cross-section is expected to be larger. Indeed, Semi-classical dipole-dipole interaction models of phycobilisome, using FRET (Forster resonance energy transfer), which assume one-dimensional phycobilisome rods, predict such an outcome. A longer antenna rod will lower the energy transfer rate 22 . Thus, the fact that the energy transfer rate increases in the larger phycobilisome systems of low light acclimated cells is surprising in view of these classical models. However, these results fit with our previous laboratory study using Synechococcus WH8102 cultures 20 , where the enhanced coupling was induced under low blue light conditions. The improved energy transfer rate was shown to be the result of enhanced coupling between the chromophores of PE: Phycourobilin and Phycoerythrobilin. Based on these results, it was suggested that the mechanism is either not purely classical, or that it involves an overlooked inter-rod transfer pathway, possibly due to the higher density of phycobilisome rods under low light conditions. The shift from a predominantly single rod one-dimensional energy transfer to coupled rods that allow multidimensional energy transfer may lead to increased efficiency 51 . In principle, a complete picture of the fate of the absorbed energy can be generated from Φ f and Φ p values (Eq. 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}$${\\Phi }_{T}={1-\\Phi }_{{\\it{f}}}-{\\Phi }_{{\\it{p}}}$$\\end{document} Φ T = 1 − Φ f − Φ p 23 . Φ f was calculated from lifetime measurements. Φ p is often estimated from variable fluorescence measurements as (Eq. 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}$$\\frac{{F}_{v}}{{F}_{m}}={\\Phi }_{{\\it{p}}}$$\\end{document} F v F m = Φ p . However, in cyanobacteria dark F m values are low 52 and F 0 values are high due to a contribution of the tail of PBS fluorescence in the chlorophyll measurement channel 53 . Nevertheless, keeping these limitations in mind we can provide an example of how such a calculation can provide insight. We can use \\documentclass[12pt]{minimal}\n\t\t\t\t\\usepackage{amsmath}\n\t\t\t\t\\usepackage{wasysym} \n\t\t\t\t\\usepackage{amsfonts} \n\t\t\t\t\\usepackage{amssymb} \n\t\t\t\t\\usepackage{amsbsy}\n\t\t\t\t\\usepackage{mathrsfs}\n\t\t\t\t\\usepackage{upgreek}\n\t\t\t\t\\setlength{\\oddsidemargin}{-69pt}\n\t\t\t\t\\begin{document}$$\\frac{{F}_{v}}{{F}_{m}}$$\\end{document} F v F m values measured for Synechococcus WH8102 20 , using DCMU to get a more accurate reading of F m . Cultures acclimated to medium or low light conditions, which correlate in the GoA, in summer, to ~60 m and ~120 m. τ avg values measured from these depths are 0.17 and 0.38, respectively. Based on these values, the calculated quantum yield of thermal dissipation will be ~80% for 60 m and ~60% for ~120 m. These values are in the range reported by Falkowski and coworkers 23 based on chlorophyll lifetime measurements. Over the past decades, variable chlorophyll fluorescence has been the most sensitive, nondestructive signal detectable in the upper ocean that reflects instantaneous phytoplankton photophysiology 54 , 55 . It has been used to estimate the biomass and physiological status of phytoplankton and has fundamentally changed the interpretation of the biological responses to ocean physics 24 . However, to obtain a complete picture of the energy budget in photosynthetic processes, two of the three competing pathways of absorbed energy (photochemistry, fluorescence, and heat) must be measured. Picosecond fluorescence lifetime measurements can complement variable fluorescence techniques and provide a complete understanding of the fate of absorbed energy. It is also crucial for the development of algorithms for remote sensing techniques (i.e., chlorophyll fluorescence measured by satellites) which are often used to estimate spatial patterns of marine primary production 56 – 58 . Such algorithms depend on the comparison with accurate in situ measurements of quantum yields 24 . Here, we demonstrate the variability in fluorescence quantum yields as a function of depth, highlighting the importance of a depth-profile fluorescence lifetime approach. To reliably estimate the cyanobacterial integrated contribution to photosynthetic activity in the ocean, their dynamics along the water column should be considered. Since our results indicate an increase in efficiency as a function of depth, which was not considered previously, it may suggest an underestimation of Synechococcus productivity by current models 59 , 60 . With the increase of ocean stratification over the past decades and a similar trend which is expected for the 21st century 61 , 62 , incorporating our results into future models may be beneficial for obtaining more precise estimations, accounting for quantum yield changes in response to the water column ambient illumination conditions."
} | 5,091 |
36331626 | PMC10345032 | pmc | 3,946 | {
"abstract": "The worldwide fossil fuel reserves are rapidly and continually being depleted as a result of the rapid increase in global population and rising energy sector needs. Fossil fuels should not be used carelessly since they produce greenhouse gases, air pollution, and global warming, which leads to ecological imbalance and health risks. The study aims to discuss the alternative renewable energy source that is necessary to meet the needs of the global energy industry in the future. Both microalgae and macroalgae have great potential for several industrial applications. Algae-based biofuels can surmount the inadequacies presented by conventional fuels, thereby reducing the ‘food versus fuel’ debate. Cultivation of algae can be performed in all three systems; closed, open, and hybrid frameworks from which algal biomass is harvested, treated and converted into the desired biofuels. Among these, closed photobioreactors are considered the most efficient system for the cultivation of algae. Different types of closed systems can be employed for the cultivation of algae such as stirred tank photobioreactor, flat panel photobioreactor, vertical column photobioreactor, bubble column photobioreactor, and horizontal tubular photobioreactor. The type of cultivation system along with various factors, such as light, temperature, nutrients, carbon dioxide, and pH affect the yield of algal biomass and hence the biofuel production. Algae-based biofuels present numerous benefits in terms of economic growth. Developing a biofuel industry based on algal cultivation can provide us with a lot of socio-economic advantages contributing to a publicly maintainable result. This article outlines the third-generation biofuels, how they are cultivated in different systems, different influencing factors, and the technologies for the conversion of biomass. The benefits provided by these new generation biofuels are also discussed. The development of algae-based biofuel would not only change environmental pollution control but also benefit producers' economic and social advancement. Graphical abstract",
"conclusion": "Conclusions The creation of third-generation biofuels, a superior type of biofuel, is the most promising application of the biomass obtained from algae species. Algae are adaptable plants that may flourish in a variety of aquatic environments, including water that contains a lot of salt or waste. Algae-producing facilities can be found in areas that are not suited for the growth of forests or agroecosystems. As a result, the production of algae does not compete with that of food, fiber, or fuel. Algae have been extensively exploited in industrial applications with the most intensive usage in the production of biofuels such as biobutanol, biodiesel, biohydrogen, or bioethanol. It has been reviewed that for the production of algal fuels, algae can be cultivated in all sorts of systems that can be closed, open, or hybrid. The production of biofuels depends upon several factors which influence the cultivation of algae. The practicality of algae biofuels will probably be most impacted in the long run by genetic engineering. Algal oil's prospects will be improved by improvements in methods for separating algal biomass from water and extracting oil from biomass. In the next 7–10 years, algae-based fuel production might be cost-effective, widely adaptable, and operational, but only if we continue improving our awareness of these magnificent species while also improving our capacity to tailor them for the specific aim of growing new energy industry. In the coming years, algae biomass could play a key role in resolving the conflict between food production and biofuel production.",
"introduction": "Introduction Commonly it is assumed that algae are photosynthetic autotrophs that mostly live in water, evolve oxygen, and are either made up of single cells or live in colonies or filamentous forms [ 1 ]. Algae are comprised of a large number of photosynthetic living beings that mostly inhabit aquatic surroundings. According to the size and morphological characteristics, algal species are usually classified into macroalgae and microalgae. Macroalgae which are also called seaweeds are made up of a large number of cells and can be seen with a naked eye. As compared to macroalgae, microalgal species can only be visualized with the help of a microscope and are highly important in the field of micro nanomedicine [ 2 ]. Based on existing pigments, brown algae, blue–green algae, and red algae are the three classes of macroalgae [ 3 ]. Although blue–green algae and bacteria share some common structural characteristics, blue–green algae were placed in the algal class because of the presence of chlorophyll and correlated complexes [ 4 ]. One more class of algae comprises the red algae. Species that belong to this class of Rhodophyta are eukaryotes that contain chloroplasts and phycobilins [ 5 ]. Brown algae named brown seaweeds are typically large macroalgae and have the comparatively immense ability to convert photons as a result of which biomass can be synthesized much more quickly. Brown algae are given more attention for the development of maintainable biofuels because their efficiency is considerably higher in contrast to that of cyanobacteria or red algae [ 6 ]. Microalgae have also arisen as a probable feedstock for the production of biofuels because a large number of microalgal strains have the ability of lipid accumulation, with a higher growth rate of biomass and greater photosynthetic production as compared to their counterparts that exist on land [ 7 ]. Difficulty to sustain and persistent debilitating of non-sustainable petroleum derivatives gave rise to the significance of inexhaustible fuel sources [ 8 ] a worldwide temperature alteration further amounts to the difficulties previously confronted [ 9 ]. These days, to move in the promising direction developed and underdeveloped countries are thinking about environmentally friendly power sources [ 10 ]. Biofuel is referred to as any fuel that is obtained from biomass that is either a plant, algae, or animal manure [ 11 ]. Biofuels are accepted to be the most natural amicable energy source. Biomass got from trees, agro backwoods buildups, marine or land plants, grasses, and harvests is the adaptable and significant sustainable feedstock for the development of biofuels [ 12 ]. The utilization of biomass as fuel is one of a handful of genuine systems to decrease the effects that greenhouse gases are causing. Contrasted with petroleum products, biomass ignition fundamentally diminishes CO 2 and CO 2 outflows and essentially lessens the debris obtained after burning [ 13 ]. As the conventional fuel resources are being depleted at a high rate, there is more focus towards the employment of alternative sources. More than 50 years ago, the concept of employing algae as a source of food, feed, and energy was first proposed. During the energy crisis of the 1970s, when programs were started to manufacture gaseous fuels (hydrogen and methane), the production of methane gas from algae received a significant boost [ 14 ]. Our knowledge of cultivating algae for fuel has greatly benefited from the researchers' work on open pond algae growth [ 15 ]. The effects of various nutrient and CO 2 concentrations were documented, the engineering difficulties of mass-producing algae were addressed, and a strong basis for algae-fuel research was established through the isolation and testing of thousands of distinct species. But in 1995, US Department of Energy’s (DOE) Office decided to end the initiative due to budgetary restrictions and low oil prices. Everything has altered in recent years. Concerns about \"peak oil\", the rising effects of atmospheric CO 2 , the United States' increasing reliance on fuel imports, and the associated hazards to energy security have all contributed to a resurgence in interest in biofuels in general and algae-based biofuels in particular [ 16 ]. Advantages in biotechnology have opened up new possibilities that were not possible during the years of former research, such as the ability to genetically modify algae to produce more oils and convert solar energy more effectively. The United States has seen the majority of activity in algae research and commercial production. Algae biofuels are currently being explored globally in both established and developing countries in Europe, Asia, and other regions. The US-based Algal Biomass Organization serves as the industry's chief spokesperson and a resource for data on the businesses pioneering the technology [ 17 ]. This article outlines the third-generation biofuels, how they are cultivated in different systems, different influencing factors, and the technologies for the conversion of biomass. The benefits provided by these new generation biofuels are also discussed. Given these factors, it is essential to remove the current bottlenecks to use microalgae for commercial purposes."
} | 2,241 |
36106689 | PMC9828436 | pmc | 3,947 | {
"abstract": "Abstract The frequency and severity of marine heatwaves causing mass mortality events in tropical and temperate coral species increases every year, with serious consequences on the stability and resilience of coral populations. Although recovery and persistence of coral populations after stress events is closely related to adult fitness, as well as larval survival and settlement, much remains unknown about the effects of thermal stress on early life‐history stages of temperate coral species. In the present study, the reproductive phenology and the effect of increased water temperature (+4°C and +6°C above ambient, 20°C) on larval survival and settlement was evaluated for two of the most representative Mediterranean octocoral species ( Eunicella singularis and Corallium rubrum ). Our study shows that reproductive behavior is more variable than previously reported and breeding period occurs over a longer period in both species. Thermal stress did not affect the survival of symbiotic E. singularis larvae but drastically reduced the survival of the non‐symbiotic C. rubrum larvae. Results on larval biomass and caloric consumption suggest that higher mortality rates of C. rubrum exposed to increased temperature were not related to depletion of endogenous energy in larvae. The results also show that settlement rates of E. singularis did not change in response to elevated temperature after 20 days of exposure, but larvae may settle fast and close to their native population at 26°C (+6°C). Although previous experimental studies found that adult colonies of both octocoral species are mostly resistant to thermal stress, our results on early life‐history stages suggest that the persistence and inter‐connectivity of local populations may be severely compromised under continued trends in ocean warming.",
"introduction": "1 INTRODUCTION In the last few decades, global climate change, and ocean warming in particular, has been recognized as one of the most important threats to marine biodiversity (Halpern et al., 2015 ). Coral communities have been severely affected by a dramatic increase in the frequency and intensity of mass mortality events linked to increased global temperatures and marine heatwaves (e.g., Baird & Marshall, 2002 ; Garrabou et al., 2022 ; Hughes et al., 2017 ). This increase in the frequency and intensity of mass mortality events is ubiquitous and has been observed in tropical, subtropical, temperate and polar seas (e.g., Barnes & Souster, 2011 ; Garrabou et al., 2019 ; Kemp et al., 2011 ). Moreover, ocean warming has also induced an earlier arrival of spring conditions affecting numerous marine ecosystems and biological processes (Loeb et al., 1997 ; Parmesan, 2006 ; Walther et al., 2002 ). An earlier arrival of spring conditions can significantly affect coral reproductive phenology and, consequently, the success of future populations (IPCC, 2007 ; Shefy et al., 2018 ; Shlesinger & Loya, 2019 ). Since many corals play a structural role increasing the diversity of marine ecosystems (Dayton et al., 1974 ; Jones et al., 1994 ; Thrush & Dayton, 2002 ), changes in their reproductive processes could also have the potential to result in a drastic loss of biodiversity at both the community and ecosystem levels. Ocean warming is predicted to increase under the expected 1–5°C rise in mean global seawater temperature by 2100 (IPCC, 2022 ). However, the distribution of excess heat will not be uniform across all the oceans. The Mediterranean Sea is recognized as a “hotspot” for ocean warming, ranking among the fastest‐warming ocean regions in the world (Garrabou et al., 2022 ; Marbà et al., 2015 ). The sea surface temperature of the Mediterranean shows a nearly continuous warming trend at a rate of 0.41°C per decade, which is three to six times higher than the warming rate of oceans globally (Cramer et al., 2018 ; Garrabou et al., 2021 ; Pisano et al., 2020 ). To our knowledge, despite this fast warming, consequences on coral reproductive phenology have never been studied before in the Mediterranean Sea. In addition, increases in the frequency and intensity of extreme heatwaves in the Mediterranean Sea have also been detected through field observations and are expected to continue in future projections (Adloff et al., 2015 ; Darmaraki et al., 2019 ), inducing mass mortality events primarily on benthic invertebrate taxa (e.g., Cerrano et al., 2000 ; Garrabou et al., 2009 ; Pérez et al., 2000 ). About 50% of all recorded mass mortality events in the Mediterranean Sea have occurred in Cnidarians, principally octocorals (Garrabou et al., 2019 ), which are the most conspicuous ecosystem engineering species in the rocky bottoms of the Mediterranean Sea (Ballesteros, 2006 ). Field and experimental studies have evaluated the immediate and delayed impacts of temperature increases on adult octocoral colonies (e.g., Coma et al., 2006 ; Ezzat et al., 2013 ; Gómez‐Gras et al., 2019 ; Linares et al., 2005 ), including sublethal impacts on their reproduction effort (Arizmendi‐Mejía et al., 2015 ; Linares, Coma, & Zabala, 2008 ). To date, however, only one study has examined the effects of thermal stress on embryonic and larval stages of a Mediterranean octocoral, Paramuricea clavata (Kipson et al., 2012 ), overlooking the possible effects on settlement rates. Studies on temperature effects on coral larvae have increased markedly in the last few years. However, most of these studies have focused on hexacoral tropical species, whereas octocoral species have been largely neglected. Although octocoral species are present across large depth ranges and in all oceans worldwide, the effects of thermal stress on their early life‐history stages have been studied in only two tropical, one temperate, and one deep‐sea species (Conaco & Cabaitan, 2020 ; Da‐Anoy et al., 2020 ; Kipson et al., 2012 ; Liberman et al., 2021 ; Rakka et al., 2021 ). These studies suggest that octocoral larvae have tolerance to thermal stress. Larvae of the deep‐sea octocoral Dentomuricea aff. Meteor exposed to high temperature (+2°C) had similar survival as larvae at ambient temperature (Rakka et al., 2021 ). Larvae from two tropical and one temperate octocoral species showed tolerance to temperature conditions expected by 2100 (+3°C; Conaco & Cabaitan, 2020 ; Da‐Anoy et al., 2020 ; Kipson et al., 2012 ; Liberman et al., 2021 ). In these studies, larval survival was only affected by thermal stress under high larval densities (Conaco & Cabaitan, 2020 ) or after a prolonged exposure (27 days; Kipson et al., 2012 ). However, this limited knowledge makes difficult to project how octocoral larvae will perform in the future under thermal stress, and more research is needed to include more species with different life‐history traits (e.g., thermal tolerance in symbiotic and brooded octocoral larvae has never been examined). The aim of this study is to provide new knowledge on the reproductive phenology and larval thermotolerance in two of the most representative octocoral species in the Mediterranean Sea, by answering the following questions: (1) How long is their breeding period? (2) How variable is the amount of larvae released during the breeding period? (3) How are larval survival and settlement affected by thermal stress? (4) How does thermal stress affect the metabolic balance of larvae in terms of biomass and energy consumption? To answer these questions, we monitored the breeding period of octocorals Eunicella singularis and Corallium rubrum , and used an experimental approach to examine the larvae performance at 20°C (control), 24°C (a temperature observed during Mediterranean heatwaves in the last few years) and 26°C (a temperature expected to occur during Mediterranean heatwaves in the near future).",
"discussion": "4 DISCUSSION Octocoral species play a paramount role as ecosystem engineers in Mediterranean benthic communities, as well as around the world (Gili & Coma, 1998 ; Velásquez & Sánchez, 2015 ; Wild et al., 2011 ), and are considered one of the main three‐dimensional constituents of the “marine animal forests” (sensu Rossi, 2013 ). Internal brooder species, such as the Mediterranean E. singularis and C. rubrum , represent more than 40% of all octocoral species with known reproductive strategies (Kahng et al., 2011 ). Despite their importance, compared with broadcast coral species, reproductive characteristics such as the duration, number of events and larval release intensity of brooders remain largely unknown. 4.1 Breeding characteristics In this study, the breeding period in E. singularis took place in one single event lasting 4–5 weeks, longer than 2–3 weeks previously reported for the same species by Theodor ( 1976 ). Initially, it had been assumed that larvae were released in approximately equal amounts during the entire breeding period, showing no correlation with the lunar phase (Weinberg, 1979 ). However, our results showed a gradual increase of larvae released, reaching its maximum 12 days after the breeding period started. The maximum release of larvae was concentrated in the period between the last quarter and the new moon. However, additional observations are needed to further explore this possible relationship between larval release and lunar phase, as has been observed in other Mediterranean octocoral species such as P. clavata (Linares, Coma, Mariani, et al., 2008 ). Similar to E. singularis , larval release of C. rubrum occurred over a single event for approximately 4–5 weeks, without any difference in the number of larvae released between day and night. The breeding period duration was longer than the 1–2 weeks previously reported for this species (Lacaze‐Duthiers, 1864 ; Vighi, 1970 ). Weinberg ( 1979 ) suggests a possible correlation between new moon and maximum intensity of larval release in C. rubrum ; however, our results show the maximum release between the last quarter and the new moon. Therefore, our study shows that breeding of E. singularis and C. rubrum occurs over a longer period and reproductive behavior (i.e., release of larvae) is more variable than previously reported. Our results could also indicate a phenological shift in the duration of the breeding season over the last 50 years. This possible phenological shift could be caused by the on‐going ocean warming, since a longer spawning season has been positively correlated with longer exposure to warmer waters in several broadcasting coral species studied across latitudinal gradients (De Putron & Ryland, 2009 ; Mangubhai & Harrison, 2009 ; Oliver et al., 1988 ). Indeed, the sea surface temperature of the Mediterranean has warmed by 1.48°C on average for the entire basin over the last four decades, corresponding to an increase of 0.41°C per decade, which is three to six times higher than the warming rate of oceans globally (Cramer et al., 2018 ; Garrabou et al., 2021 ; Pisano et al., 2020 ). During the brooding period (April to July), satellite observations from the last few decades showed a significant increase in seawater temperature with the peak increase occurring in June (0.08°C year −1 ; Nykjaer, 2009 ). Since reproductive events have evolved to occur at optimal times to maximize the survival of the next generation (Stearns, 1992 ), rapid shifts in reproductive phenology could threaten the long‐term viability of populations (Charmantier et al., 2008 ; Edwards & Richardson, 2004 ). Shefy et al. ( 2018 ) showed an increase in the duration of the larval release period of Stylophora pistillata from 2–3 to 5–6 months in the past four decades, which could be caused by anthropogenic and environmental impacts. In broadcast corals of the Red Sea, shifts in the timing of gamete release have been found due to environmental changes with potential consequences for coral reproductive success (Shlesinger & Loya, 2019 ). Our results show that the breeding periods of the studied corals span over longer time periods than previously documented which could be the result of different environmental cues in their current habitat. Future studies should focus on looking at the plasticity in reproductive breeding behavior of the adult colonies under ocean warming conditions, and performing longer larval experiments to obtain a better understanding of the resilience of parental colonies and offspring to thermal stress. 4.2 Effects of thermal stress on larval survival and energy reserves The two Mediterranean octocoral species studied here showed a contrasted response in larval survival under thermal stress treatments that simulated marine heatwave events caused by on‐going global climate change. Whereas +4°C and +6°C increases in temperature did not cause significant negative effects in the symbiotic E. singularis larvae, the survival of non‐symbiotic C. rubrum larvae was drastically reduced (Figure 3 ). Until now, the temperature effects on larvae from octocorals at shallow depths (5–30 m) have only been studied in the surface brooders Heliopora coerulea , P. clavata and Rhytisma fulvum (Conaco & Cabaitan, 2020 ; Da‐Anoy et al., 2020 ; Kipson et al., 2012 ; Liberman et al., 2021 ). In these species, larvae showed some tolerance to elevated temperature that was explained by the absence of symbionts in larvae tissues. Conversely, in the present study, the symbiotic larvae of E. singularis showed higher resistance to temperature increases than C. rubrum non‐symbiotic larvae. The thermotolerance observed in E. singularis larvae contrasts with results from several hexacoral species showing that thermal stress weakens endosymbiont interactions in coral larvae and reduces larval survivorship (Edmunds et al., 2001 ; Graham et al., 2017 ; Randall & Szmant, 2009 ; Schnitzler et al., 2012 ; Serrano et al., 2018 ). The thermotolerance of E. singularis symbiotic larvae is in line with the performance of adult colonies of the same species which do not show any evidence of coral bleaching when exposed to thermal stress (26°C; Ferrier‐Pagès et al., 2009 ). The higher resistance of E. singularis larvae observed in this study may be partly due to their large larval size (Figure 5 ; e.g., Baria et al., 2015 ; Chamberland et al., 2017 ; Conaco & Cabaitan, 2020 ). Larger larvae are more likely to contain higher endogenous energetic reserves provided maternally than smaller larvae (de Putron et al., 2017 ; Hartmann et al., 2013 ; Marshall & Keough, 2008 ). It is generally assumed that metabolic rates, and consequently energy consumption, increase with temperature, which may lead to higher mortality rates as lecithotrophic larvae deplete their endogenous energy reserves faster (e.g., Edmunds et al., 2001 ; Kipson et al., 2012 ; Pechenik, 1987 ). However, our results on larval biomass and caloric consumption of C. rubrum suggest that the earlier mortality rates caused by increased temperature were not related to depletion of endogenous energy. Although survival of C. rubrum larvae decreased as temperature increased, energy consumption was similar between all treatments (Figure 6 ). The earlier mortality of C. rubrum larvae may be related to the parental environment. Although C. rubrum colonies at ~30 m can experience high temperature variability and can experience maximum temperatures similar to those found at 5 m depth (Viladrich et al., 2016 ), colonies of C. rubrum at 25–30 m experience cooler waters than E. singularis at 13–16 m for a substantial amount of time and thus may be less adapted to high temperature stress. Other potential causes of high mortality of C. rubrum larvae, that represent important directions for future research, include disruption of routine metabolic function (Pechenik, 1987 ), damage to membrane structures disrupting transport systems into and between cells (Hofmann & Todgham, 2010 ), molecular responses such as metabolic depression (Rodriguez‐Lanetty et al., 2009 ), and increased mitochondrial ROS formation (Keller et al., 2004 ). Our results show that larval biological responses to thermal stress are complex, and they cannot only be explained by larval size, presence of symbionts in the larvae, and/or brood quality (i.e., biomass and energy consumption), as previously suggested (e.g., Cumbo et al., 2013 ; Putnam et al., 2010 ). Finally, our results also show that larval survival to thermal stress depends on the day of release in both species studied (Figures S1 and S2 ). This highlights the importance to consider the day of larvae release and shifts in reproductive phenology to better project the success and viability of future coral populations (Cumbo et al., 2012 ; Edmunds et al., 2001 ; Isomura & Nishihira, 2001 ; Putnam et al., 2010 ). 4.3 Effects of thermal stress on settlement Low recruitment rates may undermine the long‐term viability of coral populations with consequences that may scale up to the community or the ecosystem level. However, we still have a limited understanding about how thermal stress impacts on coral larvae might affect the transition from larva to juvenile. Some studies have shown negative impacts of elevated temperature on coral larvae settlement (e.g., Bassim & Sammarco, 2003 ; Conaco & Cabaitan, 2020 ), while others have found that short‐duration stressors had little to no ecological consequence for larvae (e.g., Edmunds et al., 2001 ; Ross et al., 2013 ). Our results show that coral settlement rates in E. singularis were similar between temperature treatments after 20 days, however, larvae in the high‐temperature treatment (26°C) settled faster. This suggests that if larvae are exposed to thermal stress in the field, they may settle fast and close to their native populations, decreasing their potential for long‐distance dispersal and connectivity (Costantini et al., 2016 ). Our results are consistent with other studies that have suggested that warmer temperatures can reduce coral larval swimming and facilitate settlement (Kipson et al., 2012 ; Putnam et al., 2008 ; Serrano et al., 2018 ). However, as mentioned above, the day of larvae release can also be a strong factor determining settlement probability (Figure S3 ). On the other hand, larvae of the octocoral C. rubrum did not settle in any of the treatments after 20 days despite having coralline algae L. stictaeforme to provide positive settlement cues for this species (Zelli et al., 2020 ). However, high variability in C. rubrum settlement and recruitment among years and sites have been observed in the field, suggesting settlement and recruitment rates by pulses (Bramanti et al., 2003 , 2007 ; Garrabou & Harmelin, 2002 ; Santangelo et al., 2012 ). Garrabou and Harmelin ( 2002 ) reported that the annual recruitment observed on 10 panels (4000 cm 2 ) over 22 years was limited to a single recruitment event at the beginning of the study and long‐term monitoring of red coral populations across different marine protected areas showed very low recruitment rates observed for this species (Montero‐Serra et al., 2019 ). Recently, it has been suggested that this high inter‐annual variability of settlement and recruitment rates could be related to non‐selective transfer of energy reserves (i.e., lipids) from maternal colonies to larvae in C. rubrum (Viladrich et al., 2021 ), resulting in a strong dependence of recruitment on the nutritional condition of maternal colonies (Dunstan & Johnson, 1998 ; Yoshioka, 1996 ). 4.4 Ecological consequences and management implications The ability to predict the vulnerability and resilience of corals at different life stages during extreme events is essential for understanding the effect of global climate change on species distributions (Woods et al., 2016 ), estimating the potential for adaptation and designing effective management strategies (Figueiredo et al., 2014 ). Some coral species will be able to persist; other will change their distribution or disappear due to global climate change, causing a shift in species composition. Our results on early life‐history stages of the octocoral E. singularis combined with previous experimental studies looking at the high thermal resistance of adult colonies (Ezzat et al., 2013 ; Previati et al., 2010 ) suggest that E. singularis may be a winner species under future climatic conditions in the Mediterranean Sea. However, our results also show that thermal stress can induce a faster settlement, which may result in lower larval dispersal capacity and, consequently, reduced genetic connectivity among populations (Cowen et al., 2000 ). Persistence of precious red coral population is at higher risk if heatwaves continue, and severe conservation and management plans are not applied. Although adult colonies of C. rubrum seem to be experimentally resistance to heat stress events (Previati et al., 2010 ; Torrents et al., 2008 ), the impact of recurrent heatwaves can cause collapse of their populations (Gómez‐Gras et al., 2021 ; Montero‐Serra et al., 2019 ). In addition, the present study reveals how ocean warming may have serious consequences on larval survival, limiting the introduction of new individuals in the population or the possibility to colonize new areas. The viability of C. rubrum populations is further aggravated since red coral is one of the corals most valued for use in the jewelry industry, and consequently has been and is still overexploited in several Mediterranean countries (Tsounis et al., 2010 ). Low thermotolerance of larvae, coupled with uncontrolled harvesting and the impact of recurrent marine heatwaves, could bring red coral populations to local extinction. To better understand the future of benthic communities of the Mediterranean Sea, our study provides empirical data that can be used to project population dynamics and demography of both octocoral species under the expected future global climate change scenarios based on matrix models and integral projection models (Bramanti et al., 2015 ; Doak et al., 2021 ; Linares & Doak, 2010 ; Montero‐Serra et al., 2019 ). So, management and conservation actions should be based on the outcomes of these simulations to preserve these endemic species together with their associated biodiversity."
} | 5,580 |
35269625 | PMC8910126 | pmc | 3,948 | {
"abstract": "Specialist bacteria can synthesize nanoparticles from various metal ions in solution. Metal recovery with high efficiency can be achieved by metal-tolerant microorganisms that proliferate in a concentrated metal solution. In this study, we isolated bacteria ( Pseudomonas sp. strain KKY-29) from a bacterial library collected from water near an abandoned mine in Komatsu City, Ishikawa Prefecture, Japan. KKY-29 was maintained in nutrient medium with lead acetate and synthesized hydrocerussite and pyromorphite nanoparticles inside the cell; KKY-29 also survived nanoparticle synthesis. Quantitative PCR analysis of genes related to phosphate metabolism showed that KKY-29 decomposed organic phosphorus to synthesize lead phosphate. KKY-29 also deposited various metal ions and synthesized metal nanoparticles when incubated in various metal salt solutions other than lead. The present study considers the development of biotechnology to recover lead as an economically valuable material.",
"conclusion": "5. Conclusions In this study, the strain KKY-29 was identified as a bacterium with lead resistance and the ability to synthesize nanoparticles from fresh water collected from an abandoned mine. KKY-29 belongs to the Pseudomonas genus and synthesizes pyromorphite and hydrocerussite nanoparticles from a lead acetate solution. As KKY-29 survived in lead acetate, it could be applied in fresh-water bioremediation strategies. Figure 14 presents a schematic diagram of nanoparticle synthesis by KKY-29. KKY-29 absorbed and crystallized lead ions as nanoparticles using phosphate ions generated from the metabolization of phosphate and carbonate ions.",
"introduction": "1. Introduction Elements are largely divided into metallic and nonmetallic elements. Some metallic elements, called essential elements, are necessary for maintaining metabolism and organismal homeostasis. On the other hand, both essential and non-essential elements can be toxic if ingested in excess. Since their discovery, humans have employed metals in all practices of life. Heavy metals have become particularly important in the modern industrialized landscape, where they have been utilized in various products such as paints, batteries, computers, and fertilizers. However, their widespread use exacerbates environmental pollution. Heavy metals have a tremendous impact on microorganisms, plants, and humans. For example, heavy metals accumulate in seeds and cause physiological dysfunctions and malnutrition in plants [ 1 ]. Heavy metals accumulate in plants and animals and cause irreversible damage at the cellular level, where they inhibit enzyme activity, disturb the functioning of essential elements, and promote the prevalence of reactive oxygen species [ 2 , 3 , 4 ]. Because lead has a low melting point and is easy to manipulate, it has seen frequent usage in the production of alloys since ancient times as well as batteries, paints, and water pipes more recently [ 5 ]. However, lead ions are toxic and can inhibit enzyme activity by coordinating with thiol groups in proteins [ 6 , 7 , 8 , 9 , 10 ]. Lead ions are more destructive in children, where they can cause neuropathy [ 11 ]. They have also been shown to have various detrimental effects on fetus development [ 12 ]. Microorganisms incorporate heavy metal elements in metabolic and homeostatic processes. For example, iron-reducing bacteria obtain energy by exchanging electrons with iron [ 13 ]; magnetotactic bacteria synthesize iron nanoparticles in the cells to detect magnetic fields [ 14 , 15 ]. Metal nanoparticles have unique physical properties characterized by the quantum size effect. They are used in medical and engineering fields as drug delivery agents, catalysts for chemical reactions, environmental cleanup agents, and antibacterial agents [ 16 , 17 , 18 , 19 , 20 , 21 ]. Research has also considered the value of microorganisms in synthesizing economically important metals as a novel method that requires comparatively little energy and can be incorporated in bioremediation strategies. Some bacteria with heavy metal tolerance can synthesize heavy metal nanoparticles and contribute to bioremediation [ 22 , 23 , 24 , 25 ]. Recovery of heavy metals by nanoparticle-synthesizing microorganisms has also been shown to mitigate heavy metal toxicity in plants [ 26 ]. However, the mechanism of nanoparticle synthesis remains largely unknown [ 27 ]. The present study investigated metal-resistant microorganisms able to synthesize metal nanoparticles, and aimed at describing the mechanisms involved. In order to isolate microorganisms with lead tolerance and lead nanoparticle synthesis abilities, we collected fresh water from around the Okoya mine in Komatsu City, Ishikawa Prefecture, and the Kamioka mine in Hida City, Gifu Prefecture, Japan. The Okoya mine closed in 1972. The river around Okoya mine is polluted with heavy metals, including copper (Cu), zinc (Zn), lead (Pb), and cadmium (Cd) [ 28 ]. Kamioka mine is also an abandoned Cu, Zn, Pb, and Cd mine. The pollution caused itai-itai disease, which is one of the four major types of pollution-derived diseases in Japan [ 29 ]. The comprehensive description of the mechanisms involved in nanoparticle synthesis could inform the development of bioremediation strategies and metal ion recovery.",
"discussion": "3. Discussion In this study, we used bouillon medium with beef extract as a carbon source for screening. A culture medium with a simple composition and inexpensive preparation would facilitate its application in industrial metal recovery. Some lead nanoparticles such as PbS, PbS 2 , and CsPbBr 3 have fluorescence characteristics [ 32 , 33 , 34 ]. However, microorganisms may release substances with fluorescent properties. To confirm the synthesis of CdSe quantum dots with fluorescent properties, it is necessary that the fluorescent properties of bacterial suspensions be measured as a control to verify the absence of endogenous fluorescent substances [ 35 ]. We compared the fluorescence intensity of bacterial suspensions with and without lead to screen nanoparticle synthesizing strains. Phylogenetic analysis showed that KKY-29 was closely related to P . koreensis JCM 14769, originally isolated from farmland soil in Korea. P . koreensis CPSB21 has chromium tolerance and promotes chromium uptake in plants [ 36 ]. In addition, P. koreensis AGB-1 shows tolerance to various metals, such as zinc and lead, and accumulates metals [ 37 ]. AGB-1 also accumulates lead ions, but the lead ions are fixed as metal complexes outside the cell. The differences in the presence of lead ions suggest that KKY-29 concentrates lead ions by a mechanism different from that of AGB-1. The LIVE/DEAD Biofilm Viability Kit was used to evaluate the viability of KKY-29 cells. As this kit consists of SYTO9, a membrane-permeable green fluorescent nucleic acid staining reagent, and propidium iodide, a membrane-impermeable red fluorescent nucleic acid staining reagent, the viability of bacteria can be evaluated by two-color fluorescence. In this study, KKY-29 survived after suspension in lead acetate solution. Lead is toxic not only to humans but also to bacteria, in which certain metabolic pathways are selectively inhibited. Consequently, local microbial diversity is reduced as non-resistant microorganisms are eliminated by the release of lead into the environment [ 38 , 39 ]. Pyromorphite is a mineral that is also called apatite [ 40 ]. Apatite compounds are valuable catalysts and can be used for the immobilization of toxic metals [ 41 , 42 ]. Among them, pyromorphite is considered to be the most stable mineral species for accumulating lead in the environment because of its insolubility and non-bioavailability. It is widely studied as a form of final precipitate in immobilizing lead in the environment [ 43 , 44 , 45 ]. However, the method of adding phosphate to a polluted environment shows low conversion efficiency to pyromorphite. For example, in the soil of a shooting range (a typical example of a lead-contaminated environment) about 30% of the phosphate becomes green lead ore, while the rest of the lead remains in a different chemical form [ 46 ]. Moreover, phosphate binds to aluminum and iron ions and precipitates [ 47 ]. By changing the culture conditions, the KKY-29 strain can recover a large proportion of lead from the environment and deposit it as pyromorphite, and, given further research and development, could be used in the bioremediation of lead-contaminated environments. As organic phosphate is abundant in soil, the use of inorganic phosphorus obtained by decomposition for lead deposition is being studied. Aspergillus niger precipitates lead as pyromorphite by the secretion of phytase [ 48 ]. As KKY-29 synthesizes pyromorphite, consisting of lead and phosphate, KKY-29 is considered to metabolize phosphate and use it for the synthesis of pyromorphite nanoparticles. Therefore, to investigate the effect of phosphoric acid conditions on the synthesis of nanoparticles, a synthesis medium was used. When cultured under phosphorus-free conditions, hydrocerussite nanoparticles were synthesized; pyromorphite nanoparticles were synthesized when cultured under phosphorus-rich conditions. When phosphorus was present in the medium, KKY-29 reproduced using the phosphorus carried over from the preculture medium and phosphorus in the medium, and pyromorphite nanoparticles were synthesized using excess phosphorus. When KKY-29 was cultured under phosphorus-free conditions, the phosphorus carried over from the preculture medium was used only for reproduction and was not hydrolyzed to inorganic phosphate, and the hydrocerussite nanoparticles were synthesized by combining lead ions with carbonate ions generated from the respiratory chain. The KKY-29 strain is thought to have acquired tolerance to lead ions by sequestering them through crystallization to prevent them from causing toxic damage."
} | 2,483 |
32208436 | PMC7092969 | pmc | 3,949 | {
"abstract": "Natural microbial communities contain hundreds to thousands of interacting species. For this reason, computational simulations are playing an increasingly important role in microbial ecology. In this manuscript, we present a new open-source, freely available Python package called Community Simulator for simulating microbial population dynamics in a reproducible, transparent and scalable way. The Community Simulator includes five major elements: tools for preparing the initial states and environmental conditions for a set of samples, automatic generation of dynamical equations based on a dictionary of modeling assumptions, random parameter sampling with tunable levels of metabolic and taxonomic structure, parallel integration of the dynamical equations, and support for metacommunity dynamics with migration between samples. To significantly speed up simulations using Community Simulator, our Python package implements a new Expectation-Maximization (EM) algorithm for finding equilibrium states of community dynamics that exploits a recently discovered duality between ecological dynamics and convex optimization. We present data showing that this EM algorithm improves performance by between one and two orders compared to direct numerical integration of the corresponding ordinary differential equations. We conclude by listing several recent applications of the Community Simulator to problems in microbial ecology, and discussing possible extensions of the package for directly analyzing microbiome compositional data.",
"conclusion": "Conclusions We hope that the Community Simulator will become a valuable resource for the microbial ecology community. It has already played an important role in our own work. The package initially facilitated the systematic evaluation of the robustness of results to different modeling assumptions in a study of the effects of total energy influx on community structure, diversity and function [ 24 ]. More recently, the convex optimization approach has made it possible to perform more than 100,000 independent simulations in a reinterpretation and extension of Robert May’s classic work on diversity and stability [ 32 , 33 ]. We have also employed the package to reproduce large-scale patterns in microbial biodiversity from the Human Microbiome Project, Earth Microbiome Project, and similar surveys [ 34 ]. Finally, the random matrix approach implemented in this package is amenable to analytic calculation in the limit of large numbers of species and resources, using cavity methods from the physics of disordered systems [ 35 , 36 ]. It is our belief that the Community Simulator will facilitate the further development of these mathematical techniques through efficient testing of new conjectures. One interesting future direction to explore is integrating the Community Simulator with methods for directly analyzing Microbiome sequencing data. For example, there has been a renewed interest in statistical techniques such as Approximate Bayesian Computation (ABC) for understanding ecology and evolution [ 37 ]. In ABC, the need to exactly calculate complicated likelihood functions—often a prerequisite for many statistical techniques—is replaced with the calculation of summary statistics and numerical simulations. For this reason, the Community Simulator Python package is ideally suited to form the backbone of new inference techniques for trying to related ecological processes to observed abundance patterns in microbial ecosystems."
