pmid stringlengths 8 8 | pmcid stringlengths 8 11 ⌀ | source stringclasses 2
values | rank int64 1 9.78k | sections unknown | tokens int64 3 46.7k |
|---|---|---|---|---|---|
37841953 | PMC10571103 | pmc | 1,410 | {
"abstract": "Chemically crosslinked\nelastomers are a class of polymeric materials\nwith properties that render them useful as adhesives, sealants, and\nin other engineering applications. Poly(γ-methyl-ε-caprolactone)\n(PγMCL) is a hydrolytically degradable and compostable aliphatic\npolyester that can be biosourced and exhibits competitive mechanical\nproperties to traditional elastomers when chemically crosslinked.\nA typical limitation of chemically crosslinked elastomers is that\nthey cannot be reprocessed; however, the incorporation of dynamic\ncovalent bonds can allow for bonds to reversibly break and reform\nunder an external stimulus, usually heat. In this work, we study the\ndynamic behavior and mechanical properties of PγMCL elastomers\nsynthesized from aliphatic dianhydride crosslinkers. The crosslinked\nelastomers in this work were synthesized using the commercially available\ncrosslinkers, 1,2,4,5-cyclohexanetetracarboxylic dianhydride, and\n1,2,3,4-cyclobutanetetracarboxylic dianhydride and three-arm hydroxy-telechelic\nPγMCL star polymers. Stress relaxation experiments on the crosslinked\nnetworks showed an Arrhenius dependence of viscosity with temperature\nwith an activation energy of 118 ± 8 kJ/mol, which agrees well\nwith the activation energy of transesterification exchange chemistry\nobtained from small molecule model studies. Dynamic mechanical thermal\nanalysis and rheological experiments confirmed the dynamic nature\nof the networks and provided insight into the mechanism of exchange\n(i.e., associative or dissociative). Tensile testing showed that these\nmaterials can exhibit high strains at break and low Young’s\nmoduli, characteristic of soft and strong elastomers. By controlling\nthe exchange chemistry and understanding the effect of macromolecular\nstructure on mechanical properties, we prepared the high-performance\nelastomers that can be potentially reprocessed at moderately elevated\ntemperatures.",
"conclusion": "Conclusion We\nhave demonstrated the efficient polymerization of γMCL\nusing an organocatalyst to prepare star-shaped polyesters with hydroxyl\nfunctional end groups that can readily react with commercially available\ndianhydrides without the need of an exogeneous catalyst. Internal\ncarboxylic acids generated during crosslinking facilitate in catalyzing\ntransesterification in these systems leading to interesting rheological\nproperties. Kinetics on model systems demonstrate that rapid exchange\nkinetics occur with excess alcohol moieties. We suggest a dissociative\nmechanism to be the leading mechanism for the CHDA network, due to\nthe decrease in modulus at elevated temperatures which is not observed\nby DMTA in the CBDA network. Activation energies from model reactions\nagree well with those found from stress relaxation experiments. Both\nsystems exhibit a strong temperature response with viscosity. The\ntensile properties of the networks also varied slightly when changing\nthe structure of the dianhydride crosslinker. High stresses and strains\nat break were reported for both systems, with CHDA having a higher\nmodulus. The networks investigated are promising renewable elastomers,\nwith\na biodegradation potential under appropriate conditions. Additionally,\nthe rapid transesterification at elevated temperatures makes these\nsystems the viable candidates for recycling and reprocessing. This\nchemistry can be expanded to other crosslinked systems taking advantage\nof internal catalysis of carboxylic acid groups to prepare materials\nthat are industrially competitive with incumbent materials. Varying\nthe molar mass between crosslinks, crosslink functionality and crosslink\nstructure are the mechanical properties of these systems that can\nbe tuned to afford stronger materials. The properties observed in\nthese elastomers expand the scope of catalyst free CANs. We have demonstrated\nherein a promising candidate for sustainably derived and synthesized\nelastomers taking advantage of commercially available dianhydride\ncrosslinkers. Exploring other avenues of architectural modifications\nincluding varying the number of arms in the prepolymer and exploring\nthe chemical composition of these networks with different polyesters\nor polyanhydrides and additional crosslinkers will provide a broader\nunderstanding of the impact of these systems.",
"introduction": "Introduction Chemically crosslinked polymer networks\ncan typically exhibit a\nvariety of useful properties, including toughness, solvent resistance,\nand thermal stability, which make them versatile materials for the\napplications that include elastomers, adhesives, and foams. However,\npermanent covalent bonds in these crosslinked polymer networks prevent\nthe implementation of traditional mechanical recycling strategies.\nThis leads to an accumulation of waste upon end-of-use, which is disadvantageous\nfrom a sustainability perspective. 1 In\nan effort to address this problem, replacing the permanent chemical\ncrosslinks with dynamic bonds that can be activated with external\nstimuli, such as heat or light, to rearrange network bonds has been\nwidely pursued. 2 − 6 These covalent adaptable networks (CANs) provide a pathway for accessing\ntunable materials that can be recycled by reprocessing, thus increasing\nusage lifetime, an important element of sustainability in this class\nof materials. CANs are generally divided into two categories\nbased on their dynamic\ncrosslink exchange mechanism: associative and dissociative. 7 For associative mechanisms, bond breaking and\nreformation occur simultaneously, in which a pendant or end reactive\ngroup undergoes a substitution reaction with another chain. 7 , 8 As a result, the number of crosslinks in an associative mechanism\nremains constant throughout the course of the exchange reactions. 9 Materials that operate by associative mechanisms\ntypically exhibit an Arrhenius-like (i.e., exponential) decrease in\nviscosity as a function of temperature and have been termed vitrimers. 7 , 10 , 11 Conversely, in CANs that operate\nby dissociative mechanisms, bond breakage precedes bond formation,\nand, as a result, the crosslinking density decreases as bonds are\nbroken at elevated temperature. Due to this loss of crosslink density,\ndissociative CANs typically demonstrate a more rapid decrease in viscosity\nwith increasing temperature when compared to associative networks. 9 Dissociative CANs also exhibit a gel to sol transition\nupon bond breaking that can lead to undesirable properties such as\nthe lack of solvent resistance, creep, and a loss of material integrity. 12 The consideration of the glass transition temperature\n( T g ), exchange reaction activation energy\n( E a ), identity of reactive groups, and\npolymer degradation temperatures is therefore important for selecting\nappropriate exchange chemistries. 13 Some\nof the first vitrimers investigated were crosslinked epoxy-acid and\nepoxy-anhydride networks that underwent transesterification at elevated\ntemperatures. 14 These materials exhibit\nviscosities with an Arrhenius dependence on temperature and in the\npresence of 10 mol % zinc acetyl acetonate [Zn(acac) 2 ]\nshowed an activation energy of 88 kJ/mol. 14 This transesterification reaction occurs between an existing ester\nbond and a free alcohol functional group to generate a new ester bond\nthrough a well-established associative mechanism. However, recent\nwork has demonstrated that transesterification in CANs can be achieved\nthrough a dissociative mechanism when relying on neighboring carboxylic\nacid or sulfonic acid groups. 15 , 16 For the practical\nreasons, the exchange reactions need to occur\non reasonable timescales that allow macroscopic flow at desired reprocessing\ntemperatures. To achieve this, a highly active catalyst is typically\nemployed to lower the activation energy for exchange. Lewis and Brønsted\nacids, as well as metal-based catalysts such as zinc and tin derivatives,\nhave been used in CANs. 4 , 17 − 19 However, relying\non external catalysts to facilitate the exchange reactions can lead\nto the limitations in material properties due to, for example, catalyst\nleeching over time. 20 To move away from\nthe use of external catalysts in CANs, new efforts have focused on\ncatalyst-free systems that rely on an excess of exchanging functional\ngroups, such as vinylogous urethane exchange, 3 , 21 − 23 or on internal catalysts in the as-synthesized networks. 24 − 26 Internal catalysts enable rapid exchange between reactive functional\ngroups, but necessitate close proximity to reactive functional groups\nto be effective. Also called neighboring group participation, this\nmethod increases the reaction rate of exchange by lowering the activation\nenergy as a result of this proximity effect. 27 , 28 The neighboring group participation of carboxylic acids in transesterification\nof phthalate monoesters has been reported to follow a dissociative\npathway and relies on carboxylic acid activating groups. 15 This transesterification exchange can be applied\nto aliphatic polyester elastomers as a way to prepare renewable, reprocessable,\nand degradable CANs. Aliphatic polyesters are an attractive\nalternative to hydrocarbon-based\npolymers due to their ability to be derived from renewable feedstocks,\nrecycled, and composted. 29 − 31 Polyesters can be readily prepared\nthrough the ring-opening transesterification polymerization (ROTEP)\nof cyclic ester (i.e., lactone) monomers. ROTEP allows for a high\ndegree of control over molar mass, narrow molar mass distributions,\nand the facile preparation of various molecular architectures, which\ncan in turn significantly impact the physical and mechanical properties\nof resulting materials. 29 The ester linkages\nalong the polymer backbone also allow for ready hydrolysis and polymer\ndegradation. A variety of organic catalysts including 1,8-diazobicyclo[5.4.0]undec-7-ene\n(DBU), 32 diphenyl phosphate (DPP), 33 and dimethyl phosphate 34 have been used as greener replacements for tin-based catalysts,\nsuch as stannous octoate (Sn(Oct) 2 ), 35 , 36 in the synthesis of these polyesters. We have reported the synthesis\nof high-performance thermoplastic elastomers (TPEs) 37 , 38 and chemically crosslinked elastomers from γ-methyl-ε-caprolactone\n(γMCL), 39 a seven membered lactone\nthat can be derived from p-cresol (that can in turn be sourced from\nbiorenewable lignin feedstocks). 40 Our\ngroup has also shown the efficient polymerization and control of molar\nmass of poly(γMCL) (PγMCL) with the use of such organocatalysts. 34 Synthesized elastomers exhibited impressive\nmechanical properties with high ultimate tensile strengths and elongation\nat break, both attributed to the contributions from chemical crosslinks\nand trapped entanglements. However, reported syntheses of these crosslinked\npolyester elastomers relied on catalysts such as Sn(oct) 2 and multistep syntheses for the formation of crosslinkers investigated.\nAnalogous crosslinked polyesters made from PγMCL have also been\nshown to hydrolyze enzymatically and degrade to high extents under\nindustrial composting conditions. 41 In this work, we seek to understand the behavior of transesterification\nin polyester CANs of PγMCL crosslinked with readily available\naliphatic dianhydrides. Taking advantage of the hydroxy end-functionalized\nPγMCL, crosslinking with aliphatic dianhydrides will allow for\na rapid exchange due to the presence of neighboring carboxylic acid\ngroups. A focus of this work is to understand the effect of these\nneighboring carboxylic acids in lowering the activation energy barrier\nfor exchange through kinetics studies on model systems. We also utilize\nthe facile synthesis of PγMCL using DPP as an organocatalyst\nand subsequent catalyst-free crosslinking. 34 We discuss the differences in stress relaxation behavior and material\nproperties on polyester networks synthesized from two commercially\navailable dianhydrides and connect these results to the conclusions\nabout the mechanism of exchange.",
"discussion": "Results and Discussion Model\nReaction Studies The kinetics of esterification\nbetween anhydrides and alcohols have been well studied. 42 , 43 At elevated temperatures, cyclic anhydrides can react with hydroxyl\nmoieties without added catalysts. 44 The\nesterification of a cyclic anhydride by an alcohol results in the\nformation of an ester and a carboxylic acid in close proximity, which\ncan influence subsequent transesterification of the newly formed ester\nthrough self-catalysis. However, the second esterification of this\npendant carboxylic acid with concomitant dehydration to form the corresponding\ndiester typically requires the addition of a strong acid catalyst.\nSuch a strong acid protonates and activates the carboxylic acid for\nnucleophilic attack from alcohols, allowing for the formation of the\ndiester. 43 Therefore, in the absence of\na strong acid, only the first esterification typically takes place\neven in the presence of excess alcohol. Given this, it is likely that\nthe carboxylic acid formed during crosslinking of telechelic PγMCL\nwith cyclic dianhydrides acts as a neighboring group catalyst during\nthe transesterification of these networks. To gain insight to this,\nwe investigated a model system using a cyclic aliphatic anhydride\nand long-chain alcohol to model crosslinking with a hydroxyl end-functionalized\nPγMCL chain as well as potential transesterification with free\nalcohols. Esterification kinetics were conducted on model systems\nusing 1,2-cyclohexanedicarboxylic anhydride (CHMA) and dodecanol.\nExcess dodecanol was used to promote a pseudo-first-order process\n( Figure 1 a). The melting\ntemperatures of dodecanol and CHMA are 34 and 24 °C, respectively,\nallowing for the kinetics to be readily conducted in bulk, which is\nrelevant and convenient for modeling practical polyester elastomer\nsynthesis. Aliquots taken at periodic intervals were quenched by dilution\nin CDCl 3 below room temperature, and the reaction kinetics\nwere tracked using 1 H NMR spectroscopy ( Figure S4 ). The formation of the monoester was observed at\ntemperatures of 40 °C and above; at 120 °C, the formation\nof the ester was complete after 8 min. The slopes of the pseudo-first-order\nkinetics plot in Figure 1 b were used to determine the apparent first-order rate constants, k , at each temperature (sample calculation in Supporting Information ). These rate constants\nwere then used in an Arrhenius analysis to give an activation energy\nfor this ring-opening under neat conditions of 41 ± 2 kJ/mol\n( Figure 1 c). This activation\nenergy is similar to previously reported activation energies for the\nuncatalyzed esterification of similar cyclic anhydrides including\nphthalic anhydride with 2-(2-methoxyethoxy)ethanol which gave an activation\nenergy of 28 ± 2 kJ/mol. 45 The rapid\nformation of the monoester indicates that crosslinking should occur\nrapidly between dianhydrides and hydroxyl end-functionalized PγMCL\nstar polymers at moderate temperatures. The second esterification\nis more demanding and often characterized by much higher activation\nenergies (e.g., >180 kJ/mol) to obtain the diester even with excess\nalcohol and the presence of strong acid catalysts, as is the case\nin traditional Fischer esterification. 46 Figure 1 (a)\nScheme describing the model reaction used for esterification,\nconducted in bulk (i.e., neat, no solvent). (b) Pseudo-first-order\nring-opening reaction kinetics between excess dodecanol and CHMA where\n[M] is the concentration of anhydride. (c) Arrhenius analysis of dodecanol\nand CHMA ring-opening reaction kinetics from 40 to 120 °C. While the esterification kinetics aid in understanding\nthe rate\nof crosslinking, the kinetics that describe the rate of bond exchange\nin the networks are those of transesterification between the ester\nformed from anhydride ring-opening and a free hydroxyl group. In these\npolyester systems, it is also likely that interchain esters participate\nin transesterification as also seen in the synthesis of copolyesters. 47 To explore transesterification, a similar method\nto the esterification model kinetics was employed. We first prepared\nand isolated the dodecyl ester cyclohexyl carboxylic acid product\nfrom the above reaction for the transesterification kinetics. An excess\nof benzyl alcohol was used to model this as a pseudo-first-order reaction\n( Figure 2 a). Benzyl\nalcohol was chosen due to the distinct resonances in the 1 H NMR spectrum for the protons α to the alcohol in benzyl alcohol\nand the protons α to the alcohol in dodecanol ( Figure S5 ). The formation of the benzyl monoester was observed\nat temperatures above 120 °C. A first-order kinetics plot of\nthe transesterification reaction in Figure 2 b shows the slopes at each associated temperature,\ncorresponding to the rate constants. Arrhenius analysis gave an activation\nenergy of 83 ± 6 kJ/mol ( Figure 2 c). Additionally, this transesterification reaction\nis likely internally catalyzed, with an activation energy close to\nthe previously reported transesterification in model systems using\nphthalic monoesters ( E a = 95 ± 5\nkJ/mol). 15 Figure 2 (a) Scheme describing the model reaction\nused for transesterification\n(b) Pseudo-first-order transesterification reaction kinetics between\nexcess dodecyl monoester and benzyl alcohol where [M] is the concentration\nof dodecyl monoester. (c) Arrhenius plot of dodecyl monoester and\nbenzyl alcohol exchange kinetics. The mechanism and kinetics of exchange can be correlated with the\nviscoelastic behavior of CANs. 48 We explored\nadditional model reactions to understand whether dissociative or associative\npathways were favored. 49 In a dissociative\npathway, the rate-limiting step would be the reformation of the cyclic\nanhydride through a cleavage of the ester bond with concomitant release\nof alcohol, followed by the rapid reopening of the anhydride with\nanother alcohol. An associative mechanism requires the addition of\na free alcohol to the ester near the carboxylic acid to form the\nusual tetrahedral intermediate in the rate-determining step, followed\nby the reformation of the carbonyl bond and expulsion of the alcohol\ngroup from the original ester. A dissociative pathway has been reported\nin model systems using phthalate monoesters, but has not been explored\nwith aliphatic monoesters. 15 While we anticipate\nthe transesterification mechanism to be similar between aromatic and\naliphatic anhydrides, we aimed to confirm the supposition that activation\nenergies found from our model system were consistent with those we\nfound from stress relaxation on our crosslinked materials. In an associative\nmechanism, increasing the concentration of free alcohol functional\ngroups should increase the overall rate of transesterification. While\nfor a dissociative mechanism, the rate would largely be unaffected\nif the rate-limiting step was the initial reformation of the anhydride,\nproviding that the alcohol was not involved in anhydride reformation. By 1 H NMR spectroscopy, we studied the effect on initial\nreaction rate when mono-1-dodecyl cyclohexyl carboxylic acid is reacted\nwith 0.5, 1, or 2 equivalents of monoester at 140 °C. The data\nare shown in Figure S6 , and the exact neat\nconcentrations are provided in Table S1 . The initial rate of ester exchange at approximately equimolar conditions\nis about 6.0 × 10 –4 M min –1 . At 2 equiv. of benzyl alcohol to 1 equiv. of monoester, the concentration\nof the monoester was lower by about 52% under these neat conditions.\nThe rate was lower at 3.7 × 10 –4 M min –1 , but not significantly different than that under\nequimolar conditions. Conversely, at 2 equiv. of monoester to 1 equiv.\nof alcohol, the rate increased to 9.9 × 10 –4 M min –1 . Combined, these data are consistent with\na rate-limiting step being cyclic anhydride formation early stages\nof the reaction. This is further supported with pseudo-first-order\nkinetics for both the monoester consumption and benzyl alcohol consumption\nunder the conditions that dodecyl monoester was held in a 5:1 molar\nratio to benzyl alcohol. Figure S7 shows\nthe apparent rates of the reaction with excess dodecyl monoester to\nbe greater than the reaction with excess benzyl alcohol. To\ninvestigate the effect of the carboxylic acid on the rate of\ntransesterification, a reference model reaction was studied wherein\nthe carboxylic acid at the β position was replaced with a hydrogen\nas seen in Figure 3 a. Figure 3 b,c show\nthe difference in reaction rate when the group is replaced from a\ncarboxylic acid to a hydrogen. Over an order of magnitude increase\nin the rate constant of reaction from 5.9 × 10 –4 to 6.7 × 10 –3 min –1 was\nseen for the carboxylic acid derivative, indicating that the presence\nof the internal acid facilitates the exchange reaction as expected. Figure 3 (a) Scheme\ndescribing the model reaction used for transesterification\nreplacing carboxylic acid with hydrogen (b) kinetics of transesterification\nreaction. (c) Pseudo-first-order transesterification reaction kinetics\nbetween excess dodecyl monoester with carboxylic acid or hydrogen\nat β position and benzyl alcohol where [M] is the concentration\nof dodecyl monoester. Synthesis of Crosslinked\nElastomers The results from\nthe model reactions helped inform the design of dynamic polyester\nelastomers with the rapid transesterification facilitated by the neighboring\ncarboxylic acids. Low molar mass PγMCL prepolymer was synthesized\nthrough ROTEP using trimethylol propane, a trifunctional initiator\nand DPP as an organocatalyst according to previous work. 34 The resultant star polymer with a M n of 9.7 kDa with hydroxyl functional groups at each chain\nend was subsequently reacted with a dianhydride crosslinker in an\nequimolar ratio of alcohol to anhydride functional groups ( Scheme 1 ). Two commercially\navailable aliphatic dianhydrides were chosen as crosslinking reagents\ndue to their solubility and our understanding of kinetics from model\nsystems. The differences between 1,2,4,5-cyclohexanetetracarboxylic\ndianhydride (CHDA) and 1,2,3,4-cyclobutanetetracarboxylic dianhydride\n(CBDA) were explored to understand the differences in material properties\nand mechanism of exchange. Scheme 1 Synthesis of Crosslinked Polyester Networks Two dianhydrides were investigated\nduring the catalyst-free crosslinking step. Solution casting the polymer and crosslinker from a mixture of\nTHF (2 mL) and dimethylformamide (1 mL) was required to solubilize\nthe crosslinker. The films were dried under a steady flow of nitrogen\nfor 16 h before they were reacted at 120 °C for 24 h under nitrogen\natmosphere. The resulting films were clear and colorless, void of\nany obvious macroscopic defects ( Figure S9 ). Consumption of the anhydride was determined by FTIR spectroscopy\nfor both CHDA and CBDA ( Figures S10 and S11 ). The disappearance of the anhydride carbonyl stretch at 1780 cm –1 indicates near complete consumption of the anhydride\nduring crosslinking and thus a high degree of alcohol consumption\nunder these stoichiometric conditions. The degree of crosslinking\nwas evaluated by swelling the films in THF for 24 h and extracting\nthe soluble fraction. The theoretical extent of conversion required\nfor the gelation of a stoichiometrically-balanced A 3 +\nB 2 system is 0.73. 50 The resultant\ngel fractions were greater than 0.85 showing high degrees of crosslinking\nand are listed in Table S3 . The glass transition\n( T g ) of the crosslinked elastomers was\n−56 °C, close to the prepolymer T g of −60 °C ( Tables S2 and S3 ). When the crosslinking temperature of the network\ncrosslinked with\nCHDA was increased to 180 °C and the film was left for 2 h, we\nobserved the anhydride stretch in the FTIR ( Figures 4 and S10–S12 ). This supports a dissociative cleavage of crosslink junctions leading\nto a reappearance of cyclic anhydride intermediates at these higher\ntemperatures. However, upon cooling, the gel fractions of the films\nincreased after this high temperature treatment, consistent with more\nextensive crosslinking, suggesting that dissociation may not be the\nonly reaction taking place. It is possible that at 180 °C, dehydration\nand formation of esters from esterification between neighboring carboxylic\nacid groups generated from the ring-opened anhydrides is also ocurring.\nThis would result in the formation of tri- or tetrafunctional crosslink\njunctions with no pendant carboxylic acid groups, thus limiting further\nexchange behavior ( Scheme 2 ). We postulate at elevated temperatures there are contributions\nfrom both dissociation of crosslinks and formation of anhydride junctions\nfrom dehydration of carboxylic acids. Figures S13 and S14 show a TGA isotherm of an analogous PγMCL\nelastomer held at 180 °C with material mass loss likely coming\nfrom dehydration in the network. The FTIR spectrum of this experiment\nbefore and after exposure is shown in Figure S15 . Figure 4 (a) Scheme of likely bond exchange occurring at 120 and 180 °C\n(b) IR spectrum of films crosslinked with CHDA before and after heating\nin oven under flow of nitrogen air at 120 °C for 24 h and subsequently\nexposed 180 °C. The peak at 1780 cm –1 corresponds\nto the stretch of the anhydride while the peak at 1726 cm –1 corresponds to the ester stretch of the polymer backbone. Scheme 2 Formation of Anhydride from Carboxylic Acid Dehydration\nat Crosslink\nJunction Leading to a Trifunctional Crosslink Junction The results for similar experiments with the CBDA networks\ncan\nbe found in Figure S16 , in which some evidence\nof anhydride is seen at elevated temperatures. When the networks were\ndirectly crosslinked at 180 °C, the gel fractions and glass transition\ntemperatures increased while stress relaxation times decreased, indicating\na higher degree of crosslinking ( Table S3 and Figure S17 ). Mechanical Properties of Crosslinked Elastomers The\ntemperature–modulus relationships in the PγMCL elastomers\nwere investigated by DMTA where the storage and loss moduli of both\nnetworks were monitored from −70 to 200 °C under uniaxial\nextension at a heating rate of 5 °C min –1 ( Figure 5 ). The storage modulus\nwas greater than the loss modulus at all temperatures above the glass\ntransition, consistent with crosslinked networks. The networks with\nboth crosslinkers exhibit glass transitions around −57 °C\nand the rubbery plateau modulus then remained constant until dramatic\nmaterial softening or the end of the experiment, and both networks\nmaintained mechanical integrity during testing. Specifically, the\nCBDA network maintains its plateau modulus over the entire temperature\nrange investigated ( T > 200 °C) and even\na slight\nincrease in modulus is observed at higher temperatures. However, in\nthe CHDA networks, the modulus begins to rapidly drop at approximately\n150 °C. This precipitous drop in the modulus reflects a softening\nbehavior and demonstrates that the viscosity of the system is decreasing\nlikely through a loss of crosslink density. The more rapid decrease\nin modulus observed in the CHDA system at high temperature system\nfurther supports the claim that these networks are exchanging through\na dissociative mechanism, as has been observed in Diels–Alder\npolymer networks. 51 We conclude this because\nthe energy barrier to reform the cyclohexyl anhydride is lower than\nthe cyclobutyl anhydride due to ring strain. 52 Additionally, the position of the anhydrides as cis or trans relative\nto the ring can contribute to the difficulty or ease to reform the\nanhydride. While the CBDA crosslinker is cis, the CHDA crosslinker\nused is a mixture of cis and trans isomers. The dissociative pathway\nis also supported by the recent work from DuPrez and coworkers in\npyromellitic dianhydride polyester networks. 15 Figure 5 DMTA\nplot of a three-arm 9.7 kDa PγMCL prepolymer crosslinked\nwith CHDA (blue) and CBDA (black). Samples were heated at a rate of\n5 °C min –1 from −70 to 150 °C.\nω = 6.28 rad/s. The theoretical molar\nmass between crosslinks ( M x ) for the crosslinked\nelastomers was 6.5 kg/mol. Contributions\nfrom both entanglements and crosslinks contribute to the rubbery plateau\nmodulus ( E N ′), and this value can\nbe used to estimate the effective molar mass between crosslinks M x,eff using eq 1 where E ′ is the storage modulus\nunder tension, T is the absolute temperature, taken\nat the minimum of the tan δ data, and ρ is the density\nof the PγMCL films, which was estimated to be 1.066. From this,\nthe calculated values of M x,eff were 7.1\nand 7.9 kg/mol for the CHDA and CBDA networks, respectively. These\nvalues are closer to the predicted M x than\nthe theoretical entanglement molar mass ( M e ) of 2.9 kg/mol for PγMCL, indicating that at low strains,\nmore contribution comes from crosslinks than entanglements, likely\ndue to the molar mass of the prepolymer is not high enough to lead\nto significant contributions from entanglements. 1 One of the ways to probe\nexchange mechanisms in CANs is through\nstress relaxation experiments. These stress relaxation experiments\nmeasure the stress response of a viscoelastic material placed under\na step (fixed) strain. In crosslinked elastomers, contributions from\nboth entanglements and crosslinks contribute to the modulus, and thus,\nthere may be several relaxation pathways present during stress relaxation\nexperiments. However, well above the T g in crosslinked CANs, viscous flow is limited by the rate of bond\nexchange in the networks. 53 The Arrhenius\ndependence of viscosity on temperature in CANs has been observed in\nboth networks undergoing associative and dissociative bond exchange\nmechanisms. 7 The activation energy of this\nbond exchange can be related to the terminal relaxation time found\nfrom stress relaxation experiments, as shown in eq 2 where τ 0 is the pre-exponential\nfactor, or the terminal relaxation time without thermal constraints, 7 , 54 E a S is the activation\nenergy from stress relaxation, R is the gas constant,\nand T is the temperature at which the experiment\nwas conducted. The characteristic relaxation time (τ*) is the\ntime at which the modulus ( E ′) has decreased\nto 1/ e of its initial value. Plotting the natural\nlog of these relaxation times against temperature and calculation\nof the slope allows for the estimation of the activation energy for\nstress relaxation. 2 Stress relaxation experiments were\nconducted on two crosslinked\nsystems comprised of a 9.6 kg/mol prepolymer crosslinked with an equimolar\nratio of either CHDA or CBDA to alcohol functional groups. Figure 6 a,c show the stress\nrelaxation curves of the two systems investigated at temperatures\nranging from 100 to 180 °C. All data, including samples that\ndid not fully relax past 1/ e , were fit to a stretched\nexponential decay function to obtain a value of τ* ( eq S1 , fitting parameters listed in Table S4 ). An Arrhenius plot of the characteristic\nrelaxation times allowed for the calculation of activation energies\n( E a S ) of 118 ± 8\nand 83 ± 17 kJ/mol for the CHDA and CBDA networks, respectively\n( Figure 6 b,d). This\nfalls within the range of 30–160 kJ/mol as has been observed\nin most transesterification vitrimers. 27 The stress relaxation plots in Figure 6 and the non-normalized stress relaxation\nplots seen in Figures S19 and S20 show\nmore rapid relaxation in the CHDA system compared to the CBDA systems\nat analogous temperatures. The differences between these relaxation\ntimes at various temperatures indicate that there may be more than\none relaxation pathway contributing to a higher activation energy\nfor the CHDA over CBDA system. It takes over 3000 s in the CBDA network\nfor the modulus to relax to 1/ e at 180 °C while\nit only takes 70 s in the CHDA network at the same temperature. Additionally,\nthe shape of the curves in the CBDA networks differs at higher temperatures\nthan those observed at lower probed temperatures. This could suggest\nthat different relaxation processes are occurring at different temperatures,\nmaking correlation between temperature and viscosity in this network\nmore difficult. Deviations from temperature–viscosity linearity\nhave been observed in transesterification vitrimers with two distinct\nslopes in the Arrhenius plot, indicating two mechanisms of relaxation:\nbond exchange through chemically limited transesterification at low\ntemperatures and diffusion limited dynamic rearrangement of network\nstrands at high temperatures. 55 In PDMS\nvitrimers, three regimes arise from temperature–viscosity relationships.\nAt temperatures T g + 200 K, bond exchange\ndynamics control viscosity, and at temperatures < T g + 200 K, polymer segmental dynamics control viscosity. 56 Figure 6 (a) Stress relaxation curves of 9 K PγMCL network\ncrosslinked\nwith CHDA. Samples were held at 5% strain for 3 h or until complete\nrelaxation. (b) Arrhenius plot of PγMCL-CHDA network. (c) Stress\nrelaxation curves of 9 K PγMCL network crosslinked with CBDA.\n(d) Arrhenius plot of PγMCL-CBDA network. Dashed lines indicate\nthe fits of data to eq S1 . The prefactor obtained in both these systems\nis quite\ndifferent. In the CHDA network, τ 0 is lower than\nthe CBDA network while in the CHDA network has a higher E a S . This prefactor can be understood\nas an extrapolation of the Arrhenius plot to infinite temperature\nand is what the characteristic relaxation time would be if there are\nno contributions from network rearrangements. A wider range of temperatures\nwould need to be examined to obtain a complete understanding of this\neffect. The more rapid decrease in modulus for the CHDA elastomer\nover\nthe CBDA elastomer is also consistent with a dissociative mechanism\nfor bond exchange. In a dissociative mechanism, the rate-limiting\nstep is the reformation of the cyclic anhydride. There is an increased\nenergy barrier to reform the cyclic anhydride in the cyclobutyl system\nbecause we anticipate that the ring strain would be greater than in\nthe cyclohexyl dianhydride. An associative mechanism would only be\nrate limited by the formation of a tetrahedral intermediate in the\nunstrained structures. However, above 140 °C, there could also\nbe competing effects between network dissociation and dehydration,\nleading to the formation of tri and tetra esters at crosslink junctions,\nwhich would limit bond exchange and result in longer relaxation times\n( Figures S21 and S22 ). Rheological\nAnalysis The reversibility of network dissociation\nwas studied through cyclic DMTA ramps under shear. Prior to the experiment,\ndynamic strain sweeps were conducted to determine the appropriate\nstrain in the linear viscoelastic regime for each network ( Figure S23 ). The temperature sweep experiments\naimed to probe the change in modulus of the crosslinked elastomer\nover a range of temperatures ( T = 100–180\n°C) above the temperature at which bond exchange is substantively\nactivated ( Figure 7 ). At 120 °C, the modulus of the CHDA network is initially around\n87 kPa while that of the CBDA network is approximately 100 kPa. These\ndifferences could be attributed to a greater percentage of crosslinks\nbeing dissociated upon the first heating cycle in the CHDA system.\nDuring the first heating cycle, the modulus of the CHDA system decreases\nover an order of magnitude. This is attributed to the dissociation\nof crosslinks leading to a less dense network, and thus, a drop in\nmodulus is anticipated. Upon the cooling cycles, the modulus of the\nmaterial begins to increase again, indicating the reformation of network\nbonds. However, the modulus does not return to its initial value after\nthe first heating cycle, indicating that not all the crosslinks have\nbeen reformed. More bonds may be reformed if the sample was cooled\nback to room temperature. The second and third heating cycles follow\nsimilar trends where the modulus decreases upon heating but increases\nupon cooling. More bonds may be reformed if the sample was cooled\nback to room temperature ( Figure S24 ). Figure 7 Reversibility\nof network dissociation through oscillatory shear\nexperiments for 3 cycles of heating and cooling from 120 to 180 °C.\nCHDA (left) and CBDA (right). Samples were heated and cooled at a\nrate of 5 °C/min and held at 1% strain and ω = 6.28 rad/s. These results differ in the CBDA system, particularly\nthat over\neach of the heating and cooling cycles besides the first, the modulus\nincreases. The first heating cycle results in a decrease in modulus\nbut not as much as the CHDA system. Interestingly, the subsequent\nheating and cooling cycles see the increases in the modulus. This\ncould be attributed to the dehydration in the network releasing water\nand forming ester junctions between carboxylic acids. The formation\nof more crosslinks would increase the modulus of the material. This\nwas confirmed with gel fractions on the recovered material from the\nrheological experiment, which increased from 0.86 to 0.94 after rheological\ntesting, which was not seen in the CHDA sample. Tensile Properties The elastic performance of the synthesized\nnetworks was investigated by linear and cyclical tensile tests. The\nYoung’s modulus was determined by the slope of the stress–strain\ncurve in the low strain limit. The two elastomers demonstrated low\nYoung’s moduli, as seen by the inset in Figure 8 . Though the theoretical molecular weight\nbetween crosslinks ( M x ) of both networks\nis the same, there is a difference in the Young’s modulus between\nthe CHDA and CBDA network. The lower Young’s modulus in the\nCBDA network could be attributed to the network defects and loops\nwhich are difficult to quantify. 57 It has\nbeen shown that the effect of aromatic or aliphatic crosslinker structure\nin epoxy-resins has a strong effect on the elastic modulus and there\nmay be a contribution here due to the choice of the crosslinker. 58 The strain at break was greater for the CBDA\nnetwork while the stress at break was greater than the CHDA network.\nOverall, these elastomers exhibited higher stresses and strains at\nbreak when compared to analogous elastomers crosslinked with a similar\nPγMCL prepolymer molar mass and bis-β-lactone crosslinker. 39 Figure 8 Representative stress–strain curves comparing crosslinked\nelastomers from PγMCL with a theoretical M x of 6.5 kg/mol with different crosslinkers. Samples were extended\nat a 50 mm min –1 rate until the sample broke. The\ninset shows the difference in Young’s moduli from 0 to 10%\nstrain. Cyclical loading and unloading\nexperiments on the dianhydride polyester\nnetworks were conducted to understand energy dissipation and exhibited\nlow energy loss over 20 cycles ( Figure 9 ). The largest energy loss was found for the first\ncycle. The hysteresis energy loss of the CBDA network was slightly\ngreater than that of the CHDA network. Figure 9 Cyclical uniaxial extension\ntests for a 10 kg/mol PγMCL three-arm\nstar prepolymer crosslinked with CHDA (left) and CBDA (right). Samples\nwere extended to 50% strain at a rate of 50 mm min –1 ."
} | 9,716 |
33077707 | PMC7572474 | pmc | 1,411 | {
"abstract": "The development of reliable, mixed-culture biotechnological processes hinges on understanding how microbial ecosystems respond to disturbances. Here we reveal extensive phenotypic plasticity and niche complementarity in oleaginous microbial populations from a biological wastewater treatment plant. We perform meta-omics analyses (metagenomics, metatranscriptomics, metaproteomics and metabolomics) on in situ samples over 14 months at weekly intervals. Based on 1,364 de novo metagenome-assembled genomes, we uncover four distinct fundamental niche types. Throughout the time-series, we observe a major, transient shift in community structure, coinciding with substrate availability changes. Functional omics data reveals extensive variation in gene expression and substrate usage amongst community members. Ex situ bioreactor experiments confirm that responses occur within five hours of a pulse disturbance, demonstrating rapid adaptation by specific populations. Our results show that community resistance and resilience are a function of phenotypic plasticity and niche complementarity, and set the foundation for future ecological engineering efforts.",
"introduction": "Introduction Mixed-culture biotechnological processes are essential for humankind to achieve its sustainable development goals 1 , 2 . However, in order to engineer reliable processes, fundamental insights into microbial niche ecology are necessary. Biological wastewater treatment plants (BWWTPs) represent a ubiquitous biotechnological application and occupy a central position in sustainable resource management plans 3 , 4 . Oleaginous bacterial populations are commonly found as the main constituents of floating sludge in BWWTPs and include divergent taxa such as Candidatus Microthrix parvicella or Acinetobacter spp. 5 . Storage lipids, such as triacylglycerols (TAGs), wax esters (WEs), and polyhydroxyalkanoates (PHA), derived from the lipid-rich biomass can directly be transesterified to fatty acid alkyl esters (biodiesel) 5 , whereby PHA also represents a suitable precursor for bioplastics 6 . In general terms, substrate storage provides microbial populations with a competitive advantage under rapidly fluctuating and oftentimes sparse substrate conditions 7 , 8 . Even though BWWTP operation is a controlled process, factors such as aeration cycles, seasonal changes in temperature, and composition of inflow wastewater fluctuate. These factors have a profound impact on population dynamics 9 as well as linked process efficiency 10 . For example, periods of inefficient operation have been linked to competition between polyphosphate and glycogen accumulating organisms 11 . However, for wastewater-borne lipid-accumulating populations, which have compelling potential to be used in circular economic models 3 , community shifts have been observed 12 – 14 with yet unclear links to niche ecology in situ. Integrated meta-omics approaches hold the potential to resolve the fundamental niches and realized niches of microbial populations in situ 15 . The former represents the exhaustive inventory of resource ranges and conditions sustaining viability in the absence of environmental stress, competition, or predation, while the latter represents the part of a fundamental niche that is actually utilized by a population in the presence of other species and in a particular environment. The reconstruction of the fundamental niches is possible by linking functional potential to metagenome-assembled genomes (MAGs) 16 obtained through metagenomic (MG) sequencing. Functional omics data, such as metatranscriptomics (MT) or metaproteomics (MP), allow the resolution of realized niches 16 . Meta-omics approaches have previously been used for comparative functional screening in different environments and to characterize microbial activity, e.g., by using MT/MG ratios 17 , 18 . In human gut-borne microbial communities, niche partitioning has been inferred based on transcriptional profiles 19 . Furthermore, the coupling of MT and MP to meta-metabolomic (MM) data allows the differentiation between niches of genetically closely related populations 20 . Resolving the functions of coexisting microbial populations is of particular interest in the context of the extensive functional redundancy within microbial ecosystems 21 , 22 . Based on their emergent properties 23 , microbial communities are characterized by composite metabolic capabilities and increased robustness compared to individual strains 24 , 25 . Steering these complex systems towards a desired endpoint, e.g., increased lipid accumulation, requires in-depth understanding of niche space and stability. Here, we study whether community resistance and resilience are a function of phenotypic plasticity and niche complementarity. We develop and apply a novel framework for the in situ characterization of fundamental and realized niches of individual populations providing an in-depth understanding of ecological processes within a microbial community. We delineate ecological niches by integrating longitudinal meta-omics data (MG, MT, MP, and MM) and study complementarity of the realized niches. The addition of functional omics data (MT, MP, and MM) enables the resolution of metabolic plasticity and we thereby reveal how microbial ecosystems respond to disturbance. Using ex situ experiments to simulate pulse disturbances, we assess the response of individual oleaginous populations to oleic acid addition under shifting dissolved oxygen concentrations. Our dataset and methods represent important resources for the emerging field of integrating meta-omics data to study mixed microbial communities. Our results contribute to applications beyond wastewater treatment such as informed ecological engineering or research on host-associated microbiota.",
"discussion": "Discussion The ability to reconstruct population-level genomes and infer their functional potential from metagenomes allows identification of the fundamental niches of distinct community members. Unprecedented views of realized niches are achieved by tracking functional gene expression via MT and MP analyses, as well as actual resource usage resolved via comparative metabolomics analyses of intracellular and extracellular metabolites. The joint resolution of fundamental and realized niche breadths of individual populations is key to understanding the ecological processes within microbial communities, including, but not limited to, how such consortia respond to disturbance. Here, through the application of our novel framework for the integration of multi-meta-omics datasets, we were able to track community-wide and population-resolved traits longitudinally in situ as well as ex situ. We found four distinct fundamental niche types in this ecosystem. Populations assigned to a specific type shared common functional repertoires and largely shared a similar phylogenetic background, in line with previously observed metabolic repertoires 47 , 48 . Simultaneously, some functions, e.g., related to lipid accumulation, were found to be enriched in multiple niche types. Despite our results showing a link between functional complement, realized niches, and phylogeny, we also observed distinct activities in response to the changing environmental conditions within individual niche types, e.g., some lowly abundant populations exhibited high activity. This suggests distinct adaptation strategies to variabilities in the resource space and is exemplified by the populations in the functional cluster that includes the dominant Microthrix population. Microthrix follows a strategy based on phenotypic heterogeneity for rapid adaptation to the prevailing environmental conditions 13 . Our ex situ validation experiments revealed the adaptations to changes in substrate availability and dissolved oxygen concentrations after as little as 5 h post-disturbance. This plasticity in gene expression allows the populations to be resistant to fluctuations in environmental conditions. Furthermore, this strategy was found to be unique to Microthrix as evidenced from the increased transcriptional response of other lipid-accumulating and/or lypolytic populations, e.g., Acinetobacter , Leptospira , or Anaerolinea spp., especially in the aerobic ex situ conditions. Our work highlights the requirement to account for organism-specific adaptation strategies and time-frames within mixed communities. We observed that drastically altered community composition and gene expression patterns followed a severe disturbance in substrate levels within our time-series. We hypothesize that this community shift was a consequence of excess substrate availability, and it highlights a limit to the communityʼs resistance. Individual populations recovered within ten sludge age cycles post-disturbance, which indicates that the resilience of the community is also linked to phenotypic plasticity. The overlap in realized niches reflects niche complementarity. This in turn is governed by interspecific competition over a set of substrates, such as oleic acid. Other independent work on the human gut microbiome has highlighted the importance of interspecific competition for the maintenance of stability under a constant feeding regimen 49 . How interspecific competition or lack thereof relates to resilience represents a key question for future work. Overall, our framework demonstrates that multi-meta-omics data allows an in-depth characterization of ecological niches over time. Due to the observed plasticity in activity and the recovery after a major, transient perturbation, we confirm that the relationship between resistance and resilience is a function of fine-scale competition over resources in this environment. The resulting complementarity in both the fundamental and realized niches guarantees the provision of stable ecosystem services 50 and, thus, the long-term stable operation of mixed-culture biotechnological processes. These results are particularly relevant for the future engineering of niches within mixed-culture biotechnological processes 3 , which are key to achieve humankind’s sustainability goals 1 , 2 . In more general terms, it will be important to understand if phenotypic heterogeneity and niche complementarity play similarly important roles in the stability of other microbiomes."
} | 2,580 |
26384772 | PMC4632063 | pmc | 1,412 | {
"abstract": "Many nonmodel species exemplify important biological questions but lack the sequence resources required to study the genes and genomic regions underlying traits of interest. Reef-building corals are famously sensitive to rising seawater temperatures, motivating ongoing research into their stress responses and long-term prospects in a changing climate. A comprehensive understanding of these processes will require extending beyond the sequenced coral genome ( Acropora digitifera ) to encompass diverse coral species and related anthozoans. Toward that end, we have assembled and annotated reference transcriptomes to develop catalogs of gene sequences for three scleractinian corals ( Fungia scutaria , Montastraea cavernosa , Seriatopora hystrix ) and a temperate anemone ( Anthopleura elegantissima ). High-throughput sequencing of cDNA libraries produced ∼20–30 million reads per sample, and de novo assembly of these reads produced ∼75,000–110,000 transcripts from each sample with size distributions (mean ∼1.4 kb, N 50 ∼2 kb), comparable to the distribution of gene models from the coral genome (mean ∼1.7 kb, N 50 ∼2.2 kb). Each assembly includes matches for more than half the gene models from A. digitifera (54–67%) and many reasonably complete transcripts (∼5300–6700) spanning nearly the entire gene (ortholog hit ratios ≥0.75). The catalogs of gene sequences developed in this study made it possible to identify hundreds to thousands of orthologs across diverse scleractinian species and related taxa. We used these sequences for phylogenetic inference, recovering known relationships and demonstrating superior performance over phylogenetic trees constructed using single mitochondrial loci. The resources developed in this study provide gene sequences and genetic markers for several anthozoan species. To enhance the utility of these resources for the research community, we developed searchable databases enabling researchers to rapidly recover sequences for genes of interest. Our analysis of de novo assembly quality highlights metrics that we expect will be useful for evaluating the relative quality of other de novo transcriptome assemblies. The identification of orthologous sequences and phylogenetic reconstruction demonstrates the feasibility of these methods for clarifying the substantial uncertainties in the existing scleractinian phylogeny.",
"conclusion": "Conclusion The annotated transcriptome assemblies developed in this study provide useful resources for genomic research in anthozoan species for which sequences resources were previously lacking. The searchable databases developed from these assemblies make it possible to rapidly identify genes of interest from each species. Our ortholog analysis demonstrates the feasibility of phylogenetic inference in corals using transcriptome assemblies from diverse stages and symbiotic states, highlighting a promising path toward resolving major uncertainties in the existing phylogeny of scleractinians. Future studies will benefit from the growing body of anthozoan sequence resources, including the four assemblies contributed in this study.",
"discussion": "Results and Discussion Sequencing and de novo assembly The four libraries described here were sequenced on Illumina HiSeq 2000 (each occupying 1/6th of a lane), yielding on average 26.3 million paired reads per library (range, 21.2–30.3; Table S5 ). A fraction of these (22% on average; range 14–28%) were removed during quality and adaptor filtering prior to assembly. Assembly of the remaining high-quality reads produced on average ∼170,000 transcripts. This is substantially higher than the number of genes in sequenced cnidarian genomes (23,677 in A. digitifera , 27,273 in N. vectensis ), which likely results from redundancy, fragmentation in the assemblies, and biological contamination. Assemblies included many small contigs (on average, 47% were <400 bp) that were unlikely to provide significant matches, so for analyses based on sequence homology we considered only contigs ≥400 bp (average n = 91,792). For these core transcriptome datasets used for downstream analyses, the average length ranged from 1.1 to 1.7 kb and N 50 ranged from 1.4 to 2.7 kb. These are slightly shorter than the expected size distribution for a complete cnidarian transcriptome ( e.g. , average ∼1700 and N 50 ∼2200 bp transcripts in the A. digitifera genome), suggesting incomplete assemblies. Assembly statistics of the four transcriptome references developed in this study are broadly comparable to previously published anthozoan transcriptomes ( Moya et al. 2012 ; Shinzato et al. 2011 , 2014 ; Abascal et al. 2005 ; Traylor-Knowles et al. 2011 ; Lehnert et al. 2012 ). Completeness of transcriptomes To evaluate the completeness of the transcriptome assemblies from the perspective of gene content, we conducted sequence comparisons with conserved eukaryotic genes and gene models from sequenced relatives. The core eukaryotic genes (CEGMA) ( Parra et al. 2007 ) are expected to be expressed in most eukaryotes ( Nakasugi et al. 2013 ; Sanders et al. 2014 ) and are widely used to estimate transcriptome completeness. Sequence comparisons revealed matches for 453 (98.9%) of these conserved genes in A. elegantissma and 456 (99.5%) in F. scutaria , M. cavernosa , and S. hystrix ( Figure 2A ). For a more comprehensive view of gene representation, the transcriptomes were compared with gene models from sequenced relatives (the coral A. digitifera and the anemone N. vectensis ). This analysis identified matches for more than 14,000 gene models in each genome (BLASTx, bit-score ≥50): 54–67% of gene models in A. digitifera ( Figure 2B ) and 48–49% in N. vectensis . This is comparable to the level of sequence similarity observed among anthozoans with completed genomes. BLASTp comparisons of predicted proteins from the A. digitifera and N. vectensis genomes using the same thresholds recover 35% and 42% of genes in the other genome. This is substantially lower than the optimistic estimates of representation based on CEGMA, perhaps reflecting essential functions and constitutive expression of these highly conserved genes. Comparisons with gene models of closely related taxa appear to provide a more conservative estimate of gene representation in transcriptome assemblies. Figure 2 Three metrics used to evaluate gene representation and assembly of complete transcripts in de novo transcriptome assemblies. (A) Percent of core eukaryotic genes (CEGMA) identified in each assembly. (B) Percent of A. digitifera gene models with significant matches in each assembly. (C) Median proportion of each N. vectensis protein aligned with transcripts in each assembly (OHR HITS ). Gray = our transcriptome assembly compared to the respective reference for each analysis. To evaluate the effectiveness of our assemblies in reconstructing complete transcripts, we calculated ortholog hit ratios (OHR) for each final assembly. This method estimates the amount of a de novo transcript contained in the best ortholog from a reference genome ( O’Neil et al. 2010 ), ranging from 1 (for complete transcripts) to 0 (for transcript fragments). We calculated OHR based on sequence comparisons with N. vectensis gene models using two approaches. First, a relatively stringent analysis based on the proportion of each N. vectensis gene included in regions of local similarity (OHR HITS ) produced median OHR of 63.8, 64.7, 65.7, and 58.0% for A. elegantissma , F. scutaria , M. cavernosa , and S. hystrix , respectively ( Figure 2C ). A more inclusive analysis based on the longest ORF (in BLAST defined frame) produced similar estimates (median OHR ORF : 67.4, 75.8, 77.2, and 60.3%, respectively). Each assembly included more than 5000 reasonably complete transcripts spanning at least 75% of the corresponding N. vectensis gene (range, 5262–6725). Overall, these comparisons with existing cnidarian sequence resources quantify the representation and completeness of our assemblies and provide a framework for comparison with other de novo assemblies. These estimates compare favorably with previous transcriptome completeness estimates for cnidarians ( Sanders et al. 2014 ) and several invertebrates ( O’Neil and Emrich 2013 ; Riesgo et al. 2012 ) using similar methods. Annotation of transcriptomes Transcripts were annotated using BLAST homology searches against the UniProt databases. Approximately one-third of all transcripts matched records in UniProt (range, 30–40%) ( Table S5 ). The relatively low fraction of sequences annotated is attributable in part to sequence lengths: on average, 21% of transcripts <400 bp in length were annotated as compared with 42% of transcripts 400–1000 bp in length and 78% of transcripts >1000 bp. Even among the longest transcripts (>1 kb), a substantial number of sequences lacked annotated matches in UniProt (range, 6647–12,090 sequences per assembly). This highlights the well-known bias in taxonomic composition of existing databases and the value of ongoing gene sequencing in underrepresented metazoan taxa for public sequence databases. To categorize the biological functions inferred from sequence similarity, Gene Ontology (GO) terms were assigned to transcripts matching GO-annotated records in the UniProt database. This process identified functional annotation for 77% of transcripts with BLAST matches, providing tentative gene identities for a large number of sequences in each assembly (range, 32,299–47,547 transcripts; Table S5 ). Figure 3 shows the distribution of functional categories across the four transcriptomes, visualized using the Web Gene Ontology Annotation Plotting (WEGO) application. The GO terms were broadly distributed across the three domains and the percentages of sequences mapped to a given sub-ontology were highly similar for all species and comparable to other invertebrate transcriptomes ( Riesgo et al. 2012 ; O’Neil et al. 2010 ; Lehnert et al. 2012 ; Moya et al. 2012 ; Polato et al. 2011 ; Shinzato et al. 2014 ; Stefanik et al. 2014 ; Traylor-Knowles et al. 2011 ). The similarities in functional distributions of assemblies prepared from diverse species, developmental stages, and symbiotic states highlight the constitutive expression of a broad set of genes in cnidarian transcriptomes. These core genes should facilitate comparative transcriptome studies by increasing the overlap among incomplete libraries. Figure 3 Distribution of functional categories (GO terms) in each transcriptome assembly. The percentage of transcripts with GO annotation for each category under the three main ontology domains was calculated for each assembly. To determine taxonomic origin for each transcript, we conducted a series of BLAST searches and filtering steps outlined in Figure 1 . Because our assemblies were produced from symbiotic and aposymbiotic specimens, the transcriptomes contain genes not only from anthozoans but also from their associated microbial community. To investigate the relative contributions of these sources we classified each transcript based on sequence similarity ( Figure 1 ). These analyses confirmed that metazoan sequences comprised the majority of each library as expected. Fortunately, only a small fraction of transcripts were derived from organelles (mitochondria and ribosomes): on average, 212 transcripts (range, 127–284) in each assembly matched rRNA ( N. vectensis ) and 30 transcripts (range, 16–54) matched the mitochondrial genome ( A. tenuis ). Notably, almost half of the transcripts in each assembly (range, 46.2–49.9%) lacked matches to coral or Symbiodinium spp. genes, or NCBI’s nr database ( Figure 4 ), a range that is consistent with results from other anthozoan transcriptomes ( Sun et al. 2013 ; Karako-Lampert et al. 2014 ; Polato et al. 2011 ; Traylor-Knowles et al. 2011 ). These \"unknown\" transcripts may represent lineage-specific genes (\"taxonomically restricted genes\") that require further characterization. Comparison with NCBI’s nr database revealed that the majority of sequences with matches in one or more databases (59–95%) matched a metazoan sequence better than any other taxon, suggesting they originated from the animal host rather than from dinoflagellate or prokaryotic symbionts. A negligible fraction of transcripts in each assembly (0.8–1.7%) was assigned to the “Other taxa” category, most of which matched either coral or Symbiodinium genes but were classified as “unknown” because of conflicting results in the nr search ( e.g. , transcripts that matched Symbiodinium more closely than coral but whose best matches in nr were from metazoans). Figure 4 Predicted taxonomic origin of transcriptomes based on homology searches with BLAST. The percent of transcripts that were assigned to rRNA (purple), mtDNA (blue), dinoflagellate (green), metazoan (pink), other taxa (orange), and no match (gray) are shown. The contribution of algal symbionts varied widely across samples. In nominally aposymbiotic samples of F. scutaria and A. elegantissma , 2.6% of transcripts on average were classified as dinoflagellate in origin ( Figure 4 ), which may have resulted either from unexpected presence of symbionts at low abundance in these samples or from genes lacking orthologs in the A. digitifera reference. The symbiotic samples from S. hystrix , in contrast, showed a comparable abundance of transcripts classified as metazoan (61,369) and dinoflagellate in origin (41,724). Surprisingly, the M. cavernosa library that was similarly prepared from a symbiotic sample showed only 7278 transcripts from dinoflagellates ( Figure 4 ). This striking contrast in Symbiodinium contributions from symbiotic specimens may have arisen from differing methods of RNA extraction. For S. hystrix , tissue was airbrushed off the coral skeleton directly into RNA later Stabilization Solution (Qiagen, CA), followed by complete tissue homogenization. In contrast, the M. cavernosa fragment was simply vortexed to disrupt tissue, without physical homogenization. Our findings suggest that omitting physical homogenization during lysis can minimize symbiont contamination for studies aiming to focus on the cnidarian host, whereas studies investigating both components may benefit from thorough homogenization during extraction. The gene names, functional categories, and putative origin of each transcript are annotated in Table S6 , Table S7 , Table S8 , and Table S9 . Gene searches of the database The resulting annotations and sequences are available in a set of searchable databases hosted by Oregon State University ( http://people.oregonstate.edu/∼meyere/index.html ). To illustrate the utility of databases for cnidarian researchers targeting specific genes, we compared the effectiveness of simple text searches of the databases with reciprocal BLAST (RB) analysis, a more comprehensive approach that requires additional work by the end-user. Text searches targeting a few selected genes [cell adhesion molecule sym32, green fluorescent protein (GFP), and cystathionine β-synthase (Cbs)] produced comparable results as RB searches ( Table S10 ). Text searches are obviously sensitive to query phrasing; the query “fluorescent” retrieves 51 putative GFP homologs, and functionally related synonyms (“GFP”, “chromoprotein”) retrieved an additional 10. Interestingly, the Cbs homologs identified in nominally aposymbiotic samples ( A. elegantissima and F. scutaria ) showed greater sequence similarity with Symbiodinium gene models than coral ( A. digitifera ) and were classified as dinoflagellate in our assignment procedure ( Figure 1 ), whereas Cbs homologs in symbiotic samples ( M. cavernosa and S. hystrix ) included both metazoan and dinoflagellate transcripts. This unexpected observation of apparently dinoflagellate homologs of Cbs in nominally aposymbiotic samples is noteworthy because of their variable distribution among corals and possible roles in coral nutritional dependency on symbiosis ( Shinzato et al. 2011 ). While this finding could be explained by undetected Symbiodinium harbored in these putatively aposymbiotic samples, the uncertainty introduced by these observations suggests that studies investigating the diversity of Cbs homologs across corals may require additional data ( e.g. , in situ hybridization) to confirm transcript origins. Overall, the close agreement between rigorous computational searches and simple text searches in these examples illustrates the utility of our searchable online databases for rapidly identifying genes of interest in reference transcriptome assemblies. Novel SSR markers Simple sequence repeats (SSRs or microsatellites) have been widely used to study genetic diversity, hybridization events, population structure, and connectivity in anthozoans ( Concepcion et al. 2010 ; Fernandez-Silva et al. 2013 ; Selkoe and Toonen 2006 ; Ruiz-Ramos and Baums 2014 ), and can directly influence phenotypic traits by altering DNA replication, translation, and gene expression ( Ruiz-Ramos and Baums 2014 ). SSR markers can be readily identified from de novo assemblies of NGS data and emerge as a side benefit in transcriptome assembly projects conducted for other purposes. We identified and developed primers for 52, 49, 73, and 75 candidate SSR markers in A. elegantissma , F. scutaria , M. cavernosa , and S. hystrix , respectively. Primer pairs for each species are listed in Table S11 . For three of the species studied here, varying numbers of SSR markers are already available. Previous studies of S. hystrix have developed 10 SSR markers ( Maier et al. 2001 ; Underwood et al. 2006 ) to study habitat partitioning within a single reef ( Bongaerts et al. 2010 ), dispersal and recruitment patterns across multiple reefs ( van Oppen et al. 2008 ; Kininmonth et al. 2010 ), and population changes associated with bleaching events ( Underwood et al. 2007 ). Candidate SSR markers have been identified in F. scutaria (n = 118) from the coral host and dinoflagellate symbionts ( Concepcion et al. 2010 ). SSR markers previously developed in M. cavernosa ( Shearer et al. 2005 ; Serrano et al. 2014 ) have been used to investigate the population connectivity across depth and geographic distance ( Serrano et al. 2014 ). The candidate SSR markers identified in this study provide additional markers for future studies along similar lines. To our knowledge, SSR markers have not been previously developed in A. elegantissima . Although the population structure of the host has not been described, analysis of their dinoflagellate symbionts revealed highly structured populations across their geographic range ( Sanders and Palumbi 2011 ). The markers developed in this study for A. elegantissima provide tools to investigate population structure of the host across a similar range. Orthologous groups and phylogenomic reconstructions With the increasing availability of transcriptomes and genomes, these datasets can now be mined to discover novel phylogenetic markers within Anthozoa and across the Cnidaria to resolve taxonomic uncertainties. Phylogenetic reconstruction of anthozoans has presented challenges because analyses based on morphology, life history, and molecular sequences have failed to adequately delineate taxonomic boundaries or evolutionary relationships ( Daly et al. 2003 ). To date, molecular phylogenies for anthozoans have been based on one or a small number of markers including nuclear ribosomal 28S and 18S genes ( Daly et al. 2003 ; Berntson et al. 1999 ), β -tubulin ( Fukami et al. 2008 ), mitochondrial 16S ( Daly et al. 2003 ), cytochrome b ( Fukami et al. 2008 ), and COI ( Kitahara et al. 2010 ; Fukami et al. 2008 ). Interestingly, mitochondrial sequences in anthozoans have extremely low mutation rates compared to the bilaterians and are therefore highly conserved, allowing for robust comparisons across distantly related taxa ( van Oppen et al. 2002 ; Galtier et al. 2009 ). Therefore, the mitochondrial gene COI has been used recently to define evolutionary relationships among scleractinian corals ( Kitahara et al. 2010 ; Fukami et al. 2008 ; Budd and Stolarski 2011 ) and to support the distinction of robust corals from the complex corals ( Romano and Palumbi 1996 ). One disadvantage to single gene phylogenetic inferences is that they suffer from weak phylogenetic signals, sensitivity to hidden paralogy, and spurious tree artifacts ( Philippe et al. 2004 ). Despite these potential limitations, single gene trees have advanced the field of cnidarian systematics. However, polyphyly remains a problem among several anthozoan families when using both maximum likelihood and Bayesian analyses ( Fukami et al. 2008 ; Budd and Stolarski 2011 ), which has led to recent shifts in taxonomic classification ( Budd and Stolarski 2011 ). To expand beyond previous single-gene approaches, we performed phylogenomic analyses incorporating the four new transcriptomes and other available \"omic\" resources. By simultaneously increasing taxon and gene sampling, phylogenetic inference is expected to improve ( Philippe et al. 2004 ) and may help resolve some of the challenges in reconstructing the evolutionary relationships of the Anthozoa and, more broadly, phylum Cnidaria. For phylogenomic analysis, transcripts larger than 400 bp were converted to protein with TransDecoder and clustered into orthologous groups using FastOrtho. The number of assigned orthologous groups ranged from 14,144 to 21,147 for the four transcriptomes ( Figure S1 ). Comparison of all four resulted in 6560 shared orthologs ( Figure S1 ). The three coral species shared 2045 orthologs not found in anemones and the two most closely related corals ( M. cavernosa and F. scutaria ) shared 1682 orthologs absent from the other assemblies. By incorporating 11 additional taxa for phylogenomic analysis ( Table S2 ), 443 orthologs were identified between all taxa. After setting a minimum protein length (100 amino acids), these orthologs were refined using the PhyloTreePruner analysis pipeline ( Kocot et al. 2013 ). Filtering resulted in the identification of 397 orthologs for ≥14 taxa. These were used to construct a phylogenetic tree we termed “conservative” because loci with any missing data were excluded ( Table S12 ). Missing data are a commonly encountered problem in phylogenomic analyses, from either reduced transcript length or gene absence from a transcriptome ( Philippe et al. 2004 ; Kocot et al. 2013 ; Roure et al. 2013 ). However, the sensitivity of phylogenetic inference to incomplete datasets is still under investigation, with mixed results from phylogenomic analyses on large but patchy supermatrices ( Roure et al. 2013 ; Philippe et al. 2004 ). Because the resources in this study used for ortholog identification differed in completeness, ranging from EST libraries to complete genomes ( Table S2 ), we tested the influence of missing data on our phylogenetic reconstruction. To investigate this, we lowered the required number of taxa per orthologous group to ≥10, which identified 2897 orthologs ( Table S12 ). This second set was used to create the “relaxed” phylogeny, so called because loci with some missing data were included. Both maximum likelihood phylogenomic analyses reconstructed identical and strongly supported topologies (bootstrap = 100; Figure S2 ), demonstrating that our phylogenetic inference was insensitive to missing data ( Figure 5 ). However, the relationship of the corals in the family Faviidae, containing M. cavernosa , P. strigosa , and O. faveolata , varied among the COI, ND supergene, and phylogenomic analyses. The mitochondrial ND supergene identified by Havird and Santos (2014) produced a phylogenetic tree nearly synonymous with the accepted cnidarian taxonomic relationships and phylogenomic analyses from this study ( Kitahara et al. 2010 ), except for the placement of the M. cavernosa as sister taxon to O. faveolata and P. strigosa . The analysis of single gene COI resulted in a discordant phylogenetic topology ( Figure 5 ), failing to reconstruct the complex coral clade ( P. astreoides and A. digitifera ), which was recovered by ND supergene, relaxed and conserved trees ( Figure 5 , Figure S2 ). In the COI tree, the placement of the F. scutaria , from the family Fungiidae, as sister taxon to P. strigosa and M. cavernosa from the family Faviidae, instead of O. faveolata is incongruent with current taxonomic placement ( Figure 5 ) ( Kitahara et al. 2010 ; Budd and Stolarski 2011 ). Furthermore, while the phylogenomic analyses placed O. faveolata as sister to P. strigosa with strong support (bootstrap = 100), this relationship was not recovered in either mitochondrial phylogeny ( Figure S2 ). Overall, the tree topology from the phylogenomic analyses is consistent with accepted evolutionary relationships within Anthozoa ( Budd and Stolarski 2011 ; Fukami et al. 2008 ; Kitahara et al. 2010 ). Figure 5 Discordance in maximum likelihood phylogenetic reconstruction of COI compared to a combined phylogeny of concatenated ND (2, 4, and 5) genes and two phylogenomic trees. The COI phylogeny is presented on the left and the combined phylogeny is presented on the right. Topology for the ND mitochondrial set, relaxed and conservative phylogenomic trees were nearly identical. Therefore, nodal support is summarized on the relaxed tree (right). Bootstrap support at the nodes from left to right represents ND gene set/relaxed/conservative. If topologies differed in the summary tree, then the nodal support is presented - - as next to the node. Yellow solid lines connect taxon with different positions and/or relationships between the two trees, whereas black dashed lines connect those with the same position and/or relationship. Reconstructions of groups in the class Anthozoa based on Kitahara et al. (2010) are highlighted in boxes: teal= robust corals; dark pink = complex corals; and light blue = anemones. The names of species used in this study are emphasized by bold font. Scale bars indicate the amino acid replacements per site. Conclusion The annotated transcriptome assemblies developed in this study provide useful resources for genomic research in anthozoan species for which sequences resources were previously lacking. The searchable databases developed from these assemblies make it possible to rapidly identify genes of interest from each species. Our ortholog analysis demonstrates the feasibility of phylogenetic inference in corals using transcriptome assemblies from diverse stages and symbiotic states, highlighting a promising path toward resolving major uncertainties in the existing phylogeny of scleractinians. Future studies will benefit from the growing body of anthozoan sequence resources, including the four assemblies contributed in this study."