} | 870 |
28570712 | PMC5453599 | pmc | 3,950 | {
"abstract": "Black band is a deadly coral disease found worldwide, which may become more virulent as oceanic conditions continue to change. To determine the effects of climate change and ocean acidification on black band disease virulence, Orbicella faveolata corals with black band were exposed to different temperature and pH conditions. Results showed a significant decrease in disease progression under low pH (7.7) conditions. Low pH also altered the relative abundance of the bacterial community of the black band disease consortium. Here, there was a significant decrease in Roseofilum , the cyanobacterium that typically dominates the black band mat. These results indicate that as oceanic pH decreases so may the virulence of a worldwide coral disease.",
"conclusion": "Conclusions The progression rate of black band disease was significantly reduced under low pH conditions, which was likely not a result of the physiological state or the bacterial community of the coral host. Analysis of the bacterial community of the black band mat showed a significant reduction in the Roseofilum cyanobacteria, a primary pathogenic agent of the black band disease consortium. The present study indicates that pH can significantly influence the community structure of the black band disease consortium, which results in a decreased progression rate of the disease. Therefore, at least under certain conditions, low pH could reduce the virulence of a worldwide coral disease.",
"introduction": "Introduction Black-band disease is one of the most prevalent and virulent diseases affecting contemporary corals world-wide with progression rates reaching over 2 cm day -1 [ 1 – 3 ]. Black band disease is also a generalist disease affecting at least 42 Caribbean species of corals, including both scleractinians and gorgonians [ 4 , 5 ]. The disease is characterized as a microbial assemblage that creates a dark band, ranging from black to red in appearance, which moves across healthy coral tissue, causing mortality and leaving behind bare skeleton [ 6 ]. A cyanobacterium often dominates the black band mats, and contains the light-harvesting accessory pigment phycoerythrin, giving black band disease its distinct coloration. The mat consists of several microbial functional groups including sulfide oxidizers, sulfate reducers, heterotrophic bacteria, fungi, and Archaea [ 3 , 6 , 7 ]. Black band disease functions by creating microenvironments that are anoxic, low in pH, and high in sulfide [ 8 ]. The coral animal is unable to survive in this environment and becomes the organic fuel of the black band disease microbial consortium [ 8 , 9 ]. The black band disease consortium, however, requires these microenvironment conditions to survive and thrive. Although black band has been studied for decades, Koch’s postulates, used to determine disease causation, have not been fulfilled for black band disease because of its complex microbial community. Richardson and Kuta [ 10 ] as well as Frias-Lopez et al. [ 11 ] suggest that the entire black-band community functions as a pathogenic consortium, and that there may be no individual, primary pathogen responsible for black-band disease. Many studies, however, continue to recognize that the cyanobacteria play a critical role in the creation and maintenance of black band disease [ 9 , 11 , 12 ]. Environmental conditions are often correlated with black band disease dynamics. Numerous studies show a positive correlation between seawater temperature and the prevalence of black band disease [ 13 , 14 ] and global climate change may lead to increases in black band incidence. Previous research indicates progression rates of black band disease were positively correlated with water temperatures [ 14 – 16 ] also suggesting an increase in virulence under global warming conditions. Mechanisms underlying the response of black band to increased temperature are unknown. The consortium of microbes that create black band may thrive under warm water temperature or the coral host may become immunocompromised under high temperature conditions [ 17 – 19 ]. In addition to increasing water temperatures, there is a predicted decrease in oceanic pH under future climate change scenarios[ 20 ]. The impact of decreasing pH on black band disease dynamics is unknown. The objectives of the present study were to i) quantify the effects of temperature and pH on the virulence of black band disease infecting Orbicella faveolata , ii) determine whether different temperature and pH conditions changed the photochemical efficiency of the coral-host symbiosis, and iii) characterize the change in bacterial communities within the coral host as well as the black band bacterial consortium under different pH and temperature conditions.",
"discussion": "Discussion Black band disease virulence Low pH significantly reduced the progression rates of black band disease on O . faveolata , a response which could have important implications for the prediction of coral-disease dynamics under future ocean acidification conditions. However, the water source used within the present experiment also contained high alkalinity, perhaps confounding the interpretation of the results. Seawater alkalinity is known to buffer the impacts of ocean acidification by providing an excess of proton acceptors within the water column and is often consistent through time. Within the present study, the response of reduced disease progression rates was only observed on corals within the low pH treatment indicating a true, and statistically relevant, treatment effect. Furthermore, previous research showed that most physiological change in corals associated with ocean acidification was primarily from a change in proton (H+) concentrations rather than other carbonate chemistry parameters such as total alkalinity [ 28 ]. Therefore, the reduced progression rates of black band disease may occur under low pH conditions regardless of the alkalinity; however, further research is needed to test this hypothesis. Interestingly, another study also suggests a decreased virulence of a crustose coralline algae disease under low pH conditions[ 29 ]. Statistical analyses indicated that there was no temperature effect on the black band disease progression rates. An additional short-term study also showed that temperature did not influence the progression rate of black band disease, whereas light was the primary driver [ 30 ]. However, the general trend of increasing progression rates with temperature was observed within the Sato et al. [ 30 ] seven day study. Similarly, the lack of a temperature response in the present experiment may have resulted from the short term duration of exposure for only 16 days after initial infection. Most other studies observing temperature effects have experimented with naturally occurring and well established infections [ 14 , 16 ]. The progression rates of black band disease on corals within the high temperature treatment were becoming apparent through time. We predict that a response effect to temperature would have become evident with increased exposure time. Photochemical efficiency There was no effect of either temperature or pH on the photochemical efficiency the algal symbionts within the experimental corals. Other studies often show a reduced photochemical efficiency after corals are exposed to prolonged periods of high water temperature, which results in the dysbiosis between the coral host and the photosymbiotic algae that reside within their tissues [ 31 ]. Additionally, low pH has caused reef-building corals to bleach under experimental conditions [ 32 ]. The present study, however, may not have provided the length of exposure needed to elicit a response. Alternatively, these corals may be generally more resilient to environmental change, perhaps because they were collected from the coral rescue nursery. Here, the corals were held underneath boat docks where light levels were low and turbidity was high. Although no direct water quality measurements were taken, this unique nearshore environment may have influenced the physiology of the corals within the present experiment. For example, previous exposure to high temperature has the ability to increase coral resilience to subsequent high temperature events, at least for some species [ 33 ]. The lack of a physiological response, however, indicates that the change in progression rates of the black band disease under low pH conditions was not related to the health state of the host. The photochemical efficiency did change through time, which showed a general increasing trend. These data, again, suggests that the physiological response of the symbiosis between the symbiotic algae and the coral host remained intact and may even have improved because of the experimental conditions. Bacterial community of coral tissue/mucus There was no significant difference in the bacterial community of the coral tissue/mucus among the different treatments suggesting that the change in black band virulence under low pH conditions cannot be explained by the bacterial community of the host. One particular treatment, the high temperature and control pH treatment, however, did separate out distinctly from the other three treatments within the NMDS biplot. An examination of the major bacterial classes show decreased levels of Rhodobacters under the high temperature and control pH treatment with a concomitant increase in Clostridiales and Flavobacteria. Rhodobacters have been associated with tissue-loss diseases within corals [ 34 ]. For example, an increase in Rhodobacter bacteria was detected within samples of the coral disease white plague on Orbicella faveolata , but was also found within healthy corals indicating a potential increase of opportunistic commensals under certain conditions [ 35 ]. High levels of Alphaproteobacteria were also detected within two different coral species showing signs of white plague, Diploria strigosa and Siderastrea siderea , which was primarily the result of increased abundances of Rhodobacters [ 36 ]. Alternatively, some genera of Rhodobacters, such as Roseobacter , have been identified as beneficial bacteria, important for the settlement and general health of the coral host [ 37 ]. However, within the present study Roseobacter was found in low abundances within the corals sampled regardless of the treatment condition. Within the present study, a reduction of Rhodobacters only under high temperature and control pH conditions may suggest sensitivity to temperature, which is then mitigated under low pH conditions. There were also significantly more Flavobacteria and Clostridiales bacteria within the coral tissue/mucus under control pH compared with low pH conditions. However, the increased abundance of Flavobacteria and Clostridiales under high temperature and control pH conditions is likely driving these statistical results ( Fig 3A ). Therefore, this response may be simply an increase in growth of these potential pathogenic bacteria under high temperature conditions. Flavobacteria have been implicated in a coral disease outbreak on Montipora aequituberculata [ 38 ] and associated with white plague-affected Mussismilia corals [ 39 ]. Clostridiales bacteria also have been associated with several white diseases of corals [ 34 , 35 ]. Similar to Rhodobacters, an increase in these bacteria only under high temperature and control pH conditions, and not within the high temperature and low pH conditions, may indicate that low pH mitigates temperature effects. Bacterial community of black band disease The present study showed that low pH reduced the abundance of the bacterial class Oscillatoriophycidaea, the cyanobacterium that often dominates black band disease. The microenvironment of the black band disease consortium itself creates a low pH, low oxygen, and high sulfide rich area at the interface between the band and the coral host tissue [ 6 ]. The cyanobacteria of black band, therefore, are routinely exposed to low pH conditions within the host, but appeared negatively affected by the addition of low pH conditions within the external environment of the present experiment. Cyanobacteria have the ability to regulate internal pH to an extent, although low pH leads to the acidification of the cytoplasm after thresholds are exceeded [ 40 ]. Previous research also indicates that low pH reduces cyanobacterial growth rates [ 41 ]. Generally, photosynthetic organisms, such as cyanobacteria, are predicted to thrive under ocean acidification conditions, which includes a low pH environment [ 42 ]. High levels of carbon dioxide suspended within the water may increase the rate of photosynthesis, thus providing the potential to increase productivity [ 43 ]. Although ocean acidification does include reduced pH, the high alkalinity and thus the high p CO 2 levels within our treatments (see S1 Table ), limit comparability. Still, even within the context of ocean acidification the results of several studies show variability in the physiological responses of cyanobacteria to low pH. For example, certain cyanobacterial groups such as Synechoccocus have shown higher growth rates under low pH conditions [ 44 ], as well as reduced growth rates [ 45 ]. Similarly, the Trichodesmium spp. of cyanobacteria showed reduced nitrogen fixation under low pH conditions, but only when iron was limited [ 46 ] Other studies indicate Trichodesmium spp. increase both carbon and nitrogen fixation under low pH conditions regardless of iron concentrations [ 47 , 48 ]. To complicate matters further, cyanobacteria that interact in complex assemblages, such as those found in biofilms, may be outcompeted by other photosynthetic organisms under low pH conditions [ 49 ]. In microbial biofilms, low pH has reduced the abundance and diversity of cyanobacteria [ 50 – 53 ]. For example, Hassenrück et al. [ 52 ] showed that under low pH conditions, the bacterial epiphytic community of seagrasses had a reduced abundance and diversity of cyanobacteria. In the present study, the Oscillatoriophycideae class primarily consisted of the cyanobacteria genus Roseofilum (previously named Oscillatoria ), which dominates the black band disease mat [ 54 ]. Arotsker et al. [ 55 ] showed that the most transcribed gene in the band consortium was cyanobacterial adenosylhomocysteinase, which is involved in cyanotoxin production and is a large contributor to the virulence of black band disease [ 56 – 58 ]. The cyanobacteria found within black band are indeed interacting with a complex community of other microorganisms [ 1 , 8 , 59 ]. The many other microbes that create the black band disease mat may be out-competing the cyanobacteria under low pH conditions, resulting in reduced abundances. However, whether the reduced pH is directly influencing the cyanobacteria or another component of the microbial consortium, which ultimately leads to the reduction of cyanobacteria, is unknown. Furthermore, additional studies are needed to determine whether ocean acidification, rather than just reduced pH per se , similarly affects the bacterial consortium of black band disease. A different response of the Alphaproteobacteria bacterial class was observed within the black band disease consortium compared with the bacterial community of the coral tissue/mucus. Here, significant increases in the Alphaproteobacteria were present only under the control temperature low pH treatment scenario. Again, Rhodobacteriales were the dominant group within this heterotrophic bacterial class. The differing responses of the bacterial classes from the coral tissue/mucus compared with the black band disease consortium emphasizes the important role of the environment and the complex interactions among bacterial groups found within different hosts and assemblages (i.e., within an invertebrate host versus a cyanobacterial-dominated mat). Temperature effects were detected within the bacterial class Clostridia within the black band consortium, where high temperatures increased the relative abundance of Clostridia. As mentioned previously, this bacterial class is linked to tissue-loss diseases within scleractinian corals, and flourishes under high temperatures [ 34 , 35 ]. An increase within this bacterial class was also observed within the coral tissue/mucus under the high temperature control pH treatment."
} | 4,088 |
32683192 | null | s2 | 3,951 | {
"abstract": "Engineering microbial cell factories has been widely applied to produce compounds spanning from intricate natural products to bulk commodities. In each case, host robustness is essential to ensure the reliable and sustainable production of targeted metabolites. However, it can be negatively affected by metabolic burden, pathway toxicity, and harsh environment, resulting in a decreased titer and productivity. Enhanced robustness enables host to have better production performance under complicated growth circumstances. Here, we review current strategies for boosting host robustness, including metabolic balancing, genetic and phenotype stability enhancement, and tolerance engineering. In addition, we discuss the challenges and future perspectives on microbial host engineering for increased robustness."
} | 202 |
22908009 | PMC3415069 | pmc | 3,953 | {
"abstract": "Though iron- and sulfate-reducing bacteria are well known for mediating uranium(VI) reduction in contaminated subsurface environments, quantifying the in situ activity of the microbial groups responsible remains a challenge. The objective of this study was to demonstrate the use of quantitative molecular tools that target mRNA transcripts of key genes related to Fe(III) and sulfate reduction pathways in order to monitor these processes during in situ U(VI) remediation in the subsurface. Expression of the Geobacteraceae -specific citrate synthase gene ( gltA ) and the dissimilatory (bi)sulfite reductase gene ( dsrA ), were correlated with the activity of iron- or sulfate-reducing microorganisms, respectively, under stimulated bioremediation conditions in microcosms of sediments sampled from the U.S. Department of Energy’s Oak Ridge Integrated Field Research Challenge (OR-IFRC) site at Oak Ridge, TN, USA. In addition, Geobacteraceae -specific gltA and dsrA transcript levels were determined in parallel with the predominant electron acceptors present in moderately and highly contaminated subsurface sediments from the OR-IFRC. Phylogenetic analysis of the cDNA generated from dsrA mRNA, sulfate-reducing bacteria-specific 16S rRNA, and gltA mRNA identified activity of specific microbial groups. Active sulfate reducers were members of the Desulfovibrio , Desulfobacterium , and Desulfotomaculum genera. Members of the subsurface Geobacter clade, closely related to uranium-reducing Geobacter uraniireducens and Geobacter daltonii , were the metabolically active iron-reducers in biostimulated microcosms and in situ core samples. Direct correlation of transcripts and process rates demonstrated evidence of competition between the functional guilds in subsurface sediments. We further showed that active populations of Fe(III)-reducing bacteria and sulfate-reducing bacteria are present in OR-IFRC sediments and are good potential targets for in situ bioremediation.",
"introduction": "Introduction Mining and milling of uranium for nuclear weapons production has resulted in widespread uranium contamination in subsurface environments across North America, South America, and Eastern Europe (Abdelouas et al., 1998 ). Oxidized uranium, U(VI), is highly soluble and toxic, and a potential contaminant to local drinking water supplies (Palmisano and Hazen, 2003 ). Immobilization of oxidized uranium can be achieved in contaminated groundwater through the reduction of U(VI) to insoluble U(IV) by indirect (abiotic) and direct (enzymatic) processes catalyzed by microorganisms (Wall and Krumholz, 2006 ; Kostka and Green, 2011 ). Current remediation practices for dealing with uranium contamination aim to promote U(VI) immobilization via natural attenuation or the biostimulation of indigenous microorganisms through a combination of pH neutralization and/or the addition of electron donor (Finneran et al., 2002 ; Anderson et al., 2003 ; Wilkins et al., 2006 ; Groudev et al., 2010 ). Dissimilatory Fe(III)-reducing bacteria (FeRB) and sulfate-reducing bacteria (SRB) comprise two major groups which are capable of U(VI) reduction (Tebo and Obraztsova, 1998 ; Lovley et al., 2004 ; Sani, 2004 ; DiChristina, 2005b ; Payne and DiChristina, 2006 ; Wall and Krumholz, 2006 ). Both FeRB and SRB can directly reduce U(VI) by using it as an electron acceptor, and a subset of these groups have been shown to conserve energy for growth via U(VI) reduction (Lovley et al., 2004 ). In addition, the products of microbial Fe(III) and sulfate reduction, Fe(II) and hydrogen sulfide, can react abiotically to reduce U(VI) (Liger et al., 1999 ; Hua et al., 2006 ). Therefore, FeRB and SRB are considered to have a high bioremediation potential in U(VI) contaminated subsurface sediments. Although prior research has linked the activity of these functional guilds to U(VI) immobilization in contaminated subsurface environments (Anderson et al., 2003 ; Wu et al., 2006 ; Akob et al., 2008 ; Cardenas et al., 2010 ), it remains difficult to directly relate the in situ activity of specific microbial groups to the environmental controls of the processes. In order to exploit the activity of FeRB and SRB for bioremediation, there remains a need to develop quantitative tools for monitoring the metabolic activity of these microbial groups in subsurface environments. Quantifying the in situ activity of Fe(III) reducers is particularly problematic and a molecular proxy for Fe(III) reduction has not yet been verified or calibrated with biogeochemical rate measurements in any sedimentary environment. A promising strategy for quantifying the in situ metabolic activity of SRB and FeRB would be to monitor mRNA transcript levels of key genes involved in sulfate or Fe(III) reduction. The dissimilatory (bi)sulfite reductase ( dsrAB ) gene is highly conserved among sulfate-reducing prokaryotes ( Bacteria and Archaea ) and codes for the dissimilatory (bi)sulfite reductase, which is responsible for the rate-limiting step of sulfate reduction (Wagner et al., 1998 ). Levels of mRNA for dsrAB genes were shown to increase in pure culture studies of dissimilatory SRB as rates of sulfate reduction increased (Neretin et al., 2003 ; Villanueva et al., 2008 ) and correlated with the activity of SRB in petroleum-contaminated marine harbor sediments (Chin et al., 2008 ). In the case of Fe(III) reduction, no single respiration pathway has been identified as FeRB can reduce insoluble Fe(III) oxides via direct enzymatic reduction, electron shuttling pathways, or by solubilizing metals with organic ligands (DiChristina, 2005a ). One approach is to target and monitor functional genes of important groups of FeRB known to be involved in Fe(III) reduction and to be abundant in contaminated subsurface sediments. Members of the Geobacteraceae family are one such group and cytochromes which are involved in Fe(III) reduction have been identified in pure cultures of different Geobacter species. However, comparative analysis of available Geobacteraceae genome sequences has revealed that these cytochromes are not conserved throughout the Geobacteraceae family (Butler et al., 2010 ). Furthermore, it has been known that an outer-membrane cytochrome o mcB expression patterns were largely affected by environmental fluctuations, such as changes in electron acceptor availability, suggesting that monitoring omcB transcripts in Geobacter -dominated environments would not provide an accurate indication of rates of Fe(III) reduction (Chin et al., 2004 ). In addition to these characteristics, the Geobacteraceae family does contain a phylogenetically distinct functional gene, the citrate synthase ( gltA ) gene, that codes for an enzyme involved in the incorporation of acetate into the tricarboxylic acid (TCA) cycle (Bond et al., 2005 ; Holmes et al., 2005 ). The Geobacteraceae -specific gltA gene is a good target for this group of FeRB because it is more similar to eukaryotic citrate synthase genes (Methe et al., 2003 ; Bond et al., 2005 ) distinguishing it from other prokaryotic FeRB and heterotrophs. Measurements of gltA transcripts were used as a proxy for the activity of Geobacteraceae during bioremediation of uranium-contaminated groundwater (Holmes et al., 2005 ) and sediments (Akob et al., 2008 ). However, no study of metal or sulfate reduction in subsurface sediments has directly linked transcript level with process rates determined by geochemical methods. Therefore, in this study we quantified the transcript level of functional genes as a molecular proxy for the metabolic activity of Geobacteraceae-related FeRB and SRB in parallel with determining process rates and the abundance of predominant electron acceptors in field samples.",
"discussion": "Discussion Metabolic activity of FeRB and SRB in uranium-contaminated subsurface sediments Bioremediation, via biostimulation of microbial communities by addition of large quantities of electron donor, is a proposed strategy for uranium immobilization at many uranium-contaminated sites managed by the U.S. DOE. The addition of electron donors, e.g., ethanol and acetate, has been shown previously to stimulate the reductive removal of U(VI) in the subsurface of many of these sites (e.g., Anderson et al., 2003 ; Wu et al., 2006 ; Akob et al., 2008 ; Michalsen et al., 2009 ). Iron(III)-reducing bacteria (FeRB) and SRB have been identified as the functional guilds of microorganisms likely to catalyze U(VI) reduction during in situ bioremediation experiments conducted at the DOE’s OR-IFRC site or the Old Rifle site in Colorado, in particular (Kostka and Green, 2011 ). Although a large number of studies have shown that the abundance of these microbial groups increases in response to electron donor addition and under metal-reducing conditions, the activity of specific metal-reducing bacterial populations (e.g., SRB and FeRB) remains difficult to assess under in situ conditions. Biogeochemical methods for quantifying rates of sulfate or Fe(III) reduction (Canfield et al., 2005 ) are tedious and time consuming and thus cannot be easily applied over the scales necessary for monitoring in situ bioremediation. Thus, a quantitative molecular approach for the determination of Fe(III) or sulfate reduction activity would aid in diagnosing the success of bioremediation strategies along with in situ controls of the enzymatically catalyzed processes. A method that focuses on the gene expression of metal-reducing bacterial populations would be ideal since the active populations could be identified along with the quantification of activity. The desired target should be a phylogenetically informative gene that is highly conserved and unique to a distinct group and for which expression patterns are correlated to metabolic rates. The dissimilatory (bi)sulfite reductase ( dsrAB ) gene provides such a target for SRB since it is highly conserved and codes for the enzyme responsible for the rate-limiting step of sulfate reduction. In the case of FeRB, such a robust gene target has not been identified because numerous pathways exist for metal respiration that involved a number of different proteins, which are poorly conserved (DiChristina, 2005a ). The expression of respiratory genes involved in Fe(III) reduction was shown to correlate with rates of metabolism in pure cultures of the Geobacteraceae (Chin et al., 2004 ). However, transcript level of respiratory genes was shown to respond to other parameters besides rates of metabolism and growth (Chin et al., 2004 ). It was proposed that the expression of genes linked to central carbon metabolism may provide an alternative proxy for metabolic rates of Fe(III)-reducing members of the Geobacteraceae (Holmes et al., 2005 ). Geobacter species are known to predominate under Fe(III)-reducing conditions (Bond et al., 2005 ; Holmes et al., 2005 ) and are in fact capable of outcompeting other FeRB in Fe(III)-rich environments (Rooney-Verga et al., 1999 ; Snoeyenbos-West et al., 2000 ; Röling et al., 2001 ). In addition, Geobacteraceae often predominate in uranium-contaminated subsurface sediments undergoing bioremediation, including the above mentioned DOE sites at Oak Ridge, TN, USA and in Rifle, CO, USA (Holmes et al., 2002 , 2007 ; Anderson et al., 2003 ; North et al., 2004 ). Thus in this study, we targeted the citrate synthase ( gltA ) gene that is unique to the Geobacteraceae and encodes for an enzyme involved in the incorporation of acetate in to the TCA cycle (Bond et al., 2005 ; Holmes et al., 2005 ). To date, a total of 16 Geobacter species were isolated that are all capable of Fe(III) reduction (Lovley et al., 2004 ; Nevin et al., 2005 , 2007 ; Sung et al., 2006 ; Shelobolina et al., 2007 , 2008 ; Prakash et al., 2010 ). Eight of these isolates, Geobacter daltonii, G. uraniireducens, G. metallireducens, G. sulfurreducens, G. psychrophilus, G. bemidjiensis, G. lovleyii , and G. thiogenes , have citrate synthase genes available on the NCBI sequence database that are unique as they are eukaryote-like (Methe et al., 2003 ; Bond et al., 2005 ) distinguishing them from other prokaryotic FeRB and heterotrophs. The other isolates currently do not have genome sequences available so the genes have not been identified, but their similar metabolisms supports the inference that they may harbor the same, conserved citrate synthase gene. It is important to note that many organisms that are closely related to the Geobacteraceae do not possess a complete citric acid cycle or citrate synthase or have citrate synthase genes that are prokaryote-like instead of eukaryote-like. In addition, a closely related sulfate-reducer Desulfovibrio desulfuricans lacks citrate synthase in its genome, which was suggested by Bond et al. ( 2005 ) to indicate that citrate synthase was an important requirement in the evolution of Geobacteraceae to reduce Fe(III). We are aware that citrate synthase is not a molecular marker specifically for Fe(III) reduction, however it has been shown that its expression in pure cultures correlated directly with the central metabolism that was required for electron transfer to Fe(III) (Holmes et al., 2005 ). Here we show for the first time that gene expression quantified as transcript levels of the Geobacteraceae clade of FeRB directly correlates with process rates and we identify specific microbial groups that are likely to catalyze metal reduction in situ . The competition between Fe(III) and sulfate reducers for carbon substrates could also be observed. In general, transcript levels of dsrA and Geobacteraceae -specific gltA paralleled with the extent of sulfate and Fe(III) reduction, respectively, in all of our incubations. In the acetate-amended treatments, electron-accepting processes occurred according to thermodynamic predictions (Canfield et al., 2005 ). The bulk of the more energetically favorable process, Fe(III) reduction, occurred during the first 2 weeks of incubation, whereas most sulfate reduction occurred after 2 weeks. The rapid reduction of Fe(III) in acetate treatments was correlated with an increase in the expression of the Geobacteraceae -specific gltA gene, which indicates an increase in the growth and metabolism of members of the Geobacteraceae family. A recent study reported that Geobacter spp. grow rapidly after the addition of acetate to uranium-contaminated sediments and that in conditions of excess electron donor their abundance is primarily controlled by the availability of microbially reducible Fe(III) (Barlett et al., 2012 ). In our microcosms amended with acetate and molybdate, we observed higher expression of gltA compared to the acetate treatment without molybdate. It is likely that this increase in gene expression in the presence of molybdate reflects the competition for electron donors between FeRB and SRB. With the addition of molybdate, the activity of SRB was depressed thereby removing competition for added acetate. This conclusion is supported by our observation that acetate was consumed at a faster rate in the presence of molybdate. In contrast to Barlett et al. ( 2012 ), evidence from this study does not support the concurrent growth of FeRB and SRB in the presence of excess acetate. Rather, we conclude that in addition to Fe(III) availability, competition between FeRB and SRB is a key factor in limiting the activity of Geobacter during uranium bioremediation. The choice of electron donor for bioremediation may directly impact U(VI) biotransformation by affecting the competition between FeRB and SRB. During OR-IFRC bioremediation experiments with ethanol, Fe(III) and sulfate reduction occurred simultaneously and SRB were more abundant than FeRB, suggesting that SRB play a more important role in U(VI) immobilization (Wu et al., 2007 ; Akob et al., 2008 ; Cardenas et al., 2008 ; Hwang et al., 2009 ). In corroboration of previous work, we observed concurrent Fe(III) and sulfate reduction in ethanol-amended treatments, which goes against thermodynamic predictions (Canfield et al., 2005 ). The overlap in activity suggests that separate populations of Fe(III) and sulfate reducers were successfully competing for ethanol while reducing their preferred electron acceptor. It may be that SRB populations that couple reduction to ethanol oxidation are incapable of using Fe(III) thereby making sulfate the more energetic electron acceptor for their metabolism. Ethanol was first oxidized incompletely to acetate in the initial 6–8 days of incubation and then acetate was subsequently utilized by microbial consortia. Acetate was completely consumed in the ethanol only treatments, while >1 mM acetate remained in the treatments amended with ethanol + molybdate. This likely indicates that the microcosms were depleted in Fe(III) and sulfate was not utilized due to the fact that sulfate reducers were inhibited by molybdate. Ethanol addition resulted in the enhanced removal of soluble U(VI) relative to acetate amendment. Prior to the onset of active sulfate reduction, the levels of dsrA transcripts increased in electron donor-amended microcosms and when sulfate concentrations were depleted, levels of dsr A transcripts concurrently decreased. This is not surprising because dsrAB codes for the dissimilatory (bi)sulfite reductase enzyme which is responsible for the rate-limiting step of sulfate reduction (Wagner et al., 1998 ). The increased dsrA gene expression indicates the up regulation of the DSR operon, which is necessary for the cells to synthesize and to transport the enzymes needed to reduce the available sulfate. As expected in microcosms amended with molybdate to inhibit sulfate reduction the levels of dsrA transcripts were at the lower limit of detection. Unlike the response of dsrA transcripts, Geobacteraceae -specific gltA transcript levels did not increase prior to Fe(III) reduction but increased proportionally with Fe(III) reduction activity. This follows as gltA codes for an enzyme involved in the incorporation of acetate in to the TCA cycle (Bond et al., 2005 ; Holmes et al., 2005 ). Therefore, since gltA expression is not required for Fe(III) reduction but carbon metabolism, we would expect it to correlate more closely with acetate consumption. The increase in glt A expression indeed precedes consumption of acetate and indicates up regulation of the TCA cycle prior to acetate decrease. The quantification of Geobacteraceae -specific glt A transcripts verified a direct association between Fe(III) reduction and the oxidation of acetate. Phylogeny of active metal-reducing bacteria in uranium-contaminated sediments Quantification of gltA and dsrA gene expression was successful in sediment samples from the moderately and highly contaminated OR-IFRC Areas 2 and 3, respectively. The observation that Geobacteraceae-related FeRB and SRB are metabolically active within borehole sediments without biostimulation provides evidence that these organisms are available to promote bioremediation and the long-term stability of reduced uranium in situ . The detection of transcripts in unamended subsurface sediments further demonstrates the sensitivity and specificity of the mRNA-based method. Previous studies found that Geobacteraceae - related FeRB and SRB were below detection in highly contaminated Area 3 of the OR-IFRC subsurface (Cardenas et al., 2008 ). We observed that transcript levels of gltA and dsrA were low and only correlated with the abundance of electron acceptors in Area 3 sediments. Lower transcript levels in Area 2 sediments indicated that FeRB and SRB are active only at background levels, therefore their activity is not directly reflected in the prevailing biogeochemistry of the surrounding sediment. However, we cannot ignore the possibility that transport and sampling of sediments affected microbial activity and biased transcript analysis. For Area 3 sediments the highest levels of expression were associated with sediment zones with higher Fe(II) content and low sulfate concentrations. The low sulfate concentrations and high dsr A transcript levels are indicative of on going sulfate reduction, whereas, glt A expression in regions with high Fe(II) concentrations indicated in situ activity of Geobacteraceae - related FeRB. The in situ activity of FeRB and SRB in Area 3 was surprising as this area is rich in nitrate, which serves as an alternative electron acceptor for FeRB and SRB, due to close proximity to the source zone, and is limited in electron donors or carbon substrates. However, some species of Geobacter are known to reduce nitrate, such as G. humireducens (Coates et al., 1998 ) and G. metallireducens (Lovley et al., 1993b ), and some species are known to oxidize aromatic hydrocarbons (Prakash et al., 2010 ) which are present in the contaminant plume of Area 3 ( http://www.esd.ornl.gov/orifrc/ ). Although the conditions in Area 3 are more favorable for nitrate-reducing bacteria, which are adapted to the high nitrate, low pH in situ conditions (Green et al., 2012 ), active FeRB and SRB populations are present and active at low levels in situ . Such subsurface zones with high in situ activity of FeRB and SRB are likely the areas best suited for in situ bioremediation, as active microbial populations could quickly respond to the input of supplemental electron donor. Members of the Geobacteraceae family are often detected in conjunction with metal reduction in the uranium-contaminated subsurface during biostimulation with the addition of electron donors, such as ethanol (North et al., 2004 ; Akob et al., 2008 ; Mohanty et al., 2008 ; Burkhardt et al., 2010 ; Sitte et al., 2010 ; Vishnivetskaya et al., 2010 ; Van Nostrand et al., 2011 ). The majority of cultivation-independent studies in subsurface sediments were conducted at the DNA level with SSU rRNA gene targets. In this study, the results of the phylogenetic analysis of gltA mRNA sequences demonstrated that members of the Geobacteraceae are abundant and metabolically active in biostimulated subsurface sediments. Further, our results suggest that members of the subsurface Geobacter clade, closely related to G. uraniireducens and G. daltonii , are metabolically active iron reducers that mediate metal reduction in OR-IFRC subsurface sediments. G. uraniireducens and G. daltonii were isolated from U(VI)-contaminated subsurface environments at the Rifle and OR-IFRC sites, respectively (Shelobolina et al., 2008 ; Prakash et al., 2010 ), and these two species cluster with the phylogenetically coherent subsurface clade proposed by Holmes et al. ( 2007 ). Although these two Geobacter strains share 98.1% 16S rRNA gene sequence identity, their full genome sequences are highly divergent (Prakash et al., 2010 ). Limited physiological screening has begun to reveal substantial differences in electron acceptor and donor utilization within the subsurface clade of Geobacter . For example, G. daltonii and G. toluenxydans conserve energy for growth with aromatic contaminants as the electron donor, while G. uraniireducens does not (Kunapuli et al., 2010 ; Prakash et al., 2010 ). This may be explained by the fact that both G. daltonii and G. toluenoxydans were isolated from subsurface sediments contaminated with aromatic hydrocarbons, whereas G. uraniireducens was isolated from groundwater that was not substantially impacted by organic contaminants. A number of additional features may provide a competitive advantage to Geobacter in the subsurface including the ability to utilize acetate, chemotaxis, and nitrogen fixation (Childers et al., 2002 ; Holmes et al., 2004 ). Further quantitative analysis of glt A transcripts has the potential to aid in our understanding of the environmental controls of Geobacteraceae -mediated metal reduction. Primers designed for specific strains will likely reveal niche differentiation within the Geobacter family. The active SRB observed during biostimulation and in intact core samples were related to members of the Desulfovibrionaceae and Desulfobacteraceae families within the Deltaproteobacteria and the Clostridia within the Firmicutes , and no sulfate-reducing Archaea were detected. This observation fits with previous studies at uranium-contaminated sites that correlated the activity of these organisms with the addition of carbon substrates (Chang et al., 2005 ; Akob et al., 2008 ; Burkhardt et al., 2010 ; Cardenas et al., 2010 ; Sitte et al., 2010 ; Vishnivetskaya et al., 2010 ; Miletto et al., 2011 ). The utilization of supplemental electron donors in our microcosms also fits with the known metabolism of the SRB detected, as ethanol is incompletely oxidized to acetate by Desulfovibrio and Desulfotomaculum species (Muyzer and Stams, 2008 ). In addition, both Desulfovibrio and Desulfotomaculum are known to enzymatically reduce U(VI) (Lovley and Phillips, 1992 ; Lovley et al., 1993a ; Tebo and Obraztsova, 1998 ). This suggests that the activity of these organisms was directly related to the observed decrease in soluble uranium concentrations as was seen in earlier work (Akob et al., 2008 ; Cardenas et al., 2010 ; Vishnivetskaya et al., 2010 ; Van Nostrand et al., 2011 ). In environments co-contaminated with sulfate and nitrate such as the OR-IFRC, stimulation of Desulfovibrio may have a high bioremediation potential since the presence of sulfate represses nitrate reduction in this organism (Marietou et al., 2009 ). While members of the Desulfobacteraceae are known for acetate utilization and the complete oxidation of carbon substrates (Muyzer and Stams, 2008 ), these organisms have not been shown to reduce U(VI). Thus, it follows that we observed enhanced U(VI) immobilization under sulfate-reducing conditions when ethanol was used as the electron donor, whereas the linkage was not as strong when acetate was used."
} | 6,529 |
26786050 | PMC4719670 | pmc | 3,954 | {
"abstract": "Background To deal with the increasingly severe energy crisis and environmental consequences, biofuels and biochemicals generated from renewable resources could serve as a promising alternative for replacing petroleum as a source of fuel and chemicals, among which isoprene (2-methyl-1,3-butadiene) in particular is of great significance in that it is an important platform chemical, which has been used in industrial production of synthetic rubber for tires and coatings or aviation fuel. Results We firstly introduced fatty acid decarboxylase (OleT JE ) from Jeotgalicoccus species into E. coli to directly convert MVA(mevalonate) into 3-methy-3-buten-1-ol. And then to transform 3-methy-3-buten-1-ol to isoprene, oleate hydratase (OhyA EM ) from Elizabethkingia meningoseptica was overexpressed in E. coli . A novel biosynthetic pathway of isoprene in E. coli was established by co-expressing the heterologous mvaE gene encoding acetyl-CoA acetyltransferase/HMG-CoA reductase and mvaS gene encoding HMG-CoA synthase from Enterococcus faecalis , fatty acid decarboxylase (OleT JE ) and oleate hydratase (OhyA EM ). Furthermore, to enhance isoprene production, a further optimization of expression level of OleT JE , OhyA EM was carried out by using different promoters and copy numbers of plasmids. Thereafter, the fermentation process was also optimized to improve the production of isoprene. The final engineered strain, YJM33, bearing the innovative biosynthetic pathway of isoprene, was found to produce isoprene up to 2.2 mg/L and 620 mg/L under flask and fed-batch fermentation conditions, respectively. Conclusions In this study, by using metabolic engineering techniques, the novel MVA-mediated biosynthetic pathway of isoprene was successfully assembled in E. coli BL21(DE3) with the heterologous MVA upper pathway, OleT JE from Jeotgalicoccus species and OhyA EM from Elizabethkingia meningoseptica . Compared with traditional MVA pathway, the novel pathway is shortened by 3 steps. In addition, this is the first report on the reaction of converting MVA into 3-methy-3-buten-1-ol by fatty acid decarboxylase (OleT JE ) from Jeotgalicoccus species. In brief, this study provided an alternative method for isoprene biosynthesis, which is largely different from the well-developed MEP pathway or MVA pathway. Electronic supplementary material The online version of this article (doi:10.1186/s12896-016-0236-2) contains supplementary material, which is available to authorized users.",
"conclusion": "Conclusions In this paper, isoprene was synthesized through a distinctive biosynthetic pathway harboring the MVA upper pathway from Enterococcus faecalis , the oleT JE gene from Jeotgalicoccus species and the ohyA EM gene from Elizabethkingia meningoseptica in an engineered E. coli strain. The most optimized strain, YJM33, bearing the novel MVA-mediated biosynthetic pathway of isoprene, accumulated isoprene up to 2.2 mg/L and 620 mg/L under conditions of flask fermentation and fed-batch fermentation, respectively. Despite the relatively low level of isoprene production by this novel pathway, we have reduced the complexity of the native isoprene pathway by introducing two novel enzymes to catalyze the formation of isoprene from mevalonate in only 2 steps instead of 5. To our knowledge, this is the first report of the conversion of MVA into 3-methyl-3-buten-1-ol by fatty acid decarboxylase (OleT JE ) from Jeotgalicoccus species, and it is also the first to describe the catalysis of MVA to isoprene with simultaneous heterologous expression of the oleT JE gene from Jeotgalicoccus species and the ohyA EM gene from Elizabethkingia meningoseptica . Therefore, this study supplies an unusual synthetic route for bio-isoprene production that is very different from the well-characterized MEP pathway or MVA pathway.",
"discussion": "Results and discussion Overexpression and functional analysis of OleT JE The function of the fatty acid decarboxylase (OleT JE ) from Jeotgalicoccus sp. ATCC 8456 to decarboxylate long-chain fatty acids into their corresponding terminal olefins has been previously demonstrated [ 17 ]. In this study, we determined whether OleT JE could directly catalyze the MVA decarboxylation reaction. The nucleotide sequence of the fatty acid decarboxylase (OleT JE ) generated from Jeotgalicoccus sp. ATCC 8456 was introduced into the plasmid pCOLADUet-1. The recombinant OleT JE protein carrying a N-terminal six-histidine tag was purified from E. coli , and identified by SDS-PAGE (Fig. 2b ). The enzyme activity was measured in a gas chromatography vial and a 3-methyl-3-buten-1-ol specific peak was detected by GC-MS (Fig. 2a ). No detectable 3-methyl-3-buten-1-ol was formed when the purified enzyme or MVA was omitted from the assay. These results indicated that the isolated recombinant protein possessed MVA decarboxylase activity and was able to convert MVA to 3-methyl-3-buten-1-ol. To our knowledge, this reaction has not been previously documented. Fig. 2 Enzymatic assay for 3-methyl-3-buten-1-ol production by OleT JE using GC-MS and SDS-PAGE analysis. a GC-MS analysis of a 3-methyl-3-buten-1-ol sample produced by the OleT JE assay mixtures. b SDS-PAGE analysis of OleT JE . CK: cell lysate from BL21(DE3) containing pCOLADuet-1. 1: crude cell extracts from YJM30. 2: purified OleT JE \n In the native MVA pathway, MVA is phosphorylated twice and decarboxylated to form IPP. This process requires three enzymes, including mevalonate kinase, phosphomevalonate kinase and mevalonate diphosphate decarboxylase [ 18 ]. Then, using the enzyme pyrophosphatase or phosphatase, IPP can be converted into 3-methyl-3-buten-1-ol by removing the pyrophosphates [ 19 ]. To shorten the reaction steps of this pathway, we selected the enzyme fatty acid decarboxylase (OleT JE ) from Jeotgalicoccus species, which has the ability to directly decarboxylate MVA into 3-methyl-3-buten-1-ol using only one step. The experimental results showed that the reaction catalyzed by the OleT JE enzyme shortened the pathway and was able to convert MVA to 3-methyl-3-buten-1-ol in only one step without phosphorylation. To our knowledge, this is the first report of the above-mentioned reaction. Overexpression and functional analysis of OhyA EM The capacity for cells containing oleate hydratase to transform oleic acid (OA) into 10-hydroxystearic acid (10-HSA) was first characterized by Wallen et al . in Pseudomonas sp. strain 3266 in 1962 [ 20 ]. Niehaus then showed that the reaction was reversible [ 21 ]. However, only in recent years was the gene encoding oleate hydratase in Elizabethkingia meningoseptica (formerly Pseudomonas sp. ) cloned and expressed in E. coli [ 22 ]. Marliere demonstrated that oleate hydratase has the ability to catalyze the dehydration of isobutanol to form isobutene [ 23 ]. According to the above-referenced studies, the enzyme oleate hydratase can dehydrate 3-methyl-3-buten-1-ol into isoprene. In our study, the nucleotide sequence of the ohyA EM gene from Elizabethkingia meningoseptica was altered based on the preferred codon usage of E. coli and subsequently cloned into the vector pCOLADUet-1. The protein was expressed in E. coli BL21 (DE3) and purified using a nickel-affinity chromatography column. The band of the recombinant protein was observable on coomassie-stained SDS-PAGE gel of the crude cell extracts (Fig. 3b ). The enzyme assay was conducted in a gas chromatography vial with GC-MS being used to verify an isoprene-specific peak (Fig. 3a ). No detectable isoprene was produced when the purified enzyme or 3-methyl-3-buten-1-ol was omitted from the assay. The results suggested that the enzyme OhyA EM from Elizabethkingia meningoseptica is capable of catalyzing the dehydroxylation of 3-methyl-3-buten-1-ol into isoprene. Fig. 3 Enzymatic assay for isoprene production by OhyA EM using GC-MS and SDS-PAGE analysis. a GC-MS analysis of a isoprene sample produced by the OhyA EM assay mixtures. b SDS-PAGE analysis of OhyA EM . CK: cell lysate from BL21(DE3) containing pCOLADuet-1. 1: crude cell extracts from YJM31. 2: purified OhyA EM \n Establishing a novel biosynthetic pathway for isoprene in engineered E. coli In previous experiments, the engineered strain YJM16 containing the efficient MVA upper pathway from Enterococcus faecalis was constructed, resulting in the accumulation of up to 1.31 g/L of MVA under flask culture conditions [ 9 ]. To subsequently obtain isoprene from glucose, we transformed the plasmid pYJM34 carrying the oleT JE gene from Jeotgalicoccus species and ohyA EM gene from Elizabethkingia meningoseptica into strain YJM16 harboring the MVA upper pathway. The resulting engineered strain YJM32 was inoculated in 50 ml fermentation medium and cultured at 37 °C which was further cultivated at 30 °C for 36 h with 0.5 mM IPTG addition into the broth when the OD 600 attained about 0.6. The isoprene production by the strain YJM32 reached 17.6 μg/L. While the control engineered strain YJM35 only harboring upper MVA pathway and the fatty acid decarboxylase (OleT JE ) cannot generate the isoprene. The results proved that a novel biosynthetic pathway for isoprene production containing the MVA upper pathway from Enterococcus faecalis , the oleT JE gene from Jeotgalicoccus species and ohyA EM gene from Elizabethkingia meningoseptica had been successfully constructed in E. coli . In the previous studies, several research groups, including ours, established pathways for the conversion of MVA to isoprene. This process typically requires five reactions, including a two-step phosphorylation catalyzed by mevalonate kinase and phosphomevalonate kinase, a one-step decarboxylation catalyzed by mevalonate 5-diphosphate decarboxylase, a one-step isomerization catalyzed by IPP isomerase and a one-step dephosphorylation catalyzed by isoprene synthase [ 7 , 9 , 24 ]. This study is the first to use only two-step reactions to construct a new pathway for the conversion of MVA to isoprene by combining the oleT JE gene from Jeotgalicoccus species and ohyA EM gene from Elizabethkingia meningoseptica . Accordingly, from the starting acetyl-CoA to the final product isoprene, the entire pathway containing eight reactions was shortened to five reactions. The result is a promising step in the novel MVA-mediated biosynthetic pathway for isoprene production. Optimization of a biosynthetic pathway for isoprene production To further enhance isoprene production, the expression levels of the oleT JE gene from Jeotgalicoccus species and the ohyA EM gene from Elizabethkingia meningoseptica were optimized by using different plasmid vectors containing different copy numbers and promoters. As is shown in Fig. 4 , there achieved more isoprene production of the oleT JE gene and ohyA EM gene when under the control of the T7 promoter (YJM32) than that of the ara BAD promoter (YJM34). The highest isoprene production (52.2 μg/L) was found in the strain YJM33 harboring a high copy number plasmid, which was three-fold greater than the production of isoprene by YJM32 using lower copy number plasmids. Fig. 4 Optimization of the expression levels of oleT \n JE and ohyA \n EM . The expression of oleT \n JE aand ohyA \n EM under the control of T7 promoter (YJM32) achieved much higher isoprene production than when the gene were under the control of the ara BAD promoter (YJM34). The strain YJM33 using a high copy number plasmid (pET-28a(+)) achieved the highest isoprene production (52.2 μg/L). The experiment was conducted in triplicate Enhance isoprene production through optimizing the culture conditions In our work, the “one-factor at-a-time” optimization strategy (Additional file 1 : Fig. S1) was applied to augment isoprene productivity by optimizing the organic nitrogen source, induction temperature and IPTG concentration respectively. The results showed that the highest isoprene yield (2.2 mg/L) was obtained when the YJM33 strain was cultured in fermentation medium containing 20 g/L glucose as a carbon source, 9 g/L beef powder as an organic nitrogen source and induced with 0.25 mM IPTG at 31 °C whose combined optimization effect could contribute to an approximately 42-fold increase in isoprene production. Microbial isoprene production using the novel biosynthetic pathway To assess the isoprene biosynthesis in a scalable process using the engineered strain YJM33 with the novel biosynthetic pathway encoded on the plasmids pYJM16 (pACY -mvaE - mvaS ) and pYJM35 (pET28- OleT JE - OhyA EM ), the fermentation of YJM33 under fed-batch condition was conducted on a 5-L scale. At an OD 600 of ~12, 0.25 mM IPTG was put into the broth to induce the heterologous genes of the pathway for expression. After depleting the glucose initially present in the media, glucose solution (800 g l −1 ) was added to the cultures, and the residual glucose was restrained below 0.5 g/l to reduce acetate production. The OD 600 at the end of the fermentation was ~36. As is shown in Fig. 5 , isoprene gradually accumulated over the course of the fermentation and amounted to 620 mg/L with a productivity of 6.46 mg/L/h within 32 h (Fig. 5 ). In addition, isoprene production rose dramatically from 4 h to 16 h after being induced, and the productivity of isoprene attained 8.76 mg/L/h. Fig. 5 The time course of isoprene production by YJM33. Cell growth (▲) and isoprene accumulation (■) in YJM33. Cells were induced when the OD 600 reached approximately 12 at 31 °C. Other experimental conditions are described in the section entitled “Fed-Batch Fermentation” Although much progress has been made regarding the novel biosynthetic pathway of isoprene in E. coli , the present productivity remains too low and it is economically unfeasible for large scale production. The reason for the low yield could be the low catalytic activity of the enzymes OleT JE and OhyA EM . Future studies should focus on enhancing the efficiency of the novel pathway using the following approaches: (1) The structures of both enzymes should be elucidated. Based on the structural data, it may be possible to enhance the catalytic efficiency by mutating key amino acids in the binding sites and catalytic active sites [ 25 , 26 ] or to increase enzyme expression level by optimizing the Shine-Dalgarno sequence of enzyme [ 27 ]. (2) The pathway might be dynamically regulated using the dynamic sensor-regulator system (DSRS) developed by Keasling to bio-synthesize fatty acid-based products in E. coli [ 28 ]. The DSRS utilizes a transcription factor which can sense a crucial intermediate and dynamically regulate the expression level of genes related to target product synthesis. Consequently, if the natural sensor for crucial intermediate of MVA can be found, we can develop a DSRS for isoprene production to equilibrate metabolic pathway, thus enhancing product concentration, conversion efficiency and host’s genetic stability."
} | 3,758 |
32707169 | null | s2 | 3,955 | {
"abstract": "1-octanol is a valuable molecule in the chemical industry, where it is used as a plasticizer, as a precursor in the production of linear low-density polyethylene (LLDPE), and as a growth inhibitor of tobacco plant suckers. Due to the low availability of eight-carbon acyl chains in natural lipid feedstocks and the selectivity challenges in petrochemical routes to medium-chain fatty alcohols,1-octanol sells for the highest price among the fatty alcohol products. As an alternative, metabolic engineers have pursued sustainable 1-octanol production via engineered microbes. Here, we report demonstration of gram per liter titers in the model bacterium Escherichia coli via the development of a pathway composed of a thioesterase, an acyl-CoA synthetase, and an acyl-CoA reductase. In addition, the impact of deleting fermentative pathways was explored E. coli K12 MG1655 strain for production of octanoic acid, a key octanol precursor. In order to overcome metabolic flux barriers, bioprospecting experiments were performed to identify acyl-CoA synthetases with high activity towards octanoic acid and acyl-CoA reductases with high activity to produce 1-octanol from octanoyl-CoA. Titration of expression of key pathway enzymes was performed and a strain with the full pathway integrated on the chromosome was created. The final strain produced 1-octanol at 1.3 g/L titer and a >90% C"
} | 346 |
36258069 | null | s2 | 3,956 | {
"abstract": "The cyclic di-GMP (c-di-GMP) second messenger represents a signaling system that regulates many bacterial behaviors and is of key importance for driving the lifestyle switch between motile loner cells and biofilm formers. This review provides an up-to-date summary of c-di-GMP pathways connected to biofilm formation by the opportunistic pathogen P. aeruginosa. Emphasis will be on the timing of c-di-GMP production over the course of biofilm formation, to highlight non-uniform and hierarchical increases in c-di-GMP levels, as well as biofilm growth conditions that do not conform with our current model of c-di-GMP."
} | 154 |
36364411 | PMC9654057 | pmc | 3,957 | {
"abstract": "Microbial biodiversity includes biotic and abiotic components that support all life forms by adapting to environmental conditions. Climate change, pollution, human activity, and natural calamities affect microbial biodiversity. Microbes have diverse growth conditions, physiology, and metabolism. Bacteria use signaling systems such as quorum sensing (QS) to regulate cellular interactions via small chemical signaling molecules which also help with adaptation under undesirable survival conditions. Proteobacteria use acyl-homoserine lactone (AHL) molecules as autoinducers to sense population density and modulate gene expression. The LuxI-type enzymes synthesize AHL molecules, while the LuxR-type proteins (AHL transcriptional regulators) bind to AHLs to regulate QS-dependent gene expression. Diverse AHLs have been identified, and the diversity extends to AHL synthases and AHL receptors. This review comprehensively explains the molecular diversity of AHL signaling components of Pseudomonas aeruginosa , Chromobacterium violaceum , Agrobacterium tumefaciens , and Escherichia coli . The regulatory mechanism of AHL signaling is also highlighted in this review, which adds to the current understanding of AHL signaling in Gram-negative bacteria. We summarize molecular diversity among well-studied QS systems and recent advances in the role of QS proteins in bacterial cellular signaling pathways. This review describes AHL-dependent QS details in bacteria that can be employed to understand their features, improve environmental adaptation, and develop broad biomolecule-based biotechnological applications.",
"conclusion": "5. Conclusion and the Future Prospects AHL QS is necessary for bacteria adaptation, cellular growth, cell adhesion, biofilm development, cell division, antibiotic resistance, plasmid conjugation, and virulence gene expression in Gram-negative bacteria. Biological molecules such as AHL signals, AHL synthase, and AHL receptors have diverse structural and functional diversity across bacteria. The heterogeneity of the AHL-QS system may facilitate the specificity of intracellular and intercellular bacterial signaling systems. An understanding of molecular mechanisms and structural insights, particularly conformational dynamics, is crucial for understanding AHL-QS systems. Quantifying the conformational flexibility of proteins with and without ligands and inhibitors may reveal new insights into regulatory mechanisms. Currently, the role of confirmational dynamics and protein function is poorly understood. AHL transcriptional regulators can be employed as a model system to explore conformational change and its implications on gene expression. Furthermore, the findings of our structural analysis might be relevant in developing site-specific or allosteric inhibitors. Further research is required to comprehend how the AHL-QS system interacts with plant or animal host cells. The AHL-QS system could help researchers better understand the fundamental regulatory framework for gene expression in prokaryotes and aid in developing novel anti-virulence approaches as a future antimicrobial strategy and adaptation.",
"introduction": "1. Introduction Quorum sensing (QS) is a biochemical communication system in bacteria that directs social interactions based on population density [ 1 , 2 ]. Bacteria produce extracellular signals early in their adaptation and growth phase; as the cell density increases, the signals accumulate in the environment [ 3 ]. Once a critical cell density threshold is reached, these signals interact with their cognate receptors and coordinate the expression of associated genes [ 4 ]. In Gram-positive and Gram-negative bacteria, several QS systems have been identified [ 5 , 6 ], including acyl-homoserine lactone (AHL)-based and peptide-based QS systems. In addition, many proteobacteria use the AHL-dependent QS system to regulate population-dependent behaviors, including plasmid conjugation, pigment production, virulence factor expression, and biofilm formation [ 7 ]. In the late 1960s and early 1970s, AHL-QS was first discovered in a marine bacterium called Vibrio fisheri [ 8 ]. As part of its symbiotic relationship with the Hawaiian bobtail squid, Euprymna scolopes, V. fisheri regulates its bioluminescence in a population density-dependent manner. In the V. fisheri QS system, autoinducers (AIs) and two essential proteins (LuxR and LuxI) regulate bioluminescence signal production. The LuxI enzyme produces the QS signal molecule (the N-3-oxo-hexanoyl-homoserine lactone), and LuxR is a transcription regulator that binds to the AHL signal and regulates bioluminescence via the luxCDABEG operon [ 9 ]. Following the discovery of the AHL-QS system, researchers discovered that AHL-QS is prevalent in a wide range of Proteobacteria and influences vital characteristics such as motility [ 10 ], pigment production [ 11 ], and biofilm formation [ 12 ]. Our molecular understanding of the QS system has been useful in studying plant–microbe, animal–microbe, and human–microbe interactions. The AHL molecule comprises a homoserine lactone ring and an acyl chain ranging from C4 to C18 in length [ 13 ]. Depending upon the organisms, the AHL molecules may have 3-hydroxy, 3-oxo, methyl, or varying unsaturation in some cases [ 14 ]. The first essential part of AHL signaling is the LuxI-type AHL synthases, which are involved in producing AHL molecules. Once the AHL molecules are generated, they can be transported passively and actively in and out of the cells [ 15 ]. The second component of AHL signaling is the LuxR-type AHL receptor proteins, which bind to the AHL signal molecule and govern the QS-dependent activation, repression, and depression of target genes [ 16 , 17 ]. The LuxR-type DNA-binding transcription regulators govern QS-dependent gene expression. Several reviews have summarized the molecular mechanism and phenotype regulation of AHL-QS systems. Still, there are few detailed structural and functional investigations of AHL-QS-associated proteins at the molecular level. Churchill et al. (2011) reviewed AHL signaling in exceptional detail and evaluated the molecular basis of AHL signaling in bacteria [ 18 ]. However, recent advancements in AHL-QS and the availability of new crystal structures for the AHL regulators require an updated review of the AHL-QS system. In this review, we provide a detailed overview of the AHL-QS system, the mechanism of AHL signal biomolecule synthesis, structural and functional analysis of the AHL synthase, and AHL receptors of Gram-negative bacteria, including Pseudomonas aeruginosa , Chromobacterium violaceum , Agrobacterium tumefaciens , and Escherichia coli . AHL signaling has been discovered in a wide range of bacteria, indicating that it is a widely accepted chemical language among Gram-negative bacteria for their adaptation and survival. Furthermore, biomolecules derived from bacteria can potentially be used for various applications, such as human and plant health, biofouling, bioremediation, biosensors, and cancer."
} | 1,762 |
30140114 | null | s2 | 3,958 | {
"abstract": "The future supply of platinum group metals (PGM) is at risk because of their scarcity combined with a high demand. Thus recovery of platinum (Pt) from waste is an option worthy of study to help alleviate future shortages. This research explored the microbial reduction of platinum (Pt). The ability of anaerobic granular sludge to reduce Pt(IV) ions under different physiological conditions was studied. X-Ray diffraction (XRD) and transmission electron microscope (TEM) analyses demonstrated the capacity of the microbial mixed culture to reduce Pt(IV) to Pt(0) nanoparticles, which were deposited on the cell-surface and in the periplasmic space. Ethanol supported the biologically catalyzed Pt(IV) reduction, meanwhile other electron donors; hydrogen (H This study reported for the first time the ability of an anaerobic granular sludge to reduce Pt(IV) to elemental Pt(0) nanoparticles."
} | 222 |
25192405 | PMC4156329 | pmc | 3,959 | {
"abstract": "Symbioses with the dinoflagellate Symbiodinium form the foundation of tropical coral reef communities. Symbiodinium photosynthesis fuels the growth of an array of marine invertebrates, including cnidarians such as scleractinian corals and octocorals (e.g., gorgonian and soft corals). Studies examining the symbioses between Caribbean gorgonian corals and Symbiodinium are sparse, even though gorgonian corals blanket the landscape of Caribbean coral reefs. The objective of this study was to compare photosynthetic characteristics of Symbiodinium in four common Caribbean gorgonian species: Pterogorgia anceps, Eunicea tourneforti, Pseudoplexaura porosa, and Pseudoplexaura wagenaari. Symbiodinium associated with these four species exhibited differences in Symbiodinium density, chlorophyll a per cell, light absorption by chlorophyll a, and rates of photosynthetic oxygen production. The two Pseudoplexaura species had higher Symbiodinium densities and chlorophyll a per Symbiodinium cell but lower chlorophyll a specific absorption compared to P. anceps and E. tourneforti. Consequently, P. porosa and P. wagenaari had the highest average photosynthetic rates per cm 2 but the lowest average photosynthetic rates per Symbiodinium cell or chlorophyll a. With the exception of Symbiodinium from E. tourneforti, isolated Symbiodinium did not photosynthesize at the same rate as Symbiodinium in hospite. Differences in Symbiodinium photosynthetic performance could not be attributed to Symbiodinium type. All P. anceps (n = 9) and P. wagenaari (n = 6) colonies, in addition to one E. tourneforti and three P. porosa colonies, associated with Symbiodinium type B1. The B1 Symbiodinium from these four gorgonian species did not cluster with lineages of B1 Symbiodinium from scleractinian corals. The remaining eight E. tourneforti colonies harbored Symbiodinium type B1L, while six P. porosa colonies harbored type B1i. Understanding the symbioses between gorgonian corals and Symbiodinium will aid in deciphering why gorgonian corals dominate many Caribbean reefs.",
"introduction": "Introduction Gorgonian corals (subclass Octocorallia) are abundant and important members of coral reef communities throughout the Caribbean [1] – [7] . Unlike the dramatic decline of scleractinian corals in the Caribbean [8] – [11] , gorgonian coral abundance is steady or even increasing [7] , [12] , [13] . For example, in the Florida Keys, gorgonian octocoral abundance increased significantly since 1999 [7] . And, in the US Virgin Islands, the abundance of two of three gorgonian species studied has increased since 1992 [13] . Gorgonian corals are also abundant in the Yucatan coast of México, where gorgonian species richness can exceed scleractinian coral species richness [14] . Despite being prominent members of Caribbean reef communities, studies on gorgonian coral physiology are sparse, predominantly focusing on the gorgonian coral hosts without addressing their symbiosis with the unicellular dinoflagellates in the genus Symbiodinium . Studies tracked digested material [15] , [16] , described sclerite (microscopic skeletal elements) formation [17] , [18] , or measured growth [19] , [20] , and feeding rates [21] , [22] . In addition, gorgonian secondary metabolites have been extensively studied due to their medical and economic importance [23] . Two early studies measured the photosynthetic rates of Symbiodinium in a few Caribbean gorgonian species [24] , [25] , while key photosynthetic characteristics, including light absorption efficiencies, have not been measured. Furthermore, the handful of studies that investigated the physiology of Caribbean gorgonian corals and their Symbiodinium [24] – [29] did not identify the Symbiodinium present. \n Symbiodinium are currently divided into nine phylogenetic clades, A-I [30] , although clade E may represent a single species [31] . Within each clade, Symbiodinium are often distinguished using sequences of the internal transcribed spacer regions of ribosomal DNA ( Symbiodinium types sensu [32] ). Almost all Caribbean gorgonian species associate with Symbiodinium clade B types and many harbor type B1 [33] , [34] . Within type B1, multiple lineages have been identified [35] , [36] . Hosting different Symbiodinium types can correlate with ecological [36] , [37] and physiological differences [38] between cnidarian hosts. Often, however, physiological differences are assessed when the cnidarians face stressful environmental conditions [38] . Collecting baseline physiological data on coral-algal symbioses [39] , and not just data when the symbioses are stressed, are critical to evaluating the effects of stressors on symbioses [40] . The objective of this study was to characterize the photosynthesis of Symbiodinium, in hospite and in isolation, in four common Caribbean gorgonian species: Pterogorgia anceps , Eunicea tourneforti, Pseudoplexaura porosa , and Pseudoplexaura wagenaari . Studying the physiology of the symbiosis between gorgonian corals and Symbiodinium may shed light on why gorgonian corals dominate Caribbean reefs while scleractinian coral abundance is declining.",
"discussion": "Discussion Gorgonian corals blanket the landscape of Caribbean coral reefs [1] , [2] , [7] , yet few data exist on the physiology of their mutualism with Symbiodinium \n [24] – [29] . Unlike scleractinian corals, whose abundance has dramatically declined in the Caribbean [7] – [11] , gorgonians have withstood recent environmental perturbations. Learning about gorgonian symbioses, at current ambient conditions, enhances our understanding and our ability to predict the future of Caribbean coral reefs. We therefore employed multiple tools to characterize aspects of Symbiodinium photosynthesis in four common Caribbean gorgonian species. Oxygen flux data demonstrated differences in the Symbiodinium in the four studied gorgonians. For example, the two Pseudoplexaura species had the highest average photosynthetic rates per cm 2 , probably due to the higher Symbiodinium and chlorophyll densities compared to P. anceps and E. tourneforti . On the other hand, the two Pseudoplexaura species had lower initial slopes of the P-E curves, low photosynthetic rates per Symbiodinium cell and per chlorophyll a , and lower chlorophyll a specific absorption. Taken together, these data suggest that Symbiodinium in P. porosa and P. wagenaari are less efficient in light absorbtion and utilization than Symbiodinium in P. anceps and E. tourneforti . The light levels available for Symbiodinium could differ due to Symbiodinium self-shading or to host tissue characteristics such as tissue thickness [53] , [54] or pigmentation. The possibility that Symbiodinium in P. porosa and P. wagenaari are less efficient in light absorbtion and utilization is corroborated by looking at changes in chlorophyll a specific absorption coefficient as a function of chlorophyll a density. In the two Pseudoplexaura species, the a * chl a values as a function of chlorophyll a density were very low, comparable to those values reported for phytoplankton and freshly isolated Symbiodinium from Porites banneri \n [45] , [55] . On the other hand, E. tourneforti a * chl a values fell within those previously reported for scleractinian corals. Lastly, the a * chl a values of Symbiodinium in P. anceps demonstrated a very high pigment light absorption efficiency, comparable to that of Symbiodinium in the scleractinian coral Porites banneri \n [45] . Of the four gorgonian species, Symbiodinium in P. anceps exhibited twice the photosynthetic rate per Symbiodinium cell than that of the next gorgonian species ( P. porosa ) and the highest average photosynthetic rates per chlorophyll a . The relatively high photosynthesis per Symbiodinium in P. anceps may be related to the low density of Symbiodinium and chlorophyll a per cm 2 . In addition, the thin, angular branches and low polyp density may aid Symbiodinium photosynthesis in P. anceps by maximizing gas exchange and/or reducing self-shading of Symbiodinium . The four Caribbean gorgonian species produced comparable photosynthetic and respiration rates per cm 2 to the average rates of eight shallow scleractinian coral species (P = 2.0, R = 0.64) [39] . Conversely, the Mediterranean gorgonian Eunicella singularis at 15 m depth had lower average Symbiodinium photosynthetic (∼1) and respiration (∼0.55) rates per cm 2 \n [56] . The differences could be due to E. singularis in deeper waters being exposed to lower irradiance levels than those in the current study [56] , [57] . In our study, the four gorgonian species did not exhibit photoinhibition, similar to what has been observed in other symbioses between Symbiodinium and cnidarians [58] . The lack of photoinhibition may be due to branch tissue thickness, as was seen in the octocoral Capnella gaboensis \n [53] , [59] . The gross P/R ratios in the four gorgonian symbioses, ranging from 2 to 4, were comparable to ratios for other octocorals [60] – [63] , anemones [52] , and shallow water scleractinian corals [39] . On the other hand, the ratios of diurnal integrated gross photosynthesis to respiration were below 1. It remains to be determined the extent of the contribution of the Symbiodinium autotrophic production to the energy budget of these Caribbean gorgonians. \n Symbiodinium photosynthesis within a host may also be affected by host characteristics [54] , [64] . For example, the scleractinian coral skeleton enhances light absorption by Symbiodinium \n [45] , [65] , and Symbiodinium chlorophyll a specific absorption is higher in symbiosis than in isolation [45] . Chlorophyll a specific absorption of Symbiodinium in the studied gorgonian corals was comparable to that in scleractinian corals [45] , [54] , even though gorgonian skeletal structure (sclerites and an axial rod) substantially differs from the calyx structure in scleractinian corals. The calcite sclerites within gorgonian tissues may produce the same effect as the light scattering of the scleractinian coral skeleton, perhaps similar to the influence of the siliceous spicules of sponges on light transmission [66] , [67] . Furthermore, in symbioses between cnidarians and Symbiodinium, photosynthesis is dependent upon the genetic identities of both the host and symbiont [52] . Therefore, it is imperative to identify the Symbiodinium. In the Caribbean, scleractinian corals host Symbiodinium belonging to clades A, B, C, or D [34] , [36] . On the other hand, the majority of Caribbean gorgonian species host only clade B Symbiodinium [33] , [68] . The four gorgonian species in this study were no exception; harboring clade B Symbiodinium belonging to types B1, B1i, and B1L. Types B1i and B1L have only been reported in Caribbean gorgonian species [36] , [69] . Sequencing of the B1 Symbiodinium from the four gorgonian species revealed that they harbor B1 Symbiodinium with distinct microsatellite flanking region sequences from those in scleractinian coral species. The B1 Symbiodinium obtained from most gorgonian colonies in our study formed a distinct group amongst the previously identified B1 lineages from scleractinian corals ( Figure 4 ). Although both P. anceps and P . wagenaari hosted type B1 Symbiodinium, the photosynthetic characteristics differed between these two symbioses. Photosynthetic variability within Symbiodinium type B1 has also been observed in cultures [70] , and may be associated with distinct genetic lineages within type B1 [35] , [36] . In our study, eight of the nine P. anceps colonies, and all P. wagenaari colonies, harbored symbionts from the same, highly supported, phylogenetic group within B1 Symbiodinium . Therefore, the observed photosynthetic variability between P. anceps and P . wagenaari was not due to different B1 lineages but probably due to the physiology of different host-symbiont combinations [52] . In E. tourneforti , Symbiodinium had comparable photosynthetic rates per cell in the intact symbiosis and in isolation. On the other hand, maximum photosynthetic rates per cell in Symbiodinium isolated from P. anceps, P. porosa , and P. wagenaari were lower than those measured in hospite , although the average α was higher in isolation. Diminished photosynthetic rates in isolated Symbiodinium cells may occur in the absence of host carbon concentrating mechanisms [71] or differences in carbonic anhydrase activity among Symbiodinium types [72] . Exposure to the ionic environment of seawater [73] , [74] or bacteria [75] may also reduce photosynthesis in isolated Symbiodinium. In the sea anemone Aiptasia pallida , at ambient temperatures, the photosynthetic rates of isolated Symbiodinium were also lower compared to the intact association [52] , although not to the degree measured here. Conversely, secondary metabolites released from homogenized gorgonian corals may impair isolated Symbiodinium, as reported for the homogenate of the soft coral C. gaboensis that lysed Symbiodinium cells [76] . Therefore, secondary metabolites produced by many gorgonians may limit the utility of investigating freshly isolated Symbiodinium. In conclusion, our results contribute to consequential data on Symbiodinium physiological performance in their mutualism with four Caribbean gorgonian species at ambient temperature. Given that gorgonian corals are either maintaining or increasing their abundance on Caribbean coral reefs, understanding aspects of their symbiosis is imperative to understanding the future of Caribbean coral reefs. This study demonstrates differences between Symbiodinium photosynthetic characteristics in the four gorgonian species, collected from the same site, maintained under identical conditions, and with two of the gorgonian species containing the same Symbiodinium type. The differences observed between the gorgonian symbioses emphasize the influence of the host physiology and architecture on Symbiodinium photosynthesis."
} | 3,545 |
40169382 | PMC11983688 | pmc | 3,960 | {
"abstract": "Abstract Fungal fairy rings (FFRs) significantly influence plant communities and soil microbiota. This study investigated the development of Agaricus urinascens fairy rings in a species-rich montane Mediterranean grassland. By combining vegetation analysis, soil chemistry measurements, and next-generation sequencing, we assessed fairy rings’ impact on soil properties, plants, fungi, and bacteria. Our findings reveal a fungal-driven transformation of biological communities, with significant variations across FFRs zones. At the fungal front (FF), plant biomass decreased slightly but increased more than threefold inside the ring (>1100 g m −2 ), favouring grasses like Brachypodium genuense over forbs. In addition, species richness dropped significantly in the FF (−40%) compared to surrounding grassland, particularly affecting perennials. Moreover, our findings reveal substantial alterations in soil properties at the FF, including a 534% increase in P₂O 5 , a 210% rise in electrical conductivity, and a 36% increase in soil hydrophobicity compared to the surrounding grassland. Clay content at the FF was nearly three times higher than outside the ring (162.8 versus 57.5 g kg −1 ), indicating potential structural modifications in the soil matrix. Organic carbon decreased by 10% in the FF, while the C/N ratio and cation exchange capacity dropped significantly. Distinct shifts in microbial composition were observed. Bacterial diversity declined at the FF, where Actinobacteria dominated (85%) and Proteobacteria dropped to 8%. Similarly, fungal diversity was lowest inside the ring but highest in the belt section, with Ascomycota reaching 97% at the FF. Certain taxa, such as Kribbella, Streptomyces, Trichoderma, Penicillium , and Dichotomopilus , coexisted with A. urinascens mycelium. Notably, hydrophobicity at the FF was linked to high calcium oxalate crystal coverage on fungal mycelium and plant roots. This may have accelerated root desiccation, ultimately leading to plant mortality. Overall, our findings provide strong evidence that fairy ring fungi act as ecosystem engineers, shaping the spatial patterns of biotic composition and diversity in Mediterranean grasslands.",
"introduction": "Introduction Temperate grasslands are biodiversity hotspots, with species richness maintained by a combination of abiotic disturbances and complex plant–plant and plant–microbe interactions, ranging from competition to facilitation (Habel et al. 2013 ). A clear example of a plant–microbe interaction regulating grassland diversity and functionality is that generated by fungal fairy rings (FFRs) (Shantz and Piemeisel 1917 , Edwards 1984 ). FFRs are caused by the radial expansion of basidiomycete fungi in the soil, producing concentric vegetation bands that typically alter the composition of plant communities (Edwards 1988 ). FFRs play a dual role in grassland ecosystems, acting both as natural regulators of plant dynamics and as potential stressors, depending on their impact on vegetation and soil properties. On the beneficial side, FFRs contribute to enhanced nutrient cycling, accelerating the decomposition of organic matter and making nutrients more available to plants (Van der Wal et al. 2013 ). This effect can stimulate plant growth, particularly in the belt zone, where higher nutrient availability and fungal metabolites can promote lush vegetation (Zotti et al. 2025 ). In pastures, these nutrient-enriched patches may even increase forage availability for herbivores (Duan et al. 2022a ). Additionally, FFRs increase biodiversity and habitat heterogeneity, introducing small-scale disturbance mosaics that create both resource-rich sites and gaps for plant colonization. These microsites enhance soil microbial diversity and promote ecosystem resilience (Yang et al. 2018 , Xu et al. 2023 ). However, FFRs can also have negative impacts, particularly in managed landscapes. Type-1 FFRs, characterized by vegetation dieback at the fungal front (FF), can cause hydrophobic soil conditions, leading to localized drought stress and reduced plant survival (Zotti et al. 2025 ). In agricultural and recreational grasslands, this can result in patchy vegetation, loss of forage, and invasion by undesirable weed species. Furthermore, certain fungi associated with FFRs can release phytotoxic compounds, inhibiting plant germination or root development (Salvatori et al. 2023 ). Given their dual nature, the ecological significance of fairy rings depends on the balance between their beneficial effects on nutrient dynamics and biodiversity versus their potential for plant mortality and soil degradation. In their early study, Shantz and Piemeisel ( 1917 ) categorized grassland fairy rings into three types: Type-1, characterized by a belt of dead vegetation accompanied by a contrasting parallel concentric strip of verdant plants. This pattern is commonly associated with Marasmius oreades (Cosby 1960 ) or fungi from the Agaricus genus (Halisky and Peterson 1970 ). Type-2 is identified by the presence of a single belt of grasses with enhanced growth, as observed in the rings formed by Calvatia cyathiformis (Shantz and Piemeisel 1917 ). Type-3 shows no apparent effect on the grass cover and is typically associated with the FFs of species like Macrolepiota procera and Tapinella atrotomentosa (Halisky and Peterson 1970 ). Numerous studies have attempted to explain the inhibition of plant growth upon the arrival of the FF, followed by subsequent stimulation as the front moves away. Type-1 FFRs have also been studied as phytopathogenic agents in gardens and golf courses (Fidanza et al. 2007 , Miller et al. 2012 ). The detrimental impact of Type-1 FFRs on plant communities has been linked to increased soil hydrophobicity (Gramss et al. 2005 , Fidanza et al. 2007 ), nutrient immobilization (Fisher 1977 ), direct pathogenic behaviour (Terashima et al. 2004 ), and the release of phytotoxic compounds such as cyanide (Blenis et al. 2004 , Caspar and Spiteller 2015 ). Concurrently, research efforts have focused on explaining the mechanisms driving plant community rearrangement following the passage of FFRs. These mechanisms include the disruption of preexisting pathogenic guilds, the release of nutrients from decomposing microbial biomass, the creation of empty niches (Bonanomi et al. 2012 , Zotti et al. 2020 ), and the production of plant hormones known as ‘fairy chemicals’ (Choi et al. 2010 ). Recently, an in-depth modelling study examined the development of various vegetation patterns associated with FFRs (Salvatori et al. 2023 ). This study used a process-based model to demonstrate that while the centrifugal movement of the fungal mycelium can be explained by the autotoxicity theory, linked to the accumulation of fungal self-DNA in the soil, the visible effects on plants might result from hydrophobicity or allelopathic toxicity conditions induced by the mycelial mass (Salvatori et al. 2023 ). While the impact of FFRs on soil chemistry is well-established, showing reductions in soil organic carbon, substantial decreases in soil pH, and changes in key nutrient stocks (Edwards 1988 , Gramss et al. 2005 , Fidanza et al. 2007 ), until recently, their effect on fungal and bacterial communities was less understood and often discordant. Early studies were limited by culture-based methods, but over the past decade, the spread of metabarcoding molecular techniques has provided a more comprehensive understanding of FFRs’ impact on grassland microbiota across various regions, including the Mediterranean areas of Italy and Spain (Mari et al. 2020 ), as well as the Tibetan Plateau (Xing et al. 2018 , Yang et al. 2019 ). Type-1 FFRs have a significant impact on both fungi and bacteria, often causing a temporary reduction in diversity, sometimes quite drastic (Zotti et al. 2021 ), followed by the turnover of a distinctly different microbial community. For example, Zotti et al. ( 2020 ) demonstrated that only a few bacterial species, such as Burkholderia spp., and fungi from the genus Trichoderma can coexist with the mycelium of the fungus Agaricus arvensis , which forms FFRs in the Apennines. Taken together, the available evidence indicates that the development of FFRs in grasslands acts as a travelling wave, profoundly affecting both the biotic and abiotic components of the ecosystem. Due to their impact on community diversity, structure, and functionality, FFRs can be considered ecosystem engineers, as suggested by Hastings et al. ( 2007 ). However, most studies have investigated the effects of FFRs on microbial communities (Ohara and Hamada 1967 , Kim et al. 2013 ), soil chemistry (Fisher 1977 , Gramss et al. 2005 ), or plant communities (Getzin et al. 2021 ) in isolation, with only a few attempting to assess their impact simultaneously (Zotti et al. 2020 ). To address these limitations, we studied FFRs caused by the fungus Agaricus urinascens in a montane Mediterranean grassland (Fig. 1 ). A key novelty of this study is the integration of multiple ecosystem components, e.g. soil chemistry, microbiota, and vegetation, to provide a comprehensive understanding of how FFRs function as ecosystem engineers. Additionally, for the first time, the structure of the mycelium was examined in undisturbed field samples using digital microscopy and further analysed under SEM-EDS to characterize its morphology and elemental composition. Another novel aspect of this study is the in situ monitoring of the soil hydrophobicity at the FF, by monitoring soil water content throughout the growing season using environmental sensors, allowing us to assess dynamic changes in soil water repellency at the FF. Overall, we hypothesize that: FFRs modify soil properties, including nutrient availability, pH, hydrophobicity, and water availability. The passage of A. urinascens FFRs alters soil microbial community composition. FFRs influence plant community structure, affecting biomass production and species diversity. The structural organization of the mycelium at the FF plays a key role in its ecosystem engineering effects. Figure 1. Pictures of the Aremogna study site (Central Apennines, Italy) in November 2022 (above) and July 2023 (below). During the growing summer season, the FFRs are highlighted by the presence of differently shaded vegetation bands.",
"discussion": "Discussion Fairy rings changing soil chemistry and water regime affect plant biomass and diversity The passage of the FF significantly affected grassland soil, causing a slight acidification (0.14-point decrease in pH), increased electrical conductivity, phosphorus levels, and water repellency. Meanwhile, organic carbon declined, along with a minor but significant reduction in exchangeable bases. Similar pH reductions have been linked to enzymatic decomposition releasing H⁺ ions (Yang et al. 2019 ). The decrease in organic carbon, though small, is significant given the organic nature of montane grassland soil. The rise in electrical conductivity likely results from ion mineralization at the FF, with nitrate and ammonia also contributing, as observed in C. gambosa FFRs (Zotti et al. 2021 ) and other basidiomycetes. High electrical conductivity may also result from strong hydrophobicity within the FF. Previous tests on dry soil showed that while grassland soil is hydrophilic, FF soil is extremely hydrophobic (Bonanomi et al. 2012 ), primarily due to dense mycelial growth releasing hydrophobins (Spohn and Rillig 2012 ). Additionally, calcium oxalate crystals, abundant exclusively at the FF, may further enhance hydrophobicity. Mycelial mats are known for strong hydrophobicity (Bonanomi et al. 2012 , Zheng et al. 2014 , Zotti et al. 2021 ), though their effect on grassland hydrology remains unexplored. Our seasonal water content monitoring revealed that FF soil dries faster than surrounding grassland, despite similar plant biomass. Given comparable transpiration rates, this suggests that mycelium-induced hydrophobicity limits water availability for plants, contributing to Type-1 FFRs formation. At our study site, the FFR can be classified as Type 1 due to a slight reduction in aboveground biomass and a significant decline in species richness within the FF, including the dominant grass B. genuense . However, its impact is lower than at Monte Rogedano, just 200 km away on a similar substrate, where A. arvensis creates a bare soil strip and complete species turnover, replacing perennial grasses with annuals (Bonanomi et al. 2012 , Zotti et al. 2020 ). The weaker effect of A. urinascens may stem from species-specific ecological differences or climatic variations. The Aremogna site, at a higher altitude (1820 m versus 1050 m a.s.l.), receives more summer rainfall, potentially mitigating FF-induced hydrophobicity. Future comparative studies, monitoring soil water content across sites with Type 1 and Type 2 FFRs, could clarify fungal–climate interactions. The impact of FFRs on vegetation becomes clearer when examining fungal mycelium dimensions. Early investigations by Shantz and Piemeisel ( 1917 ) and Clements ( 1928 ) illustrated a wider advancing mycelium and a thinner distal portion. Later, Couch ( 1995 ) and Fidanza ( 2007 ) classified mycelial mats as edaphic (growing in soil) and leptophilic (associated with surface litter). Our trench excavation confirmed A. urinascens as an edaphic species, growing at depths of 3–24 cm, with a mycelial mat extending over 120 cm. For the first time in FFRs, we documented a significant presence of calcium oxalate crystals in the mycelial front, previously observed under experimental conditions in basidiomycetes (Nguyen 2023 ). These crystals, first identified microscopically centuries ago (Franceschi and Nakata 2005 ), are produced by various fungi (Arnott 2020 ) and have been reported in Agaricus species in vitro (Whitney and Arnott 1987 ), but this is the first field study linking them to FFR formation. Oxalate crystallization in the presence of calcium serves multiple functions, including pH modulation (Martin et al. 2012 ), enhancing pathogenicity (Cessna et al. 2000 ), and protecting mycelia from environmental stressors or mycophagous insects (Arnott 2020 ). While their role in FFR formation remains unclear, their abundant and exclusive presence at the FF suggests functional relevance in shaping FFR dynamics. Ongoing research is quantifying this phenomenon across different FFR types to clarify their role in FFR differentiation. As suggested by our previous modelling study (Salvatori et al. 2023 ), circular FFR patterns may result from autotoxicity via self-DNA accumulation, a process also observed in plant patterns (Cartenì et al. 2012 ). However, whether FFRs affect vegetation primarily through hydrophobicity or allelopathic compounds remains unresolved. This study underscores hydrophobicity as a key factor but highlights the need for further field and experimental work to understand crystal formation dynamics and their potential contribution to soil hydrophobicity. At the Aremogna site, A. urinascens stimulates vegetation growth, doubling biomass in the belt area compared to surrounding grassland. Brachypodium genuense, C. arvense , and K. splendens dominate this zone. This biostimulation may result from increased P availability, which peaks within the FF, likely due to P accumulation in the mycelium, facilitating its conservation and redistribution across the grassland (Edwards 1984 , 1988 ). Nutrient mineralization and fungal-derived compounds may further enhance vegetation growth (Kawagishi 2019 ). These so-called ‘fairy chemicals’ were first isolated from Lepista sordida cultures (Choi et al. 2010 , Mitchinson 2014 ) and identified as 2-azahypoxanthine, a molecule that promotes root and shoot elongation at an optimal concentration of 20 μM. Their potential for agricultural use has been proposed (Ito et al. 2020 ), yet, their role in natural ecosystems remains unverified. Future research should assess and quantify these compounds in field settings. Direct and indirect effect of fairy rings on microbiota Our results show that the FF of fairy rings significantly influences soil microbiota. Advancing mycelial fronts of Agaricus species destabilize microbial assemblages, shifting the relative abundance of specific taxa. Bonanomi et al. ( 2012 ) observed that fungal passage can favour certain microbial groups while reducing overall diversity. At the FF, the bacterial dominance index peaked, while Shannon diversity index was lowest, indicating a less diverse but more dominant bacterial community. Similarly, Zotti et al. ( 2020 ) reported decreased bacterial diversity in A. arvensis FFRs, where a few taxonomic groups proliferated. Our data reveal a significant increase in Actinobacteria at the FF, alongside a reduction in Proteobacteria and Chloroflexi. Gram-positive Actinobacteria, known for their resilience and ability to degrade complex organic materials, thrived in the FF, whereas Gram-negative Proteobacteria and Chloroflexi decreased. Since Gram-negative bacteria often dominate nutrient-rich environments and associate with roots (Miller et al. 1990 , Griffiths et al. 1998 ), their reduction suggests a shift in soil nutrient dynamics. Conversely, Gram-positive Actinobacteria, typically oligotrophic specialists adapted to bulk soils (Kramer and Gleixner 2008 , Ramesh et al. 2019 ), became dominant, possibly benefiting from organic matter decomposition in the mortality zone. Zotti et al. ( 2020 ) similarly observed an increase in Actinobacteria at the FF. Among bacterial genera, Streptomyces and Kribbella were significantly enriched at the FF. These taxa contribute to soil health, organic matter decomposition, and plant growth promotion (Kim et al. 2014 , Oh et al. 2016 ) and are widely used in agroecosystems for their ability to suppress fungal plant pathogens (Holmes et al. 1998 ). Research on Streptomyces and Kribbella suggests they may form mycoparasitic relationships with certain fungi (Trejo-Estrada et al. 1998 , Govan et al. 2000 ). Their increase indicates that the FF creates an environment favouring beneficial bacteria, likely due to nutrient availability from decomposing plant and fungal matter. At the FF, Ascomycota dominated (97%), suggesting they thrive under these specific conditions. This contrasts with the belt section, where Mucoromycota peaked (13%), and the inside of the ring, where Basidiomycota were most prevalent (24%). The predominance of Ascomycota at the FF suggests adaptation to its nutrient-rich, competitive environment. Within the FF, Agaricus species, primarily A. urinascens , increased in abundance, while Trichoderma was most abundant in the belt section. Known for its antagonistic activity against other fungi, Trichoderma likely thrives in the belt, interacting with other fungi. Similarly, Wang et al. ( 2022 ) found Trichoderma as the second most dominant fungus in Leucocalocybe mongolica FFRs, suggesting its role in FFR dynamics and plant stimulation while parasitizing the dominant FF. The unique microbial dynamics at the FF indicate that FFRs create specific microhabitats that shape soil microbiota. The dominance of certain bacterial and fungal taxa in the mortality zone suggests its function as a specialized niche with distinct environmental pressures. The increase in beneficial bacteria and decomposition-related fungi highlight the complex ecological interactions within this zone."