} | 6,770 |
22126435 | PMC3339375 | pmc | 1,414 | {
"abstract": "Background Motivated by the precarious state of the world's coral reefs, there is currently a keen interest in coral transcriptomics. By identifying changes in coral gene expression that are triggered by particular environmental stressors, we can begin to characterize coral stress responses at the molecular level, which should lead to the development of more powerful diagnostic tools for evaluating the health of corals in the field. Furthermore, the identification of genetic variants that are more or less resilient in the face of particular stressors will help us to develop more reliable prognoses for particular coral populations. Toward this end, we performed deep mRNA sequencing of the cauliflower coral, Pocillopora damicornis , a geographically widespread Indo-Pacific species that exhibits a great diversity of colony forms and is able to thrive in habitats subject to a wide range of human impacts. Importantly, P. damicornis is particularly amenable to laboratory culture. We collected specimens from three geographically isolated Hawaiian populations subjected to qualitatively different levels of human impact. We isolated RNA from colony fragments (\"nubbins\") exposed to four environmental stressors (heat, desiccation, peroxide, and hypo-saline conditions) or control conditions. The RNA was pooled and sequenced using the 454 platform. Description Both the raw reads (n = 1, 116, 551) and the assembled contigs (n = 70, 786; mean length = 836 nucleotides) were deposited in a new publicly available relational database called PocilloporaBase http://www.PocilloporaBase.org . Using BLASTX, 47.2% of the contigs were found to match a sequence in the NCBI database at an E-value threshold of ≤.001; 93.6% of those contigs with matches in the NCBI database appear to be of metazoan origin and 2.3% bacterial origin, while most of the remaining 4.1% match to other eukaryotes, including algae and amoebae. Conclusions P. damicornis now joins the handful of coral species for which extensive transcriptomic data are publicly available. Through PocilloporaBase http://www.PocilloporaBase.org , one can obtain assembled contigs and raw reads and query the data according to a wide assortment of attributes including taxonomic origin, PFAM motif, KEGG pathway, and GO annotation.",
"conclusion": "Conclusions We used the 454 sequencing platform to generate a reference transcriptome for the cauliflower coral, P. damicornis . A taxonomic analysis of the sequence data indicates that we have captured some of the diversity of the coral holobiont, as many of the sequences appear to be derived from non-metazoan taxa including bacteria, fungi, viruses, and unicellular algae of the genus Symbiodinium . The data have been organized into a publicly available relational database that will be updated and expanded as new P. damicornis sequencing data become available.",
"discussion": "Utility and Discussion Sequencing yield Sequencing yielded 1, 116, 551 raw reads with an average length of 379 nucleotides (range: 29-2, 025 nt; SD = 152 nt). Reads less than 40 nucleotides in length and low quality reads that did not overlap with other reads were discarded. The remaining 955, 105 reads were assembled into 70, 786 contigs with an average length of 836 nt (range: 40-10, 512 nt; SD = 464 nt; Additional File 2 ; 3 ). These data are compared with other published scleractinian transcriptomic data sets in Table 1 . Taxonomic affinity of the sequences We used the top hit in a BLASTX [PMID: 2231712] search to characterize each of the assembled contigs according to its apparent taxonomic affinity. Overall, 47.2% (or 33, 423) of the contigs matched sequences housed at NCBI with an E-value cutoff of 0.001. The other 37, 363 contigs did not match sequences at NCBI with an E-value of ≤.001 and were excluded from subsequent analyses. Of the 33, 423 hits, 31, 271 appeared metazoan, 139 fungal, 36 viral, 764 eubacterial, and 26 archaeal (Figure 4 ). We classified 1187 hits as \"other eukaryote;\" when these other eukaryotic hits were parsed further, 142 matched a sequence from Symbiodinium , the genus of unicellular algae that are intracellular endosymbionts of hermatypic corals (Table 2 ). For a complete breakdown of BLAST matches to other eukaryotes including dinoflagellates, see Additional Files 4 , 5 . Figure 4 Taxonomic affinities of contig sequences based on BLAST . Overall, 52.9% (or 37, 423) of the contigs matched sequences from NCBI with an E-value cutoff of 0.001. Of the 37, 423 hits, 31, 271 appeared metazoan, 139 fungal, 36 viral, 764 eubacterial, 26 archaeal, and 1187 hits as \"other\" eukaryote. Table 2 Summary of Symbiodinium blast hits Gene/Protein Name NCBI ID Actin 87116473 87116475 Bacl-2 119710160 Dna J-like protein 1 75858825 Dna J-like protein 2 75858827 Dna J-like protein 3 75858829 Glyceraldehyde-3-phosphate dehydrogenase 32454981 35210444 35210448 35210454 Heat shock protein 70 75858823 Heat shock protein 90 75858821 Peridinin chlorophyll-a binding protein apoprotein precursor 23986591 23986608 23986610 23986617 23986634 23986641 1709613 23986384 23986401 23986430 23986551 Phosphoglycolate phosphatase 197091190 Polyubiquitin 75858833 Ribulose bisphosphate carboxylase 75282236 Ubiquitin ligase 2 75858845 Ubiquitin-specific protease 1 75858847 Unknown 134035981 134035977 The taxonomy of Pocillopora is currently regarded as tenuous, e.g ., some 16 species have been defined on the basis of morphological features, but cladistic groupings defined by molecular sequence data are not always congruent with these morphologically defined taxa ( e.g ., [ 31 ]). We used reciprocal blast searches to investigate whether the sequences we generated in this study most closely resemble P. damicornis sequences in NCBI, rather than sequences from other closely related corals, including other Pocillopora species. We obtained 400 \" P. damicornis \" ESTs obtained from NCBI, and we used these to query the contigs housed at PocilloporaBase (blastn with an E-value cut-off of 0.001). Because a substantial fraction of the P. damicornis sequences currently housed at NCBI represent multiple copies of the same gene (generated in population genetics studies,) many of them matched to the same contig in PocilloporaBase. Overall, the 400 sequences from NCBI matched to 21 unique contigs at PocilloporaBase. We blasted these 21 contigs back against all nucleotide sequences at NCBI classified as scleractinian (Search \"Scleractinia[Organism]\"). The results are provided in Additional File 6 . In 6 of 21 instances, the only match in the database was to a P. damicornis sequence. In 12 instances, there were matches to other corals in addition to P. damicornis , but the highest degree of sequence identity was to a sequence from P. damicornis . In one instance, a contig from PocilloporaBase exhibited equal percent identity to sequences from P. damicornis and P. meandrina . Finally, there were two contigs that exhibited a slightly higher resemblance to a sequence from a coral other than P. damicornis . On balance, these results clearly suggest that the \" P. damicornis \" populations sampled in this study exhibit greater sequence similarity to the P. damicornis sequences housed at NCBI than to any other coral species represented in that database. However, the taxonomic uncertainty of greatest concern pertains specifically to the genus Pocillopora . Here, the results are less decisive. Only seven of the twenty-one contig sequences we blasted against the NCBI database produced hits to sequences from P. damicornis and another Pocillopora species. In five of these seven instances, the top hit was to a P. damicornis sequence, but one contig matches better to P. molokensis and another matches equally well to P. damicornis and P. meandrina . Gene Ontology One or more GO annotation terms could be associated with 23, 202 of the 70, 786 contigs (see Construction and content); 7, 084 contigs had a unique GO annotation. All of the GO terms that were attributed to at least 100 contigs are summarized in Figure 5 (for a complete listing, see Additional File 7 ). Figure 5 Gene ontology terms associated with P. damicornis sequences . The number of P. damicornis contigs associated with each GO term is shown for Cellular Compartment, Molecular Function, and Biological Function. Only GO terms associated with 100 or more P. damicornis sequences are shown. KEGG Pathway Analysis The contigs were subjected to the KEGG Pathway analysis, for human and plant separately [ 30 ]. Based on this analysis, the P. damicornis sequences were mapped to metabolic pathways on the interactive tree of life [ 32 ]. Components of most metabolic pathways were identified, including photosynthesis (KEGG id: map00195 and 00196) and lipid metabolism. However, some pathways were largely or completely absent from the P. damicornis contigs, including aminosugars metabolism (00530), lipopolysaccharide biosynthesis (00540), peptidoglycan biosynthesis (00550), glycosphingolipid biosynthesis (00602), methane metabolism (00680), androgen and estrogen metabolism (00150), and biodegradation of most xenobiotics (Additional Files 8 , 9 ). PocilloporaBase: capabilities and functions PocilloporaBase http://www.PocilloporaBase.org was modeled after StellaBase, a genomic and transcriptomic database for the starlet sea anemone, Nematostella vectensis [ 33 , 34 ]. Both species-specific databases were designed to integrate with CnidBase [ 35 ], a phylum-wide database meant to facilitate cross-species comparisons among cnidarians. All of the raw sequencing reads generated in this study as well as the assembled contigs can be downloaded there. At present, unlike StellaBase, PocilloporaBase houses only transcriptome data and no genomic sequence data. This repository will grow as future mRNA and genomic sequencing projects generate additional data for Pocillopora damicornis . The assembled contigs housed at PocilloporaBase can be searched using several different modalities (Figure 6 ). The Contig Search allows you to query sequences using NCBI Protein Accession Number ( e.g .: 74000907), Nucleotide Accession Number ( e.g .: 33340018), Gene/Protein Name (e.g., 'hemoglobin'), or Species Name/Taxon ID (e.g.: 'otolemur' or 45351). The option of searching the data by organism is critical, since corals can be considered a holobiont consisting of coral host, symbiotic algae ( Symbiodinium ), bacteria and fungi. Figure 6 Functionality of PocilloporaBase . PocilloporaBase houses the raw sequencing reads generated in this study as well as the assembled contigs that were generated from the sequencing reads and the output from a number of bioinformatic analyses performed on the contigs. The contigs were used to search the non-redundant database at NCBI using BLASTX. The contigs were also used to conduct an HMM search of the Pfam database to identify conserved protein motifs. The proteins producing significant matches to P. damicornis contigs in the BLAST search were cross-referenced with biochemical pathways at KEGG and with gene ontology terms at amiGO. Users can search the Pocillopora contigs based on features of the protein they matched in the BLAST search or conserved Pfam protein domains they appear to encode. Users can also search either the contigs or the raw reads using BLAST, and all of the sequencing reads and contigs can be downloaded from the site. Steps taken to populate the database are represented by blue arrows. Actions available to the user are represented by red arrows. The Gene Ontology Search allows you to query sequences using either GO id or GO description terms [ 36 ]. Each successful query returns a table that contains the Protein ID and Protein Name, as well as GO id, GO description, and GO type. The Protein ID links to the protein's entry at NCBI. The Protein Name links to a Protein Lookup on PocilloporaBase, which returns a list of P. damicornis contigs that generated significant BLAST hits to the protein in question. The GO id links to the corresponding gene ontology page at the amiGO database. Clicking on the GO description link performs a more specific GO lookup if any gene ontology terms are children of the parent term used to conduct the original search. If too many results are returned, searches can be restricted to one of the three principal gene ontology types: biological process, cellular component, or molecular function. If the user is more interested in the number rather than the identity of the genes in the database that map to a particular gene ontology term, the \"Counts only\" box can be checked. The KEGG Pathway Search allows you to query sequences using a KEGG ID number (e.g., hsa00010) or KEGG Pathway Description ( e.g ., Glycolysis/Gluconeogenesis). Each successful query returns a table with individual contigs identified by their Contig ID, the E-value of their match to a protein in the KEGG database, the KEGG Organism ID, homologous protein name and protein accession ID. The presence of conserved protein motifs in one or more transcripts can be investigated by searching the data for matches to the conserved protein motifs housed at Pfam. The Pfam Protein Family Classification search allows you to query sequences by Pfam Accession number ( e.g ., PF00006), Motif Name ( e.g .: PAX, actin, DNA_methylase) or Protein Description key words. If there is a match to a conserved protein motif at Pfam, the search returns a table of Pocillopora damicornis contigs encoding that motif sorted by E-value. It is also possible to search for matches to a query sequence using the complete set of BLAST options. BLAST searches return contig id, the sequence for that contig, as well as the NCBI gene ID, gene name, and the gene sequence for any gene sequence found to match the original blast query. A gene search page allows for quick retrieval of gene and species information in the database."
} | 3,486 |
27649169 | PMC5037840 | pmc | 1,415 | {
"abstract": "The fabrication of cellulose-spider silk bio-nanocomposites comprised of cellulose nanocrystals (CNCs) and recombinant spider silk protein fused to a cellulose binding domain (CBD) is described. Silk-CBD successfully binds cellulose, and unlike recombinant silk alone, silk-CBD self-assembles into microfibrils even in the absence of CNCs. Silk-CBD-CNC composite sponges and films show changes in internal structure and CNC alignment related to the addition of silk-CBD. The silk-CBD sponges exhibit improved thermal and structural characteristics in comparison to control recombinant spider silk sponges. The glass transition temperature ( Tg ) of the silk-CBD sponge was higher than the control silk sponge and similar to native dragline spider silk fibers. Gel filtration analysis, dynamic light scattering (DLS), small angle X-ray scattering (SAXS) and cryo-transmission electron microscopy (TEM) indicated that silk-CBD, but not the recombinant silk control, formed a nematic liquid crystalline phase similar to that observed in native spider silk during the silk spinning process. Silk-CBD microfibrils spontaneously formed in solution upon ultrasonication. We suggest a model for silk-CBD assembly that implicates CBD in the central role of driving the dimerization of spider silk monomers, a process essential to the molecular assembly of spider-silk nanofibers and silk-CNC composites.",
"conclusion": "4. Conclusions A unique approach to harness the structural alignment and cross-bridging of spider silk proteins to CNCs for the production of silk-CNC composite materials was used; this new approach involves the concentration and sonication of silk-CBD, which seems to influence the alignment and morphology of the composite. We propose that the higher molecular order observed in these composites is due to (1) CBD dimerization (and higher order assemblies); and (2) the binding of CBD to CNCs. These new composites may be useful in a variety of applications, including medical ones. In the model depicted in Figure 12 , CBD induces spider silk molecular order at two levels. First, the ability of CBD to form dimers and to mimic the nonrepetitive spider silk terminal function encourages molecular ordering towards the formation of aligned nano-silk fibers. Second, at a higher level, CBD specifically mediates silk-CNC interactions, guiding the formation of structural and ordered composite materials. Further work is currently underway toward exploring the mechanical properties of these composites, and whether the differences seen here, e.g., morphology and thermal properties, translate into improved/altered mechanical properties.",
"introduction": "1. Introduction Silks are produced by a variety of insects and spiders and some spiders spin as many as seven different kinds of silks, each tailored to fulfill a certain biological function [ 1 ]. Dragline silk, used as the safety line and as the frame thread of the spider’s web, is composed of two proteins, each with a long repetitive sequence flanked by nonrepetitive amino and carboxy termini [ 2 , 3 ]. The repetitive sequence is characterized by stretches of poly-alanine domains that are interrupted by glycine-rich repeats. The poly-alanine domains form β-sheet crystals, whereas the glycine-rich repeats form less crystalline segments [ 4 , 5 , 6 ]. The interplay between the hard crystalline domains and the less crystalline regions gives rise to the extraordinary properties of the silk [ 7 , 8 ]. Dragline silk has unique toughness and higher tensile energy to break than any other common natural or artificial material [ 9 ]. Its strength to weight ratio is five times stronger than steel and three times tougher than top quality man-made Kevlar fiber. Due to its superb mechanical properties, dragline silk has been the focus of intense research efforts [ 1 , 10 ]. Despite the extensive knowledge that has been gained regarding the structure and properties of dragline silk, its production remains challenging. In contrast to silkworm-silk, the isolation of large quantities of silk from spiders is not feasible. Spiders produce silk in small quantities, and their territorial behavior prevents large numbers from being raised in a confined space [ 11 ]. Therefore, production of spider silk proteins through recombinant DNA techniques has been the primary path pursued by researchers. Silk-encoding genes have been cloned and expressed in a variety of heterologous hosts [ 12 , 13 , 14 , 15 ], which have allowed for production of laboratory scale quantities of silk-like protein powders. Yet, with few exceptions [ 16 , 17 ], the material properties of these cloned silks are inferior to their native counterparts. One limitation has been the lack of molecular order in the recombinant silk proteins and their tendency to aggregate in vitro, bypassing the native refolding and assembly processes [ 10 ]. Assembly of silk proteins into a solid silk fiber is extremely complex, and replication of the spider’s native spinning process is a major challenge in laboratory settings [ 18 , 19 ]. The current work addresses this challenge using a different strategy—the production of spider silk CNC bio-composites. The technique relies on the specific binding capacity of cellulose binding domains (CBDs) to CNCs and the structural properties of CNCs. Cellulose is a product of biosynthesis from plants, animals, or bacteria, while “cellulose nanocrystals” refer to cellulosic extracts or processed materials, having defined nano-scale structural dimensions [ 20 ]. CNCs are exciting biomaterials that are relevant to a number of potential applications, including polymer nanocomposites, transparent films and hydrogels. CNCs are mainly produced by acid hydrolysis/heat controlled techniques, with sulfuric acid being the most utilized acid. The crystal extraction from cellulose involves hydrolysis of amorphous cellulose regions, resulting in highly crystalline particles with dimensions of 5–20 nm in width and 100–500 nm in length for plant source CNCs [ 21 ]. The mechanical properties of CNCs is impressive with the Young’s modulus and tensile strength of a single crystal reported to be as high as 150 and 10 GPa, respectively, which make them useful for the reinforcement of polymers. In addition, the rod-like shape of the particles leads to concentration-dependent liquid crystalline self-assembly behavior [ 22 ]. Noishiki et al. [ 23 ] found that CNC-native silkworm silk films had breaking strengths and ductility about five times greater than those of the constituent materials. The authors attributed these improvements to the flat and ordered surfaces of CNCs, which served as a template for the assembly of silk β-sheets, a process that usually requires shear and elongation stress. In nature, cellulose is degraded by the concerted actions of a number of bacterial and fungal organisms, initiated by cellulolytic enzyme(s) or microorganisms that can bind cellulose substrates. The CBD, a separate, nonhydrolytic component, mediates this binding. CBDs have been cloned from different organisms such as Clostridium cellulovorans and Cellulomonas fimi [ 24 , 25 ], and enable adhesion of the water-soluble enzyme to the soluble or insoluble substrate, by bringing the catalytic module into prolonged and intimate contact with the cellulose surface [ 26 , 27 ]. In the present study, we compare synthetic 15-monomer-long dragline spider silk derived from the native sequence of MaSp1 of Nephila clavipes , referred to as “silk”, and the synthetic 15-monomer with CBD fused to its 3′ termini, referred to as “silk-CBD”. Unlike silk, upon sonication, silk-CBD dimerizes and these dimers assemble in situ into microfibrils; this mechanism appears analogous to the formation of silk fibrils in nature, which proceeds via the C-termini. Furthermore, the effects of silk-CBD dimerization upon the mechanisms of CNC self-assembly of the CNCs in sponges and films are explored. The general motivation for composites made from silk and CNCs is to produce materials with characteristics reflective of both components; for instance, silk-cellulose composites may provide a strength and toughness profile that surpasses either component. In this work, we observed that silk-CBD specifically binds CNCs and confers molecular order which is different from that of either the silk proteins or CNCs. Silk-CBD-CNC composite materials may be useful in a variety of medical and industrial applications, and this preliminary research is a necessary step toward the end goal of harnessing the attractive properties of the components into a composite with superior properties.",
"discussion": "2. Results and Discussion 2.1. Protein Expression and Purification The synthetic spider silk and the fusion spider silk-CBD genes were successfully expressed, and the resultant proteins were purified from E. coli using Ni-NTA chromatography. The pure protein yields for the expression of silk (47 kDa) and silk-CBD (65 kDa) were 60 and 40 mg/L, respectively ( Figure 1 ). 2.2. Quantitative Cellulose Binding Assay In order to characterize the binding capacity of silk-CBD to cellulose, adsorption/desorption experiments were conducted. Adsorption/desorption experiments are commonly done to test the apparent irreversible adsorption of CBD to cellulose. A reversible adsorption process is defined when the variables characterizing the state of the system return to the same values in the reverse order during the desorption stage. Therefore, in a reversible adsorption process, the ascending branch (increasing protein concentration in the solution) and the descending branch (decreasing protein concentration in the solution) of the isotherm overlap. Reversible adsorption was seen for the purified silk protein in solution with cellulose ( Figure 2 ) due to the mechanism of protein adsorption at solid/liquid interfaces [ 28 ]. In contrast, irreversible binding was observed for both CBD and silk-CBD as evident from their non-overlapping adsorption isotherms. 2.3. Composite CNC/Spider Silk Sponges Spider silk/CNC composite sponge formation was done as previously described [ 29 , 30 , 31 , 32 ]. Purified, concentrated spider silk protein was mixed with a CNC suspension and then sonicated. This procedure has a two-fold effect; in addition to homogeneous dispersion of the CNCs, the sonication process induces annealing of spider silk proteins by accelerating formation of physical cross-links, such as initial chain interactions related to β-sheet formation [ 33 , 34 ]. After sonication, three-dimensional porous structures (i.e., sponges) were generated via freeze drying. SEM pictures of the resulting sponges showed that pore architecture and alignment differed between the silk-CBD and control silk sponges ( Figure 3 ). Silk sponges had 30–100 μm pores ( Figure 3 C) of irregular shape and with no particular orientation, very similar to the CNC-silk composites ( Figure 3 D). Silk-CBD sponges featured 300–500 μm leaf-shaped pores aligned in a relatively consistent direction ( Figure 3 E). Similar characteristics were observed in sponges from native silkworm silk produced using the same conditions applied here, which were attributed to the parallel arrangements of silk fibroin crystal flakes [ 30 ]. The composite silk-CBD-CNC sponges possessed ~100 μm structurally aligned pores ( Figure 3 F–H). The glass transition and degradation temperatures of different silk and silk-CBD sponges, as determined by TMDSC analysis, are shown in Figure 4 and Table 1 . DSC analysis of the 100% silk and silk-CBD sponges gave Tg values of 140 and 172 °C, respectively, and degradation temperatures of 279 and 283 °C, respectively. Interestingly, the Tg and degradation temperatures of the 100% silk-CBD sponge were similar to those reported for natural silkworm silk and Nephilia clavipes dragline silk fibers [ 35 , 36 , 37 ]. The changes seen in the Tg , characteristic of the amorphous domains in amorphous or semi-crystalline materials, such as silks, result from increased chain interactions. The stronger the interactions between the chains, the higher the temperature required to induce a phase transition. The lower Tg of the 100% silk sponge may be due to a more disordered structure, as seen in the SEM figures. The elevation in the Tg of the 25% silk/75% CNC sponge is likely related to the significant presence of CNCs, whose crystal surfaces may serve as a template/nucleation site for the assembly of silk β-sheets, as seen in the silkworm silk-CNC composite films by Noishiki et al. [ 23 ]. As for the elevated Tg of the silk-CBD sponges, it has been well established that CBDs form types of dimers in solution [ 38 , 39 ], and this dimerization factor likely also plays a role. 2.4. Composite CNC/Spider Silk Films Films of silk-CBD and CNCs were prepared in order to further investigate the effects silk-CBD on the materials and the role of dimerization. Similar to the sponge results presented in Figure 3 , SEM cross-sectional images of CNC and silk-CBD-CNC films at mass ratios of 1:5 and 1:10 ( Figure 5 ) show differences in film morphology related to the presence and amount of silk-CBD. The CNC film shows a typical layered morphology, whereas the composite silk-CBD-CNC films appear to be more aligned and dense. Film alignment was explored using a polarized optical microscopy (POM) system equipped with an image processing module. The top images in Figure 6 shows POM images of CNC and silk-CBD-CNC composite films, where the bright and dark regions typically indicate ordered and disordered areas in the films, respectively. The bottom images in Figure 5 show the processed birefringence images. The CNC films show the multi-domain order typical to CNC films, whereas the silk-CBD films show uniform alignment. The Abrio 2.2 software (CRi, Woburn, MA, USA) provides a vector overlay tool to analyze the processed image, where the vector azimuth is measured, and the standard deviation (SD) of the vector direction gives an approximation of the degree of alignment uniformity. An average of twenty measurements per sample (same size area) and corresponding SD values were calculated; an SD of 44.25 was obtained for CNC films, 13.58 for 1:10 silk-CBD-CNC films and 1.33 for 1:5 silk-CBD-CNC films. These SD values indicate improved sample alignment with increasing silk-CBD content. Furthermore, we qualitatively observed that the films appeared more transparent when silk-CBD was present ( Figure 7 ). This may be due to the particular approach to film formation used in this work (i.e., CNC formulations cast onto hydrophobic surfaces) or may be related to the dispersion/self-assembly of the particles in the films. 2.5. Spider Silk-CBD Fusion Protein Assembly To summarize, interesting behavior is observed in the composite materials when silk-CBD is used, including an elevation in the Tg , differences in internal structure and morphology, and differences in alignment in the films. As mentioned above, CBD dimerization may play a role in the observed Tg elevation and in driving the differences observed in alignment. Gel filtration and dynamic light scattering (DLS) and small angle X-ray scattering (SAXS) analyses were conducted in order to better understand this process and whether it affects the properties of the composite materials presented in this work. Protein solutions analyzed by gel filtration and DLS before and after sonication indicated increased molecular weights and diameters of silk-CBD assemblies upon sonication, compared to the control silk protein ( Figure 8 ). The pure silk solution was eluted with a peak corresponding to 140 kDa ( Figure 8 A), whereas the silk-CBD was eluted as two major peaks, one eluted into the void volume, and the second eluted as a 230 kDa protein, which may correspond to a dimer. After sonication, most of the silk-CBD protein was eluted into the void volume ( Figure 8 B). The calculated molecular weights of the silk and silk-CBD proteins are 47 and 65 kDa, respectively. SDS-PAGE analysis under denaturing conditions supported the molecular weight calculations ( Figure 8 C,D). As reported previously, proteins rich in glycine and alanine can migrate anomalously during gel filtration, often resulting in overestimation of the protein molecular weight [ 40 ]. In addition, the gel filtration was performed under non-denaturing conditions and silk polypeptides may adopt a rod-like elongated configuration, leading to size overestimation, compared to globular protein standards. It has been established that spider silk proteins form disulfide-bridged homodimers [ 41 , 42 , 43 , 44 ]. Gel filtration of the native proteins under reducing conditions revealed 260–320 kDa protein monomers. In the absence of the reducing agent, a shift in the molecular mass to 420–480 kDa was observed, less than a two-fold gain in molecular weight [ 41 ]. Dimer formation likely leads to changes in the protein conformation, resulting in modified protein migration along the column, as compared to the monomer form. The silk-CBD dimer may act as a nucleation site for higher molecular weight assemblies, which were eluted into the void volume, even before sample sonication. After sonication, an increase in the proportion of higher molecular weight silk assemblies was observed, similar to that previously described for native silkworm silk proteins subjected to sonication [ 33 ]. As demonstrated by DLS ( Table 2 ), both un-sonicated and sonicated silk formed homogeneous solutions with 3–4 nm diameter particles, whereas un-sonicated silk-CBD had 55 and 260 nm diameter particles, which increased in size to 96 nm and 2 μm particles upon sonication. 2.6. Small Angle X-ray Scattering (SAXS) of Solutions of Silk and Silk-CBD SAXS measurements of the silk and silk-CBD solutions analyzed before and after sonication ( Figure 9 ) demonstrated a form factor closely matching that of infinitely long rods. The silk samples released a very weak signal with apparently no structurally ordered subunits formed in solution, neither before ( Figure 9 A) nor after ( Figure 9 C) sonication. In the case of un-sonicated silk-CBD, the single and rather broad correlation peak ( Figure 9 A) indicated a nematic liquid crystalline phase, with a correlation distance of 26.4 nm and a domain size of approximately 80 nm. In other words, three subunits are in positional correlation, and the rod center-to-center distance is 26.4 nm ( Figure 9 D1). After sample concentration ( Figure 9 A), the peak intensity increased and the correlation distance decreased to 24.6 nm. Namely, the nematic phase was denser and more subunits were in positional correlation with one another ( Figure 9 D2). In nature, the highly concentrated spider protein dope, much like that of the silkworm, is liquid crystalline, where the main silk protein constituent is likely to assume a compact rod-like conformation. This conformation enables silk protein processing at high concentrations. Specifically, the molecules seem to form a nematic phase in the spider gland and duct, with the long axes of neighboring molecules aligned approximately parallel to one another. Liquid crystallinity offers desirable properties, such as efficient spinning of molecules as large as silk proteins, by allowing the viscous silk protein solution to slowly flow through the storage sac and duct as complex alignment patterns are formed [ 19 ]. After sonication ( Figure 9 B), the silk-CBD correlation peaks were fitted to a two-dimensional oblique lattice with three subunits along the a-axis, forming a rod center-to-center distance of 30.8 nm and three subunits along the b-axis, forming a rod center-to-center distance of 54.98 nm. The alignment angle between the two-dimensional subunits was γ = 83.2° ( Figure 9 E1); sonication resulted in the addition of subunit interactions, which led to their two-dimensional alignment. After concentrating the solution ( Figure 9 B), the lattice parameters changed to a = 32.4 nm, b = 45.8 nm and γ = 93.4°, and namely, more subunits were in correlation with one another ( Figure 9 E2). 2.7. Cryo-Transmission Electron Microscopy (Cryo-TEM) Cryo-transmission electron microscopy (cryo-TEM) images of sonicated protein samples support the gel filtration, DLS and SAXS results ( Figure 10 ). After sonication, the silk protein did not form any orderly structures ( Figure 10 upper images), whereas silk-CBD formed 100–200 nm long fibers/microfibrils ( Figure 10 bottom images). The liquid crystals and the longer structures revealed by SAXS and DLS were not detectable by this method, as they form very thick layers that are removed from the grid when preparing the thin film (<100 nm thick) required for cryo-TEM imaging. 2.8. Silk-CBD Assembly Model Based on the gel filtration, DLS, SAXS and cryo-TEM findings, we suggest that silk-CBD forms dimers after solution concentration, which align to form liquid crystals ( Figure 11 C). After sonication, the hydrophobic spider silk protein domains are more prone to engage in further interactions, leading to higher molecular assemblies ( Figure 11 D). This molecular order likely effects the morphology and thermal properties of the silk-CBD-CNC composites, for instance by driving CNC alignment (see the mechanism presented in Figure 12 )."
} | 5,325 |
38478382 | PMC10936741 | pmc | 1,417 | {
"abstract": "Abstract Global climate changes threaten food security, necessitating urgent measures to enhance agricultural productivity and expand it into areas less for agronomy. This challenge is crucial in achieving Sustainable Development Goal 2 (Zero Hunger). Plant growth‐promoting microorganisms (PGPM), bacteria and fungi, emerge as a promising solution to mitigate the impact of climate extremes on agriculture. The concept of the plant holobiont, encompassing the plant host and its symbiotic microbiota, underscores the intricate relationships with a diverse microbial community. PGPM, residing in the rhizosphere, phyllosphere, and endosphere, play vital roles in nutrient solubilization, nitrogen fixation, and biocontrol of pathogens. Novel ecological functions, including epigenetic modifications and suppression of virulence genes, extend our understanding of PGPM strategies. The diverse roles of PGPM as biofertilizers, biocontrollers, biomodulators, and more contribute to sustainable agriculture and environmental resilience. Despite fungi's remarkable plant growth‐promoting functions, their potential is often overshadowed compared to bacteria. Arbuscular mycorrhizal fungi (AMF) form a mutualistic symbiosis with many terrestrial plants, enhancing plant nutrition, growth, and stress resistance. Other fungi, including filamentous, yeasts, and polymorphic, from endophytic, to saprophytic, offer unique attributes such as ubiquity, morphology, and endurance in harsh environments, positioning them as exceptional plant growth‐promoting fungi (PGPF). Crops frequently face abiotic stresses like salinity, drought, high UV doses and extreme temperatures. Some extremotolerant fungi, including strains from genera like Trichoderma , Penicillium , Fusarium , and others, have been studied for their beneficial interactions with plants. Presented examples of their capabilities in alleviating salinity, drought, and other stresses underscore their potential applications in agriculture. In this context, extremotolerant and extremophilic fungi populating extreme natural environments are muchless investigated. They represent both new challenges and opportunities. As the global climate evolves, understanding and harnessing the intricate mechanisms of fungal‐plant interactions, especially in extreme environments, is paramount for developing effective and safe plant probiotics and using fungi as biocontrollers against phytopathogens. Thorough assessments, comprehensive methodologies, and a cautious approach are crucial for leveraging the benefits of extremophilic fungi in the changing landscape of global agriculture, ensuring food security in the face of climate challenges.",
"conclusion": "CONCLUDING REMARKS – HARNESSING EXTREMOTOLERANT FUNGI FOR SUSTAINABLE AGRICULTURE IN A CHANGING CLIMATE Facing the challenges of global climate change Global climate changes increasingly expose agronomical land to drought, salinity, and extreme temperatures, threatening traditional agronomic practices. In this context, sustainable agricultural practices and innovative solutions cannot be overstated. Plant growth‐promoting microorganisms (PGPM): A green paradigm for agriculture Within the intricate web of the plant holobiont, the role of PGPM emerges as a promising green paradigm. These microorganisms exhibit multifaceted contributions to plant well‐being and present a holistic approach to sustainable agriculture and environmental resilience. Bacteria versus fungi: Commercialization disparity While bacteria have garnered significant attention and commercial success in the realm of PGPM, fungi, with their remarkable plant growth‐promoting (PGP) functions, remain comparatively understudied and underutilized. The unique attributes of fungi, encompassing their ubiquity, morphology, and ability to endure harsh environments, position them as exceptional plant growth‐promoting fungi (PGPF). Fungi's comparative advantages and the need for more synergistic consortia Fungi, with their diverse forms, including filamentous, yeasts, and polymorphic fungi, bring distinctive advantages compared to bacteria as PGPM. The ubiquity and endurance of fungi make them exceptional contributors to sustainable agriculture. Furthermore, the call to transition from single‐strain probiotics to synergistic, multi‐functional consortia aligns with the versatility exhibited by fungi in controlling plant‐parasitic nematodes, herbivore insects, and pathogens. This shift opens avenues for the emergence of synthetic communities (SynComs) as promising functional plant probiotics. Unlocking the potential of extremotolerant fungi In the face of escalating challenges, extremotolerant fungi stand out in relation to potential new solutions. Their beneficial properties, ranging from extreme salinity and drought tolerance to biocontrol abilities, position them as indispensable components of future agricultural strategies. Certain genera of fungi, including Trichoderma , Penicillium , Fusarium , and others, have already shown their capabilities in mitigating a spectrum of stresses. Research on their interactions with plants opened a new frontier for harnessing the benefits of extremotolerant and extremophilic fungi, populating diverse extreme environments. Fungal probiotics: Unveiling the potential Within the diverse realm of plant probiotics, fungi emerge as potent contributors with almost untapped potential. The notion of fungal probiotics challenges the prevailing emphasis on bacterial counterparts, introducing a paradigm shift in harnessing the benefits of microorganisms for plant health. Despite the commercial utilization of bacterial plant probiotics, the inclusion of fungal counterparts remains limited, opening the way for innovative applications. Research endeavours focusing on isolating and characterizing these fungal probiotics present a new way for expanding our understanding of plant‐microbe interactions and, in turn, enriching sustainable agricultural practices. Changing agricultural strategies through advanced technologies Agricultural research is marked nowadays by new technological innovations, also featuring advanced ‘omics’ technologies. The integration of genomics, transcriptomics, proteomics, and metabolomics into the study of plant‐microbe interactions offers new insights into the intricate mechanisms governing these relationships. Harnessing the power of these omics technologies will allows researchers to decipher the genomic blueprint of extremotolerant fungi, unravelling the genetic basis of their resilience to extreme conditions. Furthermore, the application of metagenomics facilitates a comprehensive understanding of microbial communities in diverse environments, aiding in the identification of novel extremotolerant fungi. The emergence of synthetic communities (SynComs), designed also by applying omics technologies, represents a pioneering approach to developing functional plant probiotics. These consortia, comprising carefully selected microorganisms, highlight the potential for precision agriculture, where the microbiome is strategically tailored to enhance plant health and productivity. Further research and caution Understanding the intricate mechanisms of fungal–plant interactions, especially in stressed environments, is important for leveraging the benefits of extremotolerant fungi. Increased research, especially on the synergies between plants and extremotolerant fungi, is urgently needed. In this way we can increase sustainable agriculture and help to ensure food security in the face of unprecedented climate challenges. Fungi, with their potential, can significantly contribute to a more sustainable future.",
"introduction": "INTRODUCTION Humanity is approaching an unprecedented catastrophe. During the last decades, different regions of the planet have been severely impacted by exceptional weather and climate extremes, including heat waves, droughts, hurricanes, and heavy precipitation (Zhou et al., 2023 ). Data from the last year indicate that the consequences of the global warming process are much more significant than previously thought (IPCC, 2023 ). Climate extremes are expected to severely harm human society's welfare and ecosystem sustainability. The recognition by the Secretary General of the United Nations of the enormous difficulties in achieving the Sustainable Development Goals (SDGs) (UN Report, 2023 ) adds an additional layer of worry to the current universal perception of an uncertain future (Figure 1 ). Consequently, it is increasingly acknowledged that the effort required to meet the aspirations of a large part of humanity to achieve a better future for our planet and for all humankind will be of a colossal magnitude. FIGURE 1 Some impacts of global warming and the threats posed on attaining the Sustainable Development Goals (SDGs). The sections, arranged sequentially from left to right, examine the principal projected future consequences of climate change on the environment, the interplay between plants, soil, and microbes (including anticipated soil properties under extreme conditions in the future), the productivity of agroecosystems, and the society, that could give rise to humanitarian crises in the near future. The SDGs that may be impacted due to the future extreme scenario are situated at the bottom. Among the SDGs, number two stands out: Zero Hunger. More than 820 million people are currently suffering from hunger, and another 250 million people will be on the brink of starvation by the end of the 2020s. According to estimates compiled by the Food and Agriculture Organization (FAO), to meet the future food demand of more than 9.3 billion people, food production should be increased by approximately 60 per cent by 2050 (FAO, 2017 ). Nevertheless, this will be out of reach if we do not introduce profound changes in the global agri‐food system to increase agricultural production and food production sustainably since (i) the consequences of climate extremes impacting agricultural regions can be disastrous in terms of global food supply and global food security (Zhou et al., 2023 ); (ii) the amount of cropland is finite and has already been taken up, and marginal lands are too stressful for plants to grow well, due to harsh physical, chemical or agronomic factors (Pancaldi & Trindade, 2020 ); (iii) intensive farming systems cannot deliver food and crops sustainably, causing massive deforestation, water scarcities, soil depletion and degradation, and high levels of greenhouse gas emissions (FAO, 2017 ); and, (iv) the capacity of ecosystems to mitigate the impact caused by the excessive use of agrochemicals (fertilizers + pesticides) has been stretched to extreme limits. Failure in this area will sentence humanity to suffer from numerous crises: more hunger, more malnutrition, more wars, more forced migrations, and more inequity (Figure 1 ). In the face of these threats, it has been suggested – based on a large body of evidence – that microbes may be part of the solution (Timmis et al., 2017 ). Indeed, scientists worldwide have insisted on considering microbes as ‘weapons for peace’ (Anand et al., 2023 ) and, among many other applications, using their abilities and ecological functions to promote plant growth for agricultural purposes (Bakker & Berendsen, 2022 ; Hu et al., 2022 ). The immediate consequence of taking advantage of beneficial microbes would be to increase food production sustainably and reduce hunger, with all the positive outcomes that derive from this. Indeed, microorganisms have been used for agriculture intensification for decades (Mitter et al., 2021 ). In addition to participating in multiple processes related to the promotion of plant health and growth and being crucial for the restoration of degraded soils (Anand et al., 2023 ), microorganisms can be used to rationally design and produce different agricultural inputs (= ‘bioinoculants’ = ‘plant probiotics’) (see next section). These products date back to the end of the 19th century and have been gradually and steadily incorporated into the arsenal of tools that allow us to achieve higher levels of agricultural yields, with the advantage of not threatening the ecological balance of agricultural and natural ecosystems. In other words, adequately used plant probiotics allow the sustainable intensification of agriculture and, therefore, the production of more food. However, when applied in the field, many plant probiotics are adversely affected by environmental factors that limit their universal and widespread use. Because they are alive, once introduced into agroecosystems, the microbes included in plant probiotics must reactivate and multiply actively to function efficiently (Sanjuán et al., 2023 ). Therefore, any biotic or abiotic factor that limits beneficial microbe's establishment and inhibits their active multiplication becomes a barrier to their utilization. At a time when soils in large regions of the planet are undergoing a process of aridification (Malpede & Percoco, 2023 ) and even desertification (Mirzabaev et al., 2019 ), factors such as high temperatures, salinization, and water scarcity – to mention a few – must be seriously considered when thinking about plant probiotics for the (near) future, because of the strong influence the former have on soil fertility, stability and biodiversity. Since many agroecosystems worldwide experience already (or will experience soon) extreme conditions, the most apparent solution to intensify agriculture with plant probiotics would be to consider extremophilic microorganisms as the ones that offer the most promise for developing these bioinputs. Many microbes are not only capable of enduring extreme conditions (in which case they are considered ‘extremotolerant’) but often multiply and perform better when such conditions are met (the so‐called ‘extremophiles’). The group is as diverse as it is broad and includes bacteria, fungi, and microscopic algae (Coker, 2019 ). Bacteria have been the most studied within the group of extremophiles able to promote plant growth. However, another group of extremophiles deserves to be considered in depth: extremophilic fungi. Beyond displaying plant‐growth‐promoting (PGP) traits, most of which resemble the ones exhibited by bacteria, fungi are fundamental components of all complex microbial communities such as the rhizosphere of plants (Pozo et al., 2021 ). It is there – but also in aerial and inner tissues – that they exert their promoting activity, providing nutrients, releasing plant hormone‐like compounds, antagonizing (and even killing) plant pathogens, protecting plants from heavy metals, and providing many other ecological functions. In the following pages, we will present the benefits that could result from using extremophilic fungi to develop future agriculture and prevent humanitarian crises. We will place extremotolerant and extremophilic fungi in the context of ‘plant probiotics’ and present illustrative examples of their use to mitigate the detrimental impacts caused by extreme environmental conditions on their plant hosts. Finally, we will discuss future perspectives, focusing on aspects that have not received sufficient attention from researchers and companies."
} | 3,815 |
34778642 | PMC8582064 | pmc | 1,420 | {
"abstract": "Direct interspecies\nelectron transfer (DIET) is a breakthrough\nthat can surpass the limitations of anaerobic digestion. Conductive\nmaterials and polarized bioelectrodes are known to induce DIET for\nmethane production but are still challenging to apply at a field scale.\nHerein, compared to polarized bioelectrodes, electrostatic fields\nthat promote DIET were investigated in an anaerobic reactor with conductive\nmaterials. As a conductive material, activated carbon enriched its\nsurface with electroactive microorganisms to induce DIET (cDIET).\ncDIET improved the methane yield to 254.6 mL/g COD r , compared\nto the control. However, polarized bioelectrodes induced electrode-mediated\nDIET and biological DIET (bDIET), in addition to cDIET, improving\nthe methane yield to 310.7 mL/g COD r . Electrostatic fields\nselectively promoted bDIET and cDIET for further methane production\ncompared to the polarized bioelectrodes. As the contribution of DIET\nincreased, the methane yield increased, and the substrate residue\ndecreased, resulting in a significant improvement in methane production.",
"conclusion": "Conclusions Activated\ncarbon in anaerobic reactors enriches the surface with\nEAMs, improving methane production through cDIET. In anaerobic reactors\nwith activated carbon, polarized electrodes enrich EAMs on the surfaces\nof the electrode and activated carbon and in the bulk solution, improving\nmethane production by eDIET, cDIET, and bDIET. The surface insulation\nof the electrode blocks eDIET, but the electrostatic field further\npromotes cDIET and bDIET selectively in the bulk solution to improve\nmethane production. When applied to large-scale anaerobic reactors,\nthe electrostatic field has many advantages over polarized bioelectrodes\nin promoting DIET for methane production. Thus, the anaerobic digestion,\ncombined with an electrostatic field, is a viable bioelectrochemical\nplatform that can apply to field-scale anaerobic digestion.",
"introduction": "Introduction Anaerobic digestion is a series of microbial\nmetabolic processes\nthat decompose and stabilize organic matter in an oxygen-free environment\nand produce methane as a byproduct. In the anaerobic metabolic processes,\nacidogenic bacteria break down organic matter and transfer the electrons\nto the intermediates, such as acetate and hydrogen. Then, methanogenic\narchaea convert the electrons from the intermediates into methane.\nThus, anaerobic digestion is an indirect interspecies electron transfer\n(IIET) process via intermediates from organic matter to methane. 1 − 3 However, methanogenic archaea have physiological properties that\nare different from those of acidogenic bacteria. 4 , 5 Thereby,\nthe intermediates in anaerobic digestions can be quickly accumulated\nby disturbing the balance between their production and consumption,\neven with minor changes in environments such as pH and temperature. 3 , 6 In addition, due to thermodynamic limitations, the intermediates\ncannot be entirely converted to methane, and some of the intermediates\nwith low solubility or high volatility can be released into the gaseous\nphase. 5 Acidogenesis and methanogenesis,\nin particular, are multiple enzymatic reactions with inevitable energy\nlosses. 2 , 5 , 7 Therefore,\nthe methane yield of organic matter is generally lower than the theoretical\nvalue. In recent years, direct interspecies electron transfer\n(DIET) between\nelectroactive microorganisms (EAM) has received considerable attention\nas a key to overcoming the limitations of anaerobic digestion. 5 , 8 The microbial groups of EAM for methane production include exoelectrogenic\nbacteria (EEB) and electrotrophic methanogenic archaea (EMA). EEB\nis the bacterial species that is capable of transferring electrons\ndirectly outside the cells through C-type cytochromes or conductive\npili. 2 , 4 , 7 , 9 EMA can produce methane by directly accepting the\nelectrons from EEB without the intermediates. 8 , 10 In\ngeneral, the types of DIET for methane production are classified as\nfollows: (i) conductive material-mediated DIET (cDIET), (ii) electrode-mediated\nDIET (eDIET), and biological DIET (bDIET) between EEB and EMA electrically\nconnected by physical contact. 1 , 5 , 9 DIET is advantageous in kinetics over IIET and conserves electrons\nbetter. 2 , 7 Therefore, the anaerobic digestion process\ncan be more stable and robust, as DIET contributes more to methane\nproduction. 5 , 11 DIET can be induced by enriching\nEAM, and providing the driving force for the electron transfer. EAM\ncan be naturally enriched when insoluble electron acceptors are outside\nthe cells in anaerobic or nutrient-limited environments. 12 , 13 However, conductive materials and polarized bioelectrodes can also\nenrich their surfaces with EAM in the anaerobic condition, driving\ncDIET and eDIET, respectively. 14 , 15 In anaerobic digestion,\nvarious conductive materials such as activated carbon, metal oxide,\ncarbon fiber, and metal-conductive polymer composites significantly\npromoted methane production through cDIET. 16 , 17 It has been understood that cDIET from EAM to EMA through conductive\nmaterials promotes methane production by reducing carbon dioxide. 4 , 8 However, recent reports suggest that conductive materials also promote\nacetate dismutation for methane production. 17 − 19 The conductive\nmaterials with electrochemical activity, such as nanocarbon compounds\nbased on a carbon nanotube and graphene, metal–organic framework,\nand metal-polyelectrolyte complexes, appear to have the potential\nto enhance methane production further. 20 − 24 Microbial electrolysis cells (MECs) are examples\nof polarized bioelectrodes that enrich the electrode surface with\nEAM, promoting eDIET for hydrogen or methane production. 25 − 27 However, the methane production in the anaerobic digestion combined\nwith MECs could only be improved satisfactorily by providing polarized\nbioelectrodes of a sufficient surface area. 5 It is worth noting that electrostatic fields move electrons,\npolar\nmolecules, and charged ions by the Lorentz force. 28 The Lorentz force may promote DIET for methane production\nunder electrostatic fields. There are several reports that electrostatic\nfields promote the redox reaction in the biological process. The polarized\nbioelectrode installed in an upflow anaerobic reactor significantly\nimproved methane production, while the eDIET contributed only a few\npercentage points. 29 , 30 The performances of aerobic composting\nand biological nitrogen removal processes were also improved under\nelectrostatic fields, and their bulk medium showed bioelectrochemical\nactivities. 31 − 33 Direct evidence that electrostatic fields promote\nDIET for methane production was observed from the conversion process\nof lignite to methane with anaerobic microorganisms. 34 , 35 It indicates that electrostatic fields enrich the bulk medium with\nEAM, and drive bDIET for methane production. 36 The electrostatic field-driven DIET in the bulk medium is not proportional\nto the electrode surface area, with fewer electrode-related issues. Interestingly, it was recently found that both polarized bioelectrodes\nand electric fields promote DIET through conductive materials in anaerobic\nreactors. 26 , 27 , 37 This implies\nthat methane production can be maximized by adding a conductive material\nto the anaerobic reactor and applying a polarized bioelectrode or\nelectrostatic field. Therefore, it is necessary to elucidate the characteristics\nof DIET promoted by polarization electrodes or electrostatic fields\nin anaerobic reactors with conductive materials and discuss their\nadvantages and limitations. In this study, in an anaerobic batch\nreactor with powdered activated\ncarbon as a conductive material, the characteristics of methane production\ndepending on the polarized bioelectrode and electrostatic field were\ninvestigated based on the electron transfer pathway, residual substrate,\nbioelectrochemical activity, substrate conversion, and microbial community.",
"discussion": "Results\nand Discussion Activated Carbon Mediated DIET for Methane\nProduction Cumulative methane production in AC, an anaerobic\nbatch reactor with\npowdered activated carbon, increased in the form of a sigmoidal curve\nwith time ( Figure 1 ). The methane production was saturated to 332.2 mL, higher than\nthe control without the activated carbon. In anaerobic reactors, conductive\nmaterials, including activated carbon, carbon cloth, carbon nanotube,\ngraphene, and magnetite, can enrich the surface with EAM and serve\nas a conduit for cDIET for methane production. 14 , 15 , 38 In AC, the initial lag time required for\nsubstantial methane production was not considerably different from\nthat of the control ( Table 1 ). This suggests that anaerobic microorganisms quickly adhere\nto the surface of activated carbon particles in AC, express their\nelectroactive genes for the thermodynamic benefits of cDIET over IIET,\nand promote methane production through cDIET. After repeating the\nbatch cycle for AC, the methane yield was approximately 254.6 mL/g\nCOD r , 21.4% higher than the control ( Table 1 ). The percentages of electron conversion\nfrom the substrate to methane for AC and control are 72.7 and 59.9%,\nrespectively. In general, DIET improves methane yield over IIET by\nconserving the electrons better. 2 , 5 , 7 In methane production through IIET, electron conversion efficiency\ncan be greatly affected by environmental factors such as temperature\nand pH and intermediate products and inhibitors. 4 , 6 , 8 In mesophilic conventional anaerobic digestion,\nthe electron conversion from the substrate to methane ranged 35–75%\nin previous studies. 3 , 5 In control, the main electron\ntransfer route for methane production may be IIET via the intermediates.\nThus, the percentage of electron conversion for IIET can be considered\nas about 59.9% obtained from the control. On the other hand, the percentage\nof electron conversion between EAMs ranged from 70 to 96%, estimated\nfrom the Coulombic efficiency for methane production via the electrode\nin MECs. 3 , 25 Thus, assuming that the contribution of\nDIET to methane production was 96% of the electron conversion, the\npercentage of cDIET contributing to methane production in AC can be\nroughly estimated as 35.5% (96 x + 59.9(100 – x ) = 100 × 72.7, x = 35.5), based\non the electron balance ( eq 2 ). However, the contribution of DIET to methane production\ncan increase further as the electron conversion (%) of DIET decreases.\nIn a previous study using 1 g/L of activated carbon, the methane yield\nwas 221.8 mL/g COD r, and the percentage of cDIET was 7.8%,\nrespectively, less than those in AC. 37 Figure 1 (a) Cumulative\nmethane production and (b) soluble COD. Table 1 Features of Methane Productions Promoted\nby Activated Carbon and the Polarized Electrode content control AC ACPB ACEF λ (d) 0.60 ± 0.03 0.67 ± 0.01 0.74 ± 0.01 0.98 ± 0.01 μ m (mL CH 4 /d) 104.2 ± 0.6 122.8 ± 0.3 157.3 ± 1.2 196.4 ± 0.9 P u (mL CH 4 ) 278.4 ± 0.1 337.8 ± 0.2 432.8 ± 0.1 462.5 ± 0.7 removed COD (g) 1.32 ± 0.02 1.32 ± 0.02 1.39 ± 0.01 1.45 ± 0.01 CH 4 yield (mL/g COD r ) 209.8 ± 0.1 254.6 ± 0.1 310.7 ± 0.1 317.3 ± 0.4 electron conversion (%) 59.9 72.7 88.8 90.7 DIET (%) - ≤35.5 ≤80.1 ≤85.3 DIET rate (10 –7 e – moles/s) 0 1.80 5.21 6.93 remained soluble\nCOD (mg/L) 1753 ± 108.5 1583.7 ± 55.0 1344.0 ± 101 1238.5 ± 84.7 It is well\nknown that various conductive materials such as a carbon\nnanotube, carbon fiber, and iron/PANI complex improve methanogenesis\nby promoting extracellular electron transfer. 14 , 16 It seems that the contribution of cDIET to methane production depends\non the properties of the conductive materials, such as specific surface\narea, electric conductivity, shape, and dose. 4 , 15 , 38 In AC, the maximum methane production rate\nwas 122.8 mL/d, faster than the 104.2 mL/d for the control ( Table 1 ). The increased methane\nproduction rate in AC appears to be related to more methane production. 4 Interestingly, at the end of each batch cycle,\nthe residual of soluble COD in AC was slightly lower than that in\ncontrol ( Figure 1 b),\nindicating that the substrate affinity of cDIET is higher than that\nof the IIET in anaerobic digestion. This confirms that cDIET is advantageous\nover IIET in kinetics and thermodynamics for methane production. 5 , 35 , 36 Polarized Bioelectrode-driven\nDIET In ACPB, the cumulative\nmethane production increased over time in a pattern similar to that\nin AC ( Figure 1 a).\nHowever, the methane production increased to 432.8 mL, and the yield\nreached 310.7 mL/g COD r , significantly higher than the\nAC ( Table 1 ). The percentage\nof electron conversion from the substrate to methane in ACPB was 88.8%.\nThe other fraction of electrons appears to be lost in the transfer\nprocess from the substrate to methane or used for microbial cell synthesis. 31 , 36 However, a small amount of methane dissolved in the liquid medium\nthat was not recovered in a gaseous form may also be included in the\nlost electrons. The high methane yield in ACPB led to the improved\nmaximum methane production rate of 157.3 mL/d, higher than the AC.\nThe DIET fraction in methane production was estimated at up to 80.1%\nor more (96 x + 59.9(100 – x ) = 100 × 88.8, x = 80.1), significantly higher\nthan 35.5% of the cDIET alone in AC ( Table 1 ). This suggests that DIET significantly\ncontributed to improving methane production in ACPB. ACPB is a kind\nof single-chamber MECs containing activated carbon. In MECs, the electric\npotential on polarized electrodes drives eDIET for methane production\nby enriching the electrode surface with EAM. 15 , 25 Herein, it is noted that the polarized electrode potentials in MECs\ncreate an electrostatic field in the bulk solution. The electrostatic\nfield may enrich the bulk solution with EAM. 5 , 32 − 37 Thus, it is believed that bDIET in the bulk solution also contributed\nto methane production in ACPB. 5 However,\nin ACPB, the methane yield of 310.7 mL/g COD r was slightly\nhigher than that of MECs without activated carbon in the previous\nstudy. 5 This indicates that the powdered\nactivated carbon in ACPB has played an essential role in improving\nthe methane yield. Conductive materials have been found to promote\ncDIET in MECs. 15 It seems that the electrostatic\nfields can polarize the surface of activated carbon particles, transforming\nthem into small bipolar electrodes. 39 , 40 The polarized\npotentials on the bipolar activated carbon particles in ACPB may have\nenriched the surface with EAM and promoted the cDIET. Although it\nis still necessary to study further to estimate the individual DIET\ncontributed to methane production in ACPB, at least three DIET types,\nnamely, eDIET, cDIET, and bDIET appear to improve the methane production\nof ACPB, compared to that of the AC. In addition, at the end of each\nbatch cycle, the residual of soluble COD in ACPB was lower than in\nAC ( Figure 1 b). This\nsuggests that the more methane production in ACPB than in AC was due\nto higher methane yield and more substrate conversion. Electrostatic\nField-Driven DIET In ACEF, the methane\nproduction was more interesting than in AC or ACPB. Cumulative methane\nproduction in ACEF was gradually saturated at 462.5 mL over time,\nhigher than the ACPB ( Figure 1 a). ACEF is an anaerobic reactor that is similar to ACPB with\npowdered activated carbon and polarized bioelectrodes. However, the\nelectrodes in ACEF were electrically insulated by coating their surface\nwith a dielectric polymer film. The dielectric film on the electrode\nsurface blocks the eDIET from transferring electrons for methane production.\nNevertheless, the methane yield of ACEF was 317.3 mL/g COD r , slightly higher than that of the ACPB, and the electron conversion\nfrom the substrate to methane was as high as 90.7% ( Table 1 ). In ACEF, although the eDIET\nwas excluded, the activated carbon and the electrostatic field were\nstill involved in the electron transfer for methane production. Based\non the electron balance, the contribution of DIET for methane production\nin ACEF was estimated as 85.3% or more (96 x + 59.9(100\n– x ) = 90.7 × 100, x = 85.3), higher than the AC or the ACPB. Compared to in AC, this\ncan be direct evidence that the electrostatic field promoted bDIET\nand cDIET for methane production in ACEF. Also, in ACEF, the contribution\nof DIET being higher than in ACPB provides detailed information on\nthe electron transfer for methane production. First, the electron\ntransfer loss for eDIET is higher than that for bDIET or cDIET. Second,\nwhen eDIET is blocked, the electric field further improves bDIET and\ncDIET. The electron transfer losses in the bioelectrochemical\nsystem are generally due to activation, ohmic, and concentration overpotentials. 5 , 35 , 41 , 42 The activation overpotential is the potential to overcome the activation\nenergy for electron transfer. The overpotentials for bDIET are primarily\nrelated to the ability of EAM species to engage in extracellular electron\ntransfer. The ohmic resistance of conductive materials, electrodes,\nor the electrode–wire interface may also interfere with the\nelectron transfer, especially eDIET or cDIET. 5 , 41 , 42 The concentration overpotential is related\nto the mass transfer limiting the electron transfer. The higher electron\nlosses in ACPB than in ACEF are likely due to the high ohmic resistances\nrelated to the electrode or the electrode–wire interface. There\nseems to be a difference in the thermodynamic potential to induce\nelectron transfer between each DIET. The residual soluble COD in ACEF\nat the end of each batch cycle was 1238.3 mg/L, slightly lower than\nin ACPB ( Figure 1 b).\nAs the contribution of DIET increases, the methane yield and substrate\nconversion appear to increase. It can be concluded that surface insulated\nelectrodes, rather than polarized bioelectrode, selectively promote\nbDIET and cDIET by providing an electrostatic field to the bulk solution,\nthereby improving methane production. Bioelectrochemical Activities\nin the Bulk Solution The cyclic voltammogram (CV) for the\nbulk solution shows one oxidation\nand two reduction peaks under non-turnover conditions, similar to\nall anaerobic reactors ( Figure 2 a). The redox peaks of the CV indicate electrochemically active\nsubstances that include biotic and abiotic redox substances. 32 , 36 , 37 , 43 EEB and EMA belong to the biotic redox substances, and there are\nseveral types of abiotic substances, including flavins, quinones,\nand humic substances. 5 , 31 The redox peak potentials for\nEEB and EMA are known to be in the range −0.20 to 0.41 V versus\nAg/AgCl and −0.07 to −0.41 V versus Ag/AgCl, respectively. 5 , 34 , 36 For AC, the oxidation peak was\n0.01 V versus Ag/AgCl, and there were two reduction peaks at 0.17\nand −0.38 V versus Ag/AgCl. The oxidation peak and the second\nreduction peak appeared to be the activities of EEB and EMA, based\non their reported potential ranges ( Figure 2 a). The peak potentials for ACPB and ACEF\nwere similar to those of AC. These similar bioelectrochemical properties\nin the anaerobic reactors appear to be caused by the activated carbon.\nIn control, the redox peaks may be due to abiotic substances that\nare unable to mediate DIET. For AC, the peak heights for the oxidation\nand reduction were 0.50 and 0.57 mA, respectively. The peak heights\nfor ACPB were higher than those for AC, and further for ACEF ( Table 2 ). The peak height\nof CV indicates the abundance of EAM in the bulk solution. 5 , 34 This indicates that in AC, ACPB, and ACEF, EAM improved methane\nproduction by promoting DIET. Figure 2 Electrochemical properties of the bulk solution:\n(a) CVs and (b)\nEIS curves. Table 2 Electrochemical Properties\nin CV and\nEIS for the Bulk Solution content control AC ACPB ACEF E p,ox / I p,ox (V/mA) –0.04/0.51 0.01/0.53 0.01/0.54 0.01/0.60 E p, red1 / I p,red (V/mA) 0.13/0.13 0.17/0.10 0.15/0.12 0.16/0.10 E p,red2 / I p,red (V/mA) –0.42/0.53 –0.38/0.57 –0.39/0.60 –0.39/0.61 equivalent\ncircuit for EIS R s (Ω) 21.5 21.1 19.1 18.1 R s – Q |( R ct – W )– C | R Q y 6.14 × 10 –4 3.33 × 10 –4 3.03 × 10 –4 3.40 × 10 –4 Q a 7.08 × 10 –1 8.24 × 10 –1 8.17 × 10 –1 8.18 × 10 –1 R ct (Ω) 3626.0 61.0 51.9 47.5 W (Ω/√ s ) 2.21 × 10 –5 1.61 × 10 –4 2.05 × 10 –4 2.31 × 10 –4 C (F) 6.80 × 10 –4 7.38 × 10 –5 8.20 × 10 –5 9.44 × 10 –5 R (Ω) 19,950.9 3.2 3.5 3.3 r 2 0.9998 0.9998 0.9997 0.9995 The biotic and abiotic redox substances are known to be electrically\nconductive. 1 , 34 , 37 Thus, the conductivity or resistance of the bulk solution of bioelectrochemical\nreactors can be another indicator of the abundance of redox substances. 37 The solution resistance estimated from the electrochemical\nimpedance spectrum (EIS) data was 21.5 Ω for control, and slightly\ndecreased in the order AC > ACPB > ACEF ( Table 2 ). The solution resistances were well matched\nwith the redox peak heights in the CV. The activation overpotential\nfor DIET can be informed from the charge transfer resistance of EAM.\nIn control, the charge transfer resistance was very high at 3626.0\nΩ ( Table 2 ).\nThis clearly shows the absence of EAM in control, and that even abiotic\nsubstances cannot mediate DIET. The charge transfer resistance\nfor the AC was 61.0 Ω, significantly\nsmaller than for control. This means that the activated carbon surface\nwas enriched with EAM in AC, and then methane production was promoted\nby cDIET. It is known that conductive materials could replace the\npili or c-type cytochrome required for DIET in anaerobic digestion. 4 , 38 However, the force driving cDIET between EEB and EMA on conductive\nmaterials is still unclear. One hypothesis that could explain the\ndriving mechanisms of cDIET is a local polarization of conductive\nmaterials. EEB transfers electrons to conductive materials for thermodynamic\nbenefits. The electrons transferred from EEB can locally polarize\nthe conductive materials. In addition, the electric potential of conductive\nmaterials can be locally positive by donating electrons to EMA for\nmethane production. These local polarizations of conductive materials\nmay induce direct electron transfer from EEB to EMA. However, further\nstudies are needed to prove the hypothesis to describe the mechanisms\nof cDIET induction. The charge transfer resistances for ACPB and ACEF\nwere further reduced to 51.9 and 47.5 Ω, respectively, compared\nto AC. ACPB and ACEF are the anaerobic reactors exposed to the electrostatic\nfield. It can be concluded that the electrostatic field of ACPB and\nACEF further enriches the bulk solution and activated carbon surface\nwith EAM, promoting bDIET and cDIET for methane production. Microbial\nCommunities The microbial samples collected\nfrom the bulk solution were taxonomically profiled based on the NGS\nplatform. The valid reads for the archaea were significantly higher\nthan that for the bacteria, but the operational taxonomic units (OTUs)\nwere higher in bacteria ( Table 3 ). For archaeal species, the richness (ACE, Chao1) and evenness\n(NPShannon, Simpson) in AC were significantly higher than in the control,\nbut slightly lower in ACPB than the AC, further in ACEF ( Table 3 ). It seems that activated\ncarbon enriches the bulk solution or its surface with the archaeal\nspecies, and the polarized bioelectrodes and the electric field select\nthe species. The effect of activated carbon, the polarized electrode,\nand the electric field on the richness of bacterial species was weaker\nthan that of archaea, but it was not evident in the species evenness. Table 3 Valid Reads, OTUs, and Diversity Indices\nfor the Microbial Samples Collected from the Bulk Solution (Control,\nAC, ACPB, and ACEF) indices valid reads OTUs Ace Chao1 NPShannon Simpson control bacteria 33,237 1674 1807.9 1724.4 4.72 0.045 archaea 73,546 95 104.3 98.9 1.32 0.446 AC bacteria 31,301 1771 1913.2 1820.3 4.97 0.033 archaea 80,177 171 182.2 173.8 2.64 0.101 ACPB bacteria 29,191 1733 1893.1 1791.0 5.22 0.023 archaea 73,372 174 178.8 175.1 2.63 0.106 ACEF bacteria 28,520 1571 1714.4 1620.8 4.71 0.049 archaea 70,969 153 163.5 156.3 2.59 0.110 Of the OTUs identified bacterial taxa, Streptococcus and BBZD_g were commonly\nabundant bacteria at the\ngenus level in all reactors ( Figure 3 a). However, Porphyromonadaceae_uc , AJ 009469 _g , and Cloacamonas were more abundant in AC, ACPB, and ACEF,\nthan in control. The biplot obtained from principal component analysis\n(PCA) shows that the bacterial communities have similarities in AC,\nACPB, and ACEF containing the activated carbon ( Figure 4 a). The species Streptococcus\nhenryi and uncultured species GQ 458215 _s , BBZD_g_uc , and AB 234269 _s were the most abundant in control. Figure 3 Abundance of microbial\nspecies (a) bacteria and (b) archaeal. Figure 4 (a) Biplot\nfor the bacterial communities and (b) Biplot for the\narchaeal communities. However, Porphyromonadaceae_uc and Cloacamonas\nacidaminovorans were more abundant bacterial\nspecies in AC, ACPB, and ACEF than control. The family Porphyromonadaceae\nis an obligately anaerobic fermenter producing acetate. 43 C. acidaminovorans is a synergistic species in an anaerobic digester degrading the\nacetate and propionate. 3 The species Porphyromonadaceae_uc and C. acidaminovorans are likely to be the EEB mediating DIET, which is frequently observed\nin bioelectrochemical systems. 5 , 43 S. henryi and CU 921187 _s were more abundant\nspecies in ACEF than the others. S. henryi is a facultative anaerobic fermenter, but appears to be an electroactive\nspecies that is commonly observed in bioelectrochemical systems. 5 , 44 CU 921187 _s is an uncultured bacterial\nspecies isolated from a mesophilic anaerobic digester for municipal\nwastewater sludge, abundant in the bioelectrochemical system for the\nmethane conversion of coal. 35 , 45 It seems that S. henryi and CU 921187 _s are also EEB species that promote DIET in the electrostatic field. Clostridium quinii was an abundant bacterial species\nin AC and ACEF and is a commonly observed EEB species in bioelectrochemical\nsystems. 5 , 36 In archaeal groups, a clear difference\nin the dominant groups appeared\nat the genus level. Genus Methanocorpusculum was the predominant group in control. However, genera Methanobacterium , Methanosaeta , LNJC_g , and Methanomassiliicoccus were abundant in AC, and more in ACPB and ACEF ( Figure 3 b). As in the bacterial community,\nthe archaeal community in AC was similar to those in ACPB and ACEF,\nbut significantly different from that in control ( Figure 4 b). At the species level, Methanocorpusculum_uc and Methanocorpusculum\nlabreanum were the predominant archaea in control.\nHowever, the five archaeal species, including Methanobacterium\npalustre , Methanomassiliicoccus_uc , Methanosaeta concilii , LNJC_s , and Methanobacterium subterraneum , were abundant in common in AC, ACPB, and ACEF. It is well known\nthat M. concilii is an acetoclastic\nmethanogen, and the other archaeal species are hydrogenotrophic methanogens. 5 , 10 However, these five archaeal species seem to be EMAs or syntrophic\nmicrobes involved in DIET commonly observed in bioelectrochemical\nanaerobic systems. 5 , 10 , 35 , 36 , 45 This suggests\nthat the activated carbon enriches its surface with EMA, and the electrostatic\nfield selects the microbial species to drive DIET for methane production. Implications The performance and stability of anaerobic\ndigestion for organic matter significantly depend on the IIET, such\nas IIET and DIET, between acidogenic bacteria and methanogenic archaea. 1 , 2 In conventional anaerobic digestion, the main electron transfer\npathway for methane production is IIET, in which various enzymatic\nreactions are involved. 2 , 5 , 29 Control\nis a conventional anaerobic digestion reactor. The methane yield of\ncontrol based on the removed COD was as small as 209.8 mL/g COD r . This indicates that during IIET for methane production,\na significant amount of electrons are lost. 2 , 5 , 7 It is well known that compared to\nIIET, DIET between EEB and EMA better conserves electrons and produces\nmore methane. 7 , 29 , 36 Conductive materials, polarized electrodes, and direct electrical\nconnection between EEB and EMA have been found to induce DIET. 5 , 25 , 38 In AC, the methane yield was\n254.6 mL/g COD r , which was higher than in control. AC is\nan anaerobic reactor that contains powdered activated carbon as a\nconductive material. This indicates that the activated carbon enriched\nEAM on its surface and mediated cDIET, consistent with the previous\nstudies. 37 , 38 The contribution of DIET to methane production\nin AC was 35.5% or more depending on the electron conversion (%) of\nDIET, but the electron loss in AC was still high at 27.3% ( Table 1 ). In anaerobic digestion,\nit has been thought that EMA improves methane production by reducing\ncarbon dioxide with electrons transferred through conductive materials. 8 , 26 , 38 However, it revealed that the\nexoelectrogenic activity of microorganisms also contributes significantly\nto acetate dismutation for methane production in anaerobic digestion\nwith conductive materials. 17 − 19 It is believed that the methane\nyield in AC could be further improved by understanding the cDIET mechanism\nfor methane production. However, how conductive materials enrich EAM\nand mediate cDIET is still not well explained. Herein, a local polarization\nof conductive materials was proposed to describe the driving mechanism\nof cDIET. In MECs, EEB oxidizes low molecular organics on the\npolarized bioanode\nsurface. 5 , 15 , 25 The electrons\nmove to the biocathode through the external circuit and combine with\ncarbon dioxide to produce methane. ACPB is a kind of single-chamber\nMECs with activated carbon in the bulk solution. In ACPB, the methane\nyield was 310.7 mL/g COD r , significantly higher than that\nof AC, and the contribution of DIET to methane production was 80.1%\nor more ( Table 1 ).\nIt is considered that activated carbon-induced cDIET and polarized\nbioelectrodes-driven eDIET in ACPB have improved the methane yield.\nHowever, in the anaerobic reactor with polarized electrodes, EAMs,\nincluding EEB and EMA, were abundant in the bulk solution and the\npolarized electrode surface. 5 , 29 bDIET also appears\nto have contributed to the high DIET in ACPB. However, MECs, including\nACPB, have some limitations in their practical applications. 36 First, the polarized bioelectrode continuously\nconsumes electric power to maintain the Faradaic current for methane\nproduction. Second, organic matter oxidation and carbon dioxide reduction\nfor methane production are expected only on the bioanode and biocathode\nsurfaces, respectively. This indicates that the DIET for methane production\ndepends on the surface area of polarized bioelectrodes. However, the\nelectrode with sufficient area can interfere with agitation and increase\nthe initial capital cost for bioelectrochemical systems. Third, polarized\nbioelectrodes can quickly deteriorate in the anaerobic digestion environment,\nincreasing the maintenance cost to replace the electrode periodically. In the case of ACEF, the electrodes were insulated by coating the\nsurface with a dielectric material. Thus, the electrode polarization\ncreates the electrostatic field in the bulk solution, but there is\nno Faradaic current through the electrode surface. The only DIETs\nthat can be expected in ACEF are cDIET and bDIET. Interestingly, the\nmethane yield in ACEF was 317.3 mL/g COD r , higher than\nthat in ACPB, and the contribution of DIET to methane production was\n85.3% or more ( Table 1 ). The electron transfer loss of eDIET appears to have been more\nsignificant than that of cDIET or bDIET. In the case of ACPB, the\nelectrode or wire connection part involved in eDIET may have high\nohmic resistance. 41 This means that the\nelectrostatic field further improved cDIET and bDIET selectively in\nthe bulk solution, consistent with the previous studies. 35 − 37 The electrostatic field-driven DIET can be explained by the Lorentz\nforce that moves electrons, polar molecules, and charged ions. 28 However, in redox reactions, the electron transfer\nunder electrostatic fields can also be described based on thermodynamics. 46 The equilibrium constant (K) for a redox reaction\ndepends on the free energy change (Δ G = − RT × ln K , where R is the ideal gas constant, and T is the absolute\ntemperature). The free energy ( G ) is a function of\nenthalpy ( H ), entropy ( S ), and temperature\n( T ) ( G = H – TS ). The electrostatic field changes the enthalpy and entropy\nby changing the molecular polarity, bridging the charged particles,\nor oscillating the permanent dipole. 46 − 48 This suggests that the\nelectrostatic field reduces the activation energy required to be the\ntransient state, providing the kinetic advantages to the electron\ntransfer for the redox reaction, promoting DIET. 48 , 49 In AC, the DIET rate, estimated from the maximum methane production\nrate, was 1.80 × 10 –7 e – moles/s\n( Table 1 ). However,\nthe polarized bioelectrode increased the DIET rate by 2.89 times compared\nto AC, and the electric field in ACEF further improved by 3.85 times.\nIn anaerobic digestion, the electrostatic field-promoted cDIET and\nbDIET have several advantages over eDIET via polarized bioelectrode.\nFirst, the electrostatic field can be created with relatively small\nsurface areas of polarized electrodes. Second, the electrode surface\ncan be coated with a durable dielectric polymer to extend its life.\nThis means that the electrostatic field in the bulk solution can be\nsomewhat free from the limitations due to electrode size or electrode\nmaterial. Third, the surface-coated electrodes do not directly consume\nDC electric power to create the electrostatic field. However, electrostatic\nfields further increase the contribution of DIET to methane production\ncompared to polarized bioelectrodes. It is noted that as the DIET\ncontribution for methane production increases, the methane yield and\nsubstrate conversion increase, significantly improving the overall\nmethane production. Therefore, the electrostatic field-promoted DIET\ncan be a new bioelectrochemical platform that economically improves\nanaerobic digestion performance."