} | 4,897 |
33502957 | PMC8885433 | pmc | 3,961 | {
"abstract": "Nature offers bionic inspirations for elegant applications of mechanical principles such as the concept of snap buckling, which occurs in several plants. Exploiting mechanical instabilities is the key to fast movement here. We use the snap-through and snap-back instability observed in natural rubber balloons to design an ultrafast purely mechanical elastomer actuator. Our design eliminates the need in potentially harmful stimulants, high voltages, and is safe in operation. We trigger the instability and thus the actuation by temperature changes, which bring about a liquid/gas phase transition in a suitable volatile fluid. This allows for large deformations up to 300% area expansion within response times of a few milliseconds. A few degree temperature change, readily provided by the warmth of a human hand, is sufficient to reliably trigger the actuation. Experiments are compared with the appropriate theory for a model actuator system; this provides design rules, sensitivity, and operational limitations, paving the way for applications ranging from object sorting to intimate human-machine interaction.",
"conclusion": "Conclusions A natural rubber balloon mounted on a sealed chamber of appropriate volume is a structure possessing two stable equilibria for a given common pressure: one with a small and one with a large balloon volume. The transition between these states happens on the timescale below 4 ms. We utilize this bistable system to create an ultrafast purely mechanical switch, or sensor, which operates without any electronic components. In a temperature-driven actuation the internal pressure change is provided by the liquid–gas phase transition of a suitable volatile fluid agent (acetone). The associated pressure change is about 1.1 kPa/K, which is readily activated by the warmth of a human hand or other comparable heat sources. Theoretical analysis of the device performance and sensitivity is provided and compared with experiments. Furthermore, the physical limitations for the maximal operational speed are discussed. In many cases the limiting factors are the internal gas dynamics and the inertia/sound effects in the added (induced) air mass, rather than the acceleration of the membrane itself. The system response times, both thermal and mechanical, decrease with the system size about quadratically. This bodes well for the further optimization, speed enhancement, and miniaturization of the setup in possible applications. The very short response time and compact design, as well as the possibilities for further improvement, make this actuator concept an attractive candidate for future applications in safe object handling, for haptic interfaces, as soft sensors and in soft robotics.",
"introduction": "Introduction In engineering, robots are designed to achieve a particular task. To fulfill their purpose, they traditionally combine basic mechanical frameworks with pneumatic and electric components. They are as rigid and inflexible as the materials they are made of and, consequently, do not adapt well to different environments and various tasks. By making parts of the robot soft, flexible, and compliant, their field of application can be extended. 1 , 2 Great possibilities exist for grabbing, moving, and sorting delicate and fragile objects. 3–5 Actuators based on soft balloons are compliant, robust, light weight, and simple in structure and have low costs. In Ref. 6 a possible application as fast sorting device, for example, for conveyor belts, the handling of sensitive objects, and catching a falling ping-pong ball, was demonstrated. 6 In this study, we introduce a new design for soft actuators based on the mechanical instability that occurs during the inflation of a balloon made out of a natural rubber membrane. 6–8 The abrupt change in the balloon size is triggered by the liquid/gas phase transition of a low-boiling point fluid. 9 This combination of instability and phase transition enables fast switching operations within a few milliseconds. In contrast to other fast responding elastomer actuators, the use of conformable electrodes and high voltages or explosives is not necessary. 10 , 11 Our approach allows safe operation and does not need insulation coatings for electrical protection, which can stiffen the structure, reduce the performance, or alter the overall behavior. 12 , 13 By carefully selecting and adjusting the operating range, one can trigger the mechanical instability by touching the liquid/gas reservoir with one's hand, as shown in Figure 1 (thermographic image and Supplementary Video S1 ). FIG. 1. Triggering the mechanical instability by the warmth of a human hand. (a) Pictures (photographs and thermal image) of the transition from the unactuated to the actuated state in < 4 ms (see the thermographic video, Supplementary Video S1 ), the balloon is marked with a circle . (b) Pictures representing consecutive frames of the high-speed video showing the change in volume during snap-through ( top ) and snap-back ( bottom ) instability. We devise a model actuator system, identify its optimal operational parameters, and validate experimental results with analytical theoretical predictions.",
"discussion": "Results and Discussions Snap-through with pneumatic pressure control In a first step, the pressure-volume response of the system was recorded without liquid in the reservoir ( Supplementary Video S2 ). Figure 2b shows the measured data in the pressure-volume plane. In this study, the pressure p was increased at a constant rate using compressed air supply through a valve, resulting in the formation of a balloon with the volume V . In the process, the membrane is subject to a purely mechanical snap-through instability during inflation. As shown in Figure 2b , the volume increases abruptly (Δ t < 4 ms) within two frames of the high-speed video from states 2 to 3 without appreciable change in pressure. Thereafter, the pressure rises again until the valve is closed at point 4 in Figure 2b . This behavior of rubber balloons is already well studied in the literature. 14 The pressure–volume dependence for an inflatable rubber membrane follows a nonmonotonous N-shaped curve, with a critical pressure reached at point 2. This is illustrated in Figure 2b , where experimental results are fitted with a theoretical curve for the equilibrium overpressure p inside a thin spherical balloon made of an incompressible elastomer. The physical origin of N-shaped dependence can be understood from the differential work-energy balance during the balloon inflation. Upon a small volume increase dV , the pressure performs work pdV , while the elastomer energy changes by V E d W , where V E is its constant volume, and W is the configurational part of the volumetric strain (free) energy density (for equal-biaxial deformation in our geometry). For a thin incompressible spherical elastomer balloon with the varying radius R = λ R 0 the lateral stretch λ is related to the volume inside the balloon by V = 4 π R 0 3 λ 3 ∕ 3 , and one obtains the following equation for the pressure: \n \n In this study, H 0 = 60 μ m and R 0 = 0 . 185 c m are the initial thickness and the (equivalent) initial radius of the membrane. Due to incompressibility λ 1 λ 2 λ 3 = 1 , the energy density W always contains nonlinearity at appreciable stretches. Already the simplest hyperelastic neo-Hookean expression W ∼ ∑ i = 1 3 λ i 2 ∼ e q u a l − b i a x i a l 2 λ 2 + λ − 4 , together with the factor λ − 2 in Equation (1), results in the pressure dependence p ∼ λ − 1 − λ − 7 . The different signs of terms stem from the rates of changes in the energy W with respect to lateral and “thickness” directions. At small stretches, the second, “thickness” term dominates and the pressure increases, while at large stretches, the lateral contribution (first term) wins and the pressure falls p e in Equation (1) is the maximum pressure at point 2 in the neo-Hookean approximation, which provides the scale for the pressure values. At very large stretches the elastomer stiffens due to the finite extensibility of the polymer chains, and p ( V ) curve in Figure 2b bends upwards at large V . In our case, this is described by the Gent hyperelastic model 15 :\n (2) W ( λ ) = − μ J l i m 2 ln 1 − 2 λ 2 + λ − 4 − 3 J l i m \n Above, μ ≈ 0.58 MPa is the small-stress shear modulus, and J lim ≈ 46.99 accounts for the stiffening at large deformation. For our parameters, the system can be considered quasi-statically (the applicability limits are discussed in section Dynamic limitations below). In the experiment, the amount of air in the system (total number of molecules N ) is slowly increased through the valve in Figure 2a , so that the gas-dynamic effects are negligible. The common overpressure p with respect to the atmospheric pressure p atm is the same everywhere in the balloon V and reservoir V R . We assume that the gas obeys the ideal-gas law at the constant temperature T :\n (3) ( p + p a t m ) ( V + V R ) = N k B T \n Here k B is the Boltzmann constant. Resolving Equation (3) for p we obtain:\n (4) p = N k B T V + V R − p a t m ≈ N k B T V R 1 − V V R − p a t m \n The second expression represents the first two terms of the Taylor expansion for V ∕ V R < < 1 . In equilibrium, this pressure should be equal to the elastic value [Eq. (1)]. The pressure p in Equation (4) slowly increases with added gas N , reaching the critical value at point 2 in Figure 2b . Further increase in pressure becomes impossible, and the very fast (< 4 ms) snap-through to a new equilibrium (point 3) occurs. During this almost instantaneous transition, the total number of air molecules in the system remains virtually constant ( N ≈ 10 22 ). As the balloon volume V increases, p decreases according to Equation (4). However, when the reservoir is large compared to the balloon, V R > > V , the pressure remains almost constant. The influence of an additional chamber was studied by Keplinger et al. , 7 to provide a loading path for a dielectric elastomer actuator averting electrical breakdown, and discussed theoretically by Zhu et al. , 16 providing guidelines for choosing an appropriate chamber size. In our case V ∕ V R ≤ 5 . 7 × 1 0 − 3 , the relevant part of the hyperbola [Eq. (4)] in the p-V plane looks like an almost horizontal line with < 0.6% change in the overall pressure (≈700 Pa). Equation (4) is sketched in Figure 2b as a dashed curve, tangential to the pressure-volume curve of the elastomer membrane according to Equations (1) and (2) at the snap-through point 2. The second intersection of these two curves defines the equilibrium state after the snap-through; the transition itself roughly follows the membrane p - V curve, 6 modified by nonisothermal and dynamic effects. The transient behavior cannot be resolved on the present time scale, due to the much smaller size and the correspondingly faster dynamics of the current setup. During the rapid snap-through, the membrane quasi-adiabatically heats by several degrees. 8 As it thermalizes back to room temperature, the elastomer softens proportionally, 17 slightly increasing the balloon volume, which can be seen in Figure 2b , after the point 3. The gas entering the balloon somewhat cools down semi-adiabatically, but the pressure corrections in Equations (3) and (4) remain small, as long as V ∕ V R < < 1 . Thus, both effects contribute a few percent at most, and the real process falls between the isothermal and adiabatic limits. If the balloon is inflated and deflated several times, the rubber membrane is cyclically loaded, revealing material-dependent effects. First, the elastomer membrane shows an intrinsic hysteresis due to stretch-induced crystallization, 8 , 18 so that different curves in the p - V plane are followed for inflation and deflation. 6 Furthermore, natural rubber softens appreciably during the first loading cycles ( Fig. 2c )—this is known as Mullins effect. 19 The measurements show that after about 10 cycles the additional changes become negligible, and the sequential cycles coincide with each other. The acetone vapors may alter elastic properties of the membrane and/or diffuse through it. We did not observe such effects during ∼1 h long experiments. Moderate acetone loss is irrelevant, as liquid is in surplus, and the vapor remains saturated. However, over the long term, these issues are of concern, and the influence of different liquid agents on the polymer membrane, as well as alternative combinations of volatile liquid and/or membrane materials, should be performed for practical implementations. 20 Thermally triggered actuation driven by liquid/gas phase transition The snap-through instability discussed above may be used to produce a fast direct mechanical response to a thermal stimulus, without any intermediate electronic control. The idea is to modulate the pressure in a sealed reservoir volume using the saturated pressure of a volatile liquid agent. A thermal stimulus can come from an external heater or the body warmth of a person and thus can be used to detect the touch. Within the applicability of Dalton's law, the total pressure inside the reservoir and balloon is the sum of the partial air pressure p a and the saturated vapor pressure p s . The partial pressure of air obeys the ideal gas law, similar to Equation (3), with a different, but constant number of molecules N :\n (5) p a ( V + V ′ R ) = N k B T \n V ′ R is the reservoir volume accessible for the gas, that is without the volume of the added liquid V L . Typically, V L < < V R , so that V ′ R ≈ V R , and the change in liquid volume upon evaporation can be neglected. The pressure of the saturated vapor is given by the Clapeyron–Clausius relation, 21 resulting in a steep exponential dependence on temperature:\n (6) p s = p 0 e − T v T = p a t m e T v 1 T b − 1 T ≈ p s ( T 0 ) e T − T 0 T F K , T F K = T 0 2 T v \n In these equations, T v = L ∕ R G is the Arrhenius exponent, related to the latent heat of vaporization per mole L through the universal gas constant R G , and p 0 being some prefactor. Both are approximately constant. The second equality is expressed through the boiling temperature T b , where by definition p s ( T b ) = p a t m applies. The third expression is a Taylor expansion of the Arrhenius exponent near the reference room temperature T 0 = 293 K , emphasizing the exponential behavior for small temperature changes [Frank–Kamenetskii (FK) approximation]. The total overpressure p = p a + p s − p a t m thus becomes:\n (7) p = N k B T V + V ′ R + p a t m e T v 1 T b − 1 T − 1 ≈ N k B T V R − p a t m + p s ( T 0 ) e T − T 0 T F K \n This expression implies that enough liquid is present, so that it never fully evaporates, both phases are in thermodynamic equilibrium (no kinetics, saturated vapor), and volume changes due to the liquid are negligible, V ′ R ≈ V R = c o n s t . The second equality uses FK approximation from Equation (6). As before, the equilibrium overpressure p given by the Equation (7) is equal to the quasi-static elastic expression [Eq. (1)]. The first term in Equation (7) depends on volume in the same way as in Equation (4), while the second term does not depend on V at all. Thus, the p ( V ) dependence [Eq. (7)] is a very flat hyperbola, which can similarly be approximated by an almost horizontal line. The multiplier in the first term is linear in temperature, while the second additive term has a steep exponential temperature dependence. In fact, the shear modulus in Equations (1) and (2) is itself proportional to temperature, 17 \n μ = μ ( T 0 ) T ∕ T 0 , so that the pressure disbalance is almost exclusively due to the p s ( T ) dependence (precise analysis should consider temperature transients). An increase in T shifts the hyperbola [Eq. (7)] up; when the intersection with the curve [Eq. (1)] near the state 2 ceases to exist, fast snap-through to the new equilibrium state 3 occurs. The only difference is that previously the hyperbola was scaled up by the influx of air (slow increase in N ), while now N = c o n s t , and the hyperbola is shifted upwards by the temperature, through the steep p s ( T ) dependence of the additive term in Equation (7). We tested several low-boiling point liquids as a phase-change agent; acetone with T b = 329 . 3 K = 56 . 1 5 ∘ C and L = 31 . 3 k J ∕ m o l near room temperature 22 showed the best results. Using these data and the Antoine equation, we deduce T v = 3764 . 5 K , T F K = 22 . 83 K , and p s ( T 0 ) = 24 . 6 k P a , which are close to the snap-through p e value in Equation (1) and Figure 2b . The exponential term in Equation (7) increases the pressure by about 1.1 kPa/K near T ≈ T 0 . The results of the experiment in which the instability was triggered purely thermally are shown in Figure 3 . In this study, a small amount (1.5 mL) of liquid acetone is added to the system through the inlet in Figure 2a , and the system is set to an initial pressure of 11.5 kPa following a pressure ramp using the compressed air supply. Thereafter, the reservoir is sealed. Resistive heating of the system with constant power P = 3 . 6 W (comparable with the heat flow from human hands) starts at the time t 1 at the bottom of the reservoir. This leads to a gradual increase in the saturated and the overall pressure according to Equation (7), which can be seen in Figure 3a for t 1 < t < t 2 , when finally snap-through occurs. The observed pressure change of about 7.5 kPa corresponds to an increase of acetone temperature by about 7 K. Although the heating is externally stopped when snap-through is observed, the balloon volume still expands further. This is because the membrane adiabatically heats in a fast expansion, 8 while the gas inside the balloon semiadiabatically cools . As thermalization of both materials toward the ambient temperature starts for t > t 2 , the elastomer softens ( μ ∼ T ), while the gas heats. Both factors slightly increase the balloon volume. Because the system is enclosed into an insulation box, the temperature inside the reservoir drops slowly after the heating is switched off. A cooling spiral around the outer walls of the reservoir with circulating refrigerating fluid cools down the system effectively. Active cooling is started at the time t 3 . As the temperature decreases, the acetone saturated pressure p s ( T ) drops, and snap-back occurs at the time t 4 . Subsequently, the process is repeated. FIG. 3. (a) Increase in the overall overpressure ( left axis ) upon resistive heating with the constant power P ( right axis ) triggering the snap-through instability at t 2 . Active cooling is started at t 3 until the snap-back occurs at t 4 . (b) Accompanying change in balloon volume, calculated from the high-speed camera recording. In the experiment, the cycle time exceeded t 2 − t 1 = 153 s and required an energy expenditure of about 550 J ( Fig. 3 ). The system was not optimized, and the results show the importance of the initial state for the overall performance. To achieve high sensitivity and a fast response, the system should be brought slightly below the verge of instability (critical pressure state 2). As a result, mechanical energy is stored in the system and the energy barrier for triggering the instability is reduced to a suitable level, so that a small thermal increase in the saturated and the overall pressure causes the snap-through. Nature exploits the same principle to enable the rapid movements of several carnivorous plants 23 : A slow accumulation of mechanical (elastic) energy is followed by its rapid release triggered by a small stimulus. For carnivorous plants, the control of elastic instabilities in geometrically slender parts of their trapping mechanisms offers an alternative to the muscle-powered movements in animals. Fine-tuning of our setup provided an even higher sensitivity of about 2 K, which can be readily provided by a human hand ( Fig. 1a ). In this study, keeping the room temperature constant is the limiting factor. Due to the steep p s ( T ) dependence, the air pressure in the device should be adjusted with respect to the actual room temperature. Dynamic limitations One of the distinct attractive features of our setup is its small size, resulting in an exceptionally fast time constant for a mechanical elastomeric device. It is instructive to discuss the physical factors limiting its operational speed. For fast snap-through and snap-back stages the quasi-static approximation should be replaced by the appropriate dynamic equations, for example, using Euler–Lagrange formalism. 24 Gas dynamics analysis of the gas flow between the reservoir and the balloon can be required as well. A first insight into dynamics of dielectric elastomer actuator was provided by Xu et al. , 24 Zhu et al. , 25 Zhang et al. , 26 Li et al. , 27 and Chen et al. 28 These works focus on the role of membrane inertia. If the balloon of radius R and thickness H expands with the speed v = d R ∕ d t , force balance results in an equivalent dynamic overpressure:\n (8) p m ≈ ρ H d v ∕ d t ≈ 34 . 9 P a \n In the numerical estimations we used experimental values Δ R ∼ R = λ R 0 ≈ 1 c m , H = λ − 2 H 0 ≈ 60 μ m , Δ t ≈ 4 m s , v ≈ 2 . 5 m ∕ s , d v ∕ d t ≈ 625 m ∕ s 2 , elastomer density ρ ∼ 0 . 93 g ∕ c m 3 . Below, we also use air density ρ a i r ∼ 1 . 225 k g ∕ m 3 and speed of sound c = 343 m ∕ s . The obtained inertial overpressure is small in comparison to values used in the experiments. However, the expanding balloon also compresses and accelerates the air in its path, producing spherical sound waves. The (over)pressure within such waves can be deduced from Equation (9) in Ref. 29 \n (9) p S ≈ 2 ρ a i r v 2 ≈ 15 . 3 P a \n This is even smaller than the inertial overpressure [Eq. (8)]. However, if the balloon is large, the situation is closer to a one-dimensional (1D) piston, which is described in Ref., 30 §99, Problem 1. For v ≪ c one obtains:\n (10) p p ≈ ρ a i r c v ≈ 1050 P a \n This is a much larger value, comparable with the saturated vapor pressure change per Kelvin, raising the question which of the estimations [Eqs. (9) and (10)] is more relevant for our case. The solution [Eq. (7)] in Ref. 29 implies that the spherical piston starts from R ( 0 ) = 0 with constant velocity d R ∕ d t = v = α c ( c is denoted as a there), and the sound wave relaxes appreciably when it reaches the radius R ( t ) . However, this spherical sound solution can be modified to describe our situation, assuming that the piston starts moving from a finite radius and time \n R ( t 0 ) = R 0 = v t 0 . In notations of Ref., 29 the initial condition for the disturbance changes from f ( 0 ) = 0 to f ( w 0 ) = 0 , where w 0 = ( α − 1 ) c t 0 < 0 . As a result, in Equation (7) of this Ref., the constant C ≠ 0 , but should be found from the condition f ( w 0 ) = 0 . The expression that becomes zero at w 0 is:\n (11) f ( w ) = c α 3 w 0 2 1 − α 2 w 2 w 0 2 − w w 0 α − 1 α \n The derivative of this expression is used to find the overpressure behind the sphere from the Equation (3) there, using r = R , R − c t = w , and therefore R = v t = α w ∕ ( α − 1 ) .\n (12) Δ p = − ρ c R f ′ ( w ) = ρ c 2 α 2 1 + α 2 + 1 − α α w w 0 − 1 α − 1 \n We further simplify this for α = v ∕ c < < 1 , which holds in the sound limit and is fulfilled for our numbers, see the estimations below Equation (8). Furthermore, as R = v t = α w ∕ ( α − 1 ) , we can replace w ∕ w 0 = t ∕ t 0 , where t 0 = R 0 ∕ v is the starting time for the expansion, where w = 0 . This results in:\n (13) Δ p ≈ ρ a i r 2 v 2 + c v t t 0 − c v − 1 \n This expression describes the transition from the initial large 1D overpressure ρ a i r c v near t = t 0 [Cf. Eq. (10)] to the small three-dimensional sound overpressure 2 ρ a i r v 2 at longer times [Cf. Eq. (9)]. Both terms in Equation (13) become equal at a time t = t 0 + t 1 , when \n \n The last ratio is ≈ 0 . 036 for our numbers. Because the snap-through time is of the order of t 0 , this implies that the overpressure is “high” during the first several percent of the expansion. After that, the sound wave detaches from the balloon surface, and the pressure there decreases. In practice, the acceleration of the balloon is more gradual, and the initial overpressures are lower than estimated in Eqs. (10) and (13). The spherical acoustic problem can be solved for an arbitrary prescribed expansion dynamic R ( t ) , using Fourier results for a given frequency ω . The reaction force on a sphere sinusoidally oscillating with velocity amplitude v is given, for example, in Ref., 30 §74, Problem 1. Recalculating it into pressure, we obtain:\n (15) p ω = ρ a i r c v e − i ω t k R i ( 2 + k 2 R 2 ) − k 3 R 3 3 ( 4 + k 4 R 4 ) ⇒ k R > > 1 , 1 3 ρ a i r c v k R < < 1 , 1 6 ρ a i r c v k R \n The upper limit corresponds to the “large” spheres and pressures. Our numbers are closer to the “small” lower values with ω ≈ 1 . 5 × 1 0 3 s − 1 , k R ≈ 0 . 05 . This results in p ω ≈ 8 P a , similar to the estimation [Eq. (9)]. The estimations [Eqs. (8)–(15)] show that the expansion speed is often limited by the inertia of the air, rather than of the membrane itself. For a balloon, these effects can be even higher by a factor of about 2, due to similar rarefaction effects on the inner side of the expanding membrane. If the overpressures [Eqs. (8)–(10)] are added to the r.h.s. of the first expression in Equation (1), one obtains dynamic equations, most conveniently in terms of λ . In a snap-through, or if pressure is instantaneously increased to the characteristic value p e , the elastic contribution is small, and the time constants can be estimated as:\n (16) p e ∼ p m ∼ ρ H 0 R 0 λ − 2 d 2 λ d t 2 ⇒ τ m ∼ ρ H 0 R 0 p e ∼ ρ R 0 2 μ ≈ 0 . 07 m s p e ∼ p S ∼ 2 ρ a i r R 0 2 d λ d t 2 ⇒ τ S ∼ 2 ρ a i r R 0 2 p e ∼ 2 ρ a i r R 0 3 μ H 0 ≈ 0 . 02 m s p e ∼ p p ∼ ρ a i r c R 0 d λ d t ⇒ τ p ∼ ρ a i r c R 0 p e ∼ ρ a i r c R 0 2 μ H 0 ≈ 0 . 03 m s \n Here, the first estimations for the time constants τ are in terms of p e [Eq. (1)]. Substituting its expression from there (omitting 12 ∕ 7 7 ∕ 6 ≈ 1 . 24 ), we obtain the dependence of characteristic times on the system parameters μ , H 0 , R 0 (second equalities for τ ). They show how the device dynamics scales with size. A smaller radius R 0 and a larger product μ H 0 (stiffer membrane) reduce the mechanical time constants. The overall response times are governed by the thermal parameters and decrease sharply for smaller sizes, according to t T ∼ R 0 2 ∕ D , where D is the effective thermal diffusivity of the system. It may depend on the reservoir size and geometry and on the combinations of thermophysical parameters of materials. The intrinsic pressure sound equilibration time τ s e ∼ R 0 ∕ c ≈ 5 μ s is usually (much) shorter. In practice, dynamics is often limited by the gas flow through the valves and in the reservoir between the valve and membrane."
} | 6,799 |
35322846 | null | s2 | 3,962 | {
"abstract": "Cellular processes and functions can be regulated by mechanical forces. Nanodevices that can measure and manipulate these forces are critical tools in chemical and cellular biology. Synthetic DNA oligonucleotides have been used to develop a wide range of powerful nanodevices due to their programmable nature and precise and predictable self-assembly. In recent years, various types of DNA-based mechanical nanodevices have been engineered for studying molecular-level forces. With the help of these nanodevices, our understanding of cellular responses to physical forces has been significantly advanced. In this article, we have reviewed some recent developments in DNA-based mechanical sensors and regulators for application in the characterization of cellular biomechanics and the manipulation of cellular morphology, motion and other functions. The design principles discussed in this article can be further used to inspire other types of powerful DNA-based mechanical nanodevices."
} | 246 |
25343307 | PMC5381366 | pmc | 3,963 | {
"abstract": "The spatial context of microbial interactions common in natural systems is largely absent in traditional pure culture-based microbiology. The understanding of how interdependent microbial communities assemble and coexist in limited spatial domains remains sketchy. A mechanistic model of cell-level interactions among multispecies microbial populations grown on hydrated rough surfaces facilitated systematic evaluation of how trophic dependencies shape spatial self-organization of microbial consortia in complex diffusion fields. The emerging patterns were persistent irrespective of initial conditions and resilient to spatial and temporal perturbations. Surprisingly, the hydration conditions conducive for self-assembly are extremely narrow and last only while microbial cells remain motile within thin aqueous films. The resulting self-organized microbial consortia patterns could represent optimal ecological templates for the architecture that underlie sessile microbial colonies on natural surfaces. Understanding microbial spatial self-organization offers new insights into mechanisms that sustain small-scale soil microbial diversity; and may guide the engineering of functional artificial microbial consortia.",
"discussion": "Results and Discussion We first considered trophic interactions between two microbial species (sp1 and sp2) with different apparent yields to two obligatory nutrients ( N1 and N2 ) present in the aqueous phase of a homogenous rough surface. The mechanistic modeling of growth, multiplication and chemotactic motility of microbial cells of the two populations spontaneously gave rise to formation of spatially segregated sectors populated by single species, as shown in Figs. 1 and 2 . The resulting spatial patterns were remarkably stable irrespective of population growth within a sector as evidenced by the persistent angular distributions of microbial populations (seeing Fig. S1, and Supplementary Movie 1 ), and by calculated segregation index 27 values of sp1 and sp2 (for consortium I as an example), which decreased rapidly with elapsed time, and remained stable since 10 hours after inoculation on homogeneous and heterogeneous surfaces ( Fig. S2 ). The segregation and spatial patterns were linked to differences in physiological yields for the two nutrients and the spatial patterns of local concentrations that resulted from preferred nutrient uptake. Interestingly, differences in nutrient utilization resulted in leaky trails of nutrients (see Supplementary Moves 2 and 3 ) indicative of consumption efficiency commensurate with the prescribed stoichiometry and complementary nutrient utilization by the dominant species in each sector. These nutrient trails, and the presence of both nutrients at sector boundaries, induced the formation of internal microbial population bands ( Fig. 3 , and Supplementary Movie 2 ). The populated bands of the two microbial species maintained a relatively stable separation distance that reflected optimal concentration peaks based on the stoichiometric relations of each species and their respective nutrient yields. In other words, each species flourished at the optimal combination of concentrations of the obligatory N1 and N2 along opposing gradients of these two nutrients in adjacent sectors. The results depicted in Fig. 3 show agreement between the ratios of local nutrient concentrations of N1 and N2 at the sector boundaries due to spontaneous self-assembly, and the theoretical values based on trophic stoichiometry illustrating the strong coupling between physiological traits, trophic dependencies and spatial organization. Next, we introduced a third species (sp3) into the virtual inoculum where sp3 relies exclusively on a by-product ( N3 ) produced by both sp1 and sp2, forming either commensal 25 or mutualistic 26 28 trophic interactions among the three species (see Table 1 ). For the commensal scenario, the addition of sp3 did not alter the original spatial self-organization of sp1 and sp2 (in the absence of sp3), and the sp3 population simply followed sp1 and sp2 ( Fig. 1 ). The simulated patterns are reminiscent of spatial self-assembly of microbial consortia observed in other microbial systems such as on moist leaf surfaces 29 or agar surfaces 30 . A scenario where the increased concentration of the by-product plays an inhibitory role 24 for the simulated growth of sp1 and sp2 (i.e., a transition into mutualistic interactions) was clearly manifested in the resulting spatial pattern of the stable consortium. A narrow and loosely populated front consisting of sp1 and sp2 only formed ahead of the expanding main population front, and the spatial segregation between trailing internal population bands of sp1 and sp2 has been reduced ( Fig. 1 ). The results illustrate the consequences of mutualistic trophic dynamics in which inhibition at high by-product concentrations depresses growth of sp1 and sp2, “dilutes” population density at the front, and forces a tighter and denser association with sp3 at the perimeter of the consortium. The role of mutualism in promoting soil microbial diversity has been considered previously 28 , and the results here illustrate a spatial manifestation of mutualistic trophic interactions. The additional constraint generally retarded consortium expansion and resulted in proximal internal population bands of sp1 and sp2 that persisted at a close distance from each other due to the depressed “leaky” concentrations ( Supplementary Movie 3 ). Additionally, to quantify the role of chemotaxis on microbial spatial self-organization, we conducted simulations with random cell motions. The results depicted in Fig. S3 showed complete mixing and limited spatial segregation among trophically interdependent populations for the same conditions listed in Table 1 . The results highlight the importance of chemotaxis for the emergence of spatially self-organized microbial consortia. Surface heterogeneity and patchy nutrient concentrations due to hydration dynamics and complex diffusion pathways are the rules in natural terrestrial systems 1 2 3 6 12 . We evaluated effects of surface heterogeneity and hydration status on consortia spatial patterns for similar trophic interactions considered above. The results illustrated emergence of spatially self-organized microbial spatial patterns similar to those forming on homogeneous surfaces ( Fig. 1 , and Supplementary Movie 1 ) albeit with suppressed population sizes and constrained dispersion even for mildly wet (−2.0 kPa) and heterogeneous surfaces ( Fig. S4 ). No organized microbial spatial patterns emerged under drier conditions (−5.0 kPa, Fig. S4c ) presumably due to suppression of microbial motility, an essential ingredient for growth and self-organization on hydrated surfaces 3 6 23 . The results suggest a narrow range of hydration conditions (a few kPa of matric potential) that support microbial self-motion in aqueous films 31 and enable self-assembly of spatially ordered consortia. Equally important, the results indicate that the drivers for self-assembly of microbial populations on heterogeneous surfaces remain dominant despite limitations to motility and fragmentation of nutrient diffusion fields. In other words, trophic interactions and nutrient supply boundary conditions are likely to support microbial consortia, whose spatial pattern optimizes nutrient interceptions even within heterogeneous and patchy nutrient fields as seen, for example, consortium II on a heterogeneous surface ( Fig. S5a ). We conducted additional tests for spatial self-organization of trophically-interacting microbial populations based on experimentally derived parameters for commensal 25 (Consortium IV) or mutualistic 26 (Consortium V) microbial consortia. The modelling results showed well-defined spatial self-arrangement for both commensal and mutualistic microbial consortia ( Figs. S5b and S5c ) lending support to the generality of the principle of emergence of spatial self-assembly of interacting microbial populations. The results reflect the importance of specific metabolic stoichiometry within multi-nutrients environments that result in trophic-induced niche partitioning 32 33 for promoting stable microbial spatial patterns. The study contributes to the understanding of mechanisms for spatial self-organization of microbial communities 34 , by elucidating the joint roles of a threshold nutrient levels 34 (in excess of cell maintenance requirements), and of microbial chemotactic motion 35 . Even without resolving the ubiquity of these conditions for microbial pattern formation in the broader ecological context, we can identify a theoretical biophysical characteristic length (maximal separation distance) for the onset of microbial community self-assembly. We derived an analytic prediction of such characteristic length termed the trophic interactions distance ( TID ) that defines the maximal initial separation distance between consortium members for the activation of trophic interactions and self-organization ( Fig. 4a ). The TID links hydration-mediated nutrient diffusivity with microbial motility ranges, and thus provides a predictive metric for the onset of consortium self-assembly on rough surfaces. Conditions resulting in large TID values (namely rapid cell migration or low threshold concentrations) suggest potential for self-assembly even for low densities or large separation distances among community partners; whereas low TID values indicate that consortium members must reside within this limited range to trigger self-assembly ( Fig. 4b ). This biophysical metric could be useful for design of bioremediation activities including selection of optimal water content, and volumetric densities and strengths of bio-stimulation. Similar estimates may be used for estimating quorum sensing ranges and thresholds 36 37 in soil and other domains. The growing interest in microbial spatial association as an important ingredient for better understanding of natural microbial ecology has not yet been fully integrated into practical models due to limited experimental information 8 9 11 . The proposed modeling framework offers an exploratory platform for guiding information gathering and systematic evaluation of trophic dependencies and potential spatial patterns in the context of natural hydrated surfaces that could contribute to anchoring environmental microbiology back in the natural concourse 11 . The results show that cell-level interactions among multispecies with different trophic dependencies induce dynamic and heterogeneous localized nutrient patterns that have been observed in natural systems 2 3 4 11 . Considering the relatively limited temporal window of favorable hydration conditions that support microbial self-organization (for many geographic and climatic regions lasting only several hours a few times per year), the resulting stable spatial patterns could be viewed as rudimentary ecological templates for the more permanent microbial colonies forming on newly inhabited soil surfaces 2 3 5 6 . The results may also offer guidance for acquiring experimental observations to enhance understanding of functioning microbial communities 11 12 . The proposed biophysical TID for triggering spatial patterns 11 may have practical applications for the design of artificial microbial consortia in the context of synthetic biology 22 , and for improving efficiency of bioremediation activities in the natural environment 12 38 ."
} | 2,882 |
27066394 | PMC4802428 | pmc | 3,964 | {
"abstract": "The development of sustainable, bio-based technologies to convert solar energy and carbon dioxide into fuels is a grand challenge. A core part of this challenge is to produce a fuel that is compatible with the existing transportation infrastructure. This task is further compounded by the commercial desire to separate the fuel from the biotechnological host. Based on its fuel characteristics, 1-octanol was identified as an attractive metabolic target with diesel-like properties. We therefore engineered a synthetic pathway specifically for the biosynthesis of 1-octanol in Escherichia coli BL21(DE3) by over-expression of three enzymes (thioesterase, carboxylic acid reductase and aldehyde reductase) and one maturation factor (phosphopantetheinyl transferase). Induction of this pathway in a shake flask resulted in 4.4 mg 1-octanol L −1 h −1 which exceeded the productivity of previously engineered strains. Furthermore, the majority (73%) of the fatty alcohol was localised within the media without the addition of detergent or solvent overlay. The deletion of acrA reduced the production and excretion of 1-octanol by 3-fold relative to the wild-type, suggesting that the AcrAB–TolC complex may be responsible for the majority of product efflux. This study presents 1-octanol as a potential fuel target that can be synthesised and naturally accumulated within the media using engineered microbes.",
"introduction": "1 Introduction The demand for diesel fuel continues to increase ( Cames and Helmers, 2013 ). In response to these demands, and amid concerns over the environmental impact of fossil fuels, intensive research efforts have been made in developing renewable and sustainable methods for the production of diesel substitutes ( Peralta-Yahya et al., 2011 , Steen et al., 2010 ). To date, biodiesel is the most extensively researched diesel fuel replacement ( Atabani et al., 2012 ). Its synthesis traditionally involves chemical or mechanical extraction of oils from plant or algal sources followed by trans-esterification to yield either fatty acid methyl esters or ethyl esters. Innovations in microbial engineering have led to bioprocesses which obviate the need for a separate esterification step and potentially allow multiple industrial waste streams to be utilised ( Kalscheuer et al., 2006 , Steen et al., 2010 ). The major downside to biodiesel or even precursor fuels such as the recently reported bisabolane ( Peralta-Yahya et al., 2011 ) is that they require energy intensive extraction procedures ( e.g. usage of organic solvents for end-product isolation, high centrifugal forces for biomass recovery, physical methods for disruption of biomass) and/or further chemical modifications, and this can present a major economic barrier for the purpose of fuel commercialisation ( Chisti, 2013 ). Alcohols and alkanes are both highly attractive biofuel candidates as they do not require further chemical modification. In diesel engines, one particularly notorious issue with the use of alkanes is the formation of particulate matter. These carbonaceous particles, also known as soot, are attributed to incomplete combustion and have been shown in several studies to exacerbate respiratory illnesses and contribute to global warming ( Bond et al., 2013 , D’Amato et al., 2013 ). In this regard, alcohols have generated a considerable amount of interest since their increased oxygenated content can significantly stimulate the completion of combustion and thereby lower the production of particulate matter. Herein, we evaluate the fuel and physicochemical characteristics of saturated fatty alcohols and conclude that the C8 fatty alcohol, 1-octanol, is a highly attractive biofuel with diesel-like properties. Previously, 1-octanol had been synthesised by (i) the reversal of beta-oxidation ( Dellomonaco et al., 2011 ), (ii) rerouting branched-chain amino acid biosynthesis ( Marcheschi et al., 2012 ) and (iii) extending the 1-butanol pathway ( Machado et al., 2012 ). In this study, we engineer a novel metabolic route for the production of 1-octanol in Escherichia coli and furthermore show that it can be naturally excreted into the media.",
"discussion": "4 Discussion On account of its diesel-like characteristics, 1-octanol is an attractive biofuel target. In comparison to petrodiesel, its lower vapour pressure could reduce transportation and storage hazards, while its higher auto-ignition temperature could raise the air:fuel compression ratios and in turn improve fuel combustion efficiency. Real simulation tests with compression ignition engines have shown that 1-octanol can reduce particulate matter by as much as 20-fold in comparison to petrodiesel ( Heuser et al., 2013 ). In the case of alternative fuels such as biodiesel, inefficient combustion can lead to the formation of hygroscopic by-products such as fatty acids, mono- and di-glycerides which can promote engine corrosion ( Atabani et al., 2012 ). Furthermore, the cold flow properties of biodiesel are inferior to 1-octanol, ( e.g. pour point; −9 °C vs −13.5 °C) making the latter a potentially better fuel to handle and operate under cold conditions, at least as a blending component. In this study, we engineered a pathway that utilised precursors from the natural process of fatty acid synthesis. The octanoic acid precursor was generated by incorporating a fatty acyl-ACP thioesterase which was previously shown to have a preference for fatty acyl-ACP chains of 6–8 carbon atoms. Although the yield (62 mg L −1 ) was found to be ~2-fold lower than the Dellomonaco et al. (2011) study, the CAR-based platform required a considerably shorter cultivation period (14 h) for optimal production of 1-octanol ( Table 2 ). Thus, the productivity (4.4 mg1-octanol L −1 h −1 ) exceeded previously obtained values of 2.1, 1.5 and 0.04 mg 1-octanol L −1 h −1 ( Dellomonaco et al., 2011 , Machado et al., 2012 , Marcheschi et al., 2012 ; Table 2 ) without resorting to host engineering or advanced bioprocessing techniques. From a stoichiometric viewpoint, since 2 moles of glucose (MW=180) would be required for the synthesis of 1 mole of 1-octanol (MW=130), the maximum theoretical yield possible is 361 mg 1-octanol per gram glucose. Clearly then, the low yields obtained in this and previous studies indicate that there is still considerable room for metabolic improvements. We expect that optimisation of the process via fatty acid engineering ( Torella et al., 2013 ), modulated levels of efflux pumps ( Wang et al., 2013 ), regulated metabolic flux ( Dahl et al., 2013 ) and in situ product removal techniques ( Baez and Shiloach, 2013 ) is likely to improve 1-octanol productivity, yields and titres even further. A major advantage of biologically producing 1-octanol as an end-product lies in its propensity for extracellular localisation. This trait is advantageous for downstream purification of the end-product. In contrast, fuels such as biodiesel become confined to the internal cellular compartments due to their rather large structure and intracellular role as a carbon storage medium ( Kalscheuer et al., 2006 ). Energy-intensive extraction methods are therefore typically required to harvest, extract and isolate biodiesel from the host organism ( Chisti, 2013 ). As suggested in this work, the extracellular localisation of 1-octanol is most likely facilitated by the AcrAB–TolC complex. The importance of TolC, though not AcrAB, for the internal production of free fatty acids (C8–C14) was recently observed in a study in which excretion unfortunately was not monitored ( Lennen et al., 2013 ). Similarly, the over-expression of native E. coli transporters or transporter components have previously been found to stimulate either excretion ( Doshi et al., 2013 ) or both excretion and production ( Wang et al., 2013 ) of native or non-native products. These studies on the role of TolC-dependent transporter complexes suggest that the elucidation of cause-and-effect may be complicated by their (1) overlapping substrate specificities and (2) complex sub-unit stoichiometry; such properties are likely to alter the dynamics of efflux in very subtle ways and make difficult the task of determining the efflux contribution of any single transporter complex. In addition, though not yet studied, it is possible that the overlapping susbstrate specificities may well serve complementary roles depending upon the physiological state of the cell. Despite the complexity of these experimental systems, the simultaneous reduction in both excretion and total production of 1-octanol in the Δ acrA mutant, as well as previous work ( Lennen et al., 2013 , Wang et al., 2013 ), suggest that efflux pumps can have a substantial impact on the total flux of metabolic pathways that lead to a product with no obvious storage role within the cell. Studies are currently underway to shed further light on the correlation between excretion and production, as well as elucidating the entire complement of efflux pumps responsible for the excretion of fatty fuels of varying chain-lengths. In summary, given the biological feasibility of its production, along with its natural excretion within the media and diesel-like fuel characteristics, we propose 1-octanol to be an attractive metabolic target for fuel-related applications."