} | 8,568 |
36194263 | PMC9592645 | pmc | 1,423 | {
"abstract": "Abstract Bioleaching of metal sulfides is performed by diverse microorganisms. The dissolution of metal sulfides occurs via two chemical pathways, either the thiosulfate or the polysulfide pathway. These are determined by the metal sulfides’ mineralogy and their acid solubility. The microbial cell enables metal sulfide dissolution via oxidation of iron(II) ions and inorganic sulfur compounds. Thereby, the metal sulfide attacking agents iron(III) ions and protons are generated. Cells are active either in a planktonic state or attached to the mineral surface, forming biofilms. This review, as an update of the previous one (Vera et al., 2013a), summarizes some recent discoveries relevant to bioleaching microorganisms, contributing to a better understanding of their lifestyle. These comprise phylogeny, chemical pathways, surface science, biochemistry of iron and sulfur metabolism, anaerobic metabolism, cell–cell communication, molecular biology, and biofilm lifestyle. Recent advances from genetic engineering applied to bioleaching microorganisms will allow in the future to better understand important aspects of their physiology, as well as to open new possibilities for synthetic biology applications of leaching microbial consortia. Key points • Leaching of metal sulfides is strongly enhanced by microorganisms • Biofilm formation and extracellular polymer production influences bioleaching • Cell interactions in mixed bioleaching cultures are key for process optimization",
"conclusion": "Concluding remarks Bioleaching has become an established technology in the mining industry, and research on further applications has expanded over the last years (reviewed in part B). Over the last decades, the knowledge about the diversity of acidophilic iron- and sulfur-oxidizing bacteria and archaea and their interaction with sulfide minerals in terms of oxidation mechanisms, chemical pathways, the relevance of EPS, and biofilm formation has been expanded greatly. The still missing knowledge about detailed interactions of cells with minerals and among themselves on the molecular level might be gained with gene networks and evolutionary traits, high throughput proteomics, development of recombinant strains, and synthetic biology. With this knowledge, several novel bioleaching applications using acidophiles will be possible as green biotechnology.",
"introduction": "Introduction\n The application of bioleaching of metal sulfides (MS) and its understanding have evolved over the last decades. The mobilization of metal cations from often almost insoluble minerals in ores by biological acidification, oxidation, and complexation processes is referred to as bioleaching, and its application is termed biomining, being now a worldwide established geobiotechnological process. Biomining is mainly employed for copper, cobalt, nickel, zinc, gold, and uranium. These are extracted either from insoluble sulfides or—in the case of uranium—from oxides. For gold and silver recovery from refractory ores, the activity of leaching microorganisms is applied only to dissolve metal sulfides like arsenopyrite bearing the precious metals prior to cyanidation treatment. For this process, the term bio-oxidation is used because the solubilized metals such as iron and arsenic are not of economic value. The term biomining covers both applied bioleaching and bio-oxidation (Schippers et al. 2014 ; Johnson 2015 ; Kaksonen et al. 2018 ; Johnson et al. 2022 ). Recent developments in the fields of molecular biology, “omics” techniques, chemical analysis, biofilm research, and nanotechnology have contributed to an improved understanding of this bioprocess. Nevertheless, which processes are actually occurring at the molecular scale at microbe-mineral interfaces is still not fully known. Unwanted leaching of metal sulfide-containing ores by leaching microorganisms generates acid mine/rock drainage (ARD/AMD). Improved AMD countermeasures can be developed only if the microbe-mineral interactions are understood thoroughly. The (bio)chemical fundamentals of the leaching reactions have been the subject of intensive research in the last decades. In this context, the sulfur chemistry behind the leaching mechanisms has been widely understood (Schippers 2004 ; Vera et al. 2013a ). The “indirect mechanism,” i.e., the non-enzymatic metal sulfide oxidation by iron(III) ions combined with an enzymatic (re)oxidation of the resulting iron(II) ions, is well accepted to explain bioleaching. In addition, two bioleaching modes exist: “contact” and “non-contact” leaching (Sand et al. 2001 ; Rawlings 2002 ). Non-contact leaching is basically exerted by planktonic microorganisms, which oxidize iron(II) ions in solution. The resulting iron(III) ions get into contact with a mineral surface, where they are reduced, and the sulfide moiety is oxidized. Thus, iron(II) ions can enter the cycle again. In a strict sense, this represents the previously designated indirect mechanism (Sand et al. 1995 ). Contact leaching takes into account that cells attach to the surface of sulfide minerals. This means that the electrochemical processes resulting in the dissolution of sulfide minerals take place at the interface between the microbial cell and the mineral sulfide surface. This space is filled with extracellular polymeric substances (EPS), a mixture of polysaccharides, proteins, lipids, and nucleic acids. However, even after several years of research, many open questions remain. In both, contact and non-contact leaching, the microorganisms contribute to mineral dissolution by the generation of the oxidizing agent, the iron(III) ions, and by subsequent oxidation of the released sulfur compounds arising from the metal sulfide to sulfuric acid. To avoid traces of other metals, which may cause defects/instabilities in the crystal lattice, synthetic minerals produced under rigorously defined conditions can be used for systematic studies on the mechanisms (Tributsch and Bennett 1981 ). Cell-to-cell communication systems of quorum sensing (QS) are present in some leaching bacteria and control biofilm development (Farah et al. 2005 ; Ruiz et al. 2008 ; Gonzalez et al. 2013 ), but their importance for bioleaching processes remains to be understood. Detailed knowledge of the interactions among the microorganisms in leaching environments, including elucidation of interaction mechanisms and identification of still unknown cell–cell communication signals, may be a future option for further process optimization. As a consequence, this review, which is based on the previous ones (Sand et al. 1995 , 2001 ; Rohwerder et al. 2003 ; Vera et al. 2013a ), required revision. While this mini-review part A focusses on microbiology, biofilm formation and bioleaching mechanisms, part B provides an overview on biomining enriched with metal production data (Roberto and Schippers 2022 )."
} | 1,718 |
32151155 | null | s2 | 1,425 | {
"abstract": "Although cyanobacteria and algae represent a small fraction of the biomass of all primary producers, their photosynthetic activity accounts for roughly half of the daily CO"
} | 43 |
31988473 | PMC7082337 | pmc | 1,426 | {
"abstract": "Anaerobic oxidation of methane (AOM) is a major biological process that reduces global methane emission to the atmosphere. Anaerobic methanotrophic archaea (ANME) mediate this process through the coupling of methane oxidation to different electron acceptors, or in concert with a syntrophic bacterial partner. Recently, ANME belonging to the archaeal family Methanoperedenaceae (formerly known as ANME-2d) were shown to be capable of AOM coupled to nitrate and iron reduction. Here, a freshwater sediment bioreactor fed with methane and Mn(IV) oxides (birnessite) resulted in a microbial community dominated by two novel members of the Methanoperedenaceae , with biochemical profiling of the system demonstrating Mn(IV)-dependent AOM. Genomic and transcriptomic analyses revealed the expression of key genes involved in methane oxidation and several shared multiheme c -type cytochromes (MHCs) that were differentially expressed, indicating the likely use of different extracellular electron transfer pathways. We propose the names “ Candidatus Methanoperedens manganicus” and “ Candidatus Methanoperedens manganireducens” for the two newly described Methanoperedenaceae species. This study demonstrates the ability of members of the Methanoperedenaceae to couple AOM to the reduction of Mn(IV) oxides, which suggests their potential role in linking methane and manganese cycling in the environment.",
"introduction": "Introduction Methane is a greenhouse gas that is ~28 times more potent than carbon dioxide [ 1 ], and its fate in environmental systems has important implications for Earth’s climate. Anaerobic oxidation of methane (AOM) is a globally important microbiological process that prevents the atmospheric release of a substantial proportion of the methane from natural sediments (>85% in some environments) [ 2 – 4 ]. Several archaeal lineages within the class Methanomicrobia have been shown to mediate AOM, including the ANME-1, ANME-2a-c, Methanoperedenaceae (formerly known as ANME-2d), and the ANME-3. In the absence of a pure culture, “omic” and single cell visualization approaches have demonstrated the ability of different anaerobic methanotrophic archaea (ANME) lineages to oxidize methane coupled to the reduction of sulfate, in concert with a syntrophic partner (ANME-1, 2a–c, 3) [ 5 – 8 ], and nitrate ( Methanoperedenaceae ) [ 9 ]. In addition to these electron acceptors, several environmental studies have provided evidence for the potential for AOM coupled to the reduction of iron (Fe(III)) and manganese (Mn(IV)) in marine and freshwater environments based on geochemical measurements [ 10 – 15 ]. Given the large amounts of iron (~730 Tg/yr) and manganese (19 Tg/yr) being deposited into continental margins [ 16 , 17 ], and that these metals can be oxidized and reduced up to 300 times before burial [ 18 ], AOM coupled to metal reduction could represent an important global methane sink [ 15 ]. The ability of microorganisms to mediate AOM coupled to the reduction of iron and manganese oxides was first demonstrated in incubation experiments with marine sediments [ 15 ]. Subsequent analyses of these sediment incubations, using fluorescence in situ hybridization coupled to secondary ion mass spectrometry (FISH-SIMS), identified archaeal populations that were morphologically similar to ANME-2 as being active and likely responsible for the observed Mn-driven AOM [ 19 ]. Similar FISH-SIMS studies have also demonstrated the ability of marine sediment ANME-2 populations to couple AOM to iron reduction [ 6 ]. In addition, freshwater sediment bioreactors dominated by a member of the genus “ Ca . Methanoperedens sp. MPEBLZ” (within the family Methanoperedenaceae ) were shown to exhibit AOM activity during short term incubations (3 days) when nitrate was substituted for either Fe(III) or Mn(IV) oxides [ 20 ]. More recently, “ Ca . Methanoperedens ferrireducens” was enriched in a long-term culture shown to couple AOM to Fe(III) reduction [ 21 ]. Based on meta -omic analysis, “ Ca . M. ferrireducens” was hypothesized to oxidize methane using a unique set of multiheme cytochromes (MHCs) for extracellular dissimilatory Fe(III) reduction [ 21 ]. Despite preliminary evidence for AOM coupled to Mn(IV) reduction [ 15 , 20 ], long-term performance and mass balance data, combined with a detailed understanding of the microbial community and pathways responsible for this metabolism are still lacking. To assess the potential for Mn(IV)-dependent AOM, a bioreactor fed with methane and Mn(IV) oxides in the form of birnessite was operated for 480 days. Bioreactor performance data and meta -omic analysis was used to identify two novel members of the Methanoperedenaceae capable of AOM coupled to Mn(IV) reduction and the likely metabolic strategies they employ to perform this metabolism.",
"discussion": "Results and discussion Establishment and long-term performance of a Mn-dependent AOM bioreactor Biomass collected from a bioreactor performing Fe(III)-dependent AOM [ 21 ], originally seeded from freshwater reservoir sediment, was used to inoculate a newly established bioreactor supplied with methane, and birnessite as the electron acceptor. An increase in dissolved Mn(II) was observed within the first 50 days, indicating manganese reduction. A relatively small increase in dissolved Fe(II) was observed in the first 50 days, indicating the reduction of residual Fe(III) during that period. From day 140 to 480, methane consumption concomitant with the increase of dissolved Mn(II) (Fig. 1a ) was observed with an average rate of methane oxidation of 56.2 μmol l − 1 d −1 , which is comparable with the average of 62.9 μmol l −1 d −1 measured in the parent Fe(III)-driven AOM system (calculated from day 200 to 1100; see Supplementary Fig. 4 of Cai et al. [ 21 ]). Birnessite was pulse fed to the bioreactor periodically. A dramatic decrease in dissolved Mn(II) (Fig. 1a ) after feeding was likely due to it being adsorbed to the added birnessite, as many amorphous oxides have been shown as strong adsorbents for metal ions [ 48 , 49 ]. To assess the stoichiometry of Mn(IV)-dependent AOM in the bioreactor, Mn(II) production and methane consumption rates were measured during two intensive analyses periods starting on days 434 and 456, respectively (Fig. 2 ; Supplementary Fig. 1 ). During these periods, the average methane consumption rate was 44.5 μmol l −1 d −1 and the average Mn(II) production rate was 184.7 μmol l −1 d −1 . The ratio between the Mn(II) production rate and the methane consumption rate in each analysis period (4.3, days 434–455; 4.0 days 456–479) was close to the calculated stoichiometric ratio of 4:1 (Eq. ( 2 )), demonstrating that AOM was most likely coupled to Mn(IV) reduction in the bioreactor. Based on the in situ conditions of the reactor, AOM coupled to birnessite (simplified as MnO 2 in Eq. ( 2 )) was estimated to yield a potential free energy of ΔG = −383 kJ mol −1 CH 4 . 2 \\documentclass[12pt]{minimal}\n\t\t\t\t\\usepackage{amsmath}\n\t\t\t\t\\usepackage{wasysym} \n\t\t\t\t\\usepackage{amsfonts} \n\t\t\t\t\\usepackage{amssymb} \n\t\t\t\t\\usepackage{amsbsy}\n\t\t\t\t\\usepackage{mathrsfs}\n\t\t\t\t\\usepackage{upgreek}\n\t\t\t\t\\setlength{\\oddsidemargin}{-69pt}\n\t\t\t\t\\begin{document}$${\\rm{CH}}_{4\\left({\\rm{aq}}\\right) +}4{\\rm{MnO}}_{2({\\rm{s}})} + 7{\\rm{H}}^ + \\to {\\rm{HCO}}_{3({\\rm{aq}})}^ - + 4{\\rm{Mn}}^{2 + } + 5{\\rm{H}}_2{\\rm{O}}.$$\\end{document} CH 4 aq + 4 MnO 2 ( s ) + 7 H + → HCO 3 ( aq ) − + 4 Mn 2 + + 5 H 2 O . The potential contribution of other electron acceptors to AOM appears to be negligible. Nitrate and nitrite were consistently low (<0.01 mmol), and residual sulfate introduced with the inoculum remained stable in the bioreactor over the course of operation (Supplementary Fig. 2 ). Substantial Fe(III)-driven AOM is also unlikely given the main product would be unreactive Fe(II)-carbonates (FeCO 3 ; accounting for >95% of the Fe(II) produced in the parent bioreactor [ 21 ]) and total Fe(II) remained stable during the mass balance analyses periods (Fig. 2 ). These data collectively support Mn(IV)-dependent AOM as the primary process in the bioreactor. Genome recovery and community structure To obtain genomes representing the abundant members of the bioreactor community, metagenomic sequencing was performed on biomass samples from days 26, 72, 152, 228, 314, and 405. A nonredundant set of 21 high-quality MAGs (>77% complete and <5% contamination, based on CheckM) (Supplementary Table 2 ) was recovered from the metagenomes. The taxonomic affiliation of these MAGs was assessed with archaeal and bacterial genome trees constructed using 122 and 120 conserved single-copy marker genes, respectively (Fig. 3 ; Supplementary Fig. 3 ). Taxonomic affiliation of the MAGs was also assessed using GTDB-Tk, which classifies genomes based on their placement in a reference genome tree, relative evolutionary distance, and FastANI distance. This classification was consistent with the genome trees (Supplementary Table 3 ). Two MAGs (>99% complete and <5% contamination) belonging to the genus “ Ca . Methanoperedens” were identified and the names “ Candidatus Methanoperedens manganicus” and “ Candidatus Methanoperedens manganireducens” are proposed based on their apparent ability to utilize manganese as an electron acceptor. Based on the archaeal genome tree, these MAGs were phylogenetically distinct from “ Ca . M. ferrireducens” and “ Ca M. nitroreducens” [ 9 , 21 ], with “ Ca . M. manganicus” forming a sub-cluster with “ Ca . M. nitroreducens sp. BLZ2” [ 50 ] and “ Ca . M. ferrireducens” [ 21 ], and “ Ca . M. manganireducens” being more closely associated with the Mizunami Methanoperedens species [ 51 ] (Fig. 3 ). The average AAI of “ Ca . M. manganicus” and “ Ca . M. manganireducens” when compared with their closest sequenced relative “ Ca . M. nitroreducens sp. BLZ2” (89.2% similar) and “ Ca . M. nitroreducens sp. IPS” (76.0% similar), respectively, indicates that both of these MAGs represent novel species within the genus “ Ca . Methanoperedens” (Supplementary Table 4 ) [ 52 ]. “ Ca . M. manganicus” and “ Ca . M. manganireducens” have an AAI of 73.3%. Fig. 3 Phylogenetic placement of the dominant “Ca. Methanoperedens” populations. Genome tree showing the phylogenetic placement of the two “ Ca . Methanoperedens” genomes. The genome tree was inferred using the maximum-likelihood method with a concatenated set of 122 archaeal-specific marker genes, and bootstrap values were calculated using nonparametric bootstrapping with 100 replicates. The “ Ca . Methanoperedens” genomes from this study are highlighted in red. Black and white dots indicate ≥90% and ≥70% bootstrap values, respectively. The scale bars represent amino acid substitutions per site. To examine the relative abundances of the MAGs over the course of reactor operation, metagenome reads from each sampling time point were mapped onto the 21 dereplicated MAGs, as well as all publicly available Methanoperedenaceae genomes (Supplementary Table 5 ). Approximately 70% of the reads from each metagenome were mapped onto this genome set for the metagenome samples taken around the time of the metatranscriptomic and mass balance analyses. A large increase in relative abundance of “ Ca . M. manganicus” was observed over time, starting from day 26 and ranging from 1.6 to 47.7%, with a gradual increase of “ Ca . M. manganireducens” from 0.01% on day 26 to 23.6% on day 314 (Supplementary Table 5 ). The potential ability of these two species to also utilize Fe(III) is unclear, noting that both “ Ca . M. ferrireducens” and “ Ca . M. manganicus” were present early in the operation of the reactor when the simultaneous reduction of Mn(IV) and residual Fe(III) was detected (Fig. 1 ). However, given neither “ Ca . M. manganicus” nor “ Ca . M. manganireducens” were detected in the seed Fe(III)-fed AOM system or the original freshwater reservoir sediment [ 21 ], and both were enriched when manganese was supplied as the electron acceptor, it seems that they have a clear preference for manganese as an electron acceptor. Conversely, a substantial decrease in relative abundance of “ Ca . M. ferrireducens,” the dominant microorganism in the parent reactor, was observed throughout the different time points starting from 23% on day 26 to 0.44% on day 405. This observation suggests that “ Ca . M. ferrireducens” is either unable to utilize manganese, or has a lower affinity for it relative to the two “ Ca . Methanoperedens” species that succeeded it, and does not make a substantial contribution to Mn-driven AOM in this system. In addition, MAGs belonging to “ Ca . Methylomirabilis” (Mn-Methylomirabilis-1), known to mediate “intra-aerobic” methane oxidation [ 24 , 53 ], and members of the family Geobacteraceae (Mn-Geobacter-1 and Mn-Geobacteraceae-1), which includes several known metal respiring species [ 54 ], were at low relative abundance (≤1.5 %) from day 152 onwards, suggesting that any contribution they made to AOM or metal reduction in the system was minimal. Analysis of 16S rRNA-based community composition profiles, generated from each metagenome using GraftM [ 29 ], showed largely congruent results with the relative abundance of the MAGs (Supplementary Table 6 ). Analysis of the anaerobic methane oxidation and energy conservation pathways in the “Ca. Methanoperedens” populations In order to identify the most active microorganisms and pathways during Mn(IV)-dependent AOM, a metatranscriptome was generated from a sample collected on day 405. A large fraction of metatranscriptomic reads mapped to the MAGs recovered from the reactor (87%; Fig. 4 ). “ Ca . M. manganicus” and “ Ca . M. manganireducens” contributed 53.1% and 23.7% of total mRNA-TPM values, respectively, while analysis of the Mn-Methylomirabilis-1, Mn-Geobacter-1, Mn-Geobacteraceae-1 MAGs revealed low transcriptomic expression (Fig. 4 ; Supplementary Dataset 1 ). These results suggest that the “ Ca . Methanoperedens” populations were the most transcriptionally active in the bioreactor at the time of sampling and collectively responsible for the bulk of the observed Mn(IV) reduction and AOM. Fig. 4 Relative expression of genes encoding microbial metabolisms of interest for the dereplicated genome set. The total transcripts per million (TPM) was calculated for each gene. KEGG annotation was used to identify ORFs coding for the methanogenesis, oxygen respiration (M00155 + M00156), dissimilatory nitrate reduction, dissimilatory sulfate reduction, and aerobic methane oxidation pathways. ORFs coding for ≥3 CXXCH motifs were used for calculating total expression levels for MHCs for each genome. Genes involved in the microbial metabolisms of interest are included in the Supplementary Dataset 1 , Sheet 4—“Community members’ genes”. Analysis of the transcriptomic data from the “ Ca . Methanoperedens” MAGs was performed to identify the genes/pathways they use to facilitate Mn(IV)-dependent AOM. Both MAGs encode and expressed a complete “reverse methanogenesis” pathway and genes encoding multiple energy converting mechanisms such as the cytoplasmic and membrane-bound heterodisulfide reductase (HdrABC and HdrDE), F 420 H 2 dehydrogenases (Fpo), Na + translocating methyltransferases (Mtr), and V-type ATPase (Fig. 5 ). These enzymes allow the recycling of soluble electron carriers such as coenzyme B, ferredoxin, and co-factors F 420 that are necessary for methane oxidation, generation of a proton gradient, and production of ATP. The F 420 H 2 dehydrogenase and the membrane-bound heterodisulfide reductase (HdrDE) are hypothesized to transfer electrons from the cytoplasm into the menaquinone pool. Genes encoding the menaquinone biosynthesis pathways (Supplementary Dataset. 1 ) were also identified in the “ Ca . Methanoperedens” genomes supporting their use of menaquinone as the membrane-soluble electron carrier, consistent with other Methanoperedenaceae [ 21 , 55 ]. The annotation of a membrane-bound formate dehydrogenase for “ Ca . M. manganireducens” suggests the potential of this species to use formate as a carbon and electron source. Unlike the Methanoperedenaceae MAGs recovered from Japanese groundwater samples [ 51 , 56 ], no membrane-bound H 2 uptake NiFe hydrogenases were identified in the “ Ca . Methanoperedens” MAGs, excluding hydrogen as an electron donor for the Mn(IV) reduction observed. Fig. 5 Metabolic construction of the putative pathway for AOM coupled to Mn(IV) reduction in the “ Ca . Methanoperedens” genomes. Electrons from methane are generated through the “reverse methanogenesis” pathway and transferred into the menaquinone pool (MK/MKH 2 ) via the Fpo and Hdr complexes, which oxidize F 420 H 2 and CoM-SH + CoB-SH, respectively. Reducing equivalents are transferred via the menaquinone:cytochrome c oxidoreductases to MHCs located outside the cytoplasm to reduce the Mn(IV) oxides. Abbreviations for enzymes and co-factors: H 4 MPT tetrahydromethanopterin, MFR methanofuran, Fwd formylmethanofuran dehydrogenase, Ftr formylmethanofuran/H 4 MPT formyltransferase, Mch methenyl-H 4 MPT cyclohydrolase, Mtd F 420 -dependent methylene H 4 MPT dehydrogenase, Mer F 420 -dependent methylene-H 4 MPT reductase, Mtr Na + -translocating methyl-H 4 MPT:coenzyme M methyltransferase, Mcr methyl-coenzyme M reductase, Fpo F 420 H 2 dehydrogenase, MK menaquinone, CoB-SH coenzyme B, CoM-SH coenzyme M, Fd ferredoxin, Hdr heterodisulfide reductase, FrhB F 420 -reducing hydrogenase subunit B, Cytb b -type cytochrome, NrfD polysulfide reductase subunit D, FeS ferredoxin iron–sulfur protein. The MHCs are colored blue and a number of hemes are indicated as red diamonds. Each potential MHC represented has a TPM above the median gene TPM for each species. Genomic and transcriptomic analysis of the “ Ca . Methanoperedens” MAGs reveals that electrons generated from methane oxidation are likely transferred into Mn(IV) through multiheme c -type cytochromes (MHCs; Fig. 5 ). MHCs have been previously hypothesized to mediate electron transport from ANME to syntrophic bacteria (for ANME-1 and ANME-2a,b,c [ 5 , 8 , 57 ]) and iron oxides (for “ Ca . M. ferrireducens” [ 21 ]). In total, 43 and 25 putative MHCs were found to be encoded in “ Ca . M. manganicus” and “ Ca . M. manganireducens,” respectively (Supplementary Dataset 1 ). Twenty MHC protein families were found to be conserved in both “ Ca . Methanoperedens” MAGs, five of which were co-located with menaquinone:cytochrome c oxidoreductase gene clusters hypothesized to transfer electrons from the menaquinone pool to MHCs outside the cytoplasmic membrane (Supplementary Dataset 1 ). Interestingly, the two species showed differential expression patterns in the complement of shared MHCs during AOM coupled to Mn reduction (Supplementary Dataset 1 ). In “ Ca . M. manganicus,” four out of six MHC containing oxidoreductase complexes were highly expressed, with (1) one operon encoding a noncanonical bc1/b6f complex adjacent to two 6-heme MHCs, (2) an operon encoding a NrfD-like transmembrane protein, a 4Fe-4S ferredoxin iron–sulfur protein, with 3- and 6-heme MHCs, and (3) two copies of an operon encoding a b- type cytochrome and a 6-heme MHC (Fig. 5 , Supplementary Dataset 1 ). In “ Ca . M. manganireducens,” an operon encoding a b -type cytochrome and a 6-heme MHC was the most highly expressed of five putative menaquinone:cytochrome c oxidoreductase gene clusters, which is similar to the oxidoreductase gene cluster highly expressed for “ Ca . M. ferrireducens” during Fe-driven AOM [ 21 ] (Fig. 5 , Supplementary Dataset 1 ). These bc and nrfD complexes are frequently found in other metal reducing microorganisms as key components in electron transport from the cytoplasm to the periplasm [ 58 – 60 ]. In total, 23 of the 33, and 9 of the 19, MHCs predicted to be extracellular were found to be highly expressed (above the median gene TPM for the species) by “ Ca . M. manganicus” and “ Ca . M. manganireducens,” respectively (Supplementary Dataset 1 ). It has previously been reported that extracellular transfer of electrons out of the S-layer may be mediated by MHC/S-layer fusion proteins [ 5 ]. “ Ca . M. manganicus” contained three copies of a 12 heme/S-layer protein and a 22 heme/S-layer protein, while “ Ca . M. manganireducens” contained three MHC/S-layer proteins possessing 113, 52, and 19 hemes. All but the 19 heme MHC/S-layer protein of “ Ca . M. manganireducens” were highly expressed indicating their importance for extracellular electron transfer (EET) for these species (Supplementary Dataset 1 ). This is in contrast to previous metatranscriptome analyses of “ Ca . M. ferrireducens” and ANME-2c where MHC/S-layer proteins had been suggested to be of minor importance to EET due to their relatively low expression during AOM [ 21 , 57 ]. The 113 heme MHC is the highest recorded number of hemes in a cytochrome in any microorganism, and almost double the size of the 69 heme MHC in HGW-Methanoperedenaceae-1 [ 56 ]. Several extracellular MHCs highly expressed by “ Ca . M. manganicus” (14 MHCs) and “ Ca . M. manganireducens (7 MHCs), that are not associated with a menaquinone reductase gene cluster and lack an S-layer protein domain, are potentially involved in facilitating the final electron transport step to birnessite (Fig. 5 ; Supplementary Dataset 1 ). Interestingly, no homologs of the extracellular MHCs that were highly expressed by “ Ca . M. ferrireducens” during Fe-driven AOM [ 21 ] were found in the “ Ca . M. manganicus” and “ Ca . M. manganireducens” MAGs (Supplementary Fig. 4 ), suggesting these unique MHCs may be linked specifically to reduction of ferrihydrite and not birnessite. The relatively high number and diversity of the MHCs and menaquinone:cytochrome c oxidoreductase gene clusters, encoded by the genomes of “ Ca . M. manganicus,” “ Ca . M. manganireducens” and other members of the family (Supplementary Fig. 4 ), likely provides these microorganisms with the metabolic flexibility to utilize electron acceptors with diverse redox potentials. Further work is required to determine the function of the MHCs encoded by members of the Methanoperedenaceae , including their specificity for the reduction of different metal oxides (Supplementary Fig. 4 ). In addition to the MHCs, conductive nanowire structures are proposed to be important for the transfer of electrons between marine ANME-1 and ANME-2c and their syntrophic sulfate reducing bacterial (SRB) partners [ 8 , 57 ] These structures are suggested to allow electron transfer over greater distances relative to MHCs alone [ 57 ]. Genes encoding archaellum-like proteins were highly expressed by ANME-1a and ANME-2c in consortia with SRB were suggested to encode for such conductive structures [ 57 ] and potentially also facilitate the transfer of electrons to metal oxides. The “ Ca . M. manganicus” and “ Ca . M. manganireducens” both encode genes involved in the formation of archaellum, including multiple genes encoding the major subunit flagellin ( flaB ) (Fig. 5 ; Supplementary Dataset 1 ). Two of the four flaB genes encoded by “ Ca . M. manganicus” were found to be highly expressed during Mn-driven AOM, while the five encoded by “ Ca . M. manganireducens” were not expressed (Fig. 5 ; Supplementary Dataset 1 ). The high expression of these genes in “ Ca . M. manganicus” suggest that these could be conductive appendages involved in electron transfer (Fig. 5 ). Overall, this study further highlights the metabolic versatility of the Methanoperedenaceae lineage. Based on meta -omic analysis, “ Ca . M. manganicus” and “ Ca . manganireducens,” like the iron reducing “ Ca . M. ferrireducens,” encode and express genes in the “reverse methanogenesis” pathway and multiple MHCs, suggesting their active role in electron transport and reduction of birnessite. Interestingly, these MHCs are differentially expressed despite their conservation in both the abundant “ Ca . Methanoperedens” MAGs, suggesting two different mechanisms for electron transport under Mn(IV)-reducing conditions. Further investigation is required to understand the roles of these differentially expressed MHCs and their specificity for different metal oxides. Genomic characterization of the first two Methanoperedenaceae representatives capable of sustained Mn-driven AOM expands the known metabolic diversity of the family and provides the foundation for important future studies into their environmental relevance to the global methane and manganese cycles."
} | 6,155 |
34833218 | PMC8618084 | pmc | 1,427 | {
"abstract": "Increasing environmental awareness among the general public and legislators has driven this modern era to seek alternatives to fossil-derived products such as fuel and plastics. Addressing environmental issues through bio-based products driven from microbial fermentation of synthetic gas (syngas) could be a future endeavor, as this could result in both fuel and plastic in the form of bioethanol and polyhydroxyalkanoates (PHA). Abundant availability in the form of cellulosic, lignocellulosic, and other organic and inorganic wastes presents syngas catalysis as an interesting topic for commercialization. Fascination with syngas fermentation is trending, as it addresses the limitations of conventional technologies like direct biochemical conversion and Fischer–Tropsch’s method for the utilization of lignocellulosic biomass. A plethora of microbial strains is available for syngas fermentation and PHA production, which could be exploited either in an axenic form or in a mixed culture. These microbes constitute diverse biochemical pathways supported by the activity of hydrogenase and carbon monoxide dehydrogenase (CODH), thus resulting in product diversity. There are always possibilities of enzymatic regulation and/or gene tailoring to enhance the process’s effectiveness. PHA productivity drags the techno-economical perspective of syngas fermentation, and this is further influenced by syngas impurities, gas–liquid mass transfer (GLMT), substrate or product inhibition, downstream processing, etc. Product variation and valorization could improve the economical perspective and positively impact commercial sustainability. Moreover, choices of single-stage or multi-stage fermentation processes upon product specification followed by microbial selection could be perceptively optimized.",
"conclusion": "10. Conclusions This paper reviewed approaches that are being investigated for the production and fermentation of syngas. Insight has been provided on polymer production from the component of syngas through pure and mixed microbial colonies in single-stage and/or multi-stage cultures. The number of microbial strains associated with syngas fermentation and PHA production along with their major biochemical pathways is being reviewed. Due to the limited choices of microbial strain for direct syngas fermentation to polymers and issues related to their productivity, mixed or multi-stage culture is strongly suggested. Moreover, a focus on product variation and valorization will improve the techno-economical aspect of PHA production through syngas fermentation.",
"introduction": "1. Introduction The exploitation of fossil deposits for fuels and plastics has been considered a bane by succeeding generations, as these are the concerning causes of the ongoing environmental crisis, and furthermore, it is believed that natural chaos is inevitable if this exploitation continues. Reported possibilities in exploring renewable sources to substitute these fossil-derived products could be a future endeavor. Fermentation technology has opened a window for the production of bioproducts that could be potential alternatives for fossil products. Biofuels and bioplastics from microbial catalysis are biodegradable and eco-friendly alternatives. Cellulosic and lignocellulosic raw materials, which are abundant in municipal and agricultural waste, can be utilized for these bioproducts [ 1 , 2 , 3 ]. These feedstocks are complex organics that can either undergo biochemical conversion through biocatalysis (mostly microbial catalysis) or be combusted into carbon-rich syngas. The obtained syngas could either be chemically converted to hydrocarbon fuels via a thermochemical process or the syngas could be fermented into valuable bioproducts [ 2 , 3 , 4 ]. The Fischer–Tropsch (FT) process (thermochemical conversion) and biochemical conversion, the conventional processes for lignocellulosic biomass conversion, are often associated with various limitations (discussed in Section 2 ). On the other hand, syngas fermentation addresses these limitations of biochemical and thermochemical conversion. Its flexibility towards the H 2 :CO ratio broadens its applicability to a variety of carbonaceous feedstocks and does not require pretreatment for lignocellulosic biomass [ 2 ]. As this process exploits the whole cell for catalysis, it is comparatively less sensitive to impurities, toxins, and inhibitors. Moreover, fermentation is accomplished at a mild temperature and pressure. Furthermore, microbial inclusion makes the process more substrate-specific and lessens byproduct formation. The overall operational cost could be minimized through the syngas fermentation process. In a review provided by Daniell et al. (2012), all three processes, i.e., biochemical, thermochemical, and syngas fermentation, were compared and it was concluded that the overall efficiency of syngas fermentation is higher [ 2 ]. Biochemical conversion had significant energy loss in converting lignocellulosic biomass to fermentable sugars and energy from lignin could not be captured through this process, whereas both lignin and cellulose could be converted to syngas by gasification. The gasification energy efficiency ranges from 75% to 80%, and it depends on the feedstock composition of carbon, moisture, and ash. Overall plant energy efficiency, which accounts for the energy stored in the feedstock converted to the final product, was 57% in the case of syngas fermentation. However, this overall plant energy efficiency was only 45% for the FT process. These comparative parameters point towards the syngas fermentation process being economically viable for the conversion of lignocellulosic biomass. Apart from waste utilization, syngas fermentation also provides a perspective for product diversification and valorization. Upon the optimal exploitation of microbes, various products, such as ethanol, higher alcohols, organic acids, and bio-polymers, could result from syngas fermentation [ 1 , 2 , 3 , 5 ]. Although most of the acetogenic bacteria can utilize CO and H 2 to produce alcohols and organic acids, some of them (most from the Clostridium species) can produce higher alcohols. Depending on the feed supplement and/or use of chain-elongating bacteria (e.g., Clostridium kluyveri ) in the culture, the carbon chain in both acid and alcohol could be extended [ 6 , 7 , 8 ]. Syngas could also be directly fermented to polyhydroxyalkanoates (PHA) through phototrophic bacteria such as Rhodospirillum rubrum . Moreover, fermentation effluent from the acetogenic culture could act as a suitable feedstock for polymer production by PHA-producing microbes [ 9 ]. Similarly, the type of polymer, i.e., short-chain-length ( scl ) PHA and medium-chain-length ( mcl ) PHA, could be optimized through the carbon chain length of organic acid in the fermentation medium [ 10 ]. This paper reviews various aspects of syngas fermentation for the production of fuel and polymers. A scheme for the conversion of lignocellulosic biomass to syngas fermentation products is given in Figure 1 ."