} | 2,328 |
39609485 | PMC11605091 | pmc | 3,965 | {
"abstract": "The rise in global energy demand has prompted research on developing strategies for transforming conventional nonrenewable sources to cleaner fuels. Biogenic methane production is a promising source that caters to increasing energy demands. Therefore, research to enhance their production is of great importance. Implementation of successful enhancement strategies requires knowledge of the factors impacting coalbed methane production. The microbial diversity of the formation water in coal seams is the crucial parameter influencing biomethane production. This study explores microbial diversity in the Producing and Marginal wells of Bokaro, India, intending to understand the potential application of microbial-enhanced coalbed methane technology in the marginal wells of this reservoir. The high throughput sequencing analysis revealed the presence of both archaeal and bacterial groups in both well types. The result showed significant differences in the diversity of the samples from the two well groups, suggesting the immense role played by the microbes in producing methane gas. Random forest analysis shows genera Gelria, Methanothermobacter, Thaurea, Youngiibacter, and Proteiniclasticum in the Producing wells while Roseomonas, Rhodobacter, Mycobacterium, Methylobacter, and Bosea in the Marginal wells as the significant contributor in differentiating the overall diversity between the wells of Bokaro. The current study is the first to show microbial uniqueness in coalbed methane wells based on gas production efficiency. It also explores the role of physicochemical factors in framing microbial community structure in the wells. The results provide salient information that will help better understand the impact of microbial diversity on the production of coalbed methane wells of studied coal seams. This knowledge will further aid in exploring the prospects of microbial-enhanced methane in the Marginal wells.",
"conclusion": "Conclusion and future prospects The enhancement of microbial-produced coal bed methane from marginal wells carries great environmental and economic benefits. Various factors like geography, coal rank, hydrology, and microbial composition influence the coal bed methane production. The present study revealed how microbial composition and environmental factors can significantly impact methane gas production in different CBM wells within a reservoir. The findings indicate a high abundance of methane-oxidizing bacteria in the Marginal wells, while Producer wells show a high abundance of hydrogen-producing and hydrocarbon-degrading microbes. A deeper knowledge of the microbial composition in CBM wells will help better strategize improvement in gas production in the marginal wells. The findings will help design optimized biostimulation and bioaugmentation jobs in the Marginal wells that can improve their CBM production efficiency. However, further investigation utilizing transcriptomics and proteomics analysis is required for a profound understanding of microbial metabolisms in CBM wells.",
"introduction": "Introduction Regardless of efforts made globally to lessen reliance on fossil fuels, coal remains a major fuel for energy production. Since coal is the primary source of CO 2 emissions and electricity generation, switching to low-carbon energy systems presents a unique challenge. Coal Bed Methane (CBM) is an unconventional form of natural gas formed inside coal seams 1 . CBM is a clean form of energy; therefore, its development and utilization carry great social and economic benefits and provide a way forward for the global energy transition 1 . Considering the surge in energy demand and environmental perspective, CBM is a better alternative to fulfill domestic and industrial energy demands 2 . Production of methane from coal beds is biogenic and thermogenic 3 . Biogenic methane production involves complex metabolism performed by a consortium of indigenous microbes 4 . The coal seam ecosystem is an example of synchronizing hydrolyzing and fermenting microbes with methanogens. The methane generation process involves the degeneration of complex hydrocarbons to lower molecular weight organic compounds, which act as a substrate for the methanogens and are finally converted into methane gas 5 . Methanogens can be divided into three categories according to the metabolic pathway followed in methane production. Hydrogenotrophic methanogens utilize hydrogen as an electron donor to reduce carbon dioxide to methane, e.g., Methanobacterium. The acetotrophic methanogens convert acetate to methane and CO 2, e.g., Methanosarcina and Methanosaeta. Lastly, the methylotrophic methanogens convert methanol and methylamine into methane 3 . Recently, the development of biogenic CBM has picked up and is gathering more attention. Researchers are focusing on microbial diversity to understand the microbial ecosystem and their functioning in the production of CBM 6 . A recent study explored seasonal variation in coal bed water in China and reported the prevalence of phyla Proteobacteria, Bacteroidetes, and Firmicutes 3 . The comparative analysis of coal bed diversity of Gunnedah, Sydney, and Surat coals shows the abundance of Firmicutes, Proteobacteria, Euryarchaeota, Bacteroidetes, and Actinobacteria 7 . Another study exploring the Jharia coal mine's formation water diversity revealed Proteobacteria, Bacteroidetes, Actinobacteria, Verrucomicrobia, and Firmicutes as the dominant phyla in the formation water 8 . These studies established the significant role of these members in the production of biogenic methane by degradation of complex polymers and providing substrate for methane production. Worldwide efforts are being made to enhance the production of biogenic CBM. Coal provides over 58% of India's energy demands, followed by hydrocarbons at 38% and nuclear and hydroelectric power at 4% each. The high reliance creates immense pressure to limit the coal dependency 9 . A recent report suggested that harnessing 10% of coal bed methane reserves can cut India's energy import bill by two billion US dollars ( https://energy.economictimes.indiatimes.com/news/coal/harnesing-10-of-coal-bed-methane-reserves-can-cut-indias-energy-import-bill-by-2-billion-experts/100237602 ). Therefore, a deeper understanding of the microbial diversity and environmental factors that drives the CBM production process is essential. Despite the various research in this area, there is a void in understanding variation in microbial diversity of CBM wells differing in their gas production 10 . The present study utilizes metagenomics to decipher microbial communities in the Producing and the Marginal groups of CBM well in Bokaro, India. The present investigation is the first to demonstrate microbial distinctiveness in the CBM wells according to the gas production performance. The study will facilitate a better understanding of the roles of the microbes in the production of biogenic methane, which may further help in implementing field jobs for its augmentation in the wells.",
"discussion": "Results and discussion Physico-chemical analysis of formation water and coal A physio-chemical analysis of the formation water collected for this study is shown in Table 2 . The pH ranges from 7.14 to 7.87. The TDS of the formation water ranges from 320 to 4438 mg/l. Also, sulphate, chloride, fluoride, and iron were found (Table 2 ). The physicochemical analysis shows differences in various parameters in the formation water of two groups. The analysis shows high sodium, TDS, and conductivity in the Producer wells. Previous studies have shown that high TDS, salinity, and conductivity are indicative of high coal bed methane in the wells 21 – 23 . The gas phase desorption of methane can be accelerated by salinity, or the TDS in the formation water, as it can compete with methane for coal surface adsorption sites 24 . The concentration of sodium was found to be significantly higher in the Producer wells. This finding relates to another study that reported high sodium concentrations in the formation water of CBM wells 25 . Research reveals that high sodium and low calcium, magnesium, and sulfate concentrations in groundwater typically point to a high CBM enrichment potential, which is favorable for high production outcomes 26 . High sodium and TDS in the Producing wells may contribute to high CBM production in the wells. Further, most of the heavy metals in the samples were found below the detection limit or in very low concentrations, suggesting ambient conditions for the growth of the microbes. The proximate and ultimate analysis of coal samples of the seams from the Bokaro CBM wells was carried out (Table 3 ). The coal contains low moisture (0.07–0.24%), medium ash (22.35–36.3%) with 21.08–23.04% volatile matter, and 40.59–56.27% fixed carbon. Ultimate analysis results indicated low sulfur content, carbon ranging from 53.19 to 67.12%, and hydrogen at 3.08–4.09% (Table 3 ). The calorific value of the coal sample was found to be 5006–6387 kcal/kg. The analysis of all the coal samples indicates that the ASTM rank of the coal is high volatile 'A' bituminous (HVAB). The analysis showed the same category of coal in all the selected wells. General statistics for 16S rRNA sequencing and microbial diversity analysis in the producer wells and marginal wells After pre-processing and quality filtering, the resulting library size included a total of 2,435,998 sequences. A total library size of 274,498 sequences was further utilized for analysis after OTU picking. The average library size of samples was 39,214, which was further rarefied to the minimum library size. The rarefaction curve based on the observed OTU shows that the produced sequence adequately represents the present microbial communities in the wells. The rarefaction curve also depicts higher species richness in Well 47 in Producing wells, while in the Marginal wells, the highest species richness was found in Well 27 (Figure S1). The amplicon sequencing reveals the abundance of the diverse microbiome in the formation water of Producers and Marginal wells. At the domain level, both bacterial and archaeal groups were present (Fig. 2 a). The Producer wells show an abundance of Bacteria —at 99% and Archaea at 1%, whereas the Marginal wells show an abundance of Bacteria —at 97.4% and Archaea at 2.6%. The archaeal groups comprise members of the class Methanobacteria belonging to the Euryarchaeota phylum, suggesting hydrogenotrophic as the primary pathway for generating biogenic methane in the wells 27 . According to the sequence read classification, the relative abundance analysis shows the variance in the abundance of dominant phyla in the two groups (Fig. 2 b). The dominant phyla in the Producer group were Proteobacteria ( recently named Pseudomonadota), Epsilonbacteraeota, Firmicutes ( recently named Bacillota), Dictyoglomi ( recently named Dictyoglomerota), and Spirochaetes ( recently named Spirochaetota) . In contrast, the dominant members in the Marginal groups were Proteobacteria, Actinobacteria ( recently named Actinomycetota), Armanimondetes, Bacteroidetes ( recently named Bacteroidota ), and Spirochaetes . Phyla Proteobacteria, Firmicutes , and Bacteroidetes are the commonly identified bacterial phyla in the anaerobic digestive systems 28 . At the class level, Proteobacteria consists of Gammaproteobacteria, Alphaproteobacteria , and Deltaproteobacteria (Fig. 3 a). Studies have shown Gammaproteobacteria's important role in mediating the utilization of methyl-, sulfur- and petroleum organic compounds in deep ocean hydrothermal plumes 29 . Members of Deltaproteobacteria have been studied for their hydrocarbon-degrading capabilities in anaerobic conditions 30 . Figure 2 b shows the abundance of phylum Firmicutes in Producer wells. This phylum comprises members of the class Clostridia, Erysipelotrichia, and BRH_c20a . Members of the class Clostridia are known to be involved in hydrogen-producing mechanisms 31 , 32 . Hydrogen acts as the limiting factor in the hydrogenotrophic methanogenesis pathway; therefore, the presence of these members may immensely impact the gas production in the wells 33 . Members of class Dictyoglomus are abundantly present in producing wells. Dictyoglomus are extremely thermophilic, chemoorganotrophic, and obligate anaerobes 34 . Members of class Spirochaetia are involved in syntrophic acetate oxidation in anaerobic methanogenesis 35 . These variations in microbial diversity and other parameters may contribute to the difference in biogenic methane production. Previous studies have shown higher coal methanogenesis in samples with a high abundance of Firmicutes. Thus suggesting that coal methanogenesis is unlikely to be limited by methanogen biomass but rather by the activation and degradation of coal constituents 36 . The phylum Actinobacteria has been recently reported for their significance as purportedly resistant organic matter decomposers, particularly lignocellulose, and consequently, their capacity to aid in the creation of bio-based goods (energy and materials) while lowering carbon emissions has been taken into account 37 . The archaeal phyla Euryarchaeota mainly consists of class Methanobacteria , thus showing hydrogenotrophic production of methane as the prominent way of methanogenesis. This group of methanogens utilizes hydrogen as an electron donor to reduce carbon dioxide to methane 38 . Figure 2 Microbial abundance of the taxonomic groups in the Producer and Marginal wells ( a ) Pie charts represent abundance at the domain level, ( b ) Stacked bar plot showing phylum level abundance. Figure 3 ( a ) Pie charts depicting microbial abundance at the class level in Producer and Marginal well, ( b ) Heatmap showing abundance of microbes in Producer and Marginal wells at the genus level. The red line indicates abundant genera in Producer wells. The blue box indicates microbes that consume methane and hydrogen. The differential abundance of microbes is further supported by the heatmap analysis at the genus level (Fig. 3 b). The red line marks the higher abundance of genus members in the Producing wells. Genus Youngiibacter belongs to Firmicutes phylum, abundant in Producer wells, and is a newly described genus of the family Clostridiaceae . It's a strictly anaerobic bacteria that ferments a range of carbohydrates to ethanol, formate, acetate, and CO 2 39 . Previous studies have reported Rhizobium in coal and formation water samples 40 , 41 . Studies have also shown that Rhizobium members have phenol and trichloroethene degrading capabilities 42 . The members of the genus Sulfurospirillum, Proteiniclasticum , produce hydrogen and acetate as bi-products of their metabolism 43 , 44 , thus responsible for methane production enhancement. Genus Soehngenia, Proteiniclasticum , and Gelria belong to the class Clostridia . Soenhgenia is a fermenting bacterium often isolated from petroleum reservoirs 45 . Gelria is an anaerobic thermophilic bacteria obligatory syntrophic bacteria; few members are often isolated from was isolated from a propionate-oxidizing methanogenic enrichment culture 46 . Genus JGI_0000079_D21, abundant in Producer wells, is associated with the degradation of phenols and N-heterocyclic compounds in anaerobic digestion 47 . Members of Coprothermobacter grow in a protein-rich environment and are associated with hydrogen and methane production. They also have syntrophic relationships with hydrogenotrophic methanogenic archaea 48 . Methanothermobacter in the Producing wells shows the presence of thermophilic methanogens, thus carrying methanogenesis at higher temperatures 49 . The analysis also reveals the abundance of Methyloversatilis, Methylococcus, Methylocystis, and Methylobacterium in the Marginal samples. These members use methane as a substrate for their metabolism 50 . Their abundance in the Marginal group might significantly contribute to the lesser methane production in the wells. Genus Hydrogenphaga are abundant in Marginal wells and are hydrogen-oxidizing bacteria 51 . Therefore, their abundance may limit the availability of hydrogen to methanogens. A previous study has also reported their presence in CBM wells. However, their function in the wells needs more exploration 52 . The heatmap analysis also reveals variation in the relative abundance of microbial groups among the wells of the Producer group. Geographical variations, location, nutrient availability, and other factors impact the relative abundance of microbes in the formation water of coal beds 53 , 54 . The analysis also revealed variations in the abundance of methanogens ( Methanobacterium and Methanothermobacter ) among the wells of the Producer group. The variation in methanogen abundance demonstrates how, in addition to methanogen abundance, various other factors also control CBM production. Therefore, CBM production in the wells is unlikely to be limited by methanogenic mass 55 . Alpha and Beta Diversity Analysis The alpha level diversity shows the highest abundance and richness in Well#47 in the Producer wells, and in the case of Marginal wells, the highest abundance and richness was found in Well#27 (Figure S2). Beta diversity projection on the PCoA plot revealed significantly different communities between the Producers and Marginal wells at the genus level (Fig. 4 , ANOSIM, p-value < 0.05). The significant variance supports the existence of diverse microbes in both wells. This signifies microbial communities' role in producing methane from the well. The distinctiveness in the microbial diversity in the two groups is also supported by the dendrogram analysis, which shows the samples of the two groups belong to two different clusters (Figure S3). Previous studies have demonstrated distinct microbial communities residing in dissimilar hydrological areas of southern Qinshui Basin coal reservoirs 56 . Another study revealed seasonal variation in the microbial communities in the Erlian basin, China 3 . Figure 4 Beta diversity evaluation at the genus level: Principal Coordinate Analysis (PCoA) based on BrayCurtis metrics shows the dissimilarity of microbial communities in Producer and Marginal wells. [ANOSIM] R: 0.318; p value < 0.05. Differentially abundant taxa in the producer and the marginal wells The differences between the microbial communities can be understood by analysis of the differentiating members of the core microbiome. A core microbiome represents those genomes or genetic markers common to all the samples studied and is critical to the genetic functions and composition of the microbial communities 57 . Therefore, it is crucial to identify the community structures and ecological processes of the core microbiome in the CBM wells (Fig. 5 a, b). At the class level, Gammaproteobacteria, Campylobacteria, Alphaproteobacteria, Clostridia, Spirochaetia, and Methanobacteria were present in the Producer well groups. In contrast, the Marginal group shows class Gammaproteobacteria, Alphaproteobacteria, Actinobacteria, Fimbriimonadia, and Spirochaetia . Microbes belonging to these classes are widely reported in Indian coal bed wells 8 , 11 . The analysis of the core microbiome suggests the presence of hydrogen-producing ( Clostridia ), syntrophic acetate oxidative microbes ( Spirochaetia ) 58 , and hydrogenotrophic methanogens in all the samples of the Producer groups. Research has shown that syntrophic acetate oxidative microbes play a key role in hydrogenotrophic methanogenesis 59 . The analysis shows the importance of synergy among these microbes and their critical role in the process of methanogenesis in the CBM wells. Figure 5 Core Microbiome Analysis at the class level in ( a ) Producer well and ( b ) Marginal wells. ( c ) Random Forest Analysis. The top 10 genera with the highest distinct abundance between both well groups are listed. Red fields show a high abundance, and blue shows a low abundance of the particular genus in the well group. Random forest analysis shows a significant association of microbes with the Producing and the Marginal wells (Fig. 5 c) . Gelria, Methanothermobacter, Thaurea, Proteiniclasticum, Youngiibacter, JGI_0000079_D21, Coprothermobacter, and Rhizobium were significantly associated with Producing wells. The Marginal wells were found to be significantly associated with the genera Roseomonas, Rhodobacter, Mycobacterium, Methylobacterium, Bosea, Bradyrhizobium, and Limnobacter. The role of Rhodobacter and Roseomonas member s is studied in modulating the anaerobic digestion of different substrates 60 , 61 . Members of Methylobacterium use methane as a substrate for their metabolism 50 . Random forest analysis further strengthens the relationship between microbial members belonging to Clostridia and Methanothermobacter (methanogens) members with the Producer group. These findings corroborate the previous studies, which show an abundance of these microbial groups in methane-producing systems 62 63 . Relationship between microbial community and physicochemical parameters of wells The relatedness among the microbial communities and the physicochemical parameters was done using CCA analysis. The CCA analysis at the genus level supports the microbial and physicochemical distinctiveness between the Producer and Marginal wells (Fig. 6 ). Axis 1 and Axis 2 account for 63.61% of the total variance. Physicochemical parameters like sodium, conductivity, TDS, fluoride, sulfate, and chloride were found to be related to Fusibacter, Proteiniclasticum, Sulfospirillum, Youngiibacter, Coprothermobacter , Magnetospirillum . Most of these bacteria are known for their hydrocarbon-degrading capabilities. Salts can modulate the hydrocarbon-degrading ability of microbes; recent findings stated higher expression of hydrocarbon-degradation-related genes with salt addition in slurry bioreactors 64 . A recent study has shown the relation between high bulk conductivities and TDS with enhanced mineral weathering, interlinked with the activities of hydrocarbon-degrading microbes in aquifers contaminated with hydrocarbons 65 . The strong relationship between chromium and bacteria suggests its importance in microbial metabolism in the ecosystem. Research has shown the role of trace metals, including chromium, in predicting methanogenic community structure and moderate concentrations of trace metals, which are essential for microbial functioning 66 . The CCA plot also reveals relatedness among members of Methanothermobacter, Dictyoglomus, Syntrophothermus, Gelria, Rhizobium, Acetothermia, Azospira, and Desulfovibrio . This indicates the presence of microbial syntrophy, which is essential for the process of methanogenesis 67 . Figure 6 Canonical correspondence analysis (CCA) ordination plot shows the effect of physicochemical parameters of formation water on microbial community structures (genus level) in the CBM wells. The CCA analysis shows an association of zinc, manganese, magnesium, copper, iron, and pH with microbial genera abundant in the Marginal wells. The analysis also revealed the relatedness among the genera Methylobacterium, Rhodobacter, Reyranella, Methylococcus, Methylocystis, Hydrogenophaga, and others. Copper and iron are widely studied to play a role in the methane oxidation mechanism of methanotrophic bacteria 68 . The analysis shows the crucial role of physicochemical parameters of the formation water in defining the microbial composition of the wells. The findings of the study showed distinct microbial communities residing in the Producer and Marginal wells of the Bokaro region. Further, it emphasizes the role of physicochemical parameters of the formation water in framing distinct microbiomes. The results of the present study also reflect a higher diversity of degrading syntrophic, hydrogen-producing, and fermentative microbes in the Producer wells. Studies have supported the crucial role of these microbes in the process of methanogenesis 52 , 69 . Further, the results indicate a high abundance of various methylotrophs, which may be one of the significant parameters for decreased methane production in the wells. Previously, methylotrophs have been studied to significantly decrease methane production in different ecosystems 70 , 71 . Further exploration on methanotrophs is required to fully understand their role in the CBM production."
} | 6,126 |
36772177 | PMC9920654 | pmc | 3,966 | {
"abstract": "Mechanically robust ferrogels with high self-healing ability might change the design of soft materials used in strain sensing. Herein, a robust, stretchable, magneto-responsive, notch insensitive, ionic conductive nanochitin ferrogel was fabricated with both autonomous self-healing and needed resilience for strain sensing application without the need for additional irreversible static chemical crosslinks. For this purpose, ferric (III) chloride hexahydrate and ferrous (II) chloride as the iron source were initially co-precipitated to create magnetic nanochitin and the co-precipitation was confirmed by FTIR and microscopic images. After that, the ferrogels were fabricated by graft copolymerisation of acrylic acid-g-starch with a monomer/starch weight ratio of 1.5. Ammonium persulfate and magnetic nanochitin were employed as the initiator and crosslinking/nano-reinforcing agents, respectively. The ensuing magnetic nanochitin ferrogel provided not only the ability to measure strain in real-time under external magnetic actuation but also the ability to heal itself without any external stimulus. The ferrogel may also be used as a stylus for a touch-screen device. Based on our findings, our research has promising implications for the rational design of multifunctional hydrogels, which might be used in applications such as flexible and soft strain sensors, health monitoring, and soft robotics.",
"conclusion": "4. Conclusions The performance of hydrogels is directly influenced by their crosslinking structures. The creation of such hydrogels is made more difficult as a result of the fact that a mix of crosslinks is often necessary to produce the three-dimensional network of hydrogels with both the autonomous self-healing and the durability required for applications involving strain sensing. In contrast, this research makes use of a straightforward method to produce a stretchy, magneto-responsive, ionic conductive ferrogel for strain sensing. This is accomplished without the use of any crosslinking agents. The ferrogel showed the ability to self-heal when exposed to air and displayed strain sensing in real time when it was activated by an external magnetic field. Ferrogel proved to be an effective touch-screen pen, as well, according to our findings. Our discoveries might pave the way for more effective techniques for developing multifunctional hydrogels that can be used for a broad variety of applications such as flexible strain sensors and health monitors, as well as soft robotics.",
"introduction": "1. Introduction As science has progressed, wearable devices have gained popularity for use in fields including human movement detection, healthcare monitoring, and soft robotics. However, equipment designed to accommodate the growing number of smart wearable devices presents a difficulty when it comes to matching mechanical deformations such as bending, folding, twisting, and stretching [ 1 , 2 , 3 , 4 , 5 ]. Thanks to their flexibility and low weight, soft strain sensors have emerged as a leading candidate for future wearable devices. In comparison, traditional devices made from metals or semiconductors have a strain sensing range of less than 5% and are so stiff that they are uncomfortable to wear [ 4 ]. On the other hand, the current available soft, flexible strain sensors tend to fail unexpectedly and irreversibly due to mechanical stress and damage, thus losing their functionality [ 6 , 7 ]. Certain natural systems, however, have evolved means to restore themselves when they fail. The human skin is a prime example of a self-repairing natural system that can gain its initial properties after damage [ 4 , 8 ]. Inspired by this unique feature, in this research, we designed and developed a self-healing hydrogel constructed of magnetic nanochitin that can self-heal faster than human skin and might be used as a soft, flexible, self-healing strain sensor. As soft, wet, 3D crosslinked, hydrophilic polymers with unique properties, hydrogels have demonstrated their versatility in many research and industry fields, including medical applications [ 9 , 10 ], sensors [ 8 , 11 , 12 ], and water treatment [ 13 ]. However, the delicate and brittle nature of hydrogels makes their potential applications in wearable devices problematic. Thus far, many attempts have been made to solve these difficulties by mechanically reinforcing hydrogels using methods such as double networks and interpenetrating networks. Hydrogel toughening research in the last several years has shown that a number of different types of hydrogels, including nanocomposite hydrogels, double crosslinked hydrogels, and double network hydrogels, all have robust mechanical properties and can withstand significant deformation [ 14 , 15 , 16 ]. However, they ultimately break at a specific strain and irreversibly damage before any apparent fractures appear because of the presence of irreversible static chemical crosslinks inside their 3D networks [ 17 , 18 , 19 ]. Hydrogels, on the other hand, may have their networks prepared to self-repair in the case of break or damage by using reversible, dynamic physical or chemical crosslinks within their 3D structures. Hence, hydrogels fabricated by reversible, dynamic physical or chemical cross-linkers seem to be promising soft materials for flexible soft strain sensing due to their capacity to self-heal, especially when this fascinating feature is combined with ductility and conductivity [ 19 , 20 , 21 ]. However, owing to the inherent unstable mobile connections of dynamic crosslinks, ductility and toughness may drastically drop while fabricating self-healing hydrogels for soft, flexible wearable strain sensors, and the fabricated hydrogels may demonstrate a compromise between static and dynamic crosslinks in terms of mechanical strength and self-healing properties. Therefore, it is necessary to use additional chemical crosslinks in order to create a conductive hydrogel with repeatable stretchability, high toughness, and autonomous self-healing ability for strain-sensing applications [ 22 , 23 ]. In this work, a magnetic-field-sensitive hydrogel containing iron oxide (Fe 3 O 4 ) nanoparticles, called ferrogel, is synthesised with a novel magnetic remote control strain sensing ability using nanochitin as a template through in situ hydrolysis of metal precursors, imparting both autonomous self-healing and toughness to a graft copolymerised acrylic acid/starch-based ferrogel without the need for additional chemical crosslinks. Fe 3 O 4 nanoparticles have recently piqued the scientific community’s interest as biosensors [ 24 ], biolabelling [ 25 ], and targeted drug delivery [ 26 ] due to their enormous surface area, nano-size, excellent biocompatibility, low toxicity, and excellent physicochemical stability. However, agglomeration happens when Fe 3 O 4 nanoparticles are utilised in polymers due to their large surface area and high interaction energy. To address this issue, templates for the manufacture of Fe 3 O 4 nanoparticles may be used [ 27 ]. Because of its highly crystallised structure, great mechanical strength, large surface area, high aspect ratio, and abundance of functional groups accessible for modification, nanochitin, the second most abundant biopolymer on the planet, appears to be an ideal template on which Fe 3 O 4 nanoparticles can be synthesised and grafted [ 28 ]. Nanochitin is a primary precursor of cationic polysaccharides found in nature and displays biocompatibility, biodegradability, minimal allergenicity, high mechanical strength, and colloidal properties in dispersed media. It is a long-chain linear polymer with repeating β(1,4)- N -acetylglucosamine units, a derivative of glucose [ 29 ]. However, nanochitin, despite its great potential, has remained yet an underutilised biomass resource compared to nanocellulose, the most abundant biopolymer in nature. This research, hence, can pave the way for its use in performance improvement and provide ideas as a building block for the creation of new materials and alternative templates onto which other nanoparticles, such as Fe 3 O 4 , can be synthesised. In order to do so, magnetic nanochitin, synthesised from shrimp shell, was incorporated into a graft-copolymerised acrylic acid-g-starch-based network as both nanofiber reinforcement and physical cross-linker to create a tough, elastic, magneto-responsive, self-healing, ionic conductive ferrogel with excellent strain sensing properties. This method endows our ferrogel with not only exceptional self-healing properties but also the mechanical strength necessary for strain-sensing applications, in addition to ionic conductivity and magneto-responsive properties.",
"discussion": "3. Results and Discussion 3.1. Nanostructure, Microstructure, Chemical Structure, and Ferrogelation Using a stoichiometric aqueous quantity of Fe(II)/Fe(III) salt solution and an alkaline solution in a non-oxidizing environment, magnetic nanochitin was synthesised, as shown in Figure 1 a. It has been observed that the magnetic characteristics of Fe 3 O 4 nanoparticles lie between 30 and 85 emug −1 , but it tends to aggregate due to van der Waals forces among the nanoparticles to decrease their surface energy [ 31 ]. This is the key difficulty in synthesising Fe 3 O 4 nanoparticles using the co-precipitation technique. In this research, we used nanochitin as a template onto which we could graft nanoparticles of Fe 3 O 4 to minimise this behaviour. Nanochitin fibres’ abundance of surface functional groups helps keep Fe 3 O 4 nanoparticles dispersed in the suspension. When the suspension’s colour changed from yellow to black, we reckoned the reaction had taken place. More importantly, after a month in an aqueous medium, the Fe 3 O 4 nanoparticles that had been grafted onto nanochitin remained in stable suspension. When tested with a permanent magnet, magnetic nanochitin was shown to have a positive reaction, suggesting that it displays a magnetic response ( Video S1 ). The images using TEM demonstrate that the diameter of nanochitin is on the nanoscale, with an average diameter of 53 nm ( Figure 1 b). As demonstrated in the TEM image ( Figure 1 c), the functional groups on nanochitin can serve as nucleation sites for Fe 3 O 4 nanoparticle deposition, confirming that Fe 3 O 4 , instead of aggregation, had efficiently grafted onto the surface of nanochitin. Magnetic nanochitin also exhibited typical superparamagnetic behaviour with a typical S-shaped hysteresis loop, and its magnetisation rises with increasing external magnetic strength, as shown in Figure 1 d, which also reveals a magnetic saturation value of 51.47 emu/g −1 . This is another confirmation demonstrating that the Fe 3 O 4 nanoparticles were successfully grafted onto the nanochitin surface. FTIR spectroscopy was used to examine the magnetic properties created by grafting Fe 3 O 4 nanoparticles onto nanochitin ( Figure 1 d). The FTIR spectra of magnetic nanochitin and nanochitin ( Figure 1 d, I and II) show multiple similar peaks. The main distinguishing feature of chitin structures is the formation of intermolecular hydrogen bonds, which can be attributed to the peak at 1658 cm −1 as the formation of intermolecular hydrogen bonds CO…HN. The peak at 1179 cm −1 can be assigned to amide I, which is caused by the flexural bending of the NH bonds, whereas the peak at 1560 cm −1 can be assigned to amide II. Additionally, the C-O-O bond vibration inside the chitin rings was linked to the 1075 cm −1 peak. The typical bandwidth of magnetic nanochitin is about 507 cm −1 , as seen in Figure 1 d, I. This sheds light on the existence of Fe 3 O 4 nanoparticles on the surface of nanochitin and the successful interactions between the functional groups of nanochitin and Fe 3 O 4 nanoparticles [ 32 ]. Finally, we fabricated the ferrogels by graft radical copolymerizing starch and acyclic acid monomers in the presence of previously synthesised magnetic nanochitin in situ. As just said, the 507 cm −1 absorption peak is identical to the Fe-O bond in magnetic nanochitin, and the adsorption of negatively charged polyelectrolytes onto iron oxide nanoparticles slightly shifts this band to a new high at 514 cm −1 . The presence of the intense bands at 1713 1 and 1743 cm −1 in the spectra of magnetic nanochitin-loaded starch-g-acyclic acid may be attributed to the presence of carbonyl groups, revealing the successful interaction of magnetic nanochitin with the polymer chains. It also functioned as a crosslinking agent in the construction of a 3D network structure that could trap water molecules through interactions between polymeric chains and Fe 3 O 4 nanoparticles. Hence, it may replace N,N′-methylene bisacrylamide, which is an irreversible static chemical cross-linker during the fabrication of starch-g-acyclic acid ferrogel. A permanent magnet was used to mimic the existence of a magnetic response in the ferrogel, and it was revealed that the ferrogel, due to the presence of Fe 3 O 4 nanoparticles, can swim in water when directed by magnetism. As a result, magnetic nanochitin gave our ferrogels magnetic characteristics that can attract a typical permanent magnet ( Video S2 ). The negatively charged carboxyl groups in poly(acyclic acid) have been theorised in the literature to be adsorbed onto the exposed Fe atoms on the surface of Fe 3 O 4 nanoparticles grafted on nanochitin, establishing stable ionic bonds [ 33 , 34 ]. Figure 1 e depicts the possible coordination of a carboxyl group with an iron atom (I, II, and III). The positively charged H + ions may also be absorbed by the oxygen atoms on the surface of the Fe 3 O 4 nanoparticles. If the remaining carboxylic/carboxylate groups in starch-g-acyclic acid/magnetic nanochitin interact with OH/OH 2+ as the attractive force, structure IV in Figure 1 e is conceivable. As a result, we hypothesise that the physical crosslinking and ferrogelation of acyclic acid-g-starch/magnetic nanochitin are caused by ionic or dipole interactions at ambient temperature and neutral pH [ 34 , 35 ]. By varying the concentration of magnetic nanochitin, the porous structure of ferrogels is exhibited in Figure 1 f–i. The magnetic nanochitin concentration was shown to be related to narrower pore diameters with an average pore size of 39 μm in ferrogel containing 1 wt.% magnetic nanochitin and 17 μm in ferrogels containing 2 wt.% magnetic nanochitin, most likely due to enhanced crosslinking in the ferrogels. 3.2. Mechanical Properties, Self-Healing, and Notch Insensitivity During the ferrogelation, we demonstrated that magnetic nanochitin might be used as a cross-linker alternative to N,N′-methylene bisacrylamide. In addition, it has been suggested that adding nanofillers to the starch network may boost the material’s tensile strength. Given that starch is naturally a brittle polymer, graft copolymerisation of vinyl monomers, such as acrylic acid, can be a straightforward and efficient way to increase the mechanical strength of starch. Here, we conducted tensile stress experiments on rectangular shape samples to demonstrate the effect of magnetic nanochitin on the mechanical characteristics of starch-g-acrylic acid ferrogels. The results of mechanical tests for the ferrogels containing 0.1, 0.5, 1, 1.5, and 2 wt.% of magnetic nanochitin are shown in Figure 2 . As seen, by increasing the concentration of magnetic nanochitin, the ultimate tensile strength ( Figure 2 a) and Young’s modulus ( Figure 2 b) has steadily increased, indicating that magnetic nanochitin has a proper impact on reinforcing the mechanical properties of the ferrogels. The elongations at break of the ferrogels have also increased, leading to a stretchability of 1341% at 1.5 wt.% magnetic nanochitin. With more than 1.5 wt.% magnetic nanochitin, however, the elongations at break decreased slightly, most likely due to the presence of some agglomerations inside the ferrogel’s network, which functions as stress concentration points and reduces the ferrogels’ stretchability under strain. Measuring toughness using the mechanical test was performed on all ferrogels at a variety of magnetic nanochitin concentrations (0.1, 0.5, 1, 1.5, and 2 wt.%) to determine which hydrogel had the highest toughness since this is one of the most important criteria for a wearable strain sensor. The toughness of the ferrogel reinforced with 1.5 wt.% magnetic nanochitin is 1.31 MJ/m 3 . However, the toughness decreases with increasing concentrations of magnetic nanochitin, which may be owing to an increased degree of crosslinking ( Figure 2 d). We hypothesised that polymer chains would be able to entangle together over the surface of a newly cut ferrogel and form inter-chain bonds with Fe atoms since no chemical static cross-linker was used in the manufacturing of our suggested ferrogels. Here, we postulate that the carboxyl groups would adsorb onto the exposed Fe atoms of the surface of Fe 3 O 4 via three discussed dynamic potential coordination bonds (I, II, and III), or that the surface oxygen atoms of the Fe 3 O 4 would be able to take in H + ions, where the surface of OH/OH 2+ species can participate with the remaining carboxylic/carboxylate groups in the ferrogel to form structure IV in Figure 1 e. It eventually leads to the dynamic behaviour of ferrogels by which the damaged section of the ferrogel can be healed and bonded together again at room temperature without the need for any external stimulation, additives, or force. Ferrogels with varying amounts of magnetic nanochitin self-healed completely and coalesced into a single structure. Figure 2 e shows the optimum mechanical strength efficiency of 93% for the ferrogel containing 1.5 wt.% magnetic nanochitin. This illustrates that after 6 h of self-healing, the new connections at the interface are robust enough to span the space between two sliced ferrogels. Furthermore, as shown in Video S3 , after 6 h of self-healing of the ferrogel containing 1.5 wt.% magnetic nanochitin with multiple damages and cuts, this ferrogel is incredibly twistable and stretchy. These results support the hypothesis that ferrogels may self-heal while maintaining great performance. Next, notches were applied to each of the ferrogels in order to evaluate the notch sensitivity of the starch-g-acrylic acid/magnetic nanochitin ferrogels at varying concentrations. This was performed before the ferrogels were stretched under tensile stress. According to the findings, the ferrogel that had been fabricated with 1.5 wt.% magnetic nanochitin had the highest level of notch insensitivity and maintained an impressive level of stability and blunting throughout the whole experiment. One of the reasons why ferrogels’ toughness behaviour may make them resistant to cracking is because of the presence of dynamic, reversible bonds in the ferrogels. It was determined that the ferrogel containing 1.5 wt.% magnetic nanochitin would be the best sample with which to continue the research based on the outcomes of the mechanical, notch insensitivity, and self-healing tests. These tests determined that the material was not sensitive to notches and may have the needed resilience for strain-sensing applications without the need for additional irreversible static chemical crosslinks. 3.3. Conductivity and Strain Sensing If a soft, notch-insensitive, tough, and stretchy ferrogel can maintain conductivity, then it might be an excellent candidate for being employed in flexible electronics. Given the large number of ionic connections that are present in ferrogels, it was reasonable to assume that our ferrogels would exhibit ionic conductivity. Ferrogels with varied concentrations of magnetic nanochitin were subjected to a current monitoring system that was equipped with an LED light indicator. This system displayed the ionic conductivity, strain sensitivity, and electrical self-healing capabilities of the ferrogels in real time. When the ferrogels were affixed to the device, the LED light began to illuminate. It was observed that applying stress to the ferrogel and stretching it caused the LED’s brightness to drop; however, once the tension was released, the brightness of the LED reverted to normal. It is possible that the occurrence might be explained by the fact that the stretching process induces a distortion in the network structure of the ferrogel ( Video S4 ). The ferrogel was then deformed and stretched, which caused the contact area between the ionic bonds that make the material conductive to decrease. This caused the material to become less conductive. As a consequence of this, the resistances of the original ferrogel and the stretched ferrogel were distinct, as shown by the ratio (R − R 0 )/R 0 in Figure 3 a. Due to the self-healing and rapid reconstruction of the 3D conductive network of the ferrogels, the electrical current in the ferrogels was able to quickly recover after being cut in half, as was demonstrated by the LED light in Figure 3 b. In particular, the cut ferrogel restored its ionic conductivity immediately due to the instantaneous self-healing of the cut area, which suggests that ferrogels are well suited for applications such as emergency circuit repair, electrical circuit building, switching functions, and electric lines ( Figure 3 b). The ferrogel also showed its efficacy as a bendable and soft strain sensor by connecting to an artificial wooden elbow at various angles and strain sensing in real-time. As the strain and bending angles were changed, the strain sensor’s resistance increased to varying degrees. Figure 3 c reveals that the internal 3D network of the ferrogels can quickly repair itself from damage caused by the stretching, and the strain sensor returned to its initial value by revealing the same values for the relative resistance (R − R 0 )/R 0 after the elbow were straightened, proving a consistent, measurable electrical stability for detecting resistance in real-time at different bending degrees. If the strain was maintained at the same level for an extended period of time, the resistance would stay at the same level ( Figure 3 c). One of the most crucial factors in establishing the sensitivity of a strain sensor is the gauge factor, often known as GF. This may also accurately reflect the sensing capability of the sensor. The GF for this ferrogel was calculated by plugging the available data into the formula GF = (R/R 0 )/ε. Strain is represented by ε, while R 0 and R indicate the resistance at zero strain and the relative resistance variation with stretch, respectively. Figure 3 d displays the GF values of 0.5 and 1.40 for a 10–120% tensile strain. The GF of the ferrogel has been shown to be within the range that is considered optimal for strain-sensing applications ( Table 1 ). Therefore, we hypothesise that the ferrogel might satisfy the needs of a strain sensor. Because the ferrogel has magnetic properties, it is possible to use it as a magnetic sensor. In order to evaluate the ferrogel’s capability of detecting magnetic strain, it was placed between the probes of a multimeter and then exposed to an external magnet ( Figure 3 e). As seen, the ferrogel was able to quickly sense strain and respond to external actuation by moving and bending as a result of this property. When an external magnetic field was applied to the ferrogel, the changing resistance, as well as the real-time resistance track, was revealed. If the strain sensor returns back to where it was before the external magnet was removed, this indicates that the electrical stability that is necessary for sensing resistance in real-time has been restored ( Figure 3 f). Ferrogel has the potential to imitate the look, feel, and function of human skin when it is employed as an electronic skin. Ferrogel’s bright future as a prosthetic finger in flexible electronics, electronic skin for robotics, and human–machine interfaces is shown by utilising it as a touch-screen pen to draw PEJMAN letters on an iPad. This demonstrates Ferrogel’s potential applications in these areas ( Video S5 )."
} | 6,023 |
26840425 | PMC4740408 | pmc | 3,967 | {
"abstract": "Marine chlorophytes of the genus Chlorella are unicellular algae capable of accumulating a high proportion of cellular lipids that can be used for biodiesel production. In this study, we examined the broad physiological capabilities of a subtropical strain (C596) of Chlorella sp. “SAG-211-18” including its heterotrophic growth and tolerance to low salt. We found that the alga replicates more slowly at diluted salt concentrations and can grow on a wide range of carbon substrates in the dark. We then sequenced the RNA of Chlorella strain C596 to elucidate key metabolic genes and investigate the transcriptomic response of the organism when transitioning from a nutrient-replete to a nutrient-deficient condition when neutral lipids accumulate. Specific transcripts encoding for enzymes involved in both starch and lipid biosynthesis, among others, were up-regulated as the cultures transitioned into a lipid-accumulating state whereas photosynthesis-related genes were down-regulated. Transcripts encoding for two of the up-regulated enzymes—a galactoglycerolipid lipase and a diacylglyceride acyltransferase—were also monitored by reverse transcription quantitative polymerase chain reaction assays. The results of these assays confirmed the transcriptome-sequencing data. The present transcriptomic study will assist in the greater understanding, more effective application, and efficient design of Chlorella -based biofuel production systems.",
"introduction": "Introduction As the market for biofuels expands, photosynthetic microalgae can be potentially employed in third-generation biofuel production at commercial scales [ 1 ]. Single-celled algae have received heightened attention because of their fast growth rates, high photosynthetic efficiencies, and rapid lipid accumulation [ 2 , 3 ]. Many algal species are capable of biosynthesizing and storing neutral lipids composed primarily of triacylglycerides (TAGs) [ 4 – 6 ], which can be extracted and industrially transesterified into biodiesel [ 7 ]. Algae bioaccumulate TAGs as a means to store excess energy when cells experience a non-carbon nutrient limitation that prevents cell doubling (i.e., phosphate or nitrate limitation) [ 8 , 9 ]. Therefore, the accumulation of TAGs is decoupled from the exponential growth of the organism, and genes encoding enzymes contributing to TAG accumulation are tightly controlled [ 10 – 13 ]. The accumulation of lipids by individual species of chlorophytes or green algae under environmentally stressful conditions is well established in the literature (e.g., [ 14 , 15 ]). Within the green algae, for which we have a well-established genetically tractable model system ( Chlamydomonas reinhardtii ), representatives of the Chlorella genus have been shown to bioaccumulate TAGs efficiently. Chlorella strains have therefore been investigated for potential biodiesel production at commercial scales [ 16 , 17 ], and several industrial Chlorella applications for biofuel production have been reported [ 18 , 19 ]. However, limited genomic information is available for the Chlorella genus. Three full genomes for Chlorella are currently available in the NCBI database: Chlorella variabilis NC64A [ 20 ], a photobiont initially sequenced to study viral/algal interactions, Chlorella vulgaris , and Chlorella pyrenoidosa [ 21 ]. One additional strain in a related genus, Auxenochlorella protothecoides , was recently sequenced [ 22 ]. Additional Chlorella sequencing projects of various levels of completeness have been deposited in other databases (such as greenhouse.lanl.gov/ ). To yield information about the physiology of other Chlorella species without a priori knowledge of the genome, advanced sequencing technologies (e.g., Illumina RNA-seq) and sequence assembly methods allow high-throughput de novo transcriptomic analyses. Most previous transcriptomic studies of various Chlorella species utilized targeted transcriptomic approaches such as reverse-transcription quantitative PCR (RT-qPCR) to monitor pre-selected targets [ 23 – 25 ]. However, analyses of the full transcriptome using next-generation sequencing technologies are more useful because they can provide a global overview of the response of Chlorella cultures to various environmental states, and potentially link the expression of specific and large sets of transcripts with phenotypic responses. This approach is an emerging and widely applied tool that has been used to investigate the differences between autotrophic and heterotrophic growth in A. protothecoides [ 22 ], the growth of C. pyrenoidosa under salt stress [ 26 ], the transition from starch to lipid synthesis in C. pyrenoidosa [ 21 ], the growth of Chlorella sorokiniana under high carbon dioxide concentrations [ 27 ], and to survey the transcriptome of Chlorella minutissima UTEX2341 in general [ 28 ]. In the current study, we have focused on a new Chlorella strain (C596) obtained from the University of Hawaii and selected from among hundreds of isolates as a promising candidate for biodiesel production (Johnson et al., in review [ 29 ]). This study had three main goals: (1) to investigate the growth characteristics of C596, (2) to elucidate key metabolic pathway genes including the genes responsible for lipid biosynthesis, and (3) to discover specific genes that are involved in TAG accumulation. To better understand the growth characteristics of this marine algae, strain C596 was grown using media of various salt concentrations (35, 17, 8.8, and 4.4 parts per thousand (ppt)) and media enriched with organic substrates (acetate, glycerol, glucose, sucrose, or succinate) in the dark. To elucidate genes involved in the biosynthesis of lipids, strain C596 was subjected to a co-limitation of nitrogen and phosphorus to promote subsequent TAG accumulation, and the transcriptome expressed by the culture was monitored during a sequential co-limitation of phosphorus and nitrogen using Illumina RNA-seq technology. To determine specific genes involved in TAG accumulation, comparisons between RNA pools highlighted differentially expressed genes responding to nutrient stresses, and follow-up RT-qPCR studies confirmed several of these findings.",
"discussion": "Results and Discussion Phylogenetics To confirm the phylogenetic assignment of this strain, the 18S rRNA sequence of Chlorella strain C596 was sequenced and aligned with the 18S rRNA sequences of other representative freshwater and marine strains in the Chlorophyta division and with one model marine diatom, Thalassiosira pseudonana . Using sequence similarities, a phylogenetic tree was constructed ( Fig 1 ). Chlorella strain C596 is closely related to other Chlorella strains within the division Chlorophyta, in the class Trebouxiophyceae (as previously defined; [ 30 , 52 ]). The most closely related strain is C. sorokiniana (98% 18S rRNA sequence identity), a free-living marine strain and a good lipid producer [ 53 ]. Also based on its position within this phylogenic tree, our strain appears to be closely related to the five other members of the Chlorella genus identified as good lipid accumulators in a recent survey of 37 lipid-rich microalgae [ 54 ]. 10.1371/journal.pone.0147527.g001 Fig 1 18S rRNA phylogenetic tree for marine and freshwater green algae and a marine diatom outlier. The asterisks indicates organisms with available genomes in the NCBI database. The representative full-length 18S rRNA sequences were aligned with MUSCLE in MEGA 6. Growth Characteristics Under optimized growth conditions (temperature and light with CO 2 supplementation), strain C596 averaged a specific growth rate of 1.64 d −1 during the nutrient-replete exponential growth phase, began to accumulate TAGs as the external reserves of nitrate (then phosphate) were exhausted, and then entered the stationary growth phase ( Fig 2a ). In the current study, the culture bioaccumulated 225 ± 3 mg L −1 (46.5 ± 1.8 mg L −1 d −1 ) of total lipids and 94 ± 2 mg L −1 (26.2 ± 0.4 mg L −1 d −1 ) of TAGs within 72 h of nutrient depletion. Estimates of total lipid and TAG content relative to dry weight in C596 for this experiment are 66% and 28%, respectively, assuming an average cellular radius of 2.5 μm ( Fig 2b–2g ) and a dry weight fraction of 20%. Previously, the maximum lipid accumulation rate in this strain was reported to be as high as 61 mg L −1 d −1 (in cells that were approximately 45% TAG by weight) [ 55 ]. These rates approach those obtained with Nannochloropsis sp. grown under optimized conditions (75–300 mg L −1 d −1 ; [ 56 ]) and are at the high end of the range of lipid and TAG levels relative to dry weight reported for algae generally and Chlorophytyes specifically (5–50% lipid by dry weight; [ 6 ]). 10.1371/journal.pone.0147527.g002 Fig 2 Lipid accumulation in Chlorella strain C596. (a) Time-course of a batch experiment culminating in N and P co-limitation (N:P = 13:1): nitrate+nitrite (black, solid), phosphate (black, dashed), cells (blue), TAG (red, solid), and total lipids (red, dashed) concentrations; error bars represent the 95% confidence interval. (b-g) Fluorescence and phase contrast microscopy of strain C596 cells grown in batch culture culminating in P limitation (N:P = 30:1). Samples were withdrawn during the exponential phase (b-d) and four days after onset of stationary (e-g) growth phase. The fluorescence microscopy reveals the chloroplast (autoflourescence, bright white) and Nile red stained lipids (in c and f; grey). The red arrows and brackets indicate intercellular TAG locations. The white arrows highlight the chloroplast. The accumulation of lipids is also apparent in the microscopy images of C596 cells from a batch culture grown with excess nitrogen (N:P = 30:1; Fig 2b–2g ). During the exponential growth phase, chlorophyll auto-fluorescence is observable as intense white portions of the cell ( Fig 2b and 2c ) which diminish in size and intensity on reaching the phosphorous-induced stationary phase of growth ( Fig 2e and 2f ). Nile red was used to stain the internal lipids, corresponding to the lighter gray droplet of TAG readily apparent in the cells during stationary growth ( Fig 2f ), at which point the chloroplasts have been pushed to the edge of the cell by the TAG-rich lipid bodies. Chloroplast migration is also evident in the phase contrast images ( Fig 2g ). Similar observations of the chloroplast being pushed to the surface of the cell during lipid accumulation have been reported for C. variabilis NC64A cells subjected to nitrogen deprivation [ 16 ] and for Nannochloropsis [ 57 ]. Heterotrophic Growth and Salt Response Chlorella strain C596 can tolerate a wide range of salinities, but grew slower at reduced salinities ( Fig 3a ). The specific growth rates achieved by C596 in the reduced-salt Aquil media were maintained for three subsequent generations (data not shown). The highest growth rate of this Chlorella species was obtained at a salinity level of 35 ppt, the salinity of our synthetic Aquil media [ 34 ]. This salinity value is higher than the optimal value of 20 ppt reported for A. protothecoides [ 58 ]. However, higher salinities (up to 60 ppt) do not substantially reduce the intrinsic growth rate of strain C596 under nutrient replete conditions (Dr. Z. Johnson, personal communication, May 23, 2015). 10.1371/journal.pone.0147527.g003 Fig 3 Specific growth rates of Chlorella strain C596. Growth rates (d -1 , based on cell density over time) are shown for (a) different salinities and (b) various carbon substrates. White bars indicate photosynthetic growth with a PPFD of 80 μmol photons m −2 s −1 ; gray bars indicate grown in the dark plus a carbon source at a final concentration of 20 mmol L −1 ; the black bar indicates growth in the dark. Values are the means of triplicate biological replicates (n = 3); error bars represent ±1 standard error (SE). In the dark, strain C596 was able to grow using 20 mmol L −1 glucose, sucrose, succinate, and glycerol at heterotophic growth rates that rivaled the photosynthetic growth rate of ∼1 d −1 ( Fig 3b ). Only acetate resulted in statistically lower growth rates of 0.6 d −1 ± 0.2 (47% less than phototrophic rates; p-value < 0.03, Student’s t-test). Additionally, glucose supported the fastest heterotrophic growth rate compared to the other carbon substrates (36% greater with glucose than succinate, p-value < 0.003, Student’s t-test), and was 23% greater than photoautotrophic growth at the identical temperature but with lower than optimal PPFDs (p-value < 0.1, Student’s t-test). Both C. sorokiniana and C. vulgaris have been shown to grow heterotrophically on glucose as the sole carbon and energy source, with 29% and 26% improved growth rates for C. sorokiniana and C. vulgaris , respectively, for cultures grown with 27 mmol L −1 glucose compared to purely photosynthetic growth [ 59 , 60 ]. The photoautotrophic growth rates of strain C596 (∼1.2 - 1.6 d −1 ) reported in the present study are at the high end of previously reported growth rates for other Chlorella species (e.g., C. sorokiniana , 1.02 - 1.6 d −1 ) [ 61 ]. Though we did not investigate the effect of heterotrophic growth on lipid accumulation in C596, C. vulgaris was noted to have a higher cellular lipid content under autotrophic growth whereas higher lipid productivity (on a mass per day basis) under heterotrophic growth [ 62 ]. Our results reveal strain C596 to be a versatile mixotrophic algae strain capable of utilizing a broad range of organic substrates over a wide range of salt concentrations—desirable characteristics in an alga for industrial biofuel applications. Transcriptome Annotation To link the accumulation of lipids to the expression of certain transcripts, we sequenced the transcriptome over a time course when nitrogen was exhausted and TAG accumulation occurred ( Fig 2 ). Our base library was constructed from 11.4 million read pairs (5.7 Gbp of sequence), and the assembled transcriptome resulted in 131,470 transcripts. Of these assembled transcripts, 13,036 were treated as detected transcripts (at least five transcripts in at least one sample; S2 File ). The sequences in the detected transcriptome were searched against NCBI’s nr database using the BLASTx algorithm. All but 254 of these sequences displayed at least one hit to the nr database with an e-value cutoff of 10 −5 . In total, 137,484 significant hits were returned, of which C. variabilis (19,027 hits), Coccomyxa subellipsoida (15,828 hits), Volvox cartei (9,772 hits), and C. reinhardtii (9,283 hits) were the top four algae in rank order of the number of hits and combined comprised 39.2% of the total significant hits ( A. protothecoides was not available in this database). In total, 8,903 transcripts in our library were identified as displaying homology to a Chlorophyta species. These 8,903 were considered as the expressed transcriptome in this study. To estimate what fraction of the full transcriptome was detected as expressed transcripts in this study, we compared the expressed transcriptome to three other fully sequenced species in a tBLASTn analysis: C. variabilis (which returned the most hits in the BLASTx analysis above), Chlorella UTEX1228 (the complete list of predicted proteins is available at greenhouse.lanl.gov ), and A. protothecoides (a closely related genus [ 22 ]). In this analysis, the expressed transcriptome displayed sequence homology to 79% (7,705 out of 9,791 predicted transcripts with an average hit length of 213 amino acids (aas)), 64% (7,760 out of 12,169 predicted transcripts with an average hit length of 302 aas), and 77% (5,752 out of 7,431 predicted transcripts; with an average hit length of 222 aas) of all of the transcripts predicted for C. variabilis , Chlorella UTEX1228, and A. protothecoides , respectively. These results suggest that our transcript library captured a large and representative fraction of genome-encode transcripts in Chlorella strain C596. We estimate that the size of the C596 genome is approximately 30 Mbp (unpublished), between that of A. protothecoides (22.9 Mbp; [ 22 ]) and C. variabilis NC64A (46.2 Mbp) or C. pyrenoidosa FACHB-9 (56.8 Mbp) [ 21 ]. Differentially Regulated Transcripts between Nutrient-Replete and Nutrient-Limited Cultures The transcriptional regulation of biochemical pathways in response to changes in environmental stimuli was monitored using the RNA-seq data; in particular we sought to ascertain which transcripts were significantly regulated during the transition to nutrient-limited growth and TAG accumulation. A global edgeR analysis was performed on the time course samples which included biological duplicates from three time-points and comprised 4.1 million read pairs (2.1 Gbp of sequence). Applying all of the filters described in the Materials and Methods, RNA-seq data from these samples yields a subset of the expressed transcriptome (6,554 of the 8,903 transcripts). When the 86 h sample (external concentrations of nitrate and phosphate still considered saturating for growth, Fig 2 ) was compared to the 109 h sample (the culture during the transition phase), 397 of these transcripts were differentially expressed (greater than two fold-ratio change and a p-value < 0.05). When comparing the transcriptome of the organism during the transition phase (109 h samples) to the nutrient-depleted (182 h) samples, the edgeR analysis identified 2,023 transcripts as significantly differentially expressed, of which 1,099 were up-regulated and 924 were down-regulated ( S3 File ). A more pronounced shift in gene expression was thus revealed for the 182/109 h comparison relative to the 109/86 h comparison (3.3 times more genes were differentially expressed). The substantial RNA turnover in later cultures (182 h) is supported by the Bioanalyzer traces of the extracted total RNA. We note that when Chlorella strain C596 transitioned from the exponential growth phase (86 and 109 h) to the stationary phase (182 h; S2 Fig ), the RNA pool displays dampened dominant rRNA peaks. However, this degradation is not noted in late stationary phase samples, suggesting that the degradation of RNA in the early stationary phase is representative of a high amount of RNA turnover in preparation for stationary phase and not an artifact of the extraction process. The 182/109 h comparison was explored further to identify transcripts that significantly respond to nutrient limitation and TAG accumulation. The sequences for the transcripts that were significantly regulated between the 182 h and 109 h time points were compared against the KEGG annotations for C. variabilis in a tBLASTn search. In total, 475 of the 2,023 differentially expressed transcripts were assigned KEGG categories in this analysis (the unmapped transcripts had no associated KEGG category or did not map to C. variabilis ). The KEGG-assigned transcripts were binned into five top-level Groups in the KEGG hierarchy (Organismal Systems, Cellular Processes, Environmental Information Processing, Genetic Information Processing, and Metabolism) which were sub-divided into 17 Categories, and 111 Sub-categories (out of a possible 118 identified for C. variabilis ). The numbers presented in Fig 4 are the number of unique differentially expressed transcripts matching a KEGG organizational level. Because of the potential multiple assignment of KEGG sub-categories per transcript (the 475 transcripts were assigned an average of 2.5 categories per transcript because enzymes map to multiple pathways), the values for the total number of unique transcripts up- or down-regulated in the category level are not the simple sum of the sub-category level. 10.1371/journal.pone.0147527.g004 Fig 4 Differentially expressed genes between the 182 h versus 109 h samples broken into KEGG categories. Genes are considered differentially expressed when the edgeR p-value < 0.05 and the fold-ratio > 2 or < 0.5. The assignment of a transcript to a category is based upon the KEGG annotation of orthologs in C. variabilis NC64A. The axis of the dot plot displays the ratio of 182/109 h on a log scale. The numbers in the category column represent the number of genes down-regulated (-), up-regulated (+), and the difference between these (Δ). The color shading is based on the Δ (positive, red; negative, blue). A number of notable metabolic shifts are reflected in the gene expression profiles after strain C596 entered into nutrient limitation ( Fig 4 ). Within the Metabolism KEGG Group, the majority of genes associated with Carbohydrate Metabolism (23 down, 53 up) and Lipid Metabolism (10 down, 27 up) categories were up-regulated. Additionally, the majority of the transcripts identified in the photosynthesis sub-category (34 down, 6 up) and the photosynthesis—antenna proteins sub-category (14 down, 0 up) within the Energy Metabolism Category are down-regulated. Taken together, these results suggest that strain C596 is scaling back the photosynthetic apparatus when substantially enhancing lipid and carbohydrate metabolism. A similar pattern was seen in a closely related species, A. protothecoides , which also displayed up-regulation of carbon metabolism-related transcripts and down-regulation of photosynthesis related transcripts on entering the TAG accumulation phase under heterotrophic growth when compared to autotrophic growth [ 22 ]. The TAG accumulation phase in algae is known to be decoupled from organismal growth ( Fig 2 ; [ 63 ]). The differential response of various categories within the Genetic Information Processing Group highlights this decoupling ( Fig 4 ). Within the Translation Category (97 down, 20 up), ribosome associated transcripts are significantly down-regulated, suggesting that fewer ribosomes are being produced and matured. Additionally, within the Folding, Sorting, and Degradation Category, transcripts related to RNA degradation are up-regulated (4 down, 15 up), corroborating a higher RNA turnover in the 182 h sample compared to the 109 h sample ( S2 Fig ). TAG Synthesis Pathway Regulation Our KEGG based analysis only focused on genes that were differentially regulated and were annotated with a KEGG descriptor based on homology with C. variabilis NC64A. Broadening the analysis to consider all transcripts detected in the strain C596 samples that displayed homology to a Chlorophyta protein in the NCBI nr database, we identified genes associated with the last steps of the TAG synthesis pathway ( Fig 5a ). In Fig 5a , the circle, square, and triangle represent the presence of sequences predicted to encode for enzymes that catalyze each of these steps in strain C596, C. variabilis NC64A, or C. reinhardtii [ 64 ], respectively. For the majority of the steps in Fig 5a , the sequences were present and displayed strong homology in all three strains. However, the Chlorella strains deviated from the Chlamydomonas in the sequence composition of the glycerol-3-phosphate acyltransferase; the Chlorella strains display a bacterial type homolog whereas C. reinhardtii displays a plant-like homolog. 10.1371/journal.pone.0147527.g005 Fig 5 Detection and monitoring of the TAG biosynthesis pathway in Chlorella strain C596. (a) The final steps of the TAG biosynthesis pathway. The symbols indicate whether one or more transcripts encode the enzyme at that step is present for Chlorella strain C596 (circle), C. variabilis NC64A (square), or C. reinhardtii (triangle). Grey shading indicates presence. Lined shading for the glycerol-3-phosphate acyltransferase indicates a bacterial-type homolog. Lined shading for the diacylglycerol acyltransferase indicates a secondary homolog in C. reinhardtii . (b) Heatmaps of the cyclophillin-normalized values for the putative transcripts for the TAG accumulation pathway and galactoglycerolipid lipase in samples taken during exponential growth (89 and 109 h) and during lipid accumulation (182 h). The intensity of the shading represents cyclophillin normalized values ranging from 0 to 1.8; note, one transcript, c10691_g3, falls outside of this range and the appropriate box is labeled. The * symbol represents those transcripts for which RT-qPCR primers were designed. The † symbol represents the transcript whose ratio could not be calculated because the transcript was not detected in the 109 h sample. (c) Comparison of the RNA-seq (top row) and RT-qPCR results also normalized to cyclophillin (bottom row) for the predicted DAGAT (left column) and galactoglycerolipid lipase (right column) for two biological duplicates (a and b). The error bars indicate the 95% confidence interval of the technical replicates run for the RT-qPCR. The cyclophillin (a housekeeping gene) normalized time-course for the transcriptomic abundance of these TAG biosynthesis pathway homologs is displayed in Fig 5b . The results obtained using actin as the housekeeping gene were quantitatively similar (data not shown). As the organism enters the TAG accumulation phase, transcripts representing the final biosynthesis steps for TAG and the galactoglycerolipid lipases are up-regulated. Multiple potential paralogs are observed in C596 for several of these steps. This is also true in C. reinhardtii which was noted to have two 1-acylglycerol-3-phosphate acyltransferases, three phosphatidate phosphatases, six DAGATs and three lipases [ 65 ]. The transcripts encoding for the enzymes of the first two steps (c5929_g1, glycerol kinase; c11859_g1, glycerol-3-phosphate acetyltranseferase) displayed a high up-regulation (18 and 13 fold increases, respectively) ( Fig 5b ). Additionally, the transcripts encoding for two enzymes often explored in other studies, c10691_g3 (a galactoglycerolipid lipase paralog) and c10184_g1 (a diacylglycerol acyltransferase paralog; DAGAT), were also up-regulated (4.5 and 14 fold, respectively). DAGAT is responsible for the final step in TAG biosynthesis, and a substantial up-regulation was also noted for DAGAT (approximately a 140 fold-change) under nitrogen limitation in another study [ 16 ]. Additionally, the galactoglycerolipid lipase is predicted to be critical in the final formation of TAG from galactolipids in C. reinhardtii [ 64 , 66 ], but has not been previously identified as playing a role in Chlorella . A similar pattern of up-regulation of lipid biosynthesis genes was noted in C. sorokiniana when the organism experienced higher carbon-dioxide partial pressures [ 27 ]. Primers for the most strongly up-regulated galactoglycerolipid lipase and DAGAT transcripts were designed, and a RT-qPCR analysis on the identical samples was run to confirm the results of the RNA-seq analysis with respect to the response of the expression of these genes to nitrogen starvation. The RT-qPCR analysis confirmed that both the galactoglycerolipid lipase and DAGAT are up-regulated in each of the biological duplicates ( Fig 5c ). Both methods displayed a higher overall transcript abundance for the galactoglycerolipid lipase than for the DAGAT, but the cyclophillin normalized ratios of the galactoglycerolipid lipase and DAGAT between the 182 h to the 109 h time points determined via the RNA-seq analysis (4.5 ± 0.5 and 14 ± 13 fold higher, respectively) were 3.6 and 5.8 times lower than the ratios obtained using the RT-qPCR analysis (16 ± 14 and 81 ± 49 fold higher, respectively). Previous studies comparing RNA-seq ratios and RT-qPCR ratios in Chlamydomonas noted that the RNA-seq ratios were, on average, approximately four times lower than the reported RT-qPCR ratios [ 67 ]. The consistent up-regulation of these transcripts in both the RNA-seq analysis and the RT-qPCR analysis emphasizes their potential role in TAG accumulation. In green algae, DAGAT is predicted to be and enzymatically shown to be the terminal enzyme in TAG biosynthesis, transferring an additional acyl-glycerol onto a diacyl-glycerol backbone [ 68 , 69 ]. By contrast, the lipase is predicted to be involved indirectly in the accumulation of TAG via the metabolism of chloroplast monogalactosyldiacylglycerol in green algae such as C. reinhardtii [ 66 ]. In previous studies that employed gene-knockout and over-expression techniques, C. reinhardtii was shown to require the expression of this lipase to accumulate TAG under nitrogen starvation conditions [ 64 – 66 , 70 ]. This function contrasts with the role played by other lipases that recycle TAG and slow the accumulation of the lipid [ 71 ]. Our results are consistent with two distinct roles for individual lipases as well: one which is expressed under nutrient-replete conditions (c12219_g3) and the other which is expressed when the TAG accumulation begins during nutrient-depleted conditions (c10691_g3) ( Fig 5 ). Confirmation of these roles would require manipulation of individual enzyme levels (e.g., by knock-out or over-expression) and a measurement of phenotypic effects similar to studies performed in C. reinhardtii [ 69 , 71 ]."
} | 7,272 |
33627644 | PMC7904764 | pmc | 3,969 | {
"abstract": "Thermal management is the most critical technology challenge for modern electronics. Recent key materials innovation focuses on developing advanced thermal interface of electronic packaging for achieving efficient heat dissipation. Here, for the first time we report a record-high performance thermal interface beyond the current state of the art, based on self-assembled manufacturing of cubic boron arsenide (s-BAs). The s-BAs exhibits highly desirable characteristics of high thermal conductivity up to 21 W/m·K and excellent elastic compliance similar to that of soft biological tissues down to 100 kPa through the rational design of BAs microcrystals in polymer composite. In addition, the s-BAs demonstrates high flexibility and preserves the high conductivity over at least 500 bending cycles, opening up new application opportunities for flexible thermal cooling. Moreover, we demonstrated device integration with power LEDs and measured a superior cooling performance of s-BAs beyond the current state of the art, by up to 45 °C reduction in the hot spot temperature. Together, this study demonstrates scalable manufacturing of a new generation of energy-efficient and flexible thermal interface that holds great promise for advanced thermal management of future integrated circuits and emerging applications such as wearable electronics and soft robotics.",
"introduction": "Introduction Heat dissipation has been a critical technology challenge for modern electronics for decades 1 – 7 . With information technology ramping up in an increasingly digitalized world, electronics cooling is scaling up rapidly in its impact on global energy consumption 8 , 9 . For instance, current data centers consumes over 200 TWh (terawatt-hour) of electricity annually but more than 50% of the total electricity is used for cooling, i.e., removing waste heat rather than for data storage or computing 10 , 11 . In all hierarchical electronic systems ranging from nanoscale transistors, smart phones, laptops, vehicle electronics, to data server farms, waste heat dissipates from the hot spots to heat sink across a series of thermal resistance of multiple device layers and their interfaces. As a result, the device performance, reliability, and energy efficiency can be strongly degraded by a large thermal resistance and a rising hot spot temperature. To address this challenge, recent key research focus for thermal management aims to develop thermal interfaces that enhances thermal coupling and minimize thermal resistance between heterogeneous components 10 . In general, high-performance thermal interface requires both high thermal conductivity (κ) and low elastic modulus ( E ). When inserted between an electronics layer and a heat sink (Fig. 1a ), high κ minimizes thermal resistance and enhance heat dissipation, and low E enables good surface compliance, thermal contact area, and thermomechanical stability. Current commercial thermal interfaces, however, are usually limited by low κ ~ 1 W/m ∙ K or high E ~ 1 GPa, which largely constrains the cooling performance. In addition, emerging applications like wearable electronics and soft robotics require their thermal interfaces to be soft and flexible, but has yet to be explored 12 – 14 . Fig. 1 High-performance thermal interface based on self-assembled boron arsenide (s-BAs) to enhance heat dissipation. a Schematic illustration of a typical thermal interface applied in electronic packaging. Heat dissipation from the chip to heat sink via the thermal interface, unusually limited by the resulting thermal boundary resistance (TBR) 33 . ΔT is the temperature drop and Q is the heat flux across the interface. b Performance comparison of s-BAs vs. the state-of-the-art. Arrow pointing to the bottom left indicates the design goal of high-performance thermal interfaces to achieve both low elastic modulus and low thermal resistivity (i.e., high κ). c Schematic of the zinc-blende crystal structure of cubic BAs and its high-resolution TEM image showing atomically resolved lattices. The arrow indicates the crystal direction of (202). d Thermal conductivity distribution of different materials, including typical fillers.",
"discussion": "Results and discussion During the last decades, intensive research efforts have been devoted in the area and progress has been exemplified by varied categories including thermal greases, gels, pads, tapes, conductive adhesives, phase change materials, metallic solders, etc., with the understanding that different thermal interface materials may have their unique applications. The state-of-the-art performances of thermal interfaces are summarized in Fig. 1b . Fundamentally, there is trade-off between high thermal conductivity and soft mechanics 15 , 16 . Strongly bonded materials such as ceramics and dielectrics usually give high κ 17 , however, their rigid structures can potentially lead to performance degradation like mechanical pump-out, delamination, cracking, and void formation. On the other hand, soft materials such as polymers can provide effective interface contact but are usually limited by an intrinsically low κ ~ 0.2 W/m ∙ K 18 , 19 . For example, metal solders (e.g., indium)-based interfaces provide good thermal conductivity but their applications are largely limited due to high E ~ 1 GPa or above; in addition, solder systems are usually not applicable when electrical insulation is required. Nanostructures such as carbon nanotubes and metal nanowires have been applied to make compromise and improve the mechanical compliance 20 – 23 . Adhesives and gels possess good mechanical compliance, but usually exhibit a low thermal conductivity; their mixtures have been studied to make improvement over poor interfaces and weak van der Waals bonding 16 , 24 – 26 . Despite many exciting progresses (Fig. 1b ), high-performance thermal interfaces with the combination of low elastic modulus, large flexibility, and high thermal conductivity have remained to be demonstrated 16 . In the meanwhile, thermal management has been calling on the development of new materials with ultrahigh thermal conductivity 27 . Recently, building on ab initio theoretical calculations 28 – 31 , a new class of boron compound semiconductors 3 – 7 , 32 , including boron arsenide (BAs) and boron phosphide (BP), has been discovered with ultrahigh thermal conductivity beyond most known heat conductors (Fig. 1c, d ). In particular, cubic BAs has a record thermal conductivity over three times that of the industrial high conductivity standards such as copper and SiC, and twice higher than cubic boron nitride 3 , 32 . With the great application promise in thermal management, the development of BAs for thermal interface, however, has not been explored due to its recent discovery. Here, we report highly flexible thermal interfaces through self-assembly based manufacturing of polymetric composites by taking advantage of the ultrahigh thermal conductivity of BAs crystals. As demonstrated through thermal and mechanical characterizations, the BAs thermal interface exhibits record-high performance with an unprecedented combination of high thermal conductivity (κ ~ 21 W/m ∙ K), excellent elastic compliance similar to that of soft biological tissues ( E ~ 100 kPa), and high flexibility that are beyond the current state-of-the-art and could lead to advanced thermal management of solid-state and flexible electronics. To achieve high performance, we first carefully examine the structural design of BAs particles to achieve efficient heat-dissipation pathways. Based on literature experience 16 , it has been shown that structural optimization is critical to the thermal conductivity of thermal interfaces: Polymer matrixes are generally soft to enable mechanical compliance, but their intrinsic low thermal conductivity (~0.2 W/m ∙ K) could reduce the overall thermal conductivity. In particular, when high conductivity fillers are randomly distributed, the heat-transfer paths in polymer could be significantly elongated and thereby minimize the contribution from fillers 16 . In addition, the organic–inorganic interfaces could result in thermal boundary resistance due to mismatch in phonon spectra and density of states between heterogeneous components 33 – 35 . As a matter of fact, this explains why typical industrial thermal interfaces have a low conductivity around 1 W/m ∙ K or below. To quantitatively evaluate the effect from structural design on the overall thermal conductivity, we performed multiscale simulation to calculate the thermal conductivity of the composite materials with varying extents of alignment of BAs fillers (Fig. 2f ) 33 . The alignment is quantified by the standard deviation of distance (σ) from the BAs particles to the centerline of the alignment pillar, with σ approaching 0 for perfect alignment and increased σ for disorders 36 . A temperature gradient is applied across the structure to compute the volume-averaged heat flux density over the whole domain using the finite element analysis (Methods and Supplementary Information ). The effective thermal conductivities of self-assembled boron arsenide (s-BAs) with varied extents of alignment are determined and plotted in Fig. 2f (pink shadowed background). The thermal conductivity and specific heat used in this simulation are all measured from experiment. The simulation results indicate that an effective design to achieve aligned fillers could effectively enhance the overall thermal conductivity of s-BAs. Fig. 2 Self-assembly based manufacturing and thermal measurement of s-BAs. a Schematic illustrating the self-assembly process through freeze-drying of BAs suspensions to form aligned BAs pillars and polymetric composites. b SEM images of as-synthesized BAs crystals. Inset indicates the crystal size distribution. c Cross-section SEM image of the s-BAs, verifying an aligned lamellar structure. d Optical image of an inch-size s-BAs sample. e Laser flash measurement of the s-BAs samples with different BAs loading ratios. f Thermal conductivity of s-BAs with different BAs loadings. The red symbols are experimental data, and the pink shadowed background represents the modeling results considering varied extents of alignment. To achieve rational alignment of BAs structures in the thermal interface, we designed a self-assembly based manufacturing method using the ice-template process (Fig. 2a ). First, BAs particles were dispersed to form an aqueous suspension. The BAs aqueous slurry was subsequently transferred into a tube mold. A directional temperature gradient (e.g., from dry ice bath at the cold side to room temperature at the high temperature side) was applied across the mold that led the slurry to freeze. At the cold side, the nucleation of the ice crystals started and led to the formation of a constitutionally super-cooled zone directly ahead of the growing ice front. Such an unstable region eventually resulted in perturbations, breaking the planar ice front into a columnar structure; consequently, ice crystals start to follow the temperature gradient direction to gradually grow into aligned lamellar pillars. In the meanwhile, as driven by the growth of ice template 37 , 38 , BAs crystals are expelled and enforced into assembled arrays to replicate the morphology and fill the interspacing between the ice pillars. After this self-assembly process is complete, we dehydrated the sample using a freeze-drying process so that the aligned BAs structures were maintained. Thermodynamically, the freeze-drying process uses a low pressure and a low temperature below the equilibrium triple points in the phase diagram of water (i.e., 273.16 K and a partial vapor pressure of 611.657 Pa) to achieve sublimation of solid ice directly to vapor without going through a liquid phase, so that the structural distortion is minimized 37 – 39 . Finally, polymer melt was infiltrated into the BAs assembly and solidifies to enhance the mechanical supporting and form composite s-BAs. The resulting structures were carefully verified by cross-section scanning electron microscopy (SEM) images before (Fig. 2b ) and after the assembly process (Fig. 2c ), indicating that the aligned lamellar network of BAs pillars can be formed and well maintained during the processing. Note that this manufacturing approach allows readily preparation of inch-size s-BAs samples and further scaling up (Fig. 2d ). The thermal conductivity of s-BAs was measured using the standard laser flash method 40 . Figure 2e shows typical temperature rise curves for s-BAs with varied BAs volumetric loadings of 5%, 10%, 20%, 30%, and 40%, respectively. We have included detailed measurement results in Supplementary Table S1 . The measurement verifies the high thermal conductivity of s-BAs (Fig. 2f ): For example, a record-high thermal conductivity of 21 W/m·K has been measured for 40 vol% s-BAs, which represents ~20 times enhancement over typical thermal epoxies and greases as current industrial thermal interface standard. Consistent with our modeling design, the experimental results also show that the thermal conductivity of s-BAs was enhanced by over 400% through the self-assembled alignment versus random distribution. We also found that the overall thermal conductivity of the assembled s-BAs is dominated by that of BAs fillers, regardless of typical polymer matrix (elastomer, epoxy, etc.). Taking a typical thermal interface thickness of 100 μm, this leads to a total thermal resistance of 0.05 K·cm 2 /W, which is below most of the literature reports. For example, traditional materials based on greases, adhesive, gels, and phase change materials typically yield a higher resistance in the ranges of about 0.2–1, 0.15–1, 0.4–0.8, and 0.3–0.7 K·cm 2 /W, respectively 24 . Elastomic pad, silicone sheet, and thermal tapes typically have the total resistance range of about 1–4 K·cm 2 /W. The demonstrated performance of s-BAs is also record-high among other composites with various fillers including metals, ceramics, semiconductors, oxides, and nanomaterials 16 . In addition to high thermal conductivity, high mechanical compliance is another critical property for high-performance thermal interface. The capability of deformability between interfaces leads to the most fundamental engineering requirements, i.e., low elastic modulus to allow shape change and conformal interfacial contact. In addition, concerning the practical application in electronic packaging, low Young’s modulus supports flexible functionality of thermal interfaces in different directions. We performed the Young’s modulus and shear modulus measurements of the s-BAs samples with varied BAs loading ratios from 0% to 40% (Methods). The shear modulus was assessed by the lap-shear adhesion test. The representative stress–strain curves from the measurements are shown in Fig. 3a, b . The Young’s modulus and shear modulus are determined by the slope of the loading curve at a nominal strain of 5% and plotted in Fig. 3c . These measurement results verify that the s-BAs remains soft with BAs loading volumes up to 40 vol%, with the shear modulus slightly increased from 47 to 148 kPa, and the Young’s modulus from 82 to 256 kPa. The s-BAs can support uniaxial strains above 500%, similar to that of a homogeneous elastomer. These results indicate that the overall BAs/elastomer composite still remains good mechanical compliance. The mechanical properties are also evaluated using the finite element method by treating the s-BAs as a composite with the experimental structures. The Young’s and shear modulus are determined by computing the structural deformations under applied force along the axial and shear directions, respectively (Methods). As shown in Fig. 3c , there is a good agreement between the simulations (shadowed backgrounds) and experiments (solid symbols), indicating that the BAs particles are uniformly distributed in the composite. Fig. 3 Mechanical measurements and high flexibility of s-BAs. Representative stress–strain experimental curves for: a Young’s modulus and b shear modulus measurements, with varied loading ratios. c The Young’s modulus and shear modulus of s-BAs. The solid symbols are experimental data, and the shadowed backgrounds represent the modeling results considering varied extents of alignment. d Optical images of the highly flexible s-BAs. Inset on the bottom left, indicates the original size. e Bending tests and thermal conductivity measurement of s-BAs in response to the cyclic bending. Further, we demonstrate the high flexibility of the s-BAs. A highly flexible thermal interface with both high thermal conductivity and high flexibility is required for thermal management applications in flexible electronics, soft robotics, and other emerging areas, which however remains to be demonstrated 12 – 14 . As shown in Fig. 3d , the s-BAs can be highly deformable to support uniaxial strains more than 500% stretching over its original size. In addition, the s-BAs can be compressed to random geometries such as a heart-shape circle (right, Fig. 3d ) without leading to a mechanical breakdown, which is impossible for standard thermal interface materials. To further explore the potential application in flexible devices, we have performed thermal measurement of the s-BAs under cyclic mechanical bending of the sample (Fig. 3e ), verifying the preserved high thermal conductivity. The thermal conductivity of our s-BAs sample maintains stability over at least 500 bending cycles with a maximum fluctuation within 7%. The persistent high thermal conductivity actually indicates the robust structures during bending tests, as verified by the cross-section SEM images taken after bending cycles (Supplementary Figure 1 ). The retention of highly efficient heat dissipation after mechanical bending underscores the promise of using s-BAs for thermal management of flexible devices. As a further step, we demonstrated a proof of concept experiment to verify the superior device-cooling performance of the s-BAs, through the integration and in situ characterizations of a LED during its operation (Fig. 4 ). To make direct comparison, three types of thermal interfaces, i.e., the commercial thermal epoxy, silicone sheet, and our s-BAs (Methods) are integrated as sandwiched between a 10 W LED chip and a copper heat sink (Fig. 4a, b ). Note that all thermal interfaces were chosen to be with the comparable size, thickness, etc. settings. An infrared camera was used to record the surface temperature of the LED chips, with the Cu heat sink maintained at the room temperature (23 °C). Figure 4c shows a series of infrared images after lighting up the LED chips, measuring the transient temperature dependence. With thermal epoxy and commercial silicone sheet, the chip surface temperature increased up to ~110 and 95 °C, respectively. In contrast, the stable temperature is much lower (~65 °C) when the BAs composite was used as the thermal interface. Quantitatively, the time-dependent surface temperature of the LED chip was measured based on the infrared images and plotted in Fig. 4d , showing a dramatic increase for the devices integrated with thermal epoxy and silicone sheet comparing to that with s-BAs. The large contrast in hot spot temperature difference clearly demonstrates the superior cooling capability of the developed s-BAs for future thermal management applications. Fig. 4 Device demonstration of using s-BAs for high-performance thermal management. a Optical image of a light emitting diode (LED) and b schematic illustration of its integration with a thermal interface and heat sink. c Time-dependent infrared images of the LED integrated with different materials (thermal epoxy, silicone thermal pad, and s-BAs), indicating temperature distributions near the hot spot. d Comparison of the LED hot spot temperatures using different thermal interface materials. In conclusion, we developed a high-performance thermal interface material fabricated through a scalable self-assembly based manufacturing of the recently developed high-thermal-conductivity BAs for advanced thermal management. The s-BAs exhibits an unprecedented combination of high thermal conductivity (21 W/m · K) and an excellent elastic compliance similar to that of soft biological tissues (elastic modulus ~100 kPa). Our thermal and mechanical experiments together with multiphysics modeling show that, upon the designed alignment of BAs crystals, the thermal interface preserves efficient heat-transfer paths while maintaining the high mechanical compliance of polymer matrix. Moreover, the s-BAs shows high flexibility that could be applied to emerging applications such as efficient thermal management of flexible electronics and soft robotics."