} | 1,764 |
37123371 | PMC10130579 | pmc | 1,428 | {
"abstract": "To maximize the performance and energy efficiency of Spiking Neural Network (SNN) processing on resource-constrained embedded systems, specialized hardware accelerators/chips are employed. However, these SNN chips may suffer from permanent faults which can affect the functionality of weight memory and neuron behavior, thereby causing potentially significant accuracy degradation and system malfunctioning. Such permanent faults may come from manufacturing defects during the fabrication process, and/or from device/transistor damages (e.g., due to wear out) during the run-time operation. However, the impact of permanent faults in SNN chips and the respective mitigation techniques have not been thoroughly investigated yet. Toward this, we propose RescueSNN, a novel methodology to mitigate permanent faults in the compute engine of SNN chips without requiring additional retraining, thereby significantly cutting down the design time and retraining costs, while maintaining the throughput and quality. The key ideas of our RescueSNN methodology are (1) analyzing the characteristics of SNN under permanent faults; (2) leveraging this analysis to improve the SNN fault-tolerance through effective fault-aware mapping (FAM); and (3) devising lightweight hardware enhancements to support FAM. Our FAM technique leverages the fault map of SNN compute engine for (i) minimizing weight corruption when mapping weight bits on the faulty memory cells, and (ii) selectively employing faulty neurons that do not cause significant accuracy degradation to maintain accuracy and throughput, while considering the SNN operations and processing dataflow. The experimental results show that our RescueSNN improves accuracy by up to 80% while maintaining the throughput reduction below 25% in high fault rate (e.g., 0.5 of the potential fault locations), as compared to running SNNs on the faulty chip without mitigation. In this manner, the embedded systems that employ RescueSNN-enhanced chips can efficiently ensure reliable executions against permanent faults during their operational lifetime.",
"conclusion": "6. Conclusion We propose RescueSNN, a novel methodology for mitigating permanent faults in SNN chips. RescueSNN leverages the fault map of compute engine to perform fault-aware mapping for SNN weights and operations, and employs efficient hardware enhancements for the proposed mapping technique. The results show that RescueSNN improves the SNN fault tolerance without retraining. As a result, faulty SNN chips can be rescued and used for reliable SNN processing during their operational lifetime.",
"introduction": "1. Introduction In recent years, SNNs have shown a potential for achieving high accuracy with ultra-low power/energy consumption due to their sparse spike-based operations (Putra and Shafique, 2020 ). Moreover, SNNs can perform unsupervised learning with unlabeled data using spike-timing-dependent plasticity (STDP), which is highly desired for real-world applications (e.g., autonomous agents like UAVs and robotics), especially due to two reasons: these systems are typically subjected to unforeseen scenarios (Putra and Shafique, 2021b , 2022a ); and gathering unlabeled data is cheaper than labeled ones (Rathi et al., 2019 ). An SNN architecture supporting unsupervised learning is shown in Figure 1A . To maximize the performance and energy efficiency of SNN processing, specialized SNN accelerators/chips are employed (Painkras et al., 2013 ; Akopyan et al., 2015 ; Davies et al., 2018 ; Frenkel et al., 2019 ). However, these SNN chips may suffer from permanent faults, which can occur during: (1) chip fabrication process due to manufacturing defects, as fabricating an SNN chip with millions-to-billions of nano-scale transistors with 100% correct functionality is difficult, and even worsen due to the aggressive technology scaling (Hanif et al., 2018 , 2021 ; Zhang et al., 2018 ); and (2) run time operation due to device/transistor wear out and damages, that are caused by Hot Carrier Injection (HCI), Bias Temperature Instability (BTI), electromigration, or Time Dependent Dielectric Breakdown (TDDB) (Radetzki et al., 2013 ; Werner et al., 2016 ; Hanif et al., 2018 , 2021 ; Baloch et al., 2019 ; Mercier et al., 2020 ). Figure 1 (A) An SNN architecture that achieves high accuracy in unsupervised learning scenarios, i.e., a single-layer fully-connected (FC) network (Putra and Shafique, 2020 ). (B) Permanent faults in the weight memory of the SNN compute engine may exist in form of stuck-at 0 and stuck-at 1 faults. Permanent faults can affect the functionality of the compute engine of SNN accelerators/chips, including the local weight memory/registers and neurons, by corrupting the weight values and neuron behavior (i.e., membrane potential dynamics and spike generation), thereby degrading the accuracy, as shown in Figure 2 . For instance, permanent faults can change the weight values through stuck-at 0 and 1, as shown in Figure 1B . Simply stopping the executions on faulty chips at run time will lead to a short operational lifetime, while discarding the faulty chips at design time will lead to low yield and increase the per-unit cost of the non-faulty chip. Therefore, alternate low-cost solutions for mitigating permanent faults in the SNN compute engine 1 \n are required . These solutions will prolong the operational lifetime of SNN chips. Moreover, such solutions also increase the applicability of wafer-scale chips for SNNs where embracing permanent faults is important to maintain the yield. Figure 2 (A) The typical SNN accelerator architecture employs crossbar-based synaptic connections (Basu et al., 2022 ). Synapses and neurons can be affected by permanent faults, whose detailed discussion is provided in Section 2.2. (B) The stuck-at faults in the local weight memory (synaptic weight registers) can decrease accuracy. Targeted Problem: \n How can we efficiently mitigate permanent faults in the SNN compute engine (i.e., the local weight registers and neurons) on the accuracy, thereby improving the SNN fault tolerance and maintaining the throughput . The efficient solution to this problem will enable reliable SNN executions on faulty chips without the need for retraining for energy-constrained embedded systems, such as IoT-Edge devices and autonomous agents. 1.1. State-of-the-art and their limitations Besides discarding the faulty chips, the standard VLSI fault tolerance techniques like Dual Modular Redundancy (DMR) (Vadlamani et al., 2010 ), Triple Modular Redundancy (TMR) (Lyons and Vanderkulk, 1962 ), and Error Correction Code (ECC) (Sze, 2000 ), may be used for mitigating permanent faults. However, they require extra (redundant) hardware and/or executions which incur huge area and energy overheads. State-of-the-art works have studied different design aspects for understanding faults in SNNs and devising mitigation techniques, as follows. SNN fault modeling: Possible faults that can affect an SNN have been identified in Vatajelu et al. ( 2019 ). In the analog domain, fault modeling for analog neuron hardware and its fault tolerance strategies have been investigated in El-Sayed et al. ( 2020 ) and Spyrou et al. ( 2021 ), which are out of the scope of this work as we target SNN implementation in the digital domain. SNN fault tolerance: Previous works studied the impact of faults on SNN weights without considering the underlying hardware architectures and processing dataflows (Schuman et al., 2020 ; Rastogi et al., 2021 ) and each discussing a specific fault, like bit-flip or synapse removal. Recent works devised mitigation techniques for faults in the weight memories of an SNN hardware (Putra et al., 2021a , b , 2022a ), while work of Putra et al. ( 2022b ) aimed at addressing transient faults. In summary, the state-of-the-art works still focus on permanent fault modeling and injection only on the weight memories of SNN hardware. Hence, the impact of permanent faults in the SNN compute engine (i.e., synapses and neurons) and the respective fault mitigation techniques, especially with a focus on avoiding retraining costs, are still unexplored . To study the challenges of mitigating permanent faults, we perform an experimental case study in Section 1.2. 1.2. Case study and research challenges In this case study, we consider an SNN accelerator architecture in Figure 2A . We assume all neurons are not faulty, and inject permanent faults (i.e., stuck-at 0 or 1) on the weight registers with random distribution and different rates of faulty memory cells to see the significance of faulty registers on accuracy. Details on the experimental setup are discussed in Section 4. From the experimental results in Figure 2B , we make the following key observations. Classification accuracy decreases as the rate of faulty memory cells increases for both stuck-at 0 and stuck-at 1 scenarios, thereby showing the negative impact of permanent faults in the synapses. In the stuck-at 0 case, the stored weight value will either stay the same or decrease from the original value. In the case of decreased weight value, the corresponding neuron will require more stimulus (input spikes) to increase its membrane potential and reach the threshold potential for generating a spike, which represents recognition of a specific class. However, in an SNN model, multiple neurons may be responsible to recognize the same class. Therefore, if the neuron with faulty weight bits cannot recognize the input class, then other neurons may recognize it. As consequence, the accuracy degradation caused by stuck-at 0 in memory cells is relatively small and negligible in some cases. In the stuck-at 1 case, the stored weight value will either stay the same or increase from the original value. In the case of increased weight value, the corresponding neuron will require less stimulus (input spikes) to increase its membrane potential and reach the threshold potential for generating a spike, which represents recognition of a specific class. Therefore, this neuron may become more active to generate more spikes for any input classes, which leads to more misclassification. As consequence, the accuracy degradation caused by stuck-at 1 in memory cells is more significant/noticeable than the stuck-at 0 case. Combinations of fault types and fault rates lead to different accuracy, which represents different fault patterns in real-world chips, rendering it unpredictable at design time. Based on these observations, we outline the following research challenges to devise an efficient solution for the targeted problem. The mitigation technique should not employ retraining , as retraining needs huge compute/memory costs, processing time, and a training dataset that may not be available in certain cases due to the restriction policies. Moreover, retraining is not a scalable approach considering the large number of fabricated chips, as it needs to consider a unique fault map from each chip thereby retraining per chip. Note, the fault map information can be obtained through the standard wafer/chip test procedure after fabrication, hence this test does not introduce new cost and only incurs a typical cost for chip test (Xu et al., 2020 ; Fan et al., 2022 ). The mitigation should have minimal performance/energy overhead at run time as compared to that of the baseline design without fault mitigation technique, thereby making it applicable for energy-constrained embedded systems. The technique should not avoid the use of faulty SNN components (i.e., synapses and neurons) , as it means omitting the entire computations in the respective columns of the SNN compute engine, which leads to throughput reduction. SNNs require a specialized permanent fault mitigation technique as compared to other neural network computation models (e.g., deep neural networks), since SNNs have different operations and dataflows. 1.3. Our novel contributions To address the above challenges, we propose RescueSNN, a novel methodology that enables reliable executions on SNN accelerators under permanent faults . To the best of our knowledge, this work is the first effort that mitigates permanent faults in the SNN accelerators/chips. Following are the key steps of the RescueSNN methodology (the overview is shown in Figure 3 ). Figure 3 Overview of our novel contributions. Analyzing the SNN fault tolerance to understand the impact of faulty components (i.e., synapses and neurons) on accuracy considering the given fault rates. Employing the fault-aware mapping (FAM) techniques to safely map SNN weights and neuron operations on faulty compute engine, thereby maintaining accuracy and throughput. Our FAM techniques leverage the fault map of the compute engine to perform the following key mechanisms. Mapping the significant weight bits on the non-faulty memory cells of the synapses (weight registers) to minimally pollute/change the weight values. Selectively employing faulty neurons that do not cause significant accuracy degradation at inference, based on the behavior of their membrane potential dynamics and spike generation. Devising simple hardware enhancements to enable efficient FAM techniques. Our enhancements shuffle the weight bits from the synapses by employing simple combinational logic units, such as multiplexers, so that these weight bits can be used for SNN computations. Key Results: We evaluate our RescueSNN methodology using Python-based simulations (Hazan et al., 2018 ) on a multi-GPU machine. The experimental results show that our RescueSNN improves the SNN accuracy without retraining by up to 80% and 47% on the MNIST and the Fashion MNIST respectively, from the SNN processing without fault mitigation.",
"discussion": "5. Results and discussion We evaluate different design aspects including accuracy, throughput, energy consumption, and area as discussed in the following. 5.1. Maintaining accuracy Figure 14 presents the experimental results for the accuracy of different fault mitigation techniques, i.e., the baseline and our FAM-based strategies including FAM1, FAM2, and FAM3. We observe that the baseline suffers from a significant accuracy degradation as shown by ❶, because it does not mitigate faults in synapses and neurons, thereby leading to unreliable SNN executions. The significant accuracy degradation is mainly due to the fault model for faulty ‘ V mem reset' operation that makes the corresponding neuron generate spikes continuously once its membrane potential V mem reaches the threshold potential V th , thereby dominating the classification activity and leading to high misclassification. We also observe that FAM1 significantly improves the SNN fault tolerance as compared to the baseline, because FAM1 avoids the use of faulty neurons, especially for faulty ‘ V mem reset' operations, as shown by ❷. Our FAM2 improves the SNN fault tolerance even more as compared to FAM1, since FAM2 also mitigates faults in the weight registers in addition to avoiding the use of faulty neurons, as shown by ❸. Meanwhile, our FAM3 also significantly improves the SNN fault tolerance from baseline and FAM1, and obtains comparable accuracy to FAM2, since FAM3 mitigates faults in weight registers and selectively uses faulty neurons. It achieves up to 80% accuracy improvement compared to the baseline on the MNIST dataset, as shown by ❹. We also observe that the same reasons are also applicable to different workloads, thereby leading the accuracy profiles for Fashion MNIST to have similar trends to the accuracy profiles for MNIST. These results show that our FAM strategies (FAM1, FAM2, and FAM3) are effective for mitigating permanent faults in the compute engine without retraining, across different model sizes, fault rates, and workloads . Figure 14 Accuracy profiles for different mitigation techniques (i.e., baseline, FAM1, FAM2, and FAM3), different model sizes (i.e., N400, N900, N1600, N2500, and N3600), different fault rates, and different workloads: (A) MNIST and (B) Fashion MNIST. 5.2. Maintaining throughput Figure 15A presents the experimental results on the throughput of different mitigation techniques, i.e., the baseline and our FAM strategies (FAM1, FAM2, and FAM3). We observe that the baseline has the highest throughput across different model sizes and fault rates, as it uses all synapses and neurons for performing SNN executions, as shown by ❶. Meanwhile, FAM1 and FAM2 may suffer from throughput reduction because they avoid the use of faulty neurons, thereby omitting the corresponding columns of the SNN compute engine. For instance, FAM1 and FAM2 may suffer from 30% throughput reduction for N1600 with 0.1 fault rate, as shown by ❷. Meanwhile, our FAM3 can maintain the throughput close to the baseline (e.g., keeping the throughput reduction below 25% in a 0.5 fault rate), thereby improving the throughput significantly as compared to FAM1 and FAM2. The reason is that, FAM3 omits the columns of compute engine only if the corresponding neurons have faulty ‘ V mem reset' operations. For instance, FAM3 has less than 15% throughput reduction for N1600, as indicated by ❸. These results show that our FAM3 is effective for maintaining the throughput across different model sizes, fault rates, and workloads . Figure 15 (A) Throughput across different mitigation techniques, different model sizes, and different fault rates for both MNIST and Fashion MNIST, as they have a similar number of SNN weights and operations. (B) Energy consumption across different mitigation techniques, different model sizes, and different fault rates for both MNIST and Fashion MNIST, as they have a similar number of SNN weights and operations. 5.3. Energy consumption and area overheads Figure 15B shows the experimental results on the energy consumption of different mitigation techniques, i.e., the baseline and our FAM strategies (FAM1, FAM2, and FAM3). We observe that different techniques have comparable energy for small fault rates, as shown by label-❹. The reason is that small fault rates have a low probability of faulty neurons, hence the resource utilization for different techniques is similar. For large fault rates, FAM1 and FAM2 have higher energy consumption than the baseline and FAM3, as shown by label-❺. The reason is that large fault rates have a high probability of faulty neurons, hence the resource utilization for different techniques is different, i.e., FAM1 and FAM2 avoid the use of faulty neurons, thereby incurring higher compute latency and energy consumption. The baseline and FAM3 have comparable energy since FAM3 employs simple hardware enhancements: (1) multiplexing operations in each HEB which are shared for all synapses in the same column of compute engine, and (2) registers accesses in ECU, thereby minimizing the energy consumption overhead for FAM3 (i.e., within 30%). For area footprint, the original compute engine consumes around 6.27 mm 2 of area, while the one with proposed enhancements consumes around 8.56 mm 2 of area. Therefore, the proposed enhancements incur about 36.5% of area overhead, which encompasses about 36.2% of ECU and about 0.3% of HEBs. The area of ECU dominates the total area of enhancements since it mainly employs a set of 3-bit registers (i.e., 256x256 registers), which incurs a larger area as compared to HEBs (i.e., 256x25 multiplexers). These results show that our FAM3 achieves minimum overheads in terms of energy consumption and area across different model sizes, fault rates, and workloads . In summary, the above discussions show that our RescueSNN methodology can effectively mitigate permanent faults in the SNN chips without retraining . Since our RescueSNN addresses permanent faults during both the design time and the run time, it increases the yield of SNN chips, as well as enables efficient and reliable SNN executions during their operational lifetime. Furthermore, our RescueSNN also avoids carbon emission as it does not need any retraining, thereby offering an environment-friendly solution (Strubell et al., 2019 , 2020 ). 5.4. Further discussion In general, we observe that a faulty ‘ V mem reset' operation can cause significant accuracy degradation as it deteriorates the neuron from the expected behavior. The reason is that, the generated (faulty) spikes will affect how the SNN model understands the information, since an SNN model typically employs a certain spike coding scheme, i.e., rate coding in this work. Therefore, a neuron with faulty ‘ V mem reset' operation will generate a high number of spikes and dominate the classification activity, thereby leading to high misclassification and significant accuracy degradation. We also observe that, a higher number of spikes generated by faulty ‘ V mem reset' operation also indicates that the SNN model performs more frequent neuron operations that correspond to spike generation. This condition leads to higher power/energy consumption for SNN processing, which has been observed and studied in previous works (Krithivasan et al., 2019 ; Park et al., 2020 ; Putra and Shafique, 2023 ). Comparisons with Retraining Technique: In a standard chip fabrication process, manufactured chips are evaluated in a wafer/chip test procedure (i.e., wafer acceptance test and chip probing test). This test procedure aims at evaluating the quality of each chip, including any faults in the chip (Xu et al., 2020 ; Fan et al., 2022 ). In this step, the permanent faults and the corresponding fault map information from manufacturing defects are identified. Therefore, this step does not introduce new cost, and only requires a typical cost for a standard wafer/chip test procedure (Xu et al., 2020 ; Fan et al., 2022 ). In the retraining technique, the fault map information is then incorporated in the retraining process considering how the weights and neuron operations are mapped on the SNN compute engine, i.e., so-called fault-aware training (FAT). In this manner, the SNN model is expected to adapt to the presence of faults, hence maintaining high accuracy. This indicates that, the retraining technique requires (1) fault map information from the chip test procedure, and (2) a full training dataset, which may be unavailable due to restriction policy. Furthermore, each chip has a unique fault map which requires its own retraining process, thereby incurring huge time and energy costs . Otherwise, the retraining technique will not be effective. Meanwhile, our proposed FAM technique in RescueSNN methodology leverages the fault map information to safely map the weights and neuron operations on the SNN compute engine. It ensures that the SNN processing is not negatively affected by permanent faults, thereby maintaining high accuracy. Although each chip has a unique fault map which requires a specific mapping, the cost for devising the mapping strategy is significantly lower than the cost of retraining. Furthermore, our FAM technique does not require any training dataset, hence it is highly applicable to a wide range of SNN applications. Benefits and Limitations of Pruning: Neurons in the fully-connected (FC)-based SNN architecture shown in Figure 1A can be pruned while keeping the accuracy close to that of the original network, considering that a high rate of faulty ‘ V mem increase' operations does not significantly degrade accuracy. The benefits of pruning in FC-based architecture have been demonstrated in previous work (Rathi et al., 2019 ), including reduction of memory footprint and energy consumption. The pruning technique is suitable if we rely on offline training, i.e., an SNN model is trained offline with the training dataset, and the knowledge learnt from the training phase is kept unchanged during inference at run time. However, the pruning technique is not suitable if we consider SNN-based systems that need to update their knowledge regularly at run time to adapt to different operational environments (i.e., so-called dynamic environments ) such as autonomous mobile agents, e.g., unmanned ground vehicles (UGVs). The reason is that, SNN-based systems may encounter new input features in different environments and the offline-trained knowledge may not be representative for recognizing the corresponding classes, thereby leading to low accuracy at run time and requiring online training to update their knowledge (Putra and Shafique, 2021b , 2022a ). Therefore, SNN models with unpruned neurons and unsupervised learning capabilities are beneficial for learning and recognizing new features in (unlabeled) data samples from the operational environments during online training. In summary, users can select which SNN model to employ depending on the design requirements. An alternative is employing the FC-based SNNs shown in Figure 1A with/without pruning since they can enable multiple benefits, such as high accuracy when employing STDP-based learning under unsupervised settings, and efficient online training capabilities.\n\n5.4. Further discussion In general, we observe that a faulty ‘ V mem reset' operation can cause significant accuracy degradation as it deteriorates the neuron from the expected behavior. The reason is that, the generated (faulty) spikes will affect how the SNN model understands the information, since an SNN model typically employs a certain spike coding scheme, i.e., rate coding in this work. Therefore, a neuron with faulty ‘ V mem reset' operation will generate a high number of spikes and dominate the classification activity, thereby leading to high misclassification and significant accuracy degradation. We also observe that, a higher number of spikes generated by faulty ‘ V mem reset' operation also indicates that the SNN model performs more frequent neuron operations that correspond to spike generation. This condition leads to higher power/energy consumption for SNN processing, which has been observed and studied in previous works (Krithivasan et al., 2019 ; Park et al., 2020 ; Putra and Shafique, 2023 ). Comparisons with Retraining Technique: In a standard chip fabrication process, manufactured chips are evaluated in a wafer/chip test procedure (i.e., wafer acceptance test and chip probing test). This test procedure aims at evaluating the quality of each chip, including any faults in the chip (Xu et al., 2020 ; Fan et al., 2022 ). In this step, the permanent faults and the corresponding fault map information from manufacturing defects are identified. Therefore, this step does not introduce new cost, and only requires a typical cost for a standard wafer/chip test procedure (Xu et al., 2020 ; Fan et al., 2022 ). In the retraining technique, the fault map information is then incorporated in the retraining process considering how the weights and neuron operations are mapped on the SNN compute engine, i.e., so-called fault-aware training (FAT). In this manner, the SNN model is expected to adapt to the presence of faults, hence maintaining high accuracy. This indicates that, the retraining technique requires (1) fault map information from the chip test procedure, and (2) a full training dataset, which may be unavailable due to restriction policy. Furthermore, each chip has a unique fault map which requires its own retraining process, thereby incurring huge time and energy costs . Otherwise, the retraining technique will not be effective. Meanwhile, our proposed FAM technique in RescueSNN methodology leverages the fault map information to safely map the weights and neuron operations on the SNN compute engine. It ensures that the SNN processing is not negatively affected by permanent faults, thereby maintaining high accuracy. Although each chip has a unique fault map which requires a specific mapping, the cost for devising the mapping strategy is significantly lower than the cost of retraining. Furthermore, our FAM technique does not require any training dataset, hence it is highly applicable to a wide range of SNN applications. Benefits and Limitations of Pruning: Neurons in the fully-connected (FC)-based SNN architecture shown in Figure 1A can be pruned while keeping the accuracy close to that of the original network, considering that a high rate of faulty ‘ V mem increase' operations does not significantly degrade accuracy. The benefits of pruning in FC-based architecture have been demonstrated in previous work (Rathi et al., 2019 ), including reduction of memory footprint and energy consumption. The pruning technique is suitable if we rely on offline training, i.e., an SNN model is trained offline with the training dataset, and the knowledge learnt from the training phase is kept unchanged during inference at run time. However, the pruning technique is not suitable if we consider SNN-based systems that need to update their knowledge regularly at run time to adapt to different operational environments (i.e., so-called dynamic environments ) such as autonomous mobile agents, e.g., unmanned ground vehicles (UGVs). The reason is that, SNN-based systems may encounter new input features in different environments and the offline-trained knowledge may not be representative for recognizing the corresponding classes, thereby leading to low accuracy at run time and requiring online training to update their knowledge (Putra and Shafique, 2021b , 2022a ). Therefore, SNN models with unpruned neurons and unsupervised learning capabilities are beneficial for learning and recognizing new features in (unlabeled) data samples from the operational environments during online training. In summary, users can select which SNN model to employ depending on the design requirements. An alternative is employing the FC-based SNNs shown in Figure 1A with/without pruning since they can enable multiple benefits, such as high accuracy when employing STDP-based learning under unsupervised settings, and efficient online training capabilities."
} | 7,484 |
37966799 | PMC10832527 | pmc | 1,430 | {
"abstract": "Abstract Microbial electrochemical technology (MET) has proven to be a promising solution to overcome the redox and energy metabolic constraints, enabling high yields of biosynthesis beyond stoichiometric limits. While there is room for improvement in extracellular electron transfer rates and productivity of the target compounds, it is crucial to think in advance about which bioprocess could be electrified and what would face major challenges. In this opinion paper, I presented and addressed interfacial electron transfer capacity of MET, whether built on biofilm or planktonic cells, and also discussed the upper limits of the MET system for biosynthesis of chemicals accordingly. Potential future application scenarios of different MET were also briefly addressed. This opinion paper aims to encourage the community to rethink the design and development of microbial electrochemical technologies for potential future applications in industrial biotechnology."
} | 241 |
26017575 | null | s2 | 1,432 | {
"abstract": "Scalable computational modelling tools are required to guide the rational design of complex hierarchical materials with predictable functions. Here, we utilize mesoscopic modelling, integrated with genetic block copolymer synthesis and bioinspired spinning process, to demonstrate de novo materials design that incorporates chemistry, processing and material characterization. We find that intermediate hydrophobic/hydrophilic block ratios observed in natural spider silks and longer chain lengths lead to outstanding silk fibre formation. This design by nature is based on the optimal combination of protein solubility, self-assembled aggregate size and polymer network topology. The original homogeneous network structure becomes heterogeneous after spinning, enhancing the anisotropic network connectivity along the shear flow direction. Extending beyond the classical polymer theory, with insights from the percolation network model, we illustrate the direct proportionality between network conductance and fibre Young's modulus. This integrated approach provides a general path towards de novo functional network materials with enhanced mechanical properties and beyond (optical, electrical or thermal) as we have experimentally verified."
} | 310 |
36238089 | null | s2 | 1,435 | {
"abstract": "The rich structures and hierarchical organizations in nature provide many sources of inspiration for advanced material designs. We wish to recapitulate properties such as high mechanical strength, colour-changing ability, autonomous healing and antimicrobial efficacy in next-generation synthetic materials. Common in nature are non-covalent interactions such as hydrogen bonding, ionic interactions and hydrophobic effects, which are all useful motifs in tailor-made materials. Among these are biobased components, which are ubiquitously conceptualized in the space of recently developed bioinspired and biomimetic materials. In this regard, sustainable organic polymer chemistry enables us to tune the properties and functions of such materials that are essential for daily life. In this Review, we discuss recent progress in bioinspired and biomimetic polymers and provide insights into biobased materials through the evolution of chemical approaches, including networking/crosslinking, dynamic interactions and self-assembly. We focus on advances in biobased materials; namely polymeric mimics of resilin and spider silk, mechanically and optically adaptive materials, self-healing elastomers and hydrogels, and antimicrobial polymers."