} | 5,202 |
28114423 | PMC5256989 | pmc | 3,970 | {
"abstract": "Type VI secretion systems (T6SSs) are widespread in Gram-negative bacteria, including Pseudomonas . These macromolecular machineries inject toxins directly into prokaryotic or eukaryotic prey cells. Hcp proteins are structural components of the extracellular part of this machinery. We recently reported that MFE01, an avirulent strain of Pseudomonas fluorescens , possesses at least two hcp genes, hcp1 and hcp2 , encoding proteins playing important roles in interbacterial interactions. Indeed, P . fluorescens MFE01 can immobilise and kill diverse bacteria of various origins through the action of the Hcp1 or Hcp2 proteins of the T6SS. We show here that another Hcp protein, Hcp3, is involved in killing prey cells during co-culture on solid medium. Even after the mutation of hcp1 , hcp2 , or hcp3 , MFE01 impaired biofilm formation by MFP05, a P . fluorescens strain isolated from human skin. These mutations did not reduce P . fluorescens MFE01 biofilm formation, but the three Hcp proteins were required for the completion of biofilm maturation. Moreover, a mutant with a disruption of one of the unique core component genes, MFE01Δ tssC , was unable to produce its own biofilm or inhibit MFP05 biofilm formation. Finally, MFE01 did not produce detectable N-acyl-homoserine lactones for quorum sensing, a phenomenon reported for many other P . fluorescens strains. Our results suggest a role for the T6SS in communication between bacterial cells, in this strain, under biofilm conditions.",
"conclusion": "Conclusion We describe here several characteristics of the T6SS of the environmental P . fluorescens strain MFE01. We show that Hcp1, Hcp2, and Hcp3 may be involved in T6SS-mediated competition with other Pseudomonas strains on solid media, depending on the prey cell. MFE01 also impaired the formation of biofilms by P . fluorescens MFP05, a strain isolated from skin. Our findings highlight the potential of MFE01 as a biocontrol agent for preventing biofilm formation by another Pseudomonas strain isolated from skin. The protective barrier provided by human skin can be breached by burn wounds, which render the patient susceptible to bacterial infection. P . aeruginosa is frequently isolated from burn patients and causes serious infection [ 72 , 73 ], accounting for more than 50% of all deaths of these patients from infection [ 74 ]. Infections with multidrug-resistant strains of P . aeruginosa may also be untreatable with antibiotics, highlighting the need to develop new therapies [ 75 ]. For this reason, the ability of Lactobacillus to inhibit the growth of harmful bacteria in burn wounds has been investigated [ 76 ]. Although the inhibition of biofilm formation was related to T6SS integrity, we were unable to demonstrate a predominant role for any of the three Hcp proteins. Moreover, a mutant defective in the tss C gene of the unique T6SS core component could not inhibit biofilm formation or produce a biofilm under our conditions. We hypothesise that T6SS effectors may be used as a cell-to-cell signal among sister cells in this strain, which seems to be devoid of the classical AHL quorum-sensing pathway.",
"introduction": "Introduction The type VI secretion system (T6SS) is a widespread macromolecular machinery in Gram-negative bacteria [ 1 ]. This multiprotein complex delivers effectors into eukaryotic and/or bacterial cells [ 1 – 5 ]. This secretion system has a structure similar to that of the contractile tails of bacteriophages [ 6 , 7 ]. Hcp (hemolysin-coregulated protein) and VgrG (valine-glycine repeat protein G) are structural proteins of this machinery with structural similarities to the gp19 and gp5-gp27 proteins of bacteriophage T4, respectively [ 8 , 9 ]. The structural homologue of the phage tube is built from rings of Hcp hexamers with a tip complex composed of VgrG trimers, Paar protein, and effectors [ 10 – 15 ]. This tube is ejected by the contraction of a tubular sheath consisting of the conserved T6SS-associated cytoplasmic proteins, TssB and TssC [ 16 – 19 ]. Hcp proteins appear to be directly involved in effector recognition [ 13 , 20 – 23 ], acting in synergy with VgrG proteins. Moreover, many VgrG proteins possessing a variety of effector domains at their C-termini, are effectors as well as structural components [ 4 ]. The T6SS is associated with several phenotypes, including biofilm formation [ 24 ]. In natural, industrial, and clinical environments, bacteria live predominantly in biofilms [ 25 ]. Biofilms act as reservoirs of infection, protecting the bacteria they contain against diverse external aggressions. They contribute to many problems, particularly in hospitals, where they may form on catheters or implants (e.g. hip prostheses) [ 26 ]. Pseudomonas aeruginosa biofilms also aggravate bacterial infections in a human chronic wound mouse model [ 27 ]. The transition between the planktonic and biofilm states is linked to the production of adhesins and extracellular matrix (ECM) [ 28 ]. Adhesins, including flagellin, play a crucial role, by fixing the bacteria to the support, allowing biofilm formation to occur [ 28 – 30 ]. The ECM surrounding the bacteria may consist of exopolysaccharides (EPS), proteins, and DNA [ 31 ]. EPS act as adhesins on inert and living surfaces, promoting biofilm and microcolony formation. They are also involved in protection against antibacterial compounds [ 32 – 34 ]. In pseudomonads, quorum-sensing regulates the production of extracellular DNA, lectins, and biosurfactants, all of which play a role in biofilm formation [ 25 ]. We previously described P . fluorescens MFE01, a strain that secretes large amounts of Hcp proteins (mainly Hcp2 proteins) into the culture medium [ 35 ]. MFE01 is nonvirulent against various eukaryotic cell models, but has antibacterial activity against a wide range of competitor bacteria, including rhizobacteria and clinical bacteria [ 35 ]. The Hcp2 protein has been directly implicated in the killing activity of MFE01. Another Hcp protein, Hcp1, is encoded by hcp1 , mutation of which has pleiotropic effects on the phenotype of MFE01, affecting its mucoidy and motility [ 36 ]. However, hcp1 mutation has no effect on bacterial competition during incubation on solid medium. Moreover, MFE01 and its hcp2 mutant can sequester a clinical P . fluorescens strain, MFN1032, under swimming and swarming conditions, whereas the hcp1 mutant of MFE01 cannot. Hcp1 appears to reduce the motility of prey cells, to facilitate killing by Hcp2 [ 36 ]. We carried out a genomic analysis to investigate the genetic determinants of the T6SS of MFE01. This analysis revealed the existence of a unique T6SS core component locus and another orphan hcp gene, hcp3 . We investigated the possible role of the T6SS in MFE01 biofilm formation and bacterial competition. We also studied the T6SS-mediated antibacterial activity of P . fluorescens MFE01 and biofilm formation by a cutaneous isolate of P . fluorescens in contact with MFE01.",
"discussion": "Results and Discussion The MFE01 strain has a single T6SS cluster and at least three hcp genes We previously identified two hcp genes, hcp1 and hcp2 . The expression of hcp2 contributes to the killing activity of MFE01 against many Gram-negative bacteria, whereas hcp1 expression is involved in the immobilisation of prey cells. The competitor population was nevertheless reduced slightly by the double mutant, MFE01 Δhcp1Δhcp2 , suggesting the involvement of another factor in T6SS-mediated MFE01 killing activity [ 36 ]. Exhaustive studies of the MFE01 draft genome confirmed that it contained only one T6SS cluster. All the T6SS core component genes ( tssA to tssM ) were observed in classic synteny, except tssD genes (or hcp ), which were distally located from other T6SS core genes, and tssI gene (or vgrG ), which was present as two copies in opposite orientations ( vgrGc1 and vgrGc2 ), on either side of the cluster. This cluster contains six conserved T6SS-associated genes ( paar , sfa2 , tagU , tagH , pppA , and ppkA ) and two predicted ORFs ( unknown1 and unknown 2 ) ( Fig 1A ). We identified a third orphan hcp gene, hcp3 , during annotation of the MFE01 genome. The hcp3 gene is co-localised with a vgrG gene , vgrG3 , and several putative ORFs ( Fig 1B ). The first ORF downstream from hcp3 encodes a putative N-acetylmuramoyl-L-alanine amidase. N-acetylmuramoyl-L-alanine amidases have been identified as T6SS effectors or Tae (Type 6 amidase effector), and are involved in antibacterial activity [ 46 ]. The next gene downstream, duf2333 , belongs to a conserved but uncharacterised superfamily. The third gene encodes a putative NUDIX hydrolase. Duong-Ly et al . showed that a member of the NUDIX hydrolase family from Streptococcus pneumonia , UDP-X diphosphatase, hydrolysed UDP-N-acetylmuramic acid and UDP-N-acetylmuramoyl-L-alanine, two substrates for peptidoglycan construction via the Mur pathway [ 47 ]. This second putative toxin may be a new type of T6SS effector targeting substrates for peptidoglycan construction, although this remains to be confirmed. The next ORF is homologous to sui1 , encoding a translation initiation factor [ 48 ]. A gene encoding a putative arginine decarboxylase was found upstream from vgrG3 . The tec3 gene, homologous to the tec gene described by Liang et al ., was identified downstream from vgrG3 . Tec proteins (members of the DUF 4123 family) have been shown to act as chaperone proteins [ 49 ]. These results suggest that the hcp3 cluster may be involved in the T6SS-mediated antibacterial activity of MFE01. 10.1371/journal.pone.0170770.g001 Fig 1 Genomic organisation of the T6SS core component locus and the hcp3 locus in MFE01. A. Genomic organisation of the T6SS core component locus in MFE01. Genes are represented as arrows, indicating the direction of transcription. The sequences of the T6SS core component genes have been deposited in GenBank under the following accession numbers: vgrGc1 : KX941475, tssA : KX941476, tssB : KX941477, tssC : KX941478, tssE : KX941479, paar-motif : KX941480, unkown1 : KX907122, unknown2 : KX941481, tssF : KX941482, tssG : KX941483, tssH : KX941484, sfa2 : KX941485, tagU : KX941486, tagH : KX941487, tssJ : KX941488, tssK : KX941489, tssL : KX941490, tssM : KX941491, pppA : KX941492, ppkA : KX941493, vgrGc2 : KX941494. B. Genomic organisation of the hcp3 locus. Genes are represented as arrows, indicating the direction of transcription. The sequences of the hcp3 locus genes have been deposited in GenBank under the following accession numbers: hcp3 : KX941495, vgrG3 : KX941496, arginine-decarboxylase : KX941497, sui1 : KX941498, nudix-hydrolase : KX941499, duf2333 : KX941500, amidase : KX941501, tec3 : KX941502. Hcp3-mediated T6SS killing activity is prey strain-dependent We investigated the contribution of Hcp3 to competitor killing by co-culturing MFE01 or various MFE01 T6SS mutants and prey cells on a filter on solid medium for 4 h at 28°C. The strain MFN1032 is a clinical isolate of Pseudomonas fluorescens [ 50 ]. The population of P . fluoresc ens MFN1032 prey cells was significantly smaller (4 log) when cultured with MFE01 than when cultured alone ( Fig 2A ). The size of the MFN1032 population did not differ significantly between co-cultures with MFE01 and co-cultures with MFE01 Δhcp1 . The MFE01 Δhcp2 mutant had significantly weaker antibacterial activity than MFE01 and MFE01 Δhcp1 . We detected no significant killing during co-culture with MFE01 ΔtssC , indicating that the T6SS was responsible for the observed antibacterial activity. The MFE01 Δhcp3 mutant was less bactericidal than MFE01, but nevertheless significantly reduced the MFN1032 population. Hcp3 is thus responsible for some of the T6SS-mediated killing activities of MFE01 against MFN1032. 10.1371/journal.pone.0170770.g002 Fig 2 Killing activity of MFE01 and mutant strains against P . fluorescens MFN1032, P . aeruginosa H103, and P . fluorescens MFP05 during contact on solid media. Quantitative co-culture assays. A. Prey cells (MFN1032 carrying pSMC2.1 gfp) were cultured alone or mixed with P . fluorescens MFE01 or various MFE01 T6SS mutants in a 1:5 ratio. After incubation for 4 h at 28°C, the MFN1032 colonies were counted ( n = 6, the error bars represent the standard error of the mean). ** indicates a significant difference in the number of MFN1032 cfu ( p -value < 0.01) relative to MFN1032 co-cultured with MFE01; * indicates a significant difference in the number of MFN1032 cfu ( p -value < 0.05) relative to MFN1032 co-cultured with MFE01; ns indicates no significant difference. ⧫ indicates a significant difference in the number of MFN1032 cfu ( p -value < 0.05) relative to the MFN1032-alone control assay. EV means empty pPSV35 (plasmid control). B. Prey cells ( P . aeruginosa H103 carrying pSMC2.1 gfp) were cultured alone or mixed with P . fluorescens MFE01 or various MFE01 T6SS mutants in a 1:5 ratio. After 4 h at 28°C, H103 colonies were counted ( n = 6, the error bars represent the standard error of the mean). ** indicates a significant difference in the number of H103 cfu ( p -value < 0.01) relative to H103 co-cultured with MFE01; * indicates a significant difference in the number of H103 cfu ( p -value < 0.05) relative to H103 co-cultured with MFE01; ns indicates no significant difference. ⧫ indicates a significant difference in the number of H103 cfu ( p -value < 0.05) relative to the H103-alone control assay. EV means empty pPSV35 (plasmid control). C. Prey cells (MFP05 carrying pSMC2.1 gfp) were cultured alone or mixed with P . fluorescens MFE01 or various MFE01 T6SS mutants in a 1:5 ratio. EV indicates empty vector for the pPSV35 control. After 4 h at 28°C, MFP05 colonies were counted ( n = 6, the error bars represent the standard error of the mean). ** indicates a significant difference in the number of MFP05 cfu ( p -value < 0.01) relative to MFE05 co-cultured with MFE01; * indicates a significant difference in the number of MFE05 cfu ( p -value < 0.05) relative to MFE05 co-cultured with MFE01; ns indicates no significant difference. ⧫ indicates a significant difference in the number of MFE05 cfu ( p -value < 0.05) relative to the MFE05-alone control assay. We investigated the predation of MFE01 on Pseudomonas aeruginosa H103 ( Fig 2B ) [ 51 ]. The results were similar to those obtained for MFN1032, except for co-culture with MFE01 Δhcp3 . The MFE01 Δhcp3 mutant had a similar level of killing activity to the wild-type MFE01 strain. This result suggests that Hcp3 is not required for the killing of H103 mediated by the T6SS of MFE01. In our previous studies, P . fluorescens strain MFP05, isolated from human skin, was killed on a solid surface by MFE01, and was the strain giving the highest assay sensitivity [ 36 , 52 ]. We studied the effect of tssC and hcp3 mutations on this predation ( Fig 2C ). These mutations totally abolished MFE01 killing activity, which was partially restored by the reintroduction of tssC or hcp3 . Thus, Hcp3 may be critical for the MFE01 T6SS-mediated killing of MFP05. The involvement of T6SS in the killing activity was confirmed by analysis of competitive index ( Fig 3 ). The MFE01 strain outcompeted MFP05, H103 and MFN1032 (log 10 competitive index around -5). On the contrary, the MFE01 ΔtssC mutant was unable to outcompeted the three prey strains (log 10 competitive index around 0). Competitive index of MFE01 Δhcp3 indicated that Hcp3 is not responsible for the MFE01 T6SS-mediated killing of H103, in contrast to that was obtained for MFN1032 and MFP05. 10.1371/journal.pone.0170770.g003 Fig 3 Competitive index of MFE01, MFE01Δt ssC and M FE01Δ hcp3 . MFE01, MFE01Δt ssC or M FE01Δ hcp3 (attackers) were mixed with MFP05, MFN1032 or H103 (preys). The mixture was incubated on nutrient agar for 4 h, and survival were enumerated by plating survivors on appropriate selective plates. The competitive index is calculated using the equation: (input attacker/input prey)/(output attacker/output prey). Horizontal bars indicate the arithmetic mean of log-transformed data. ** indicates statistical significance (Wilcoxon signed rank test, p-value < 0.05), n = 6. MFE01 abolishes the formation of biofilms by competing bacteria during co-inoculation During the first step of biofilm formation, bacteria must adhere to a solid surface [ 53 ]. MFE01 may kill competitor bacteria in two-species biofilms through the action of the T6SS during this adhesion step. We tested this hypothesis by incubating MFE01, MFE01Δ hcp1 , MFE01Δ hcp2 , MFE01Δ hcp3 , or MFE01 ΔtssC with MFP05 on a glass surface in flow chambers. Contact with MFE01 strongly reduced the biovolume of the MFP05 biofilm. By contrast, co-inoculation with MFE01 ΔtssC did not affect MFP05 biofilm biovolume, which was similar to that obtained with MFP05 alone ( Fig 4 ). Mutation of hcp1 , hcp2 , or hcp3 did not significantly modify the reduction in MFP05 biofilm biovolume relative to co-inoculation with the wild-type MFE01 strain. These results suggest that the T6SS is involved in the decrease in MFP05 biofilm biovolume, but that none of the three hcp genes is indispensable for this competition. This result also suggests that another factor leading to antibacterial activity is activated only under biofilm conditions. 10.1371/journal.pone.0170770.g004 Fig 4 Effect of MFE01 and T6SS mutants on MFP05 biofilm formation. Biofilms were grown on a glass surface, for 48 h at 28°C, under a flow of LB medium. Biovolumes of fluorescent bacteria were determined by COMSTAT analysis after confocal laser scanning microscopy observation. P . fluorescens MFP05 bearing pSMC2.1 gfp , encoding green fluorescent protein, was co-cultured alone or with MFE01 or derivatives, in a 1:5 ratio. Each histogram represents the biovolume of fluorescent MFP05, relative to that of fluorescent MFP05 when MFP05- gfp is cultivated alone. Comparisons were made with the control MFP05- gfp ; ** p -value < 0.01; * p -value < 0.05; ns = non-significant; n = 6 (the error bars represent the standard error of the mean). A similar inhibition of biofilm formation has already been described for Pseudoalteromonas , which was found to predominate over strains of Paracoccus sp. or Vibrio sp. in two-species biofilms. The supernatant of the Pseudoalteromonas liquid culture was devoid of antibacterial activity against free-living Paracoccus and Vibrio cells, but it impaired the ability of these species to grow as single-species biofilms. It also impaired biofilm formation by Pseudomonas aeruginosa [ 40 ]. We treated MFP05 with MFE01 supernatant during the adhesion step of the biofilm experiment in flow chambers to determine whether exoproducts secreted by MFE01 inhibited biofilm formation. The MFE01 supernatant had no significant effect on MFP05 biofilm formation ( S1 Fig ). The T6SS-mediated inhibition of biofilm formation has been described in two-species biofilms containing Burkholderia thailandensis [ 54 ]. The threonine phosphorylation pathway (TPP) of P . aeruginosa activates the P . aeruginosa H1-T6SS cluster, which is involved in antibacterial activity, during culture on a solid surface [ 55 ]. The Gac/Rsm pathway upregulates H1-T6SS and biofilm formation, but downregulates H2-T6SS and H3-T6SS [ 24 ]. The tight regulation of the T6SS suggests that all hcp clusters may be regulated under biofilm conditions, working together to inhibit the formation of biofilms by other species. T6SS is involved in the structure of MFE01 biofilms Polysaccharides are an important component of the biofilm matrix [ 56 ]. We have already shown that the hcp1 mutation in P . fluorescens MFE01 halves exopolysaccharide (EPS) accumulation and impairs motility (with an absence of flagella), whereas the mutation of hcp2 does not [ 36 ]. We assessed the capacity of MFE01 mutants to form a biofilm in flow cell chambers, to determine whether these characteristics, associated with hcp1 gene expression, influenced biofilm formation. There was no significant difference in biofilm biovolume between MFE01Δ hcp1 , MFE01Δ hcp2 , or MFE01Δ hcp3 and the wild-type MFE01 ( Fig 5 ). MFE01Δ hcp1 retained the ability to form a biofilm, despite the decrease in EPS accumulation and the absence of flagella. This absence of flagella remains unexplained and is under investigation. Flagella do not seem to be essential for the adhesion of MFE01, and the smaller amount of EPS produced by MFE01Δ hcp1 was still sufficient to permit adhesion. 10.1371/journal.pone.0170770.g005 Fig 5 Effects of hcp and tssC gene mutations on biofilm biovolume in P . fluorescens strain MFE01. Biofilms were grown on a glass surface, for 48 h at 28°C, under a flow of LB medium. Bacteria were visualised with the Syto 9® green fluorescent nucleic acid stain. Biovolumes were determined by COMSTAT analysis, after confocal laser scanning microscopy observation. The values shown are biofilm biovolumes relative to that of wild-type MFE01. The data presented are the mean values for at least five independent experiments and the error bars represent the standard error of the mean. Statistical analyses were performed with non-parametric Mann-Whitney tests (two-tailed): ns indicates no significant difference in biovolume ( p -value > 0.05) relative to the MFE01 biofilm, * and ** indicates significant difference in biovolume relative to the MFE01 biofilm: * p -value < 0.05, ** p -value < 0.01. ⧫ indicates a significant difference in biovolume ( p -value < 0.05) relative to the MFE01Δ tssC biofilm. EV means empty pPSV35 (plasmid control). Surprisingly, the biofilm biovolume of MFE01Δ tssC was significantly smaller than that of wild-type MFE01 (by a factor of 20). However, the wild-type MFE01 and MFE01Δ tssc growth curves were similar, indicating that the tssC mutation had no effect on growth kinetics. The reported impacts of T6SS gene cluster mutations on biofilm formation are diverse. In APEC (avian-pathogenic Escherichia coli ), a mutation of the gene encoding IcmF (TssM), a structural protein of the T6SS, leads to a loss of adhesion to HeLa cells. This T6SS mutant also displays defective biofilm formation on glass [ 30 ]. T6SS mutants of Acidovorax citrulli also display poor biofilm formation [ 57 ]. Conversely, a deletion of tssM , encoding the T6SS structural protein TssM in Acinetobacter baumanii , reduces Hcp secretion, but does not alter biofilm formation [ 58 ]. In Vibrio alginolyticus , the phosphatase PppA, encoded by one of the genes of the T6SS gene cluster, downregulates hcp gene expression and biofilm formation [ 59 ]. Finally, a deletion in icmF3 ( tssM -like gene) in P . aeruginosa has been shown to enhance biofilm formation [ 60 ]. These studies show that the T6SS can affect biofilm formation, but that its role may depend on context. The measurement of biofilm biovolume provides information about the amount of biofilm, but not its level of maturation. We therefore observed cross-sections of biofilm from MFE01 or the mutants by confocal microscopy, to assess the structure of the biofilm and its degree of maturation ( Fig 6 ). Separately, the mutations of individual hcp genes did not abolish biofilm formation. MFE01 biofilms were heterogeneous in appearance, with a slightly hairy surface, and contained cell aggregates but no mushroom-like structures. Although not flat, the MFE01 Δhc p1 biofilms were more homogeneous, with less evident cell aggregates. MFE01 Δhcp2 biofilms did not have a hairy surface, but they contained early mushroom-like structures. MFE01 Δhc p 3 biofilms were flat, homogeneous, and thicker. However, the loss of biofilm structure did not lead to a loss of biomass. Indeed, the biofilms produced by the mutant strains appeared to be more compact than the MFE01 biofilm, possibly explaining the lack of decrease in the biovolumes of the biofilms in the mutant strains. These results indicate that the expression of hcp1 , hcp2 , and hcp3 influences the maturation of the biofilm, but not its biovolume. The expression of these genes is required for the formation of a mature biofilm, suggesting synergy between the three hcp clusters in the maturation process. These results show that none of the secreted T6SS proteins considered, Hcp1, Hcp2, or Hcp3, is the major factor involved in biofilm formation. The MFE01Δ tssC mutant displayed the lowest level of biofilm maturation and the greatest decrease in biofilm biovolume. Our results demonstrate that the adhesion step is unchanged in the mutant strains, contrary to the biofilm maturation. 10.1371/journal.pone.0170770.g006 Fig 6 Effect of hcp and tssC gene mutations on the maturation of P . fluorescens MFE01 biofilms. Biofilms were grown on a glass surface for 48 h at 28°C, under a flow of LB medium. Bacteria were visualised with the Syto 9® green fluorescent nucleic acid stain. A 3D shadow representation and a side-view projection are shown at the top and bottom, respectively, for each strain. Images show representative data from at least five independent biofilm assays. Bars, 10 μm. MFE01 may lack the conventional Pseudomonas quorum-sensing pathway Numerous factors regulate biofilm formation. In P . aeruginosa , biofilm formation and T6SS expression are linked to chronic infection and are highly regulated. The sensor RetS downregulates these phenotypes via c-di-GMP signalling [ 61 ]. Biofilm formation and T6SS expression in Burkolderia cenocepacia are upregulated by quorum sensing and downregulated by the sensor kinase AtsR [ 62 ]. Many studies in P . fluorescens have described the regulation of biofilm formation through quorum sensing (QS), biosurfactants, and the C-di-GMP or Gac/Rsm pathway [ 25 ]. Contrary to Pseudomonas aeruginosa and Pseudomonas putida species, Pseudomonas ffuorescens species seems devoid of quinolone-based QS system [ 63 ]. As described by Martins et al . [ 64 ], quorum sensing systems based on N-acyl-homoserine lactone (AHL) signalling molecules have been identified in a few P . fluorescens strains [ 37 , 65 – 68 ]. Several studies have shown that other strains of P . fluorescens do not produce AHLs [ 64 , 69 , 70 ]. Standalone blast analyses of the MFE01 draft genome indicated that it did not contain any genes encoding a protein corresponding to the conserved AHL synthase (WP_044464955.1). We screened for AHL production, using biosensor strains to verify the absence of AHL synthesis. Pectobacterium atrosepticum 6276 and Pectobacterium atrosepticum 6276 ΔexpI were used as positive and negative controls for AHL production, respectively [ 71 ]. We screened for short-chain AHL production with the biosensor Chromobacterium violaceum CV026. This biosensor responds to C4-C8 AHLs by producing a purple pigment (violacein) after overnight incubation at 28°C. No purple pigmentation was observed around CV026 streak when MFE01 was cross-streaked with CV026 ( Fig 7A ). We screened for AHL production with Agrobacterium tumefaciens NTI. This biosensor responds to C6-C12 AHLs by producing ß-galactosidase, the activity of which generates a blue colour on X-gal plates after overnight incubation at 28°C. No blue colour was observed around Agrobacterium tumefaciens NTI streak after overnight incubation following cross-streaking with MFE01 ( Fig 7B ). This preliminary screening indicates that, like many other P . fluorescens strains, MFE01 does not produce short- or long-chain AHLs recognised by the two used biosensors. The T6SS system may act as a quorum-sensing pathway in MFE01. Vettiger et al . demonstrated that T6SS substrates ( i . e . Hcp, VgrG, and effectors) were transferred between and reused in sister cells [ 71 ]. The amount of T6SS substrate received by a cell depends on cell density and seems to follow the same principle as the auto-inducer accumulation mechanism responsible for quorum sensing. 10.1371/journal.pone.0170770.g007 Fig 7 Detection of AHLs with biosensor strains. P . a indicates Pectobacterium atrosepticum 6276 strain, which produces C 8- NAHLs, used as positive control. P . aΔexp1 means Pectobacterium atrosepticum 6276 mutant strain, which does not produce AHLs, used as negative control. A. Detection of short-chain AHLs with Chromobacterium violaceum CV026 (CV026). CV026 indicates C . violaceum CV026 strain, which produces vioalacein in contact with C 4 -C 8 NAHLs, used as biosensor strain. Black arrow indicates violacein production by C . violaceum CV026 strain. LB plates were incubated for 72 h at 37°C ( n = 3). B. Detection of long-chain AHLs with Agrobacterium tumefaciens NTI (ANTI1). ANT1 means Agrobacterium tumefaciens NT1 strain, which produces β-galactosidase in contact with C 6 -C 12 NAHLs, used as biosensor strain. Red arrow indicates β-galactosidase production by Agrobacterium tumefaciens strain. LB plates containing X-Gal (40 μg/mL) were incubated for 72 h at 28°C ( n = 3)."
} | 7,293 |
23717304 | PMC3653055 | pmc | 3,971 | {
"abstract": "Anaerobic microorganisms play key roles in the biogeochemical cycling of methane and non-methane alkanes. To date, there appear to be at least three proposed mechanisms of anaerobic methane oxidation (AOM). The first pathway is mediated by consortia of archaeal anaerobic methane oxidizers and sulfate-reducing bacteria (SRB) via “reverse methanogenesis” and is catalyzed by a homolog of methyl-coenzyme M reductase. The second pathway is also mediated by anaerobic methane oxidizers and SRB, wherein the archaeal members catalyze both methane oxidation and sulfate reduction and zero-valent sulfur is a key intermediate. The third AOM mechanism is a nitrite-dependent, “intra-aerobic” pathway described for the denitrifying bacterium, ‘ Candidatus Methylomirabilis oxyfera .’ It is hypothesized that AOM proceeds via reduction of nitrite to nitric oxide, followed by the conversion of two nitric oxide molecules to dinitrogen and molecular oxygen. The latter can be used to functionalize the methane via a particulate methane monooxygenase. With respect to non-methane alkanes, there also appear to be novel mechanisms of activation. The most well-described pathway is the addition of non-methane alkanes across the double bond of fumarate to form alkyl-substituted succinates via the putative glycyl radical enzyme, alkylsuccinate synthase (also known as methylalkylsuccinate synthase). Other proposed mechanisms include anaerobic hydroxylation via ethylbenzene dehydrogenase-like enzymes and an “intra-aerobic” denitrification pathway similar to that described for ‘ Methylomirabilis oxyfera .’",
"conclusion": "CONCLUSION The anaerobic oxidation of alkanes plays an important role in the biogeochemical cycling of methane and the bioremediation of hydrocarbon-impacted environments. As we look to the future, advances in next-generation sequencing and annotation will facilitate genome-enabled transcriptomic and proteomic investigations of anaerobic alkane oxidation. The complete genome sequences of several model alkane utilizers are now publicly available and include: Desulfatibacillum alkenivorans AK-01, Desulfococcus oleovorans Hxd3, ‘ Candidatus Methylomirabilis. oxyfera ,’ and the Gammaproteobacterium, strain HdN1. Future work will rely on these model organisms for the purification and characterization of relevant enzymes.",
"introduction": "INTRODUCTION Alkanes are saturated hydrocarbons that are derived from both natural and anthropogenic sources. Due to their apolar C-H σbonds, alkanes are considered to be among the least chemically reactive organic compounds. The activation or functionalization of alkanes is initiated via cleavage of a C-H bond. Aerobic microorganisms achieve this step via monooxygenase or dioxygenase enzymes, in which oxygen serves as both the physiological terminal electron acceptor and as a reactant (for review of mechanisms and enzymes see Austin and Groves, 2011 ). The role of oxygen in the functionalization of alkanes led to the belief for many years that anaerobic microorganisms would be unable to activate and utilize these compounds as growth substrates. However, research during the last 25 years has demonstrated that anaerobic microorganisms have their own novel mechanisms of activating alkanes."
} | 806 |
35539672 | PMC9080629 | pmc | 3,972 | {
"abstract": "The industrial contamination of marine sediments with chromium, copper and nickel in Penang, Malaysia was addressed with bio-remediation, coupled with power generation, using in situ sediment microbial cells (SMFCs) under various conditions. The efficiency of aerated sediment microbial fuel cells (A-SMFCs) and non-aerated sediment microbial fuel cells (NA-SMFCs) was studied. The A-SMFCs generated a voltage of 580.5 mV between 50 and 60 days, while NA-SMFCs produced a voltage of 510 mV between 60 and 80 days. The cell design point for A-SMFCs was 2 kΩ, while for NA-SMFCs it was 200 Ω. In both SMFCs, the maximum current values relating to forward scanning, reverse scanning and oxidation/reduction peaks were recorded on the 80 th day. The anode showed maximum additional capacitance on the 80 th day (A-SMFC: 2.7 F cm −2 ; and NA-SMFC: 2.2 F cm −2 ). The whole cell electrochemical impedance using the Nyquist model was 21 Ω for A-SMFCs and 15 Ω for NA-SMFCs. After glucose enrichment, the impedance of A-SMFCs was 24.3 Ω and 14.6 Ω for NA-SMFCs. After 60 days, the A-SMFCs reduced the maximum amount of Cr( vi ) to Cr( iii ) ions (80.70%) and Cu( ii ) to Cu( i ) ions (72.72%), and showed maximum intracellular uptake of Ni( ii ) ions (80.37%); the optimum remediation efficiency of NA-SMFCs was after 80 days toward Cr( vi ) ions (67.36%), Cu( ii ) ions (59.36%) and Ni( ii ) ions (52.74%). Both SMFCs showed highest heavy metal reduction and power generation at a pH of 7.0. SEM images and 16S rRNA gene analysis showed a diverse bacterial community in both A-SMFCs and NA-SMFCs. The performance of A-SMFCs showed that they could be exercised as durable and efficient technology for power production and the detoxification of heavy metal sediments. The NA-SMFCs could also be employed where anaerobic fermentation is required.",
"conclusion": "4. Conclusions The focus of this study is to compare the performances of A-SMFCs and NA-SMFCs in terms of power generation and heavy metal remediation. The other aim is to optimize parameters such as external and internal resistance, capacitance, electrochemical impedance and construction design to enhance the performances of both A-SMFC and NA-SMFCs. The results indicate that the identified bacteria could transfer electrons to the electrodes to promote power generation and the oxidation/reduction of heavy metals. Furthermore, the amount of electricity produced by the SMFCs could be used efficiently to power minute monitoring devices in addition to enhancing the remediation of contaminant sediments. However, insufficient clues are present to confirm the exact mechanisms of the oxidation/reduction of heavy metals. Further studies into the electrotrophs (microbes that accept electrons from the electrodes) will be more useful to understand the exact mechanisms of the oxidation/reduction of heavy metals.",
"introduction": "1. Introduction Sediments are necessary constituents of marine environments. The quality of the marine bed and substances transferred between water and soil greatly affect the quality of water. Sediment surface layers consist of indicative quantities of pollutants like heavy metals and organic matter, probably impeding ecosystem integrity. 1,2 The oxidized layer at the surface of sediments prohibits the dispersion of heavy metals into marine water. One of the critical pollution dilemmas emerging from industries like electroplating and electronics is the production of wastewater containing heavy metals, which pose a severe risk to humans, animals and the environment. 3 Therefore, it is compulsory to treat industrial wastewater containing heavy metals ahead of its discharge. Several conventional techniques are used to remediate sediments, like the dredging of sediments (stabilization/solidification technology), thermal treatment, bio-chemical stabilization, sediment washing (including chemical leaching/washing, physical separation and bioleaching), in situ capping, natural attenuation (natural recovery), and waterway confinement ( in situ confinement). These are effective but they encounter some considerable drawbacks, such as high energy requirements, excessive chemical utilization and the production of heavy waste sludge in high amounts. 4 Recently, microbial fuel cell (MFC) systems have been attracting attention as the most promising approach to treat industrial wastewater, including power generation. Sediment microbial fuel cells (SMFCs) or benthic microbial fuel cells (BMFCs) are a particular class of MFCs for electricity generation and sediment remediation, utilizing the electro-potential contrast between oxic and anoxic compartments of SMFCs. 5,6 The general prototype SMFC consists of an anode enclosed in marine sediment and a cathode positioned in the surface water. Microbes break down the organic and inorganic compounds present in the sediment and release protons and electrons. Electrons are transferred from the anode to the cathode via an outer circuit and protons flow to the cathode terminal from the sediment and fuse with oxygen at the cathode to form water. 7,8 Reimers et al. 9 applied for the first time electrodes made up of platinum mesh to generate power from both estuarine and salt-marsh sediments. They used platinum mesh electrodes and so obtained a lower amount of energy and a low rate of sediment remediation. The marine sediments near Bayan Lepas, Penang, Malaysia are highly contaminated with chromium, copper and nickel. So, in the competitive global environment today, it has become necessary to clean the marine sediment contaminated by industrial wastewater to minimize aquatic pollution in the aquatic ecosystem, which directly affects human health. Thus, in this study, in situ SMFCs were used to remediate Penang marine sediments, coupled with power generation. The main difference in the SMFCs used in this study is that a bigger modeling size was used compared to previous SMFC studies, so optimizing the external parameters of these SMFCs will lead one step closer to SMFC scale-up. The optimization of SMFC parameters (pH and external resistance) was investigated through electricity generation and heavy metal reduction with and without cathode aeration, because oxygen is the main parameter that highly affects SMFC performance. 10 A comparative analysis of the electro-microbiology and biofilm morphology of both types of SMFC was also conducted to understand the effects of the microbial community on the performance of SMFCs.",
"discussion": "3. Results & discussion 3.1 SMFC voltage output The voltage generation from the SMFCs (replicates) over the experimental duration of 120 days is shown in Fig. 2 . Both the A-SMFC and NA-SMFC started to produce electricity on the first operational day, A-SMFC and NA-SMFC generated 0.60 mV (0.0006 mA) and 0.38 mV (0.00038 mA) after loading an external resistance of 1000 Ω. A maximum difference of only 0.00022 mA was noted in the beginning. In the A-SMFCs the voltage increased and reached a maximum value of 580.5 mV (0.580 mA) between 50 and 60 days, while for the NA-SMFCs, the voltage reached a maximum of 510 mV (0.51 mA) between 60 and 70 days. The maximum current difference was recorded as being about 0.07 mA at these voltage points. Fig. 2 Voltage generation trends for A-SMFCs and NA-SMFCs. Above a certain value, the current output was stable using an external resistance of 1000 Ω. Then, the current declined sharply after day 70. Near these turning points, Cr( iii ) ions, Cu( i ) ions and Ni( ii ) ions were still detected, which suggests that these heavy metals were nearly completely reduced at these turning marks. The generation of higher voltages by A-SMFCs compared to NA-SMFCs could be ascribed to: (i) the presence of high amounts of dissolved oxygen at the cathode, with an improvement in cathodic potential rather than in cell voltage; (ii) mass transfer to the anode being limited, because large numbers of H + ions moved to the anode chamber from the cathode due to the absence of oxygen, which lowered the pH of the anode chamber and resulted in a lower concentration of electron donors; and (iii) the earlier adoptability of exoelectrogens to the environment in the A-SMFCs compared to the NA-SMFCs. 15 This was perhaps due to the consumption of all substrates in the SMFCs by exoelectrogens. Another reason for the high current production in the A-SMFC may be due to the presence of filamentous bacteria detected in SEM analysis, because mostly filamentous bacteria have conductive pili, which are the dominant mechanism in exoelectrogens for electron transfer to the electrodes. Oxidation and reduction phenomena may also be involved in lowering the voltage output due to the production of basic and acidic by-products at both terminals during later operational days. 16 3.2 Polarization and internal resistance Polarization slopes were plotted to find out the relationship between resistance and current during SMFC operation. Various external resistances from 60 Ω to 2 kΩ were used to carry out the polarization study, as shown in Fig. 3a and b . Current production was negatively correlated to the external resistance values. The same power production trend was observed by Abazarian et al. 17 for A-SMFCs; when the resistance was increased from 60 Ω to 2 kΩ, the voltage was reduced from 150 mV (2.5 mA) to 90 mV (0.0045 mA). Whereas, when the external resistance was reduced from 800 to 100 Ω the voltage speedily increased from 20 mV (0.025 mA) to 50 mV (0.5 mA). A maximum power density of 450.5 mW m −2 and current density of 0.75 mA m −2 were measured in the A-SMFC, at the cell-design point of 2 kΩ (external resistance), with 900 Ω internal resistance. At lower resistances, the potential stabilization was not quick, but the power generation trend was increasing. The voltage destabilization was very fast at lower resistances, and more stabilized at higher external resistances. The slow potential drop and stabilization at lower resistances may be due to effective electron discharge. At lower resistances, the electrons move more easily through the circuit, giving higher currents and power densities with low stabilization. A polarization curve was also plotted in the case of NA-SMFCs. A maximum power density of 3781.25 mW m −2 and current density of 2.752 mA m −2 were noted at an external resistance of 200 Ω. The internal resistance was 550 Ω. The NA-SMFCs showed less internal resistance than the A-SMFCs. It has been observed that aeration at the cathode is the main cause for the voltage stabilization at higher resistances. In the A-SMFCs, oxygen was available at the cathode, which helped to increase the cathode reaction rate, resulting in the stabilization of voltage at higher resistances than in NA-SMFCs. Fig. 3 Polarization plots for (a) A-SMFCs and (b) NA-SMFCs operated over the external resistance range of 60 Ω to 2 kΩ. The low availability of oxygen at the cathode terminal results in low stabilization. The difference in internal resistance between A-SMFCs and NA-SMFCs may be due to the pH difference between the anode and cathode compartments. In the case of A-SMFCs, the pH of both chambers was negligible, but in the case of NA-SMFCs, the pH was important. The optimum pH in the NA-SMFCs was 3.0 because H + ions diffused from the cathode chamber to the anode chamber, lowering the pH. As the pH difference between both chambers increased, the internal resistance also decreased. The greater the pH difference between both chambers, the greater the power density destabilization. 10 3.3 Cyclic voltammetry The impact of A-SMFC and NA-SMFC electrode exhibited bacterial biomass (0.5 mg) (40 th 80 th and 120 th days) on voltammetric analysis was determined through characterizing cyclic voltammograms, as shown in Fig. 4a and b . A-SMFCs showed the following maximum currents in the forward scan: 40 th day, 17.11 μA; 80 th day, 25.23 μA; and 120 th day, 14.38 μA; and the reverse scan: 40 th day, −15.21 μA; 80 th day, −21.23 μA; and 120 th day, −11 μA. NA-SMFCs showed the following maximum currents in the forward scan: 40 th day, 13.23 μA; 80 th day, 17.34 μA; and 120 th day, 11.43 μA; and the reverse scan: 40 th day, −12.54 μA; 80 th day, −15.32 μA; and 120 th day, −8.45 μA. Relatively higher current output was measured during the forward scan, irrespective of experimental variation, suggesting higher oxidation rather than reduction. Fig. 4 Cyclic voltammograms for (a) A-SMFCs and (b) NA-SMFCs operating over the range of −0.8 to +0.8 μA. Cell potentials (solid lines show cell potentials, dashed dots show anode potentials, and dashed arrows show cathode potentials) for A-SMFCs (c) and NA-SMFCs (d). In both SMFCs, maximum currents in the forward scan and reverse scans were recorded as follows: A-SMFCs: FS, 25.23 μA; RS, −21.23 μA; and NA-SMFCs: FS, 17.34 μA; RS, −15.32 μA. Clear oxidation and redox peaks were recorded for both SMFCs on the 40 th , 80 th , and 120 th days. A-SMFCs and NA-SMFCs showed oxidation peaks on the 40 th ( E oxi , 0.41 V), 80 th (0.52 V), and 120 th days ( E oxi , 0.40 V), and reduction peaks on the 40 th ( E Red , −0.55 V), 80 th day ( E Red , −0.64 V) and 180 th days ( E Red , −0.43 V). On the 80 th day, the voltammogram showed the highest oxidation and reduction peaks, which is attributed to reversible e − transfer with the highest faradic current. The anodic oxidation and cathodic reduction peaks showed single electron transfer on the 80 th day with highest current production. The e − transfer to the anode surface from a redox-protein can possibly participate in a mediated electron transfer (MET) mechanism. The highest oxidation and reduction rates in the A-SMFCs might be due to the presence of oxygen, which results in high e − discharge and neutralizes e − before reaching the anode. So in the A-SMFCs, there was a negligible change in pH. The presence of glucose in both SMFCs was the main cause of the oxidation peaks. Normally, the voltammogram values increase with time, due to the exoelectrogen density increasing at the electrode. As the exoelectrogen density increases, the amount of metabolites in the feedstock also changes, affecting the capacity and conductivity of the electrolyte solution. So the disturbance caused by the addition of glucose resulting in the extra voltammogram values can be obtained for immobile phase exoelectrogens or those suspended in the physiological marine solution. Not all voltammogram peaks appeared continuously during various tests, thus indicating the presence of different electronic mediators used by electroactive biofilms during SMFC operation. 18 The variations in the anode, cathode and cell potentials were measured over the range of 60 Ω to 2 kΩ in both SMFCs, operating under all experimental conditions, as shown in Fig. 4c and d . The cathode potentials at 2 kΩ (A-SMFCs: 100 mV; and NA-SMFCs: −110 mV) show that the current produced during the running of the SMFCs was not only confined by the anode reaction. The anode potentials (at 2 kΩ: A-SMFCs: 196 mV; and NA-SMFCs: −160 mV) were significantly decreased upon decreasing the resistance. The anode potential regulates the kinetics of electron transfer from microorganisms to the anode. The anode potential in both SMFCs decreased below 1 kΩ and 2 kΩ, suggesting effective electron discharge below these external resistances. CV curves are shown for the 40 th , 80 th and 120 th days. Based on these curves, the capacitance values for the anode and cathode were measured, and are shown in Table 2 . The capacitance of the anode was increased more than the cathode on the 40 th day (A-SMFC: 1.5 F cm −2 ; NA-SMFC: 1.0 F cm −2 ) and the 80 th day (A-SMFC: 2.7 F cm −2 ; NA-SMFC: 2.2 F cm −2 ) indicating the growth of biofilm on the anode. The capacitance was decreased on the 120 th day. As biofilm grew on the anode, the total capacitance of the anode increased by 2.5–4.9 F cm −2 for both SMFCs. This increase in anode capacitance is possibly due to the transient charge storage capacity of reductive/oxidative enzymes in the bacterial cytoplasm and on the bacterial cell membrane. 19 This additional anode capacitance is also due to the acclimatization of exoelectrogens. Hong et al. 20 reported that at external low resistances with a respective number of oxidative and reductive peaks, new electron transfer pathways, such as new oxidative and redox/enzymes and self-produced mediators, could emerge in SMFCs and increase the capacitance of the anode. This relatively higher anode capacitance led to the elimination of power overshoot over a short time. The higher anode capacitance compared to the cathode capacitance has the advantage of alleviating the power overshoot when the cathode capacitance was insufficient. The capacitances of the anode and cathode with exoelectrogen colonization on the 40 th , 80 th and 120 th days Day Capacitance (F cm −2 ) Anode Cathode A-SMFC NA-SMFC A-SMFC NA-SMFC 40 1.5 1.0 0.5 0.4 80 2.7 2.2 0.7 0.3 120 0.7 0.4 0.3 0.1 3.4 Electrochemical impedance spectroscopy Equivalent circuit model fitting The ECM was applied in this study to determine the 3 series resistances: anode ( R anode ); solution ( R S ); and cathode ( R cat ). It was observed that the SMFC impedance spectra model was reversible among the cathode and anode. 21 An estimation of the individual impedance of the anode and cathode was used to characterize the related impedance of these components. The analysis was carried out on mature 120 day-biofilm to ensure that this model was correctly fit to latter periods. Previous studies only focus on how impedance varies during the earlier stage of exoelectrogen growth. Fig. 5a and b shows the impedance spectra for the cathode, anode and whole cell. The first junction on the x -axis of the Nyquist plot represents the electrolyte solution resistance (whole cell resistance). The electrolyte resistance was 21 Ω and 15 Ω in the A-SMFC and NA-SMFC, respectively. The projected points in the Nyquist model represent the anode resistances of the A-SMFC and NA-SMFC, which were 0.4 Ω and 0.2 Ω, respectively. The cathode resistances of the A-SMFC and NA-SMFC were 35 Ω and 15 Ω, respectively. This shows that the cathode and electrolyte impedances were higher than the anode impedance. This model can easily be applied to later SMFC operation to reproduce the results. Fig. 5 Nyquist curves for the cathode, anode and whole SMFC: (a) A-SMFC; and (b) NA-SMFC. The insets represent the anode response. The effect of enrichment on the impedance of a SMFC The Nyquist model was applied to the EIS spectra to determine the electrochemical impedances of both SMFC electrolytes. Fig. 6 shows impedances at operating loads through fitting the experimental data to the Nyquist model. The total impedance values (electrolyte + cathode + anode) after 40 days for the A-SMFC and NA-SMFC were 70 Ω and 33 Ω (external R : 200 Ω). These decreased to 49 Ω in the A-SMFC and 27 Ω in the NA-SMFC after 120 days (external R : 100 Ω). Fig. 6 Nyquist plots indicating EIS data fitted to the ECM model on days 40, 80 and 120 for the A-SMFC and NA-SMFC. The individual impedances of the anode, cathode and electrolyte solution dramatically changed, as shown in Fig. 7a and b . The EIS measurements were begun at day 20 with an external R of 50 Ω (A-SMFC: voltage output, 0.160 V; current density, 0.008 mA m −2 ; NA-SMFC: voltage output, 0.100 V; current density, 0.005 mA m −2 ). Fig. 7 The behavior of anode, cathode and solution impedances for the (a) A-SMFC and (b) NA-SMFC during the embellishment of exoelectrogenic microbes in SMFCs over time. The total impedance values at day 60 for the A-SMFC and NA-SMFC were 24.3 Ω and 14.6 Ω, respectively. These values decreased on day 120 to 18.5 Ω and 6.8 Ω. The anode impedance associated with R anode in both the A-SMFC and NA-SMFC decreased from 10.1 Ω to 4.5 Ω and from 6.1 Ω to 2.1 Ω from day 20 to day 120. The cathode impedance ( R cathode ) was also decreased in both the A-SMFC and NA-SMFC from day 20 to day 120, from 9.1 to 10.7 Ω and from 5.2 to 2.1 Ω, respectively. The decrease in the anode and cathode impedances was most probably due to the enrichment of exoelectrogen biofilm on the electrodes. The same impact of enrichment on electrode impedance was reported by Borole et al. 22 3.5 Sediment heavy metal remediation Previous studies have found that the aerobic remediation of certain contaminants (organic and inorganic pollutants) could be improved by supplying a solid-state anode (electron acceptor) to bacteria in a SMFC. However, these studies were only focused on in situ remediation using a two chamber MFC consisting of a permeable exchange membrane. This system is not suitable for open environment remediation due to the closed MFC. Both the A-SMFCs and NA-SMFC were run for 120 days. The heavy metal profile, such as for Cr( vi ) and Cr( iii ) ions, Cu( ii ) and Cu( i ) ions, and Ni( ii ) ions, was detected through XPS before SMFC operation, as shown in Fig. 8a–c . Fig. 8 X-ray photoelectron spectra of Cr (a), Cu (b) and Ni (c). The A-SMFC and NA-SMFC were operated at their optimal external resistances: 2 kΩ and 200 Ω, respectively. The heavy metal remediation efficiency was compared with Sediment Management Standards USA, as shown in Table 3 . The A-SMFCs reduced a maximum amount of Cr( vi ) ions to Cr( iii ) ions, about 80.70%, and Cu( ii ) ions to Cu( i ) ions, about 72.72%, and had a maximum uptake of Ni( ii ) ions, about 80.37%, after 60 days. The optimum detoxification efficiency using the NA-SMFCs was achieved after 80 days, and was about 67.36% for Cr( vi ) ions, 59.36% for Cu( ii ) ions, and 52.74% for Ni( ii ) ions. Performances of A-SMFCs and NA-SMFCs for the remediation of heavy metals compared with Sediment Management Standards USA A-SMFC NA-SMFC Heavy metal Concentration (mg kg −1 ) Remediation efficiency (%) Concentration (mg kg −1 ) Remediation efficiency (%) Days of operation Sediment Management Standards USA Cr( vi ) ions 390.3 ± 3.2 0 390.3 ± 3.2 0 0 260–270 mg kg −1 300.5 ± 3.9 23.19 320.2 ± 3.6 17.95 20 180.9 ± 1.8 53.65 210.8 ± 3.1 45.97 40 75.30 ± 2.4 80.70 165.9 ± 1.5 57.48 60 190.1 ± 3.7 51.27 127.3 ± 2.3 67.36 80 260.5 ± 3.7 33.25 177.2 ± 2.1 54.59 100 324.5 ± 4.1 16.85 260.6 ± 3.2 33.23 120 Cu( ii ) ions 480.1 ± 3.3 0 480.1 ± 3.3 0 0 390–390 mg kg −1 420.4 ± 2.0 12.43 450.5 ± 2.7 6.170 20 310.3 ± 3.4 35.36 360.5 ± 1.3 24.91 40 130.9 ± 2.7 72.72 250.9 ± 2.7 47.73 60 270.4 ± 1.7 43.67 195.1 ± 4.5 59.36 80 390.1 ± 2.2 18.75 220.6 ± 2.3 54.05 100 430.6 ± 3.0 10.31 380.3 ± 2.8 20.78 120 Ni( ii ) ions 180.5 ± 2.5 0 180.5 ± 2.5 0 0 26–110 mg kg −1 145.3 ± 3.0 19.48 165.3 ± 2.4 8.384 20 90.66 ± 2.2 49.77 124.9 ± 2.6 30.78 40 35.41 ± 1.8 80.37 100.4 ± 3.3 44.35 60 134.5 ± 3.3 25.44 85.29 ± 2.4 52.74 80 156.3 ± 3.0 13.38 124.1 ± 3.0 31.22 100 167.2 ± 2.8 7.331 145.4 ± 2.3 19.44 120 The detection of less toxic heavy metals ions like Cr( iii ) ions and Cu( i ) ions after SMFCs operation was achieved via reduction peaks in the cyclic voltammetry studies. Enzymatic redox reactions are normally part of microbial metabolism. Cr( vi ) and Cu( ii ) ions can also be reduced at exoelectrogen surfaces via nonmetabolic pathways. Intracellular precipitation is also another mechanism to reduce these metals, but the first one is the dominant reduction mechanism. 23 Cr( vi ) and Cu( ii ) ions are reduced to the less toxic Cr( iii ) and Cu( i ) ionic forms in the presence of electron donors like redox-active proteins. Cr( vi ) and Cu( ii ) ions are biologically reduced aerobically (A-SMFCs) and anaerobically (NA-SMFCs), but aerobic reduction is dominant because the reduction rate is very slow under anaerobic conditions. The oxygen concentration boosts the rate of reduction. In NA-SMFCs, due to the absence of oxygen, more CO 2 is produced via anaerobic fermentation and more H + ions are moved to the anode chamber, raising its pH. So by raising the pH, genes that express the surface binding proteins for Cr and Cu are suppressed. In anaerobic bacteria (NA-SMFCs), the toxic effects of heavy metals are associated with a disruption in enzyme structure and function due to metals binding with thiol groups and other groups on protein molecules, or replacing naturally occurring metals in enzyme prosthetic groups. Some studies have proved that the presence of Cr, Cu and nickel in sediment mixtures also produced antagonistic and synergistic effects on anaerobic bacteria (NA-SMFCs). 24 Ni( ii ) ions are mostly absorbed inside cells by reducing c-type cytochromes at the surfaces of exoelectrogens and transferring the electrons to the electrodes. OmcT, OmcB, OmcS, OmcZ, OmcF, OmpB and OmpC are dominant c-type cytochromes present at the surfaces of exoelectrogens. 25 In anaerobic bacteria (NA-SMFCs), nickel stress activates bacterial intracellular detoxification genes, which are mostly located on plasmids. Intracellular defense systems also mediate chelation, bio-methylation and exocytosis in anaerobic bacteria for nickel absorption. 26 These heavy metal remediation efficiencies fulfill the standards for sediments determined by Sediment Management Standards USA. The remediation efficiencies were lower in the earlier days due to worse adjustment with the environment, and decreased in the last days of operation, very likely as a consequence of the consumption of substrates for microbial metabolism. In the last operational days, electrons were only produced in the cathode chamber and they were transferred to the anode chamber, which altered the pH of the anode chamber and finally lowered the remediation efficiency of the exoelectrogens. This decrease in pH might also be due to the fermentation of organic matter in the sediment. The heavy metal reduction rate decreased with time due to the specific growth patterns of bacteria. The bacteria entered into the death phase after a specific time and then the number of dead bacteria was greater than the number of those alive, so the reduction rate of heavy metals decreased with time. Zhang et al. 27 detoxified copper at about 99.9% during a catholyte reaction, with a maximum power density of 7.2 W m −2 but they used synthetic copper solution, which may render a problem for microbes trying to combat the natural environment, due to the presence of other compounds. Wang et al. 28 reported that external resistances influenced the reduction of Cr( iv ), the formation of biofilm on electrodes and power generation, and therefore the overall performance of SMFCs. 27 Previous research also reported that environmental factors like pH, as mentioned above, and external resistances strongly influenced the biological remediation of contaminated sediments. 29 So to optimize SMFC performance, the influence of these factors on bioremediation and power generation should be addressed in the future. 3.6 Effect of pH on the performance of SMFCs The effects of pH on voltage generation and heavy metal reduction are shown in Fig. 9 . In the acidic pH range (1.0–6.0), A-SMFCs reduced about 10–55% of Cr( vi ) ions to Cr( iii ) ions, 9–50% of Cu( ii ) ions to Cu( i ) ions, and 5–45% of Ni( ii ) ions, with power generation of 20–360 mV. In the basic pH range (8.0–13.0), A-SMFCs reduced about 20–50% of Cr( vi ) ions to Cr( iii ) ions, 15–45% of Cu( ii ) ions to Cu( i ) ions, and 13–41% of Ni( ii ) ions, with power generation of 50–410 mV. The NA-SMFCs showed a reduction of 5–35% of Cr( vi ) ions to Cr( iii ) ions, 7–38% of Cu( ii ) ions to Cu( i ) ions, and 3–42% of Ni( ii ) ions, with power generation of 10–280 mV in the acidic pH range. The NA-SMFCs reduced about 4–45% of these heavy metals with 30–320 mV power production in the basic pH range. Both SMFCs showed maximum heavy metal reduction and power generation at pH 7.0. Fig. 9 The effects of pH on the performance of SMFCs: (a) A-SMFCs; and (b) NA-SMFCs. So the exoelectrogens in both SMFCs were neutrophils. The microbes in both SMFCs can easily tolerate acidic pH values of about 1.0–6.0, but their metabolic activities are low. Venkhata Mohan et al. 30 reported that acidic pH values affect SMFC organisms by affecting the substrate metabolism, resulting in H + ion and e − release. Exoelectrogens under acidic conditions are more susceptible to methanogenic metabolism, due to the combination of H + ions and e − with CO 2 , forming methane. So a neutral pH suppresses the methanogenic activity, increasing the metabolic rate of exoelectrogens, resulting in the generation of high power. García-Muñoz et al. 31 also reported higher power generation trends at a neutral pH: about 505.5 mV. Yuan et al. 32 also reported higher exoelectrogen biofilm activity at neutral pH. They noted that a larger number of bacteria attached to the anode at neutral pH, rather than at acidic or basic pH. 3.7 Biofilm morphology SEM images of the mixed cultures on the electrodes show the dispersed and dense cultures of exoelectrogens compared with untreated graphite, as shown in Fig. 10 . Fig. 10 Scanning electron micrograph images taken on day 120 from A-SMFCs and NA-SMFCs: (a) anode graphite; (b) cathode graphite; and (c) untreated graphite. The morphologies of the mixed cultures on the aerated anode and cathode took the form of filamentous shaped cells, while the non-aerated anode and cathode were loaded with rod-shaped exoelectrogens. Many previous studies reported SEM images of electrode biofilms. 33 However, very few studies have reported the presence of filamentous appendages in the biofilms of electrodes. Many factors can be the cause of this contradiction. First, there may be a difference in the operating conditions of SMFCs, such as the presence of inorganic substrates, organic ingredients and microbial inoculum. Second, the high SEM resolution may detect the conducting wires in the metals being reduced and the power generating Shewanella and Geobacter species. 34 From this study we can formulate the hypotheses that these conducting appendages are not necessary for electron transfer to the electrodes, especially to the anode, because other mechanisms like electron redox-active shuttles are also responsible for electron conduction, but these conducting pili affect the rate of electron transfer and the dominant mechanism in the exoelectrogens. 3.8 Electro-microbiology The diversity of bacterial exoelectrogens in the A-SMFCs and NA-SMFCs, as shown in Table 4 , is an encouraging sign that SMFC technology may be fruitful for the remediation of a wide range of heavy metals and power generation. The anodes of both A-SMFCs and NA-SMFCs mostly consist of Proteobacteria strains. Sideroxydans lithotrophicus is an exoelectrogen and a metal reducing bacterium, as previously reported by Shi et al. 35 Proteobacteria are mostly fermenting bacteria and are most suitable for electricity production and heavy metal remediation. 36 The cathodes of A-SMFCs and NA-SMFCs mostly contained Actinobacteria and Proteobacteria . It is necessary to note that not all exoelectrogen bacteria interact with electrodes as electron acceptors; they also interact with each other in interspecies electron transfer and enhance the performances of SMFCs. These bacteria help in disposing the inhibitory by-products produced by fermentation. The more diversified microbes in SMFCs produce more power, due to the availability of redox mediators like c-cytochromes (omcZ omcB, omcS, omcT, omcE and omcZL). 37 Thioalkalivibrio sp. was also reported in this study and it is a suitable exoelectrogen and suitable for heavy metal detoxification due to its production of the cytochrome cbb 3 . 38 Summary of 16S rRNA gene sequences recovered from the NCBI clone library for electrodes in the A-SMFCs and NA-SMFCs Accession no. of 16S rRNA gene Name of bacterium Percentage homology \n Anodic bacterial community \n A-SMFC NR_115756.1 \n Sideroxydans lithotrophicus \n 100% JN377592.1 \n Gallionellaceae bacterium \n 98% DQ839562.1 \n Candidatus nitrotoga \n 97% NR_025455.1 \n Propionivibrio limicola \n 93% CP002159.1 \n Gallionella capsiferriformans \n 93% NA-SMFC LT556085.1 \n Citrobacter sp. strain 92 99% FN433034.1 \n Citrobacter farmeri \n 99% DQ490332.1 \n Enterobacteriaceae bacterium \n 99% EU652047.1 \n Pseudomonas stutzeri strain aa-28 99% EF599310.1 \n Gamma proteobacterium B12 99% \n Cathodic bacterial community \n A-SMFC CP002917.1 \n Corynebacterium variabile \n 99% GU735087.1 \n Thioalkalivibrio sp. 89% NR_102486.1 \n Thioalkalivibrio nitratireducens \n 89% AJ627387.1 \n Methylocaldum szegediensis \n 89% NA-SMFC JX458392 \n Bosea sp. 99% U87773.1 \n Afipiageno sp. 99% HM136777.1 \n Bradyrhizobiaceae bacterium 98% JX219400.1 \n Starkeya sp. 99% 16S rRNA gene sequences are ideal to characterize multiple conserved metabolic functions. Bundles of clones with high similarity to bacteria are capable of heavy metal oxidation/reduction. This represents a sign of the significant preference for the oxidation/reduction of heavy metals by the diverse electrode bacterial community, and backs the hypothesis that the oxidation/reduction of heavy metals in SMFCs by the bacterial community is a default metabolic pathway."
} | 8,309 |
24113651 | null | s2 | 3,974 | {
"abstract": "Electrochemically active biofilms have a unique form of respiration in which they utilize solid external materials as terminal electron acceptors for their metabolism. Currently, two primary mechanisms have been identified for long-range extracellular electron transfer (EET): a diffusion- and a conduction-based mechanism. Evidence in the literature suggests that some biofilms, particularly Shewanella oneidensis, produce the requisite components for both mechanisms. In this study, a generic model is presented that incorporates the diffusion- and the conduction-based mechanisms and allows electrochemically active biofilms to utilize both simultaneously. The model was applied to S. oneidensis and Geobacter sulfurreducens biofilms using experimentally generated data found in the literature. Our simulation results show that (1) biofilms having both mechanisms available, especially if they can interact, may have a metabolic advantage over biofilms that can use only a single mechanism; (2) the thickness of G. sulfurreducens biofilms is likely not limited by conductivity; (3) accurate intrabiofilm diffusion coefficient values are critical for current generation predictions; and (4) the local biofilm potential and redox potential are two distinct parameters and cannot be assumed to have identical values. Finally, we determined that simulated cyclic and squarewave voltammetry based on our model are currently not capable of determining the specific percentages of extracellular electron transfer mechanisms in a biofilm. The developed model will be a critical tool for designing experiments to explain EET mechanisms."
} | 407 |
36062952 | PMC9827892 | pmc | 3,975 | {
"abstract": "Abstract Sustainable sources are key to future chemicals production. Microalgae are promising resources as they fixate carbon dioxide to organic molecules by photosynthesis. Thereby they produce unsaturated fatty acids as established raw materials for the industrial production of chemical building blocks. Although these renewable feedstocks are generated inside cells, their catalytic upgrading to useful products requires in vitro transformations. A synthetic catalysis inside photoautotrophic cells has remained elusive. Here we show that a catalytic conversion of renewable substrates can be realized directly inside living microalgae. Organometallic catalysts remain active inside the cells, enabling in vivo catalytic olefin metathesis as new‐to‐nature transformation. Stored lipids are converted to long‐chain dicarboxylates as valuable building blocks for polymers. This is a key step towards the long‐term goal of producing desired renewable chemicals in microalgae as living “cellular factories”.",
"conclusion": "Conclusion Reported examples of in cell transition metal catalysis to date were limited to bacterial or mammalian cancer cells. \n [18] \n Unlike these cells, the unicellular microalgae studied here are autotrophically grown single cell organisms that do not require external carbon and energy sources except for carbon dioxide and sunlight. Thereby, they produce unsaturated fatty acids, compounds which serve as a feedstock today for in vitro transformations to valuable chemicals. Our findings reveal the feasibility of intracellular non‐natural transition‐metal catalysis in photoautotrophic organisms, and also show that unprotected small molecule olefin metathesis catalysts can operate inside cells. The complex microalgae cell walls can be overcome by the catalysts studied, and particularly the novel HUC‐BDP catalyst does not adversely affect cell viability, at the same time it is also tolerant towards the cell environment. This enables high intracellular conversions of unsaturated fatty acids on par to reference in vitro transformations. The in cell olefin metathesis approach demonstrated here converts natural intracellular fatty acid substrates into value‐added chemicals. Specifically, the generated dicarboxylates are valuable building blocks for polymers. \n [3d] \n They originate from efficient natural carbon dioxide fixation complemented by bioorthogonal synthetic catalysis. The implementation of synthetic catalysis in photoautotrophic organisms demonstrated here enables a potentially game‐changing approach toward a sustainable chemical industry—as a first key step towards the long‐term goal of using microalgae as cell factories for the direct production of renewable chemicals. An obvious further challenge is a simultaneous release of desirable products from the living cells, by biological or simple physical pathways. We anticipate the feasibility of non‐natural catalytic transformations in unicellular microalgae revealed here will additionally be inspiring and useful to other fields beyond the production of chemicals.",
"introduction": "Introduction To achieve carbon neutral production schemes, renewable feedstock sources are a key for the production of chemicals. Microalgae are a particularly promising renewable resource as they fixate atmospheric carbon dioxide to organic compounds by photosynthesis, with sunlight as energy and without the need for additional carbon sources. Unlike higher plants, microalgae do not require arable land and fresh water resources, but can be cultivated with minimal space requirements also in brackish or salt water. By photosynthesis, microalgae build up fatty acids as intracellular storage substances. This photoautotrophic generation of lipids is one of many examples of the extreme efficiency of the chemical machinery of cells. Yet, the scope of cellular chemistry is restricted and synthetic catalysts can allow for transformations complimentary to those found in nature. Traditionally, such synthetic reactions are carried out in vitro. Thus, established schemes for the valorization of the naturally produced lipids as feedstocks for chemicals and polymers employ tedious and energy‐consuming extraction, the extracts being catalytically upgraded separate from the cellular sources of the substrates. Implementing a bio‐orthogonal synthetic catalysis in vivo could be a key step to enable new concepts for producing renewable‐sourced chemicals directly in “cellular factories”. However, a synthetic catalysis inside living photoautotrophic cells has remained elusive to date. Unicellular microalgae can generate high amounts of unsaturated fatty acids. Such fatty acids, sourced today from palm oil or other seed oils, are an important feedstock of the chemical industry. \n [1] \n For their conversion to desirable products, catalytic olefin metathesis has emerged as an advanced industrial process. The redistribution of fragments of carbon‐carbon double bonds of unsaturated substrates (cf. Figure 1 a) in olefin metathesis occurs by a unique mechanism, \n [2] \n unparalleled in transition metal as well as bio‐catalysis. The versatility and synthetic capability of olefin metathesis have opened a wide field of applications.[ \n 1 \n , \n 3 \n ] Especially during the past decade, the conversion of renewable raw materials by olefin metathesis has been implemented industrially for the production of biodiesel, polymers or chemical intermediates.[ \n 1 \n , \n 3c \n ]\n Figure 1 Concept of the in vivo synthetic olefin metathesis in microalgae. a) In microalgae cells, unsaturated fatty acids susceptible to olefin metathesis are stored in intracellular organelles, so‐called lipid bodies. These compartments are surrounded by a phospholipid monolayer with associated proteins (depicted in shades of yellow). For intracellular catalysis, olefin metathesis catalysts need to be internalized by the cells through their cell wall and membrane and targeted to these intracellular lipid organelles while retaining their catalytic activity. b) Olefin metathesis catalysts applied in this work are modified at the benzylidene ligand (HUC‐BDP and HGII‐BDP) or at the N‐heterocyclic carbene ligand (HGII‐NHC‐BDP) by tagging with a green fluorescent BODIPY dye as transport and labelling agent. Solid‐state structures of HUC‐BDP and HGII‐BDP obtained from single crystal X‐ray diffraction, hydrogen atoms are omitted for clarity. Despite its broad range of applications in vitro, only two examples of olefin metathesis in living cells have been reported: Michel et al. used metathesis catalysts as sensors for the detection of ethylene in C. reinhardtii , aiming at a stoichiometric rather than a catalytic process, releasing a fluorophore via one cycle of metathesis with ethylene. \n [4] \n Additionally, Ward et al. assembled artificial metalloenzymes in the periplasm of E. coli . \n [5] \n The activity of these enzymes was confirmed by ring‐closing metathesis of an added reactive substrate that generates a fluorescent product, which enables detection of conversion by virtue of a high sensitivity. In contrast to E. coli cells, microalgae can serve as source of valuable fatty acid feedstocks that provide substrates for olefin metathesis and can thereby be converted into the desired renewable chemicals. However, microalgae are unicellular eukaryotic organisms that are surrounded by a complex cellular barrier consisting of a cell wall and membrane. Therefore, intracellular uptake of catalysts into these organisms is much more challenging than for the highly permeable bacterial cells. Notably, neither catalytic olefin metathesis, nor any other type of synthetic catalysis has to date been realized in living photoautotrophic cells producing renewable feedstocks. This is however a key prerequisite for the aforementioned vision of cellular factories producing renewable chemicals. Therefore, this work aims at the ambitious implementation of catalytic olefin metathesis in photoautotrophic microalgae. We report that small molecule \n 1 \n organometallic catalysts can cross the cell wall of unicellular microalgae and remain active in the intracellular environment, enabling in vivo catalytic olefin metathesis as new‐to‐nature transformation. This converts stored fatty acids to desirable non‐natural long‐chain alkenes and dicarboxylates with high conversion.",
"discussion": "Results and Discussion Catalyst Design and Cell Viability Inside the microalgae, the unsaturated fatty acid components are stored in specific organelles surrounded by a phospholipid monolayer with associated proteins, so‐called lipid bodies or lipid vesicles (Figure 1 a). For intracellular catalysis of the valuable fatty acid substrates, the olefin metathesis catalyst has to be transported past the cellular barrier (wall and membrane), through the whole aqueous cytoplasm as intracellular medium containing potentially detrimental components, to these lipid storage compartments. Since the fatty acid substrates for the conceived in vivo olefin metathesis approach are stored in these intracellular organelles, the lipid bodies were identified as target for the abiotic intracellular reaction. To enable intracellular uptake and targeting to these lipid organelles, \n [6] \n we furnished single component metal alkylidene catalysts with lipophilic fluorescent BODIPY (boron difluoride‐dipyrromethene) motifs (Figure 1 b). For many other intracellular organelles like mitochondria or the nucleus, targeting motifs are known that direct substances to this specific location. For example, a triphenylphosphonium group on a compound of interest leads to a directed transport of this substance into mitochondria. However, no such targeting signal has yet been identified for the lipid bodies of interest in this work. \n [7] \n \n The choice of BODIPY motifs was therefore based on their use for staining microalgal lipid organelles for the purpose of lipid quantification and of studying lipid body growth and dynamics by microscopy. \n [8] \n \n Ruthenium‐based and phosphine‐free catalysts were employed due to their general tolerance towards functional groups as well as their anticipated cell compatibility. \n [9] \n The state‐of‐the‐art cyclic alkyl amino carbene (CAAC) catalyst motif was endowed with a BODIPY vehicle (HUC‐BDP, Figure 1 b) via ligand exchange of the chloride‐bridged dimer \n [10] \n with a BODIPY‐substituted styrene (for synthesis and characterization data see the Supporting Information Figure S1–S5). As a reference, the well‐established benchmark Hoveyda‐Grubbs II N‐heterocyclic carbene (NHC) catalyst motif \n [11] \n was also employed (HGII‐BDP \n [4] \n and HGII‐NHC‐BDP, Figure 1 b, synthesis and characterization see Supporting Information Figure S6–S13). For the NHC‐type complexes a fluorophore tag was introduced on the benzylidene as well as the NHC ligand, since only the latter remains strongly coordinated to the ruthenium center throughout the entire catalytic cycle of olefin metathesis (see below, cf. Figure 3 b). The catalytic performance of these modified catalysts was confirmed by in vitro experiments with extracted microalgae oil as a test substrate \n [12] \n (Supporting Information Figure S16). All complexes are active for self‐metathesis, with a decreased activity of HGII‐BDP vs. its non‐modified parent analogue while for HUC‐BDP the activity is increased compared to the parent unmodified cyclic amino alkyl carbene‐substituted catalyst (Supporting Information Table S5). As model organism, the unicellular microalga Phaeodactylum tricornutum is widely used in research on renewable feedstocks due to its high growth rates, robust cultivation with minimum space requirements and production of high amounts of fatty acids, therefore being a promising target for in vivo catalysis. Please note that although this microalga is a diatom, no siliceous cellular barrier is present in these organisms if cultivated in absence of an appropriate Si source, as performed in this work (see Supporting Information for details on cultivation conditions). \n [13] \n For the approach pursued here, tolerance of the microalgae towards the catalyst is essential. The viability of Phaeodactylum tricornutum cells exposed to the catalysts was studied by MTT assays (colorimetric viability assays based on methylthiazol tetrazolium bromide). No adverse effect on microalgal cell viability was found for HUC‐BDP or HGII‐NHC‐BDP (Figure 2 a), while for unmodified HGII and HGII‐BDP slightly reduced cell viability was observed. The modification of the benzylidene ligand with BODIPY does not affect cell viability (HGII and HGII‐ BDP comparable), whereas the nature of the NHC/CAAC ligand appears to have a decisive influence. Catalysts with carbene ligands differing from the parent HGII complex (HUC‐BDP and HGII‐NHC‐BDP) show significantly higher cell viabilities compared to catalysts with unmodified carbenes (HGII and HGII‐BDP), which is in accordance with previous findings for other NHC‐Ru complexes where the structure of the NHC ligand was identified as a decisive factor for viability of mammalian cells. \n [14] \n In microalgae, the HUC‐BDP catalyst is particularly well tolerated and does not adversely affect cell viability over a broad range of concentrations, an effect also confirmed for prolonged incubation (Supporting Information Figure S18).\n Figure 2 Cell viability and in cell catalytic activity of modified olefin metathesis catalysts. a) Cell viability of microalgae treated with BODIPY‐labelled catalysts. Determined by MTT assays, referenced to controls without added catalyst. b) Intracellular catalytic activity of labelled catalysts. For cells incubated with olefin metathesis catalysts, fluorescent umbelliferone is formed as indicated by an increase in fluorescence intensity over time. Ex.: 322 nm, Em.: 440 nm. Error bars represent mean±standard deviation of three independent biological replicates (N=3). Two‐way ANOVA (with Tukey′s multiple comparisons test for viability assays). p<0.0002 (***), p<0.0021 (**), p<0.0332 (*), p>0.1234 (ns); p‐values <0.0001 for catalytic activity (relative to control sample with substrate only, no catalyst). Catalyst Uptake and in Cell Catalytic Activity To investigate the intracellular catalytic activity of the labelled metathesis catalysts in the microalgae cells, an established pro‐fluorescent substrate was chosen (Figure 2 b, black). \n [15] \n This non‐fluorescent precursor is very reactive in ring‐closing metathesis, yielding umbelliferone as product (Figure 2 b, blue) whose formation can readily be monitored via fluorescence intensities. For catalyst uptake, the microalgae were incubated with aqueous catalyst solutions (containing 0.5 v % of water‐miscible co‐solvent to dissolve the catalyst) and washed thoroughly prior to addition of the non‐fluorescent substrate to avoid formation of fluorescent product by potential extracellular catalyst in the surrounding medium (cf. Supporting Information for details). Monitoring of the fluorescence intensity over time reveals an increase for microalgae precedingly incubated with catalyst, relative to control samples without added catalyst (Figure 2 b, results for HGII‐NHC‐BDP, other catalysts see Supporting Information Figure S21). The olefin metathesis catalysts promote ring‐closing metathesis in microalgae cells, resulting in the formation of the fluorescent umbelliferone. Notably, even though olefin metathesis catalysts are well known to be sensitive to cellular components like water and nucleophiles, \n [16] \n the modified catalysts remain active in the intracellular environment of microalgae in the presence of the manifold cellular components. Elucidation of the uptake and localization of the catalysts by confocal fluorescence microscopy reveals that the BODIPY‐modified catalysts are internalized by the microalgae cells. No additional permeabilization is required and the catalysts are located inside the intracellular lipid bodies (Figure 3 a HUC‐BDP and HGII‐NHC‐BDP, green: fluorescence of catalysts, red: chlorophyll autofluorescence, for HGII‐BDP and comparison with the free BODIPY dye see Supporting Information Figure S22). The intracellular lipid organelles treated with HGII‐BDP show a higher fluorescence intensity compared to lipid bodies with HGII‐NHC‐BDP (Supporting Information Figure S23), which is expected taking into account the catalytic cycle of olefin metathesis (Figure 3 b). \n [2] \n The fluorescence of the BODIPY‐modified catalyst precursors is quenched due to the vicinity of the dye moiety to the ruthenium center. \n [17] \n Catalyst initiation by reaction with olefinic substrates results in the release of the labile benzylidene ligand and therefore in higher fluorescence intensities for benzylidene‐modified catalysts like HGII‐BDP. In contrast, in HGII‐NHC‐BDP the modified NHC ligand is strongly coordinated to the ruthenium center throughout the entire catalytic cycle. Therefore, particularly the observation of intracellular fluorescence for this permanently labelled HGII‐NHC‐BDP (Figure 3 a), showing the same distribution as the HUC‐BDP catalyst, clearly demonstrates catalyst uptake into the lipid storage compartments.\n Figure 3 Cellular uptake of modified olefin metathesis catalysts. a) Microalgae cells of Phaeodactylum tricornutum (Control) treated with modified olefin metathesis catalyst HUC‐BDP and HGII‐NHC‐BDP. Green=BODIPY, red=chlorophyll, scale bar=20 μm, Ex.: 488 nm, brightness and contrast adjusted on each color channel individually. b) Mechanistic scheme of initiation and catalytic cycle of olefin metathesis with HGII/HUC catalysts labelled at the benzylidene ligand. The BODIPY fluorescence is quenched by the proximity of the dye to the ruthenium center and upon initiation of the catalyst the BODIPY is released and the fluorescence quantum yield increased. Green=fluorescence, blue=olefinic substrate, for example intracellular unsaturated fatty acids. Olefin Metathesis of Intracellular Lipid Substrates With the catalysts taken up in lipid bodies, they are localized in the storage compartment of intracellular fatty acids. In the investigated microalga Phaeodactylum tricornutum , mainly unsaturated palmitoleic acid (16 : 1), oleic acid (18 : 1) and eicosapentaenoic acid (20 : 5) are stored in these lipid compartments (Figure 4 a, black), as well as saturated fatty acids not susceptible to olefin metathesis (myristic (14 : 0) and palmitic acid (16 : 0), grey in Figure 4 a). Conversion of the unsaturated compounds via self‐metathesis would result in 7‐tetradecene (A14 : 1), 7‐hexadecene (A16 : 1) and 9‐octadecene dioate (DE18 : 1) as main products (Figure 4 a, blue).\n Figure 4 In vivo olefin metathesis in living photoautotrophic microalgae for the conversion of intracellular substrates. a) Unsaturated fatty acids in the lipid bodies of Phaeodactylum tricornutum microalgae are converted into value‐added chemicals by olefin metathesis. Saturated fatty acids myristic acid (14 : 0) and palmitic acid (16 : 0) are not susceptible to olefin metathesis (grey), whereas self‐metathesis of palmitoleic acid (16 : 1), oleic acid (18 : 1) and eicosapentaenoic acid (20 : 5, black) yields mainly the long‐chain alkenes 7‐tetradecene (A14 : 1), 7‐hexadecene (A16 : 1) as well as the unsaturated diester 9‐octadecene dioate (DE18 : 1, blue). [Ru]=ruthenium‐based olefin metathesis catalyst, R=triglycerides in lipid organelles (storage of fatty acids), R=Me for fatty acid methyl esters (derivatization for analysis via gas chromatography). b) Gas chromatogram of extracted microalgae after in vivo olefin metathesis (top) indicates the formation of self‐metathesis products. The chromatogram of the control (bottom) is shown with relatively amplified intensity to demonstrate the absence of product peaks. 6 mol% catalyst loading, 72 % conversion. Samples transesterified with MeOH/1 v % H 2 SO 4 for analysis via gas chromatography. To probe the catalytic activity of the internalized catalysts for the conversion of these intracellular substrates, the cellular components were extracted after catalyst incubation to verify product formation (note that the stored fatty acids and their corresponding self‐metathesis products do not possess any tags for in vivo investigation). Gas chromatographic analysis of microalgae treated with HUC‐BDP catalyst reveals substantial formation of the self‐metathesis products of the intracellular fatty acids (Figure 4 b, cf. Supporting Information section 7 for full experimental details including reference experiments). All expected major products are identified in the extracts by GC‐FID and GC‐MS analysis (Supporting Information Figure S24–S25). Remarkably, the self‐metathesis products are formed with high conversions (up to 79 %, Figure 4 b, for details on calculation see Supporting Information), similar to the conversions observed in in vitro self‐metathesis (cf. Supporting Information). The small molecule catalysts remain active in the intracellular environment, even though the catalysts are exposed to the incubation medium, internalized by the cells, transported through the microalgal cytoplasm and targeted to the lipid body storage organelles. In terms of different catalysts, the benzylidene modified derivatives HUC‐BDP and HGII‐BDP show comparable in vivo catalytic activity (cf. Supporting Information), but in general the HUC‐BDP catalyst is nevertheless favored due to its lower cytotoxicity (see above). The catalytic activity of HGII‐NHC‐BDP is lower (cf. Supporting Information), which is in accordance with in vitro self‐metathesis experiments of extracted algae oil as model substrate (Table S5, entry 3 vs. 4)."
} | 5,423 |
27231875 | PMC5250801 | pmc | 3,976 | {
"abstract": "The green microalga Dunaliella salina survives in a wide range of salinities via mechanisms involving glycerol synthesis and degradation and is exploited for large amounts of nutraceutical carotenoids produced under stressed conditions. In this study, D. salina CCAP 19/30 was cultured in varying photoperiods and light intensities to study the relationship of light with different growth measurement parameters, with cellular contents of glycerol, starch and carotenoids, and with photosynthesis and respiration. Results show CCAP 19/30 regulated cell volume when growing under light/dark cycles: cell volume increased in the light and decreased in the dark, and these changes corresponded to changes in cellular glycerol content. The decrease in cell volume in the dark was independent of cell division and biological clock and was regulated by the photoperiod of the light/dark cycle. When the light intensity was increased to above 1000 μmol photons m −2 s −1 , cells displayed evidence of photodamage. However, these cells also maintained the maximum level of photosynthesis efficiency and respiration possible, and the growth rate increased as light intensity increased. Significantly, the intracellular glycerol content also increased, >2-fold compared to the content in light intensity of 500 μmol photons m −2 s −1 , but there was no commensurate increase in the pool size of carotenoids. These data suggest that in CCAP 19/30 glycerol stabilized the photosynthetic apparatus for maximum performance in high light intensities, a role normally attributed to carotenoids.",
"introduction": "1 Introduction Microalgae are a source of a variety of natural products ( Priyadarshani and Rath, 2012 , Spolaore et al., 2006 ) including high value nutraceuticals, the exploitation of which started in the 1970’s, with the use of Dunaliella salina for the production of β-carotene, an antioxidant and a precursor of vitamin A ( Raja et al., 2008 ). Carotenoids are essential for photosynthesis within algal chloroplasts: they are involved in both the light harvesting and photoprotecting processes, and in stabilizing the structure of photosynthetic pigment-protein complexes and aiding in their function ( Mimuro and Akimoto, 2003 , Mulders et al., 2014 ). The halotolerant Dunaliella genus comprises green microalgae that have been intensively studied for the production of β-carotene as a valuable compound for the health food industry ( Ben-Amotz et al., 1982a , Davidi et al., 2014 , Tafreshi and Shariati, 2009 ). Dunaliella algae differ from many green algal species as their cells lack a rigid polysaccharide cell wall and are bounded only by a cytoplasmic membrane, which allows them to adjust their volume and shape rapidly in response to hypo- or hyper-osmotic changes ( Sadka et al., 1989 , Zelazny et al., 1995 ). Dunaliella also produces glycerol in a salt medium via photosynthesis and the level of intracellular glycerol has been found to be proportional to the extracellular salt concentration, reaching above 50% of the dry cell weight ( Ben-Amotz et al., 1982b ). In the natural environment, all life is exposed to a daily cycle of light and dark fluctuation of light intensities and seasonal oscillation of daylight length as a result of the rotation of the planet. Eukaryotic and prokaryotic cells have evolved to respond to the rhythmic changes in environmental conditions and synchronize their cellular processes to the most appropriate time of the day ( Dixon et al., 2014 , Duanmu et al., 2014 ). Research on the green microalga Chlamydomonas reinhardtii shows that a wide range of biological processes including cell division, phototaxis, chemotaxis, cell adhesion, and nitrogen metabolism can be regulated by the natural clock and environmental conditions ( Matsuo and Ishiura, 2011 ). Apart from the regulation of biological processes, both the yield and the composition of algal biomass are dependent on environmental light conditions. For example, starch synthesis and degradation in the marine microalga Ostreococcus tauri showed a diurnal pattern with maximum starch content obtained towards the end of the day when cultured under a 12 h light/12 h dark cycle ( Sorokina et al., 2011 ). In industrial algal cultivations, illumination conditions such as continuous light or a light dark cycle, the length of the photoperiod and the light intensity affect both the growth of microalgae and the biomass composition ( Wahidin et al., 2013 ). Continuous illumination in a photobioreactor system is often used to maximize the biomass production; however, excess light energy that cannot be converted into chemical energy induces photoinhibition damage to the algal photosynthetic apparatus, and inhibits the growth of algal cells ( Baroli and Melis, 1998 , Mulders et al., 2014 ). The provision of appropriate light and dark periods is therefore likely to be essential for both the growth of D. salina and optimum yields of target products. So far, despite the fact that intensive research has been carried out on the response of various Dunaliella strains to changes in environmental salt concentrations ( Alkayal et al., 2010 , Goyal, 2007a , Goyal, 2007b , Lin et al., 2013 , Zhao et al., 2013 ), limited information has been reported on the strain D. salina CCAP 19/30 in terms of its growth under varying light cycles and intensities. To explore the effect of light period and light intensity on the regulation of growth, photosynthesis and biomass composition of CCAP 19/30, cells were cultured under a range of light/dark periods within a 24 h cycle and different light intensities and different markers of growth compared with those under continuous light. Photosynthesis of the cells was monitored in relation to light conditions along with the cellular content of photosynthesis-related biomass compounds including chlorophyll, carotenoids, protein, glycerol and starch. Understanding how regulation of light conditions and diurnal control helps to improve the production of algal biomass and desired bioproducts will guide treatment with suitable stressors and determine the time for treatment or harvesting of the algal biomass.",
"discussion": "4 Discussion The work presented here shows that D. salina CCAP 19/30 cells alter their cell diameter in response to a light/dark cycle of light, and that the periodicity of change in cell diameter corresponds to change in cellular glycerol content. Results in Fig. 2 illustrate that the cells increase in volume in the light periods and decrease in volume in the dark periods. To our best knowledge, this is the first report of cell volume oscillation in D. salina when growing under diurnal conditions. Since D. salina cells lack a rigid polysaccharide cell wall, their cytoplasmic membranes allow the cells to adjust their volume and shape rapidly in response to the environmental changes ( Maeda and Thompson, 1986 ). Although early studies have reported rhythmic changes in cell shapes of Euglena gracilis ( Lonergan, 1983 ), the mechanism in E. gracilis is different to that in D. salina reported in this study. E. gracilis cell shape is under direct control of the biological clock and thus even under continuous light, the daily rhythm of cell shape remains. However, when D. salina was grown under continuous light, the oscillation in cell shape ceased ( Fig. 1 ), indicating that it is under the control of diurnal change rather than circadian rhythm, or due to cell division. The oscillation in cell volume and cellular glycerol content of CCAP 19/30 is not found in several other Dunaliella species maintained in the laboratory, including Dunaliella parva , Dunaliella quartolecta , and Dunaliella polymorpha nor in green microalga Chlamydomonas reinhardtii (data not shown); and since no previous report has been found, it may be a species specific property. The synthesis or degradation of glycerol in response to osmotic pressure change is commonly assumed to be triggered by cell volume change ( Ben-Amotz and Avron, 1981 , Zelazny et al., 1995 ). However in this study, we show by using light/dark cycles that the change in cell volume can also be the result of change in cellular glycerol content. Glycerol is produced by Dunaliella via photosynthesis and the adjustment in glycerol concentration is achieved by regulating the carbon flux between either the synthesis of starch or glycerol ( Goyal, 2007b , Goyal, 2007a ). In the dark, there is no carbon fixed to glycerol from photosynthesis, and starch is respired to produce energy and metabolites: the pool size of glycerol is thus reduced. This causes the cell volume to decrease in the dark to maintain the osmotic pressure. With further time in the dark an apparent increase in cellular glycerol and starch contents was observed: this seemingly counter-intuitive observation may nevertheless be due to additional dark-related events such as the effects of lipid catabolism in photoautotrophic algae that have been deprived of an external carbon and energy source, releasing fatty acids and glycerol. It is well documented that exposure of chloroplasts to high light leads to PSII photodamage when the rate of photodamage exceeds that of the repair cycle, leading to photoinhibition and reduced photosynthetic efficiency ( Melis, 1999 , Yokthongwattana and Melis, 2008 ). For the higher plant Arabidopsis thaliana , the rate of photodamage dominates in light intensities greater than 500 μmol photons m −2 s −1 and leads to photoinhibition ( Havaux et al., 2000 ). In our study, light intensity at all values above 500 μmol photons m −2 s −1 resulted in a decrease in the ratio of Fv/Fm by about 34% compared to that at 200 μmol photons m −2 s −1 indicating PSII damage. The decreased photosynthetic rate observed for cultures acclimated to light intensities of 200 and 500 μmol photons m −2 s −1 is therefore likely to be due to photoinhibition. At higher light intensities, however, photosynthetic rate increased to a maximum level ( Fig. 6 A and 6B). This finding implies that CCAP 19/30 might have evolved an efficient repair cycle that allows it to turnover damaged PSII at a much faster rate at high light intensity to allow it to maintain maximum photosynthetic efficiency. Indeed, it has been shown that Dunaliella tertiolecta was able to recover the PS II efficiency by 80% from photodamage within just 1 min of dark adaption ( Casper-Lindley and Björkman, 1998 ). The enhanced rate of photosynthesis at high light intensities in this study contributed to the tolerance of CCAP 19/30 to photodamage. The rate of photodamage is dependent upon Q A redox state, occurring at low probability when Q A is oxidized and excitation energy is utilized in the electron transport chain at a much faster rate ( Melis, 1999 ). Increase in the photosynthetic rate therefore leads to rapid oxidation of the PQ-pool, which in turn drains electrons at a much faster rate from the Q A site of PSII, reducing the possibility of PSII-photodamage. The ability of D. salina to enhance photosynthetic activity under stressed conditions has been previously reported by Liska et al. ( Liska et al., 2004 ). In their study, the enhanced photosynthesis was found to contribute to salinity tolerance of D. salina and cells grown at high salinity showed enhancement of CO 2 assimilation, starch mobilization as well as up regulated key enzymes in photosynthesis. When exposed to high light, the Dunaliella cells are reported to use the carotenoid synthesis pathway as a protective mechanism against photodamage ( Kim et al., 2013 , Mulders et al., 2014 , Park et al., 2013 , Salguero et al., 2003 ). Different Dunaliella strains may vary significantly in their response to light stress and show different sensitivity to the light intensities. In this study, the strain CCAP 19/30 only shows a slight increase in the carotenoids/chlorophyll ratio with both carotenoids and chlorophyll content decreased due to photoinhibition ( Table 2 ). This suggests that carotenoid synthesis in this strain may not be the main functioning mechanism to protect cells from high light stress. Instead, the >2-fold increase in cellular glycerol content at high light indicates glycerol may act as a chemical chaperone to maintain photosynthetic efficiency at high light. In a previous study by Yilancioglu et al. ( Yilancioglu et al., 2014 ), a strong correlation between glycerol production and the maximum and effective photosynthetic yield parameters showed that glycerol plays an important role not only in regulating the osmotic balance but also determining the yield and biochemical composition of the biomass under oxidative stress. Glycerol synthesis in Dunaliella species as a response to osmotic stress has been intensively studied. Upon hyperosmotic stress, Dunaliella cells respond immediately by reducing their cell volume due to water efflux across the cell membrane ( Chen and Jiang, 2009 ). Plasma membrane sterols sense the cell volume change and trigger the synthesis of glycerol ( Zelazny et al., 1995 ). In the present study, increasing cellular concentration of glycerol correlated positively with photodamage to the cells cultured under different light intensities, as indicated by the Fv/Fm values ( Fig. 5 ). At the same time results in this study show that with high light stress, photosynthesis is enhanced and that the increase in photosynthesis was accompanied by an increase in dark respiration rate ( Fig. 6 C and D) and glycerol synthesis ( Fig. 7 A and B). The faster growth rate in high light ( Fig. 4 ) shows energy demand is higher, and was probably met by the faster carbon assimilation and respiration ( Fig. 6 ). These findings accord with a previous study which found that Dunaliella cells contained higher glycerol contents at higher light intensities ( Davis et al., 2015 ). The higher cellular glycerol content of cells grown at high light intensities also indicates larger cell diameters at high light, as shown in Table 2 , the average cell volume of cells indeed increases with the light intensity, which is in line with the study on D. salina CCAP 19/18 by Park et al. ( Park et al., 2013 ). However, despite the significant increase in cellular glycerol content at 1000 and 1500 μmol photons m −2 s −1 (about 2-fold of that at 50, 200 and 500 μmol photons m −2 s −1 ), the cell volume did not show such a significant increase with light ( Table 2 ), indicating that glycerol functions as more than an osmolyte to balance the osmotic pressure in CCAP 19/30, but also a protecting mechanism when under high light stress. In Fig. 8 we propose a working model in which CCAP 19/30 uses glycerol in metabolic homeostasis because glycerol synthesis is able to reduce the possibility of photoinhibition by draining electrons from the Q A site from photosynthesis ( Fig. 7 ). Thus in light, dihydroxyacetone phosphate (DHAP), the precursor of glycerol is isomerised rapidly and reversibly by triose phosphate isomerase from glyceraldehyde-3-phosphate (GAP), the export product of the Calvin cycle in photosynthesis placing an energy demand on the cell, whilst glycerol synthesis from DHAP requires reducing equivalents (NADH or FADH 2 ) for glycerol phosphate dehydrogenase functionality. The parallel increase in respiration ( Fig. 6 C and D) suggests that synthesized glycerol might be used for anabolic metabolism via oxidative respiration using the mitochondrial citric acid cycle. The actual pool size of glycerol however could be a reflection of multiple survival strategies under high light intensity since glycerol can also be used for membrane (triglyceride) regeneration, or stored in the form of starch. In some lipid-storing green algae ( Combe et al., 2015 , Yilancioglu et al., 2014 ) oxidative stress either caused by nitrogen depletion or by exposure to excess light or by application of exogenous H 2 O 2 is correlated with an increase in triglycerides content (but also see ( Ben-Amotz et al., 1985 )). Glycerol can also serve as a biocompatible solute or chemical chaperone to assist in refolding damaged proteins ( Lamitina et al., 2006 ). Clearly glycerol has multiple functions that serve to protect and maintain growth of CCAP 19/30 cells not only in conditions of high salinity but also under high light intensities."
} | 4,090 |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.