} | 309 |
35835899 | PMC9283346 | pmc | 1,436 | {
"abstract": "Quorum quenching (QQ), a mechanism which inhibits, interferes or inactivates quorum sensing, has been investigated for control of biofilms instigated by quorum sensing process. Application of quorum quenchers (QQs) provides the possibility to investigate how different phenotypes of Pseudomonas aeruginosa (non-mucoid, mucoid, and heavily mucoid strains) modulate their gene expression to form biofilms, their quorum sensing (QS) mediated biofilm to be formed, and their virulence expressed. The mRNA expression of the AHL-mediated QS circuit and AHL-mediated virulence factors in P. aeruginosa was investigated in presence of QQs. qPCR analysis showed that farnesol and tyrosol actively reduce the expression of the synthase protein, LasI and RhlI, and prevent production of 3OC12-HSL and C4-HSL, respectively. Also, the use of farnesol and tyrosol significantly moderated gene expression for exo-proteins toxA, aprA, LasB , as well as rhlAB , which are responsible for rhamnolipid production. Our findings were promising, identifying several suppressive regulatory effects of furanone and Candida albicans QS signal molecules, tyrosol, and farnesol on the AHL-mediated P. aeruginosa QS network and related virulence factors. Supplementary Information The online version contains supplementary material available at 10.1007/s11274-022-03339-9.",
"conclusion": "Conclusions This study demonstrated the effect of quorum quenchers derived from C. albicans and the subsequent changes in gene expression of biofilm formation and virulence factors in P. aeruginosa . Our findings have shown promise in identifying several suppressive regulatory effects of C. albicans QS signal molecules on AHL mediated P. aeruginosa QS network and related virulence factors. It is well known that conventional antibiotics cause resistance in bacteria. This study provides a novel option for the control of P. aeruginosa biofilm related infections and diseases, as well as a foundation for future research on the mechanisms of biofilm inhibition in different P. aeruginosa phenotypes from the perspective of QQ. The QQs used in this study provide the opportunity for application of these molecules as an alternative to traditional antibiotics. Future challenges for the research include studying the effects of the QQs in vivo. This brings the knowledge obtained from this research closer to real-life application. Further research will extend the understanding of mechanistic action of the QQ. Quantification of P. aeruginosa virulence factors such as rhamnolipid, pyoverdine, pyocyanin, and elastolytic activity, as well as the use of the crystal violet method to measure biofilm formation upon treatment with quorum quenchers, may help to improve the findings.",
"introduction": "Introduction In humans, Pseudomonas aeruginosa strains can cause infections and diseases such as wound infections, periodontitis, keratitis and chronic pneumonia during cystic fibrosis (CF). These usually occur in individuals with compromised immune systems (Gellatly and Hancock 2013 ). The diversity of genetic variation amongst P. aeruginosa strains is highlighted by the presence of non-mucoid, mucoid and heavily mucoid phenotypes (Workentine et al. 2013 ). An important factor contributing toward the pathogenesis of P. aeruginosa is its remarkable ability to switch between planktonic and sessile modes of growth in the onset of chronic infections. The sessile cells, compared to planktonic cells, do not show an enhanced resistance towards antibiotics. However, they show an increased threshold in their minimum inhibitory concentration (MIC) against the antibiotic of interest when compared as they are enveloped within an extracellular matrix (Lebeaux et al., 2014 ). The basic assumption regarding the protective role played by the biofilm is that of reduced diffusion of antibiotics into the inner layers of the biofilms and their effective access to the sessile bacteria (Mah et al. 2003 ; Singh et al. 2021 ). P. aeruginosa produces numerous virulence factors that aid in its colonisation during an infection. Virulence factors such as mucoid exopolysaccharides, rhamnolipids, haemolysins, proteases, lipopolysaccharides, pili, and lipases are common tools used by P. aeruginosa to invade, adhere and colonise the host. The arsenal of virulence factors makes P. aeruginosa a potent pathogen (Schaber et al. 2004 ; Schroeder et al. 2017 ). Many processes involving virulence are modulated by the hierarchical quorum sensing (QS) circuits documented in P. aeruginosa . These traits are involved in the pathogenicity of the bacteria (Lee and Zhang 2014 ). It has been proposed that targeting the QS system of P. aeruginosa by interrupting bacterial communication instead of killing the bacteria by antibiotics would have an anti-pathogenic effect and may aid in the fight against biofilm forming pathogens and antibiotic resistance (Barlow and Nathwani 2005 ; Rémy et al. 2020 ). Conventional P. aeruginosa antibiotics such as ceftazidime, ciprofloxacin, and azithromycin display QQ activity (Skindersoe et al. 2008 ; Swatton et al. 2016 ). However, even though the mechanism is not fully understood, it has been speculated that their QQ activity involves inhibiting bacterial protein synthesis which prevents the expression of the inducer protein from synthesising the QS signal molecules (Reuter et al., 2016 ). Azithromycin is also known to inhibit alginate production in mucoid strains of P. aeruginosa (Imperi, Leoni and Visca, 2014). In a recent study, the newly synthesised P. aeruginosa quorum-sensing autoinducer analogues (AIA-1, -2) were demonstrated to boost the bactericidal efficacy of azithromycin by altering the cell surface hydrophobicity of P. aeruginosa , reducing antibiotic tolerance (Abe et al. 2021 ). The use of the said antibiotics has been known to induce resistance amongst bacteria, hence they cannot be used as potential QQs. As Acyl-Homoserine Lactones (AHLs) represent the primary QS signal molecules in P. aeruginosa , utilising small molecules that mimic AHLs to inhibit QS is a promising strategy. Furanones were first identified as QQs that inhibit by mimicking AHL molecules by attaching to the LasR receptor of P. aeruginosa . This interfered with the binding of the AHL molecule, thereby preventing QS mediated gene regulation in P. aeruginosa (Manner and Fallarero., 2018 ). Since the discovery of the role of furanone in inhibition of the QS system, numerous synthetically produced, structurally diverse furanone derivatives have been synthesised (Irie et al. 2017 ). Similarly, an AHL analogue, meta-bromo-thiolactone, competitively inhibits QS in P. aeruginosa and prevents biofilm formation and pyocyanin production and protects lung cells against the antagonistic activity of P. aeruginosa (O’Loughlin et al., 2013 ). While there are numerous environmental and physiological cues for dispersal of biofilm, chemicals produced by microorganism themselves can be utilised in a strategy to induce biofilm dispersal and inhibit QS in P. aeruginosa . Candida albicans produces two QS molecules, namely farnesol and tyrosol (Kaplan 2010 ). Farnesol was shown to inhibit the morphological shift from yeast to hyphae at high cell densities while tyrosol was shown to accelerate the transition from yeast to hyphae in C. albicans (Decanis et al. 2011 ). As P. aeruginosa and C. albicans are known to coexist in numerous nosocomial opportunistic infections (Méar et al. 2013 ; Doing et al. 2020 ), the influence of farnesol and tyrosol as potential QQs may be investigated against P. aeruginosa biofilm formation and subsequent virulence production between the non-mucoid and mucoid phenotypes along with other QQs and biofilm dispersal agents. This study investigated the effect of quorum quenching (QQ) on biofilm formation and virulence factor secretion of three strains of P. aeruginosa : P. aeruginosa NCTC 10,662 (non-mucoid), P. aeruginosa PAO1 (mucoid) and P. aeruginosa RBHi (a heavily mucoid CF isolate).",
"discussion": "Discussion In recent years, QS has been the primary focus of research involving treatment of biofilm mediated chronic infections. It has been well established that QS plays a vital role in development and formation of biofilms which is a recalcitrant mode of growth and aids in the onset of bacterial resistance towards conventional antibiotics (Pletzer et al. 2018 ). This has led to an inexorable rise in formation of superbugs-related infections that are extremely hard to treat due to the dearth of effective therapies (Fernández et al. 2011 ). Therefore, recent research has been focused on elucidating novel therapeutic strategies in order to combat biofilm-related infections by attenuating the ability of cell-to-cell communication by targeting the QS signalling system present in bacteria (Römling and Balsalobre., 2012 and Bi et al. 2021 ). This in turn would aid in arresting biofilm formation and biofilm related chronic infections. Novel therapies that target the QS system in pathogenic bacteria could provide the foundation for the development of next generation anti-virulence therapies. Three primary strategies can be adopted to combat biofilm formation by inhibiting QS. The most frequently studied strategies are degradation and modification of QS signals to prevent bacterial communication along with competitive inhibition of the receptor protein of the QS circuit as well as impeding the function of synthase protein responsible for producing the QS molecules (LaSarre and Federle., 2013 , Bhatt et al. 2021 ). In the natural environment, blocking communication of ecological niche adversary is essential for survival. This is a promising avenue to explore in order to get a better understanding of the process of social interaction of individual bacterial species or of groups aiming to dictate the niche (Bhagirath et al. 2016 ). As vital QS is for bacterial coordination and survival, it is also essential for bacteria to interfere with QS of other microbes in order to gain an advantage for survival (Li and Tian., 2012 ). This naturally existing process of communication interference to gain an advantage over competitors can be exploited to develop novel therapies targeting the QS system of pathogenic bacteria. Gene expression studies, involving the individual effect of farnesol and tyrosol, were conducted along with furanone. All the three strains showed a significant down-regulation in both the AHL mediated synthases ( LasI and RhlR ) when treated with furanone (Fig. 1 ). However, LasR receptor was up-regulated in NCTC 10,662 using farnesol and in PAO1 using farnesol and tyrosol. Similar up-regulation of RhlR receptor was observed in RBHi when treated with tyrosol. This shows that farnesol and tyrosol actively reduce the expression of the synthase protein ( LasI and RhlI ) and prevent production of N-3-oxo-dodecanoyl homoserine lactone (3OC12-HSL) and C4-HSL respectively. lasR gene is linked to lasI and is considered to be the cognate receptor for 3OC12-HS. Signal synthase, RhlI , generates C4-HSL, and the C4-HSL receptor is called RhlR (Muh et al. 2006 ). Significant down-regulation in gene expression was observed with the use of farnesol and tyrosol for exo-proteins toxA, aprA and LasB as well as rhlAB which are responsible for the production of rhamnolipid. Since the introduction of antibiotics as novel treatment against infections in the 1940s, multi drug resistance traits have become synonymous with biofilm forming opportunistic pathogens that dominate the nosocomial setting (Bryers 2008 ). This necessitates the identification and investigation of novel antimicrobial therapies and their mode of action. Exploiting interspecies and interkingdom interactions/ competition as well as naturally occurring products and their use in synergy could provide a potential solution to combat/ eradicate biofilm related chronic infections and drug resistance. Conclusions This study demonstrated the effect of quorum quenchers derived from C. albicans and the subsequent changes in gene expression of biofilm formation and virulence factors in P. aeruginosa . Our findings have shown promise in identifying several suppressive regulatory effects of C. albicans QS signal molecules on AHL mediated P. aeruginosa QS network and related virulence factors. It is well known that conventional antibiotics cause resistance in bacteria. This study provides a novel option for the control of P. aeruginosa biofilm related infections and diseases, as well as a foundation for future research on the mechanisms of biofilm inhibition in different P. aeruginosa phenotypes from the perspective of QQ. The QQs used in this study provide the opportunity for application of these molecules as an alternative to traditional antibiotics. Future challenges for the research include studying the effects of the QQs in vivo. This brings the knowledge obtained from this research closer to real-life application. Further research will extend the understanding of mechanistic action of the QQ. Quantification of P. aeruginosa virulence factors such as rhamnolipid, pyoverdine, pyocyanin, and elastolytic activity, as well as the use of the crystal violet method to measure biofilm formation upon treatment with quorum quenchers, may help to improve the findings."
} | 3,351 |
39831382 | PMC11744494 | pmc | 1,438 | {
"abstract": "Abstract Large‐scale restoration projects are an exciting and often untapped opportunity to use an experimental approach to inform ecosystem management and test ecological theory. In our $10M tidal marsh restoration project, we installed over 17,000 high marsh plants to increase cover and diversity, using these plantings in a large‐scale experiment to test the benefits of clustering and soil amendments across a stress gradient. Clustered plantings have the potential to outperform widely spaced ones if plants alter conditions in ways that decrease stress for close neighbors. Here, we test whether intraspecific facilitation improves restoration outcomes using a suite of seven high marsh species native to central California salt marshes. We also applied a biochar treatment to test whether soil amendment boosts restoration success. We compared the performance of clustered and uniform plantings across the high marsh elevation gradient for 3 years. There was a strong effect of elevation on plant performance and clear signs of plant stress related to soil conditions. Clustering slightly improved the survival of one species out of seven, although clustering did not benefit that species in a follow‐up experiment under more stressful conditions. By contrast, clustering had strong negative effects on the growth and/or cover of all species tested. The stressors in this system—likely related to compaction and soil salinity—were not mitigated by neighbors or biochar. The prevailing negative effect on seven species from distinct evolutionary lineages lends strong generality to our findings. We therefore conclude that for this and similar high marsh systems, intraspecific facilitation confers no benefits and practitioners should space plants widely to minimize competition. To take full advantage of the learning opportunities provided by large‐scale restoration projects, we recommend including experimental treatments and monitoring the response of multiple species across years to refine best practices and inform adaptive management.",
"introduction": "INTRODUCTION Restoration can greatly improve biodiversity and the provision of ecosystem services in degraded habitats (Liu et al., 2024 ; Rey Benayas et al., 2009 ), but outcomes are strongly influenced by choices made during project design (Ehrenfeld, 2000 ; Shimamoto et al., 2018 ). Once harmful sources of disturbance have been removed, resource managers must determine whether desired species can be relied upon to colonize restored habitat naturally, or whether more active efforts will be required to meet project aims (Baur, 2014 ; Chazdon et al., 2021 ; Holl & Aide, 2011 ; Meli et al., 2013 ). Practitioners often focus on dominant foundation species as a linchpin to restore critical ecosystem function and stimulate further recovery of biotic communities (Bangert et al., 2013 ; Liu et al., 2024 ; Yando et al., 2019 ), and such species may recruit easily at restored sites when propagules are plentiful and disperse readily (Armitage et al., 2006 ; Lindig‐Cisneros & Zedler, 2002 ). Yet, evidence that biodiversity can enhance ecosystem functions and services continues to accumulate (Cardinale et al., 2012 ; Rey Benayas et al., 2009 ), suggesting that practitioners should broaden the scope of restoration projects beyond the establishment of foundation species alone (Hughes et al., 2018 ). To restore diversity and function in compromised ecosystems, land managers need proven strategies that support the establishment of multiple species under real‐world project conditions. Stress is a strong driver of restoration outcomes (Bayraktarov et al., 2016 ), and even “benign” habitat may become stressful after site preparation activities. For example, large‐scale grading to restore hydrological regimes or remove resident communities creates bare habitat where erosion, soil temperatures, and evaporation rates are elevated. Because exposure stress increases with the size of a bare patch, plot‐scale experiments may not identify practices that are effective at mitigating stress on a landscape scale (Bertness, 1991 ; Zedler & Kercher, 2005 ). Yet, large‐scale restoration projects are rarely designed as experiments to shed light on underlying drivers of plant performance, or to compare the effect of different restoration treatments on stress. Biotic interactions influence stress and restoration success, and restoration practices have tended to focus on limiting competition among transplants—perhaps a legacy of ecology's early preoccupation with competition as the dominant driver of community structure (Goldberg & Barton, 1992 ; Hairston et al., 1960 ; Schoener, 1983 ). More recently, facilitation has been recognized as an important driver of community structure (Fowler, 1986 ; Stachowicz, 2001 ), with stress mediating the balance of positive and negative interactions in the Stress Gradient Hypothesis (SGH) framework (Bertness & Callaway, 1994 ). The SGH proposes that facilitation will be most important where stress is greatest, and this framework underlies much of the work demonstrating facilitation in natural communities to date (He et al., 2013 ). Based on this evidence, practitioners have been encouraged to incorporate facilitation into restoration designs for stressful systems (Halpern et al., 2007 ; Padilla & Pugnaire, 2006 ). Planting seedlings in the shelter of established “nurse plants” or clustering groups of transplants together has reduced stress and improved plant performance in some restoration projects (Duggan‐Edwards et al., 2020 ; Gómez‐Aparicio et al., 2004 ; Silliman et al., 2015 ). However, the proportion of studies that have tested whether positive interactions improve restoration outcomes is very low (Zhang et al., 2018 ). Salt marshes have strong gradients in abiotic conditions across elevation and support a limited flora, making them excellent systems to test restoration designs intended to mitigate stress for a suite of species. Studies documenting positive interactions between plants in natural marsh systems (Bertness & Hacker, 1994 ; Bertness & Leonard, 1997 ) suggest the potential value of facilitation as a marsh restoration tool. Drastic loss of salt marsh habitat also makes these systems high‐priority targets for restoration that will benefit from experiments to establish best practices (Barbier et al., 2011 ; Gedan et al., 2009 ). Several restoration projects have demonstrated that close neighbors can relieve exposure stress in intertidal habitat (Clausing et al., 2023 ), and anoxia or erosion stress in subtidal and low marsh habitat (Bos & Van Katwijk, 2007 ; Silliman et al., 2015 ). These stressors weaken at higher elevation where inundation periods and wave action are reduced, so restoration designs that incorporate facilitation may not be beneficial in the high marsh (Bertness & Ellison, 1987 ; Bertness & Hacker, 1994 ; Silliman et al., 2015 ). However, stress related to desiccation and evaporative salt concentration can increase at higher elevations—particularly in dry climates. Climate has a strong effect on plant interaction patterns and restoration outcomes, and neighbor interactions in the high marsh may play out differently in Mediterranean or arid climates compared with mesic systems (Bertness & Ewanchuk, 2002 ; Silliman et al., 2015 ). In Mediterranean high marsh, desiccation and hypersaline conditions during dry summers (Callaway et al., 1990 ; Mahall & Park, 1976 ) may be more important stressors than the tidally‐driven anoxia that is prevalent at lower elevations. In addition, most studies testing facilitation in marsh restoration design have focused on single species that are strong dominants at lower marsh elevations (e.g., Spartina or Salicornia ). Plant communities in the high marsh are more diverse (Peinado et al., 1995 ; Wasson & Woolfolk, 2011 ), and there is a need to identify restoration strategies that support the establishment of a suite of species in restored habitat. On the central California coast, several high marsh species tend to occur as patches in a Salicornia pacifica matrix, suggesting that intraspecific facilitation may help these species establish and spread. Earlier work in this system did not find that clustering improved the survival or growth of two high marsh specialists (Tanner et al., 2022 ). However, this study was carried out in a particularly mesic year, and in small‐scale plots where topsoil was retained. Outcomes may differ on a vast expanse of bare, low organic soil in a constructed marsh, especially if rainfall is limited. Soil amendments may also play an important role in the success of marsh restoration plantings. Restoration of wetlands that have been degraded by fill or by subsidence usually requires the addition or removal of sediment, which can result in substrates that differ from native marsh soils (Langis et al., 1991 ; Mendelssohn & Kuhn, 2003 ; Stagg & Mendelssohn, 2010 ). For example, soils on constructed marsh habitat can have more sand and less organic content than reference marsh, slowing the recovery of plant and invertebrate communities (McAtee et al., 2020 ). In dry climates like southern California, restored marsh soils may also become hypersaline, contributing to the failure of restoration plantings (Zedler et al., 2003 ). Soil amendments have the potential to mitigate plant stress on restored habitat, and biochar is viewed as a particularly promising soil amendment for the marsh because it has improved plant performance in other systems where drought or salinity are prominent stressors (Agegnehu et al., 2017 ; Ali et al., 2017 ; Luo et al., 2017 ). A stable form of carbon‐rich charcoal, biochar can increase the water‐holding capacity of soil and competitively bind Na+ ions, leading to lower salt concentrations in plant tissues and improving plant performance (Ali et al., 2017 ; Hammer et al., 2015 ). To date, research on biochar effects in coastal salt marsh restoration projects remains limited. In this work, we tested different restoration treatments in a landscape‐scale high marsh restoration project situated in Monterey Bay, California (USA), which has a Mediterranean climate. Growing conditions at this site may be particularly stressful due to the large bare site footprint, its high position in the tidal frame, low organic soil content, construction‐related soil compaction, and the warm, dry conditions associated with Mediterranean summer. To ensure the generality of our findings and provide insight that can inform the design of other high marsh restoration projects, we tested seven plant species, including all taxa (other than the marsh dominant, S. pacifica ) that make substantial contributions to high marsh cover in this watershed and other central California estuaries. This experiment employed over 17,000 transplants, making it one of the largest marsh restoration experiments to date, and its 3‐year duration allowed us to assess restoration treatment effects on long‐term cover as well as early survival and growth. We asked: (Q1) Does early survival vary across elevation (i.e., moisture and salinity gradients), and do close neighbors improve survival where plants performed less well? (Q2) Do close neighbors or biochar addition treatments improve (a) plant growth or (b) later survival at high versus low elevation? (Q3) Do restoration plantings with close neighbors lead to greater restoration success in terms of (a) native or (b) exotic cover? Do patterns vary across elevation and/or species? (Q4) How do rainfall and seasonal trends in soil water potential, soil percent moisture, and soil temperature influence abiotic stress gradients in this system? (Q5) During summertime drought, do soil metrics differ under plant canopies compared with the open, suggesting a potential mechanism for facilitation between neighbors? The answers to these questions will contribute to our understanding of plant interactions in the high marsh and their influence on restoration success, informing the design and adaptive management of salt marsh restoration projects.",
"discussion": "DISCUSSION Intraspecific interactions in restoration Facilitation has been advanced as an important tool to reduce stress at restoration sites, but we found little evidence of positive interactions between plants in our restored salt marsh. In our main restoration planting, we found that close neighbors slightly improved survival of a single species ( E. californica ), but this benefit was strongly outweighed by competition for space as plants grew. For the remaining species, we found that clustering did not benefit plant performance in any way, and in fact close neighbors strongly suppressed growth and cover for all five species in planted blocks, as well as growth of two additional species planted in block wings. Our supplemental restoration planting at Yampah the following year was designed as a second test of neighbor effects on E. californica under stressful conditions, but in this case, clustering conferred no survival benefit (all transplants died in clustered as well as uniform plantings). We also found no benefit of clustering on F. salina and J. carnosa performance in an earlier plot‐scale experiment at a restored lagoon in the Elkhorn Slough complex (Tanner et al., 2022 ). Given our efforts to detect intraspecific facilitation in restoration plantings that collectively cover seven species, three independent experiments in separate years, and tests at both the plot and landscape scale, we conclude that close conspecific neighbors do not improve high marsh restoration outcomes in this estuary—and likely will not benefit restoration efforts in similar systems. Our results were surprising because positive interactions between plants that mitigate stress have been demonstrated in both natural and restored systems. Once the SGH (Bertness & Callaway, 1994 ) laid out a relationship between neighbor interaction patterns and stress, this framework was validated by theoretical modeling efforts (Travis et al., 2005 , 2006 ) and empirical studies in alpine systems (Callaway et al., 2002 ; Cavieres et al., 2006 ), deserts (McAuliffe, 1986 ; Nobel, 1980 ), grasslands (Greenlee & Callaway, 1996 ), oak woodlands (Callaway, 1992 ; Callaway & D'Antonio, 1991 ), salt marshes (Bertness, 1991 ; Bertness & Shumway, 1993 ; Bertness & Yeh, 1994 ), and intertidal zones (Bertness, 1989 ; Leslie, 2005 ). The broad evidence for positive neighbor effects in a range of natural systems (He et al., 2013 ) prompted scientists to propose that facilitation should be incorporated into restoration designs for stressful habitat (Halpern et al., 2007 ; Padilla & Pugnaire, 2006 ). Positive interactions that improve restoration outcomes have previously been demonstrated in coastal systems. In a test of interspecific facilitation including three taxa studied here ( F. salina , J. carnosa , and L. californicum ), O'Brien and Zedler ( 2006 ) found that more transplants tended to survive sediment smothering and salinity stress when planted in tight clusters. Studies focused on intraspecific interactions have also found positive effects on restoration outcomes; Bos and Van Katwijk ( 2007 ) found that clustered eelgrass plantings were able to withstand greater hydrodynamic stress, and Silliman et al. ( 2015 ) demonstrated that clumping buffers anoxia and erosion stress for Spartina transplants in the low marsh. At our restoration site, we hypothesized that conspecific neighbors with the same stress tolerance and resource needs could still drive a benefit via shading effects on soil moisture and plant water relations. Yet, plants in the uniform treatment performed best, suggesting that any benefit of shading in the clustered treatment was outweighed by competition for belowground resources (e.g., water) or aboveground resources (e.g., space). As a result, intraspecific facilitation does not appear to be an effective tool for improving restoration success in this central California high marsh. Many of the species studied here are widely distributed in marsh communities on the US West Coast (Janousek et al., 2019 ), and previous work has identified strong similarities between marsh communities on the California coast and in the Mediterranean region (Peinado et al., 1995 ). Where high marsh floristic composition and climate regime are similar to the system studied here, we predict that close conspecific neighbors will suppress restoration success. Clustered plantings using different species may yield better outcomes in the high marsh, particularly when those species differ in abiotic tolerance and competitive ability (Maestre et al., 2009 ). Physical stressors affecting restoration outcomes Stress can make revegetation of restoration sites challenging, and the ability to identify and mitigate stressors is key to enhancing restoration success (Beheshti et al., 2023 ; Brooks et al., 2015 ; O'Brien & Zedler, 2006 ). Stress mitigation may be particularly important on constructed habitat, where heavy equipment can compact soil and grading creates large swathes of bare earth where exposure stress can be high (Mossman et al., 2012 ; Thomsen et al., 2022 ; Zedler et al., 2003 ). We observed clear evidence of stress in some planted blocks, and in the supplemental experiment at Yampah. In particular, early survival of E. californica was highly variable across planted blocks (ranging from 35% to 91%) and appeared to be related to local soil conditions. S. macrotheca also experienced patchy mortality, particularly in one block where 24% of transplants died during the first 10 weeks. At the Yampah site where we planted only E. californica , all transplants died during the first few weeks after planting. For effective adaptive management of this and other similar restoration projects, it is critical to characterize the stressors present and explore solutions for decreasing them (Zedler, 2017 ). We observed clear differences in plant performance across elevation, with impacts on mortality and/or growth and cover depending on species. Transplant performance generally declined downslope, and natural recruitment of S. pacifica was also lower near the bottom boundary of the high marsh (Thomsen et al., 2022 ). Although hypoxia is known to impair salt marsh plant performance (Bertness & Ellison, 1987 ; Davy et al., 2011 ; Janousek & Mayo, 2013 ), it seems unlikely to be the culprit here; lower elevations of the high marsh are inundated more frequently, but still rarely at this site because the entire marsh was constructed to sit high in the tidal frame (Fountain et al., 2019 ). However, microtopography and sediment properties could still create waterlogged conditions at unexpected elevations (Crooks et al., 2002 ). Measurements of soil water potential declined at lower elevations while soil moisture content increased, indicating that salinity may have a stronger influence than moisture on soil water potential measurements. Although this finding suggests that salinity may be an important driver of stress in this system, a study of soil salinity at this site carried out in 2019 found that soil salinities in the high marsh were not extreme (Thomsen et al., 2022 ). We were also surprised to find an apparent lack of stress at upper elevations in the high marsh, where we thought drought would have a strong influence on plant performance in this Mediterranean system. However, plentiful rainfall in late winter and spring of 2019 may have allowed transplants sufficient time to establish and become resilient to dry summertime conditions. Although soil cores spanned the full rooting zone for seedlings at the time of planting, root growth beyond that zone may have allowed transplants to tap into sources of moisture that we could not detect—including a potential influx of groundwater from the adjoining scraped hillside (Montalvo et al., 2024 ). We also observed high mortality in localized areas of the main restoration planting, particularly in low‐elevation areas with visibly compacted soil that also accumulated salt deposits during the dry season. Compaction can limit establishment of salt marsh vegetation (Callaway, 2001 ), and we speculate that soil compaction in these locations may have exacerbated flooding or salinity stress, with patchy but strong negative effects on early survival of E. californica and S. macrotheca . Subtle variations in microtopography can have a strong influence on salt marsh plant performance (Xie et al., 2019 ), and soil texture interacts with tidal regime to drive changes in soil water content and salinity that can favor or suppress particular species (Moffett et al., 2010 ). Clearly, environmental conditions near the bottom of some blocks were not favorable for survival of E. californica or S. macrotheca . At Yampah, where sediment was added to build up lost elevation, conditions appeared to be uniformly stressful, and all supplemental plantings of E. californica died. This area has remained much more bare than planted blocks on the western side of the site—and areas that have remained bare are also more saline (Pausch, 2024 ; Thomsen et al., 2022 ), further hindering plant colonization. Managers should consider actions that could mitigate these stressors in similar systems. None of the strategies that we tested at this large‐scale restoration site were effective at reducing stress. In our main restoration planting, biochar soil amendment did not improve plant survival or growth, aligning with findings from a separate sediment treatment study carried out at Elkhorn Slough and seven other reserves (Raposa et al., 2023 ). In that study, the authors found that biochar addition did not affect vegetation cover or sediment salinity but did promote drainage and oxygenation. However, such benefits may be of limited value in the high marsh, where inundation (and anoxic conditions) is relatively rare. Our clustering treatment had a mild positive effect on early survival of E. californica , but it ultimately had a strong negative effect on growth or cover of all species. These outcomes may be explained if stress is primarily driven by salinity or compaction, which is unlikely to be buffered by close neighbors or the modest soil amendment treatments applied here. However, mortality of E. californica and S. macrotheca within the first 9 weeks of planting occurred during a period of heavy rainfall, which could be expected to mitigate excess soil salinity. Moreover, given the elevated position of the constructed marsh in the tidal frame and breaching of the site in August of 2018, the high marsh had undergone relatively few cycles of inundation by the time of restoration planting. We therefore suspect that soil compaction drove stressful conditions in areas of localized mortality. Regardless of the exact mechanisms at play in constructed marsh habitat, smaller test plantings across broader areas and multiple years will mitigate the risks posed by variable weather and spatially patchy soils. Active intervention may be required to jump‐start colonization in persistently bare areas; the combination of freshwater addition and decompaction has proved effective in other restored salt marsh systems that have been slow to recover (Beheshti et al., 2023 ). Learning from large‐scale restoration experiments Large‐scale restoration projects provide invaluable opportunities to test and refine restoration practices (Zedler, 2017 ). Guidance on the most effective strategies is lacking in many systems, and this makes restoration practice inherently experimental—yet restoration projects are rarely designed as experiments (Zedler, 2005 ). Experiments that test different restoration treatments can shed light on mechanisms that govern the assembly and structure of restored communities (Beheshti et al., 2023 ; Doherty et al., 2011 ; Zedler, 2017 ). Identifying these mechanisms allows development of best practices to enhance restoration success in similar systems, and to fine‐tune these practices for particular sites—where outcomes are influenced by the specific combination of topography, abiotic characteristics, and species present. Long‐term monitoring of restoration experiments is also needed to distinguish successful strategies from unsuccessful ones (Wolters et al., 2005 )—yet monitoring is typically limited to a brief window following restoration efforts. We designed the Hester Marsh restoration project as a suite of restoration experiments coupled with long‐term monitoring on an unprecedented scale. We learned from these experiments while also revegetating a formerly bare marsh. Using multiple species lends powerful generality to our findings. We tested the effects of conspecific clustering on seven species from different plant families, including all species that make substantial contributions to high marsh cover (aside from the dominant S. pacifica , which tends to recruit well at restored sites in this system). We also tested biochar addition effects on five of these species in the main restoration planting. This approach allows us to make a clear and strong recommendation that practitioners should space plants apart to minimize competition in similar marsh restoration projects, and that biochar amendment is unnecessary. However, we still found differences in plant performance among species. F. salina and S. macrotheca were fast growers and reached >30% cover in the first 5 months, when other species remained at 10% or less cover. Once established, L. californicum appeared to flourish in this system, while none of the T. concinna transplants survived. D. spicata and J. carnosa grew very slowly and contributed little cover in the first year, but steadily increased to >20% cover on average by the third year. F. salina was clearly best suited to growing conditions at our site, consistently providing the most cover among planted species in each year and increasing in cover year over year. We recommend this species as a top performer for this and similar systems, as did a related study (Shikuzawa et al., 2024 ). Recognizing that it can be expensive and time‐consuming to incorporate monitoring into restoration projects, we tracked different metrics to assess which provided the most insight into restoration outcomes over our 3‐year period of study. Given that restoration projects often set goals that relate directly to cover, and that clustered planting designs may trade off survival and subsequent growth and cover (Duggan‐Edwards et al., 2020 ), it is important to monitor more than early survival. Cover across elevation is the most important variable to monitor for tidal marsh restoration (Wolters et al., 2005 ), and doing this just once a year is more informative for managers than focusing on time‐consuming parameters like individual plant size or survival during the first growing season. If our experiments had considered a more limited time period or spatial area, we likely would have missed strong spatial patterns in plant performance in response to elevation and local soil characteristics over time. If we had just tracked outcomes in the main restoration planting during the first growing season, we would have ranked S. macrotheca as much more successful than D. spicata or J. carnosa . However, despite excellent performance in the first growing season, S. macrotheca cover dropped drastically in subsequent years while the cover of D. spicata and J. carnosa continued to increase steadily. We also likely would have concluded that neighbor facilitation can improve restoration outcomes for E. californica —but in the long run, neighbors were not beneficial. Moreover, the supplemental E. californica planting installed in a different year and area of the restoration site suffered total mortality, showing how variable success can be. Taken together, these results highlight that it is wise to avoid drawing conclusions from restoration outcomes over limited temporal and spatial scales (Witman et al., 2015 ). Planning restoration work in phases that span different growing seasons and areas of a restoration site provides more opportunities to learn about the factors that govern restoration success (Zedler, 2017 ). Our experiments were designed to test the effectiveness of different restoration strategies over broad spatial and temporal scales that are relevant for large‐scale restoration projects. This design is important because much guidance for restoration comes from small plot‐scale experiments that focus on early performance and may not be informative for practitioners who need to operate at landscape scales. The insights gained by building well‐designed experiments and long‐term monitoring into large restoration projects can be applied to enhance outcomes of future restoration phases and to identify best practices for other, similar systems."
} | 7,259 |
30847089 | PMC6392346 | pmc | 1,439 | {
"abstract": "Abstract Understanding the factors that determine invasion success for non‐native plants is crucial for maintaining global biodiversity and ecosystem functioning. One hypothesized mechanism by which many exotic plants can become invasive is through the disruption of key plant–mycorrhizal mutualisms, yet few studies have investigated how these disruptions can lead to invader success. We present an individual‐based model to examine how mutualism strengths between a native plant ( Impatiens capensis ) and mycorrhizal fungus can influence invasion success for a widespread plant invader, Alliaria petiolata (garlic mustard). Two questions were investigated as follows: (a) How does the strength of the mutualism between the native I. capensis and a mycorrhizal fungus affect resistance (i.e., native plant maintaining >60% of final equilibrium plant density) to garlic mustard invasion? (b) Is there a non‐linear relationship between initial garlic mustard density and invasiveness (i.e., garlic mustard representing >60% of final equilibrium plant density)? Our findings indicate that either low (i.e., facultative) or high (i.e., obligate) mutualism strengths between the native plant and mycorrhizal fungus were more likely to lead to garlic mustard invasiveness than intermediate levels, which resulted in higher resistance to garlic mustard invasion. Intermediate mutualism strengths allowed I. capensis to take advantage of increased fitness when the fungus was present but remained competitive enough to sustain high numbers without the fungus. Though strong mutualisms had the highest fitness without the invader, they proved most susceptible to invasion because the loss of the mycorrhizal fungus resulted in a reproductive output too low to compete with garlic mustard. Weak mutualisms were more competitive than strong mutualisms but still led to garlic mustard invasion. Furthermore, we found that under intermediate mutualism strengths, the initial density of garlic mustard (as a proxy for different levels of plant invasion) did not influence its invasion success, as high initial densities of garlic mustard did not lead to it becoming dominant. Our results indicate that plants that form weak or strong mutualisms with mycorrhizal fungi are most vulnerable to invasion, whereas intermediate mutualisms provide the highest resistance to an allelopathic invader.",
"conclusion": "4.3 FUTURE RESEARCH DIRECTIONS AND CONCLUSIONS Our IBM indicates that the strength of the mutualism between native plants and mycorrhizal fungi may be an important factor influencing invasion success of an allelopathic plant invader such as garlic mustard. In particular, we show that plant–mycorrhizal mutualisms that are either weak or strong are less resistant to invasion than intermediate‐strength mutualisms. Though our model results may provide practical applications for identifying native plants at risk of displacement by an allelopathic plant invader, empirical research is needed to validate if these patterns are true in nature. One such study might be to establish, in a greenhouse, a gradient of plant communities with known dependencies on mycorrhizal fungi (i.e., a plant community that is highly dependent upon mycorrhizal mutualisms, a community with intermediate dependencies, and so forth). Once this gradient of plant communities is established, experimentally introducing allelopathic plant invaders would test which communities are most resistant to invasion. Another study might be to identify a plant with plastic responses to mycorrhizal fungi; for example, a plant that is less dependent on fungi in high‐nutrient conditions but more dependent in nutrient‐poor conditions. Performing experimental introductions with allelopathic invaders across different conditions of mycorrhizal dependencies would shed light on how mutualism strengths can mediate invasion success in nature. Given that many other environmental factors, such as natural enemies, may influence how plants interact with mycorrhizal fungi, conducting the above studies as field experiments would provide the strongest tests of the predictions of our IBM. Due to the increasing disruptions to ecosystems by invasive plants, this research provides timely insights into how an allelopathic invasive plant may successfully invade native plant communities. We show that plant–mycorrhizal mutualisms of intermediate strengths are most resistant to garlic mustard invasion. We also show there might be invasion thresholds even for the plants most resistant to plant invasion. Thus, forest plants that are of conservation interest should be tested for their reliance upon mycorrhizal fungi, as our model suggests their level of dependence on this mutualism will likely determine how vulnerable they are to displacement by an allelopathic invader. In addition, forest ecosystems that have experienced degradation that weakens existing plant–mycorrhizal mutualisms, or that gives a sudden numerical advantage to an allelopathic invader, are likely to be more susceptible to invasion. Hence, we call for future field studies to evaluate the “Intermediate Mutualist‐Strength Hypothesis” for plant invaders that negatively impact the mutualism between native plants and mycorrhizal fungi.",
"introduction": "1 INTRODUCTION The introduction of invasive plants into local ecosystems is one of the most significant threats to natural plant and animal populations (Moser et al., 2009 ; Vilá et al., 2011 ). Invasive plants modify ecosystems by changing vegetation structure and productivity (Asner et al., 2008 ), resulting in habitat and biodiversity loss. Invasive plants also disrupt the functioning and energy flow of a system via changes to food‐web dynamics (McCary, Mores, Farfan, & Wise, 2016 ; Smith‐Ramesh, Moore, & Schmitz, 2017 ), soil chemistry and nutrient availability (Ehrenfeld, Kourtev, & Huang, 2001 ), and disturbance regimes (Brooks et al., 2004 ; Mack & D'Antonio, 1998 ). Thus, a current goal in ecology is to understand which factors most strongly influence invasion success for non‐native plants (Hejda, Chytrý, Pergl, & Pyšek, 2015 ; Higgins & Richardson, 2014 ; Moravcová, Pyšek, Jarošík, & Pergl, 2015 ). One proposed mechanism by which plant invaders can negatively impact natives is through their disruption of key mutualisms between native plants and mycorrhizal fungi (Grove, Haubensak, Gehring, & Parker, 2017 ; Hale & Kalisz, 2012 ; Hale, Tonsor, & Kalisz, 2011 ). Mutualism disruption is an extension of the novel weapons hypothesis (Callaway et al., 2008 ; Callaway & Ridenour, 2004 ) and may explain why some invasive plants are so pervasive. Mycorrhizal fungi form important symbioses with most plants by increasing the root's ability to absorb nutrients and water available in the soil (Smith, Facelli, Pope, & Smith, 2010 ). The fungus, in return, benefits by receiving photosynthetically derived carbohydrates from the plant (Parniske, 2008 ). Because 80%–90% of all plant species associate with mycorrhizal fungi in some manner (Smith & Read, 2010 ), a non‐native plant that interferes with this interaction will likely have a competitive advantage over most native plants. For instance, several invasive plants can indirectly limit seedling and mature‐plant growth by releasing toxic fungicidal chemicals (i.e., allelochemicals) into the soil (Brouwer, Hale, & Kalisz, 2015 ; Lankau, 2012 ), which reduce fungal densities, thereby weakening the mutualism and lowering the per‐capita growth rate of the native plant (Hale, Lapointe, & Kalisz, 2016 ). Although prior research has indicated that plant–mycorrhizal disruptions can lead to invasion success for many plant invaders, previous studies have ignored how the strength of the mutualism affects resistance to invasion. Plant–mycorrhizal mutualisms function on a continuum from facultative to obligate (Johnson & Graham, 2013 ). The strength of this mutualistic interaction is influenced by several factors, including the plant species, local environmental conditions, and the presence/absence of parasitic microorganisms (Hale & Kalisz, 2012 ; Hoeksema et al., 2010 ). For example, in high‐nitrate soils, the mutualism may become parasitic, with the fungus removing photosynthates without providing any benefits to the plant (Treseder & Allen, 2002 ). In contrast, in low‐phosphorus environments, the interaction typically functions as a mutualism in which both partners benefit (Ji & Bever, 2016 ). Therefore, to fully understand how plant–mycorrhizal disruptions influence invasive success, we need to evaluate the disruption across a gradient of mutualism interaction strengths. The degree to which a given plant species depends on this mutualism should determine its susceptibility to plant invaders that release fungicides; however, creating a natural gradient to test this prediction under field or laboratory conditions presents many logistical challenges. One approach to answering this important, yet experimentally challenging, question is to use an individual‐based model (IBM). IBMs simulate individual components in a complex system, based on simple rules and treating individuals as unique and discrete entities (Grimm & Railsback, 2005 ; Railsback & Grimm, 2012 ). In IBMs, individuals (also termed “agents”) simultaneously interact with one another and with the environment (Railsback & Grimm, 2012 ). To facilitate realistic representations of complex ecological systems, agents can adapt to altered environmental conditions and therefore influence future generations of agents (Wilensky & Rand, 2015 ). Thus, IBMs provide an ideal framework to examine how a continuum of interaction strengths might affect the success of a plant invader. In this study, we present an IBM that models how mutualism strengths can influence the invasiveness of an allelopathic plant invader. We used garlic mustard ( Alliaria petiolata) as the example because it is the best‐documented case of an invasive plant indirectly affecting the growth and abundance of native plants by suppressing mycorrhizal fungi (Brouwer et al., 2015 ; Hale et al., 2011 ). It is a pervasive invader with the potential to create dense monocultures in forest understories and has been a prominent challenge for land managers since it was first introduced into North America in the 1860s (Anderson, Dhillion, & Kelley, 1996 ; Meekins & McCarthy, 1999 ; Nuzzo, 1999 ). Garlic mustard releases glucosinolates that disrupt the plant–mycorrhizal mutualism by reducing the abundance of both arbuscular (Roberts & Anderson, 2001 ; Stinson et al., 2006 ) and ectomycorrhizal fungi (Wolfe, Rodgers, Stinson, & Pringle, 2008 ). To better understand how disruptions to plant–mycorrhizal mutualisms could explain variation in garlic mustard's invasiveness, we developed an IBM to answer two questions: (a) How might the strength of the mutualism between a native plant ( Impatiens capensis ) and a mycorrhizal fungus affect the plant's resistance to garlic mustard invasion? (b) Can there be a non‐linear relationship between initial garlic mustard density and establishment? We hypothesized that strong mutualisms, that is, higher nutrient exchanges between the mycorrhizal fungus and the native plant, would exhibit the highest resistance to plant invasion. Increased nutrient exchanges should ultimately lead to higher fecundity for both fungi and plants, resulting in lower garlic mustard densities. Furthermore, we hypothesized that there would be a starting density that garlic mustard must reach before becoming invasive, which we defined as garlic mustard eventually representing more than 60% of the total plant cover.",
"discussion": "4 DISCUSSION 4.1 Effects of mutualism strength on invasion success Plant–mycorrhizal mutualisms, which are important belowground symbioses in forest ecosystems (Read, Leake, & Perez‐Moreno, 2004 ), can influence plant and animal populations (Fitter & Garbaye, 1994 ; Teste et al., 2017 ). Simulations with our IBM reveal that disruptions to this mutualism can explain invasion success for an allelopathic plant invader. Furthermore, the strength of the plant–mycorrhizal mutualism played a critical role in determining garlic mustard's ability to invade the native plant in our model. Surprisingly, and contrary to our prediction, the mutualism of intermediate strength was the most resistant to garlic mustard invasion, whereas strong and weak mutualisms were both highly susceptible. This pattern was most pronounced at high initial densities of garlic mustard. The intermediate‐strength mutualism afforded I. capensis increased fitness when the fungus was present, but the cost of not having fungi was not severe enough to markedly decrease the native's equilibrium population size. Thus, once the native plant established numerical dominance over garlic mustard, the production of offspring was sufficiently high to maintain large population sizes even when the mycorrhizal fungus had decreased to low levels in response to garlic mustard. Although the native plant and mycorrhizal fungus had higher fitness when the mutualism was strong, garlic mustard was still able to dominate the equilibrium plant community. This outcome was unexpected given that increased fecundity due to the mutualism should have made the native plant a better competitor for space. Early in the simulation, our prediction appeared to be true—the native plant dominated plant density in the first 30–40 time steps. However, once the garlic mustard population was large enough to cause a substantial drop in mycorrhizal fungal densities, the size of the native plant population rapidly declined. Thus, disruption of the symbiosis favored garlic mustard over time since fitness of the native plant without the obligate mutualist was too low to overcome competition for space with the invader. This finding supports a body of literature that shows members of obligate (i.e., strong) mutualisms are more susceptible to environmental changes (Takimoto & Suzuki, 2016 ). For example, obligate plant‐pollinator mutualists are constrained from switching partners, rendering them susceptible to extinction due to partner loss (Pellmyr, Thompson, Brown, & Harrison, 1996 ; Sachs & Simms, 2006 ). Despite the fact that in our model, in the absence of an invader, high levels of mutualism strength had a higher fitness than intermediate levels, a strong plant–mycorrhizal mutualism is more likely to become dominated by an allelopathic plant invader because the invader eventually has a too ‐ strong negative effect on the relative fitness of the symbiotic relationship. Weak plant–mycorrhizal interactions were also more susceptible to garlic mustard invasion compared to intermediate symbiotic strengths. Unlike the situation for obligate mutualisms, under weak facultative mutualisms, the presence of the fungus had only a minor effect on plant growth and reproduction; nevertheless, garlic mustard still became a numerical dominant under this condition. Two factors can explain this result. First, the fitness of both plant and fungus was not sufficient to outcompete garlic mustard, and because there is little exchange of nutrients/resources between the native plant and the fungus, the mutualism had a negligible impact on competitive ability. This lack of resource exchange effectively lowered the fitness of the native plant, leading to garlic mustard's dominance. Second, once garlic mustard establishes in the model, the native plant does not have a secondary trait (such as high seed production or fast generation times) that can overcome the fitness advantage of the plant invader. Thus, as garlic mustard colonizes more patches, the intrinsically low fitness of the native plant leads it toward extinction. 4.2 Invasion threshold for garlic mustard Even though the plant–mycorrhizal mutualism of intermediate strength was least impacted by garlic mustard invasion under certain combinations of parameters, the native plant did not always dominate garlic mustard numerically. If an environmental condition not explicitly included in our IBM were to allow garlic mustard to reach a threshold of ca. 6,000 plants per 400 m 2 (perhaps even lower, given the size of the SE's for some densities), our model predicts that garlic mustard would become numerically dominant, and perhaps could limit the abundance of the native plant even further if the simulation were to continue. This finding indicates that initial site conditions and non‐native plant densities could have profound effects on the success of a plant invader. Several studies have suggested that a habitat must be degraded or substantially altered before a non‐native plant can become established as an invasive (Dassonville et al., 2008 ; Davis, Grime, & Thompson, 2000 ; Suding et al., 2013 ). Carduus pycnocephalus (calflora), for example, became a more aggressive invasive herbaceous plant when it was sown under conditions where native mycorrhizal fungi were suppressed (Vogelsang & Bever, 2009 ). In a situation where the system has been previously degraded, an invasive plant like garlic mustard can establish itself and further erode the system through its persistent allelopathy, though for garlic mustard, the effects appear most pronounced in early invasions. Our IBM indicates that even native plants with substantial but not obligate symbioses with mycorrhizal fungi will have difficulty resisting invasion by plants such as garlic mustard if environmental conditions give invasives an initially large numerical advantage. 4.3 FUTURE RESEARCH DIRECTIONS AND CONCLUSIONS Our IBM indicates that the strength of the mutualism between native plants and mycorrhizal fungi may be an important factor influencing invasion success of an allelopathic plant invader such as garlic mustard. In particular, we show that plant–mycorrhizal mutualisms that are either weak or strong are less resistant to invasion than intermediate‐strength mutualisms. Though our model results may provide practical applications for identifying native plants at risk of displacement by an allelopathic plant invader, empirical research is needed to validate if these patterns are true in nature. One such study might be to establish, in a greenhouse, a gradient of plant communities with known dependencies on mycorrhizal fungi (i.e., a plant community that is highly dependent upon mycorrhizal mutualisms, a community with intermediate dependencies, and so forth). Once this gradient of plant communities is established, experimentally introducing allelopathic plant invaders would test which communities are most resistant to invasion. Another study might be to identify a plant with plastic responses to mycorrhizal fungi; for example, a plant that is less dependent on fungi in high‐nutrient conditions but more dependent in nutrient‐poor conditions. Performing experimental introductions with allelopathic invaders across different conditions of mycorrhizal dependencies would shed light on how mutualism strengths can mediate invasion success in nature. Given that many other environmental factors, such as natural enemies, may influence how plants interact with mycorrhizal fungi, conducting the above studies as field experiments would provide the strongest tests of the predictions of our IBM. Due to the increasing disruptions to ecosystems by invasive plants, this research provides timely insights into how an allelopathic invasive plant may successfully invade native plant communities. We show that plant–mycorrhizal mutualisms of intermediate strengths are most resistant to garlic mustard invasion. We also show there might be invasion thresholds even for the plants most resistant to plant invasion. Thus, forest plants that are of conservation interest should be tested for their reliance upon mycorrhizal fungi, as our model suggests their level of dependence on this mutualism will likely determine how vulnerable they are to displacement by an allelopathic invader. In addition, forest ecosystems that have experienced degradation that weakens existing plant–mycorrhizal mutualisms, or that gives a sudden numerical advantage to an allelopathic invader, are likely to be more susceptible to invasion. Hence, we call for future field studies to evaluate the “Intermediate Mutualist‐Strength Hypothesis” for plant invaders that negatively impact the mutualism between native plants and mycorrhizal fungi."
} | 5,107 |
38650280 | PMC10992918 | pmc | 1,440 | {
"abstract": "Photorespiration consumes fixed carbon and energy generated from photosynthesis to recycle glycolate and dissipate excess energy. The aim of this study was to investigate whether we can use the energy that is otherwise consumed by photorespiration to improve the production of chemicals which requires energy input. To this end, we designed and introduced an isoprene synthetic pathway, which requires ATP and NADPH input, into the cyanobacterium Synechocystis sp. 6803. We then deleted the glcD1 and glcD2 genes which encode glycolate dehydrogenase to impair photorespiration in isoprene-producing strain of Synechocystis . Production of isoprene in glcD1/glcD2 disrupted strain doubled, and stoichiometric analysis indicated that the energy saved from the impaired photorespiration was redirected to increase production of isoprene. Thus, we demonstrate we can use the energy consumed by photorespiration of cyanobacteria to increase the energy-dependent production of chemicals from CO 2 . \n Supplementary Information The online version contains supplementary material available at 10.1186/s40643-021-00398-y.",
"conclusion": "Conclusions In summary, the production of isoprene in photorespiration-impaired strain doubled, indicating the energy consumed by photorespiration was reused for isoprene production. We further demonstrated that the introduced isoprene synthetic pathway can recover the photoinhibition of both PSII and PSI that normally results from impairing of photorespiration, thus providing an alternative strategy for avoiding or engineering photorespiration (Kebeish et al. 2007 ; Shih et al. 2014 ) in photosynthetic organisms. By designing and introducing an energy consuming pathway, we can, therefore, make the photorespiration of cyanobacteria dispensable, and put the otherwise consumed energy into the photosynthetic conversion of CO 2 to useful chemicals.",
"introduction": "Introduction Photorespiration refers to the metabolism of 2-phosphoglycolate (2-PG), which is derived from the oxygenase activity of ribulose-1,5-bisphosphate carboxylase/oxygenase (Rubisco) (Tolbert 1997 ). Usually one-third to one-fourth of the generated ribulose-1,5-bisphosphate (RuBP) is channeled into photorespiration in higher plants (Hagemann and Bauwe 2016 ). The energy consumed by photorespiration accounts for up to one-third of the total energetic cost of CO 2 fixation (Hagemann and Bauwe 2016 ). Avoiding photorespiration is, therefore, a key target for improving crop yields and effort for removing or reducing photorespiration has never been stopped. If the oxygenase activity of Rubisco can be deactivated, the photorespiration could be removed, as there will be no 2-PG available. To increase photosynthetic efficiency, engineering Rubisco, including increasing the carboxylase activity and/or decreasing the oxygenase activity of Rubisco, has been one of the holy grails of photosynthetic research. However, due to the extremely complex structure of Rubisco, little improvement was achieved (Cai et al. 2014 ; Shih et al. 2014 ). Since photorespiration cannot be removed from deactivation of the oxygenase activity of Rubisco, impairing the metabolic pathway of 2-PG might prevent loss of the fixed carbon and energy from the photorespiratory 2-PG cycle. However, impairing photorespiratory 2-PG cycle in higher plants (Kozaki and Takeba 1996 ; Engel et al. 2007 ) or in cyanobacteria (Eisenhut et al. 2008 ) all resulted in high-CO 2 requirement for normal growth, which is termed as the high-CO 2 -requiring (HCR) phenotype. The HCR phenotype means cells could not grow under ambient conditions and require more CO 2 or low light for growth. To date, effort for increasing photosynthesis efficiency through impairing the photorespiration has been unsuccessful (Hagemann and Bauwe 2016 ). Engineering photorespiration into photorespiratory bypass might partially reduce the loss of fixed carbon and photochemical energy consumed by the photorespiratory 2-PG cycle. In 2007, a bacterial glycolate catabolic pathway was introduced into Arabidopsis thaliana chloroplasts. Part of the photorespiratory glycolate was converted into CO 2 and glycerate, the latter was then channeled into the Calvin–Benson–Bassham cycle (Kebeish et al. 2007 ). A reduced photorespiration and an increased biomass production were observed under ambient conditions (Kebeish et al. 2007 ). In 2014, a synthetic CO 2 -fixing photorespiratory bypass was introduced into cyanobacterium. However, no significant improvement on photosynthesis was observed (Shih et al. 2014 ). Recently, introducing alternative glycolate metabolic pathways into tobacco chloroplasts resulted in a more than 40% increased biomass productivity (South et al. 2019 ; Fernie and Bauwe 2020 ). It has long been considered that photorespiration is superfluous or incomplete in cyanobacteria because of their CO 2 -concentrating mechanism (CCM) (Colman 1989 ; Eisenhut et al. 2008 ). However, it was shown recently that photorespiration is essential in cyanobacteria (Eisenhut et al. 2019 ). Photorespiration-impaired cyanobacterium strain required high CO 2 for normal growth and 2-phosphoglycolate phosphatases, a key enzyme involved in photorespiration, were identified in Synechocystis sp. PCC 6803 (hereafter termed Synechocystis ), indicating photorespiration is essential in cyanobacteria (Eisenhut et al. 2008 ; Dyo and Purton 2018 ). Since recently, cyanobacteria has been widely engineered for production of chemicals from CO 2 , and some chemicals are derived from pathways that require strong input of energy and reducing equivalent, like isoprene. We therefore wonder if we could impair the photorespiration of cyanobacteria, can we use the energy that is otherwise consumed by photorespiration to improve the production of chemicals whose synthetic pathway is dependent on energy supply? To this end, we chose cyanobacterium Synechocystis as a model strain, to which an isoprene biosynthesis pathway was introduced and optimized. We then impaired the photorespiration and investigate the consequence.",
"discussion": "Results and discussion Design and construction of an energy consuming isoprene synthetic pathway in Synechocystis In addition to detoxification of 2-PG and salvage of organic carbon (Bloom 2015 ), it is generally accepted photorespiration plays an important physiological function that is to protect the host from photoinhibition by dissipating excess photochemical energy (Kozaki and Takeba 1996 ). When the capacity of photorespiration in higher plant was improved or decreased, the tolerance to high-intensity light was increased or decreased, respectively (Kozaki and Takeba 1996 ). We, therefore, hypothesized if we could introduce a photochemical energy consuming pathway into cyanobacteria, we might be able to impair photorespiration without generating HCR phonotype, as the physiological function of photorespiration can be replaced by the introduced energy consuming pathway. In this way, the energy that is otherwise consumed in photorespiration can be saved and used to increase the production of target chemicals. We noticed that Synechocystis has a methylerythritol phosphate (MEP) pathway, which consumes 3 ATP and 3 NADPH (Fig. 1 ). If we could engineer this pathway to allow it produce isoprene from CO 2 , we might be able to test whether the above hypothesis would work. Fig. 1 Design and creation of an isoprene synthetic pathway in Synechocystis . The isoprene synthetic pathway is a MEP-dependent and energy consuming pathway. An exogenous isoprene synthase (IspS, from Pueraria montana ) was introduced to convert DMAPP to isoprene. To optimize the isoprene-synthetic pathway, dxs , idi , ispD and ispF from E. coli were overexpressed in Synechocystis . RuBP ribulose-1,5-bisphosphate, 3PGA 3-phosphoglycerate, 2PG 2-phosphoglycolate, G3P glyceraldehyde-3-phosphate, DXP 1-deoxy- d -xylulose-5-phosphate, MEP methylerythritol phosphate, CDP-ME diphosphocytidylyl methylerythritol, CDP-MEP diphosphocytidylyl methylerythritol 2-phosphate, MEcPP methylerythritol 2,4-cyclodiphosphate, HMBPP hydroxymethylbutenyl 4-diphosphate, IPP isopentenyl pyrophosphate, DMAPP dimethylallyl pyrophosphate. Enzymes: DXS DXP synthase, DXR DXP reductoisomerase, IspD 4-diphosphocytidyl-2C-methyl- d -erythritol synthase, IspE 4-(cytidine-5′-diphospho)-2-C-methyl- d -erythritol kinase, IspF 2C-methyl- d -erythritol 2,4-cyclodiphosphate synthase, IspG 4-hydroxy-3-methylbut-2-enyl-diphosphate synthase, IspH HMBPP reductase, IDI IPP isomerase, IspS isoprene synthase To engineer the energy consuming MEP pathway into an isoprene synthetic bypass, we introduced the isoprene synthase encoding gene ispS from the higher plant kudzu ( Pueraria montana ) (Lindberg et al. 2010b ) into Synechocystis , to catalyze the conversion of dimethylallyl diphosphate (DMAPP) into isoprene (Fig. 1 ). The codon-optimized ispS gene was placed under the control of the strong P cpc560 promoter (Zhou et al. 2014 ), and inserted into the pta site (Zhou et al. 2012 ) of Synechocystis , generating the first-generation isoprene-producing strain, designated as IspS (Fig. 2 a). Fig. 2 Construction of mutant strains. a Genetic modifications to impair photorespiration and introduce the isoprene synthetic pathway. b Whole-cell PCR for the recombinant cassette harboring the ispS gene at the pta site; one primer was set 100 bp external to the recombinant cassette, which demonstrated the complete segregation of the Isps mutant. c Whole-cell PCR with specific primers demonstrating the integration of each gene into the chromosome of the recombinant strain MEP*-IspS. d Whole-cell PCR with specific primers for complete segregation of knocking out glcD1 and glcD2 gens in photorespiration impairing mutantsΔglcD1/D2, IspSΔglcD1/D2 and MEP*-IspSΔglcD1/D2 Since the activity of the MEP pathway determines the consumption of photochemical energy, we further optimized the MEP pathway with the aim to generate a second generation isoprene-producing strain. Previous studies on engineering the MEP pathway for the production of isoprene or isoprenoid compounds in microorganisms have demonstrated that the intracellular concentration of DMAPP can be increased by increasing the expression levels of dxs , encoding 1-deoxy- d -xylulose-5-phosphate synthase, idi , encoding isopentenyl pyrophosphate (IPP) isomerase, ispD , encoding 4-diphosphocytidyl-2C-methyl- d -erythritol synthase, and ispF , encoding 2C-methyl- d -erythritol 2,4-cyclodiphosphate synthase (Gao et al. 2016 ; Ajikumar et al. 2010 ). We therefore cloned dxs , idi , ispD, and ispF from E. coli and assembled these genes into an artificial operon under the control of a strong promoter P cpc560 (Zhou et al. 2014 ), and inserted the artificial operon into the phaCE site of the IspS strain. The resulting second-generation isoprene-producing strain was designated as MEP*-IspS (Fig. 2 a). Complete segregation and correct gene insertion were verified by PCR and sequencing (Fig. 2 b, c). Metabolite spectrum analysis indicated that the mutant IspS strain constructed in this work produced 65.4 ± 1.7 µg L −1 of isoprene when incubated in BG11 medium at 30 °C under constant white light for 6 days, while the mutant MEP*-IspS produced 95.4 ± 1.0 µg L −1 (Figs. 2 a, 3 a). These results confirmed that two mutants equipped with energy consuming MEP pathways of different activities were obtained. Fig. 3 Phenotype analysis of all strains. a Time-course of isoprene productions by the strains IspS, MEP*-IspS, IspSΔglcD1/D2 and MEP*-IspSΔglcD1/D2. b High-CO 2 -requiring (HCR) phenotype analysis of the photorespiration-impaired strains ΔglcD1/D2, IspSΔglcD1/D2 and MEP*-IspSΔglcD1/D2. The strains were grown on BG11 plates with (HC) or without (LC) 50 mM NaHCO 3 under 100 μmol photons m −2 s −1 light condition. c Quantification of intracellular and extracellular glycolate in WT and all mutant strains grown on HC condition Impairing photorespiration increased isoprene production in IspS and MEP*-IspS strains To test whether we can use the energy that is otherwise consumed by photorespiration to improve the photosynthetic conversion of CO 2 to isoprene, we tried to impair the photorespiration in the IspS and MEP*-IspS strains. A previous study had shown that inactivation of the key genes of the 2-PG cycle, glcD1 and glcD2 , encoding glycolate dehydrogenase, completely impaired photorespiration in Synechocystis but resulted in an HCR phonotype (Eisenhut et al. 2008 ). The glcD1 and glcD2 were, therefore, inactivated in the Synechocystis wild-type (WT) strain, as well as the IspS and MEP*-IspS mutant strains. The resulting mutants in which the glcD1 gene was replaced with the erythromycin resistance cassette Em r and the glcD2 gene was replaced with the spectinomycin resistance cassette Spec r were designated as ΔglcD1/D2, IspSΔglcD1/D2, and MEP*-IspSΔglcD1/D2, respectively (Fig. 2 a). Complete segregation and correct gene insertions were verified by PCR and sequencing (Fig. 2 b–d). We then investigated whether the wild-type (WT) and the mutant strains ΔglcD1/D2, IspSΔglcD1/D2, and MEP*-IspSΔglcD1/D2 would exhibit the HCR phenotype. As shown in Fig. 3 b, the strain ΔglcD1/D2 did not grow under normal conditions and required the supplementation of additional inorganic carbon. By contrast, strains IspSΔglcD1/D2 and MEP*-IspSΔglcD1/D2 grew as fine as that of the WT. This demonstrates that introducing an additional energy consuming pathway can indeed make photorespiration dispensable, and the mutant can grow under the ambient conditions. Although strains IspSΔglcD1/D2 and MEP*-IspSΔglcD1/D2 did not grow better than the WT (Additional file 1 : Fig. S1), the isoprene production of strains IspSΔglcD1/D2 (130.64 ± 11.26 μg L −1 ) and MEP*-IspSΔglcD1/D2 (184.18 ± 8.68 μg L −1 ) were approximately twofold higher than that of the strains IspS and MEP*-IspS, respectively (Figs. 2 a, 3 a). The isoprene productivity of strain MEP*-IspSΔglcD1/D2 newly developed in this work was approximately 2.3 µg L −1 h −1 , which was lower than the highest isoprene productivity in Synechocystis (12.8 µg L −1 h −1 ) reported previously (Chaves et al. 2016 ). Furthermore, the isoprene production rate to biomass was analyzed. On the sixth day, the cells density (OD 730 ) reached 2.6 (Additional file 1 : Fig. S1), which is about 0.96 g dry cell weight L −1 according to a predetermined correlation factor of 0369 g L −1 per OD 730 in a previous report (Gao et al. 2016 ). Therefore, the isoprene production rate to biomass of strain MEP*-IspSΔglcD1/D2 was about 0.19 mg g −1 and previously reported the highest one was 0.73 mg g −1 , indicating driving more carbon flux to isoprene from biomass will greatly contribute to isoprene production in Synchocystis (Chaves et al. 2016 ). This suggests that the photochemical energy consumed by photorespiration could be redirected towards isoprene synthesis. Glycolate is accumulated and secreted extracellularly in photorespiration-impaired mutants Isoprene biosynthesis can rescue the HCR phenotype of impairing photorespiration (Fig. 3 b). This indicates that the isoprene biosynthesis in strains IspSΔglcD1/D2 and MEP*-IspSΔglcD1/D2 can completely consume the excess ATP and NADPH that are otherwise consumed by photorespiration. To determine the fate of photorespiratory glycolate after impairing the photorespiratory glycolate metabolism, the intracellular and extracellular concentration of glycolate of the WT and all mutant strains were determined and comparatively analyzed (Fig. 3 c). Figure 3 c showed that trace amount of intracellular glycolate accumulation (< 1.5 µM) was observed in all strains tested. Extracellular glycolate was at undetectable level in strains WT, IspS, and MEP*-IspS. However, large amount of extracellular glycolate was detected in photorespiration-impaired mutants IspSΔglcD1/D2, MEP*-IspSΔglcD1/D2, and ΔglcD1/D2. Moreover, the extracellular glycolate concentration of isoprene-producing strains IspSΔglcD1/D2 and MEP*-IspSΔglcD1/D2 was 40% lower than that of the ΔglcD1/D2 strain, indicating that the introduced energy consuming isoprene synthetic pathway can reduce photorespiration, but cannot avoid the oxygentation of Rubisco. When the photorespiratory pathway was impaired upon knocking out of glcD1/2 , glycolate was accumulated and the energy for glycolate metabolism might be saved and used for other biochemical process, such as isoprene biosynthesis. To further confirm if the energy saved from blocking glycolate metabolism can contribute to isoprene production, stoichiometric analysis was conducted. Based on the stoichiometry formula containing ATP and reducing equivalents (Additional file 1 : Table S3), the ATP and NAD(P)H amounts containing in isoprene biosynthesis and glycolate metabolism were calculated and analyzed (Additional file 1 : Table S4). As shown in Additional file 1 : Table S4, upon blocking the photorespiratory cycle, isoprene synthesis could consume a large part of the energy that otherwise is associated with glycolate metabolism. The stronger the isoprene pathway was, the higher the energy consumption ratio reached. The energy calculation also showed that there was a negative correlation between the flux of isoprene synthesis and photorespiratory cycle, which is consistent with what was shown in the Fig. 3 c. These data further indicated that the photosynthetic conversion of CO 2 to isoprene can consume the energy saved from impaired photorespiration and the impaired photorespiration can contribute isoprene production. It is interesting that the isoprene-producing and photorespiration-impaired mutant strains still produced considerable amount of glycolate but did not exhibit the HCR phenotype, nor did the glycolate produce affect cell growth. This further demonstrated that glycolate metabolism is not essential, nor is glycolate toxic, for cyanobacteria. These data also support our hypothesis that the primary physiological role of photorespiration is to dissipate excess energy rather than detoxifying 2-PG, as photorespiration can be impaired once an alternative energy consumption pathway is available. To further determine whether the extracellularly secreted glycolate could be metabolized via an unknown pathway, 1 mM 13 C-labeled glycolate was added to the cultures of strains WT, IspS and MEP*-IspS grown for 48 h, and then cultivated for another 48 h. The initial and final extracellular concentration of 13 C-labeled glycolate were analyzed by LC–MS (Additional file 1 : Fig. S2). The data showed that no significant change of the extracellular concentration of 13 C-labeled glycolate was observed in all three strains. Photosynthetic microorganisms can only take up and metabolize very little extracellular glycolate (Hess and Tolbert 1967 ), our result, thus, demonstrated isoprene synthesis did not drive the uptake of the extracellularly supplemented glycolate. This also suggests the extracellularly secreted glycolate in glcD1 / glcD2 disrupting strains cannot be re-assimilated either. Furthermore, no significant change of 3-P-glycerate (Additional file 1 : Table S5) and not detectable pyruvate of these strains indicated that isoprene synthesis did not drive the metabolism of intracellular glycolate either. The secretion of the accumulated glycolate was also observed in a previous study, where an inhibitor was supplemented to impair the oxidation of glycolate in cyanobacteria (Norman and Colman 1988 ). In that study, the extracellular concentration of glycolate can reach a level 20-fold that of the intercellular concentration of glycolate, when incubated with 100% O 2 which enhanced photorespiration (Norman and Colman 1988 ). Therefore, our data, together with this previous study (Norman and Colman 1988 ), demonstrated that impairing glycolate oxidation would result in glycolate accumulation. The fact that the excess glycolate was secreted extracellularly indicates the photorespiratory metabolism of glycolate is not essential for cyanobacteria, so metabolizing the glycolate should not be considered as the primary physiological function of photorespiration in cyanobacteria. Improved photosynthetic performance in IspS and MEP*-IspS strains To investigate the effect of impairing photorespiration and/or introducing the energy consuming isoprene synthetic pathway on the activity of PSII, chlorophyll fluorescence kinetics were investigated. The light response curve of rETR(II), the relative electron transport rate of PSII (Fig. 4 a) and Y (II), and the effective quantum yield of PSII (Additional file 1 : Fig. S3a) were measured under light intensities ranging from 18 to 2292 μmol photons m −2 s −1 . Fig. 4 Analysis of PSII and PSI activity. a Light response curve of rETR(II), the relative electron transport rate of PSII. b O 2 evolution rate under 300 μmol photons m −2 s −1 . c Light response curve of rETR(I), the relative electron transport rate of PSI. The cells grown under HC and measurements were done at LC conditions. Error bars indicate standard deviations (SD) of the data from three independent experiments. For each experiment, three technical replicates were performed Under low light intensities (18–37 μmol m −2 s −1 ), there were almost no difference in rETR(II) and Y (II) among all strains (Fig. 4 a; Additional file 1 : Figs. S3a, S4a, S5a, S6a, b, and S7a, b). That means impairing glycolate metabolism did not result in decrease of photosynthetic activity of PSII at low light intensities. As no photoinhibition occurs at low light intensities, there is no need to initiate photorespiration to dissipate excess energy; therefore, impairing photorespiration does not affect photosynthesis. When light intensity was increased from 60 to 2292 μmol photons m −2 s −1 , the rETR(II) (Fig. 4 a; Additional file 1 : Figs. S4b, S6a, S7a) and Y (II) (Additional file 1 : Figs. S3a, S5b, S6b, S7b) of strains ΔglcD1/D2, IspSΔglcD1/D2, and MEP*-IspSΔglcD1/D2 exhibited decreases of the photochemical activity of PSII ranging from 5.6% to 60%, 2.2% to 30%, and 0.45% to 13% (Table 1 ), respectively. This means that impairing photorespiration resulted in a more severe photoinhibition of PSII along with increasing light intensity, while the introduced and optimized isoprene-synthetic pathway was able to recover approximately 60% and 80% of the photoinhibition of PSII, respectively. The PSII photoinhibition recovery in strain MEP*-IspSΔglcD1/D2 was approximately 1.3-fold higher than that of the strain IspSΔglcD1/D2, indicating that further enhancing the flux of the introduced isoprene synthetic pathway might potentially completely recover the photoinhibition of PSII under higher light intensities. Table 1 Analysis of PSII and PSI activity decrease in the photorespiration-impaired mutants compared to the WT Light intensity Parameters ΔglcD1/D2 (%) IspS MEP*-IspS ΔglcD1/D2 (%) ΔglcD1/D2 (%) Normal rETR(II) and Y (II) 5.6–10 2.2–4.2 0.45–2.2 rETR(I) and Y (I) 0–14 − 1.47 to 3.8 − 2.6 to 1.4 High rETR(II) and Y (II) 14–60 5.5–30 3–13 rETR(I) and Y (I) 20–73 8.1–32 2.7–11.6 Normal light intensity, 60–100 μmol m −2 s −1 High light intensity, 142–2292 μmol m −2 s −1 Moreover, the introduced isoprene synthetic pathway was able to recover the decrease of the saturation light point that resulted from impairing photorespiration. As Fig. 4 a shown, the rETR(II) of strainΔglcD1/D2 reached the highest value (approximately 14) under around 300 μmol photons m −2 s −1 , whereas the rETR(II) of the WT kept increasing and reached the highest value of 22 under approximately 600 μmol photons m −2 s −1 , indicating that the impairing of photorespiration resulted in a decrease of the saturation light point from 600 to 300 μmol photons m −2 s −1 . Interestingly, the rETR(II) of strains IspSΔglcD1/D2 and MEP*-IspSΔglcD1/D2 reached their respective highest values (19.0 and 20.5, respectively) under around 600 μmol photons m −2 s −1 , which was similar to the saturation light point of the WT (Fig. 4 a). Furthermore, the oxygen evolution rate, another sensitive indicator of PSII function, was investigated under 300 μmol photons m −2 s −1 , the half-saturation point of the WT. Figure 4 b shows that the oxygen evolution rate of strainΔglcD1/D2 was approximately 42% lower than that of the WT ( P < 0.01). Interestingly, there were no significant differences between the oxygen evolution rates of the strains IspSΔglcD1/D2, MEP*-IspSΔglcD1/D2 and WT, under 300 μmol photons m −2 s −1 . This shows that impairing photorespiration inhibited photosynthetic oxygen evolution, and isoprene synthesis was able to recover the resulting decrease of oxygen evolution. To further investigate the effects of impairing photorespiration and/or introducing the isoprene synthetic pathway on the energy conversion efficiency of PSI, rETR(I) (Fig. 4 c), Y (I) (Additional file 1 : Fig. S3b) and the decrease of rETR(I) and Y (I) compared with the WT (Additional file 1 : Fig. S6c, d) were also analyzed (Klughammer and Schreiber 2008 ). Similar to the effect on PSII, under the low light conditions from 18 to 37 μmol m −2 s −1 , rETR(I) and Y (I) were similar, if not identical, in all strains (Fig. 4 c; Additional file 1 : Figs. S3b, S4c, S5c, S6c, d, S7c, d). That means impairing photorespiration did not result in the decrease of photosynthetic activity of PSI at such low light intensities. The data listed in Table 1 show that the introduced isoprene pathway in the mutants IspSΔglcD1/D2 and MEP*-IspSΔglcD1/D2 was able to completely recover the PSI inhibition under normal light conditions (60 μmol m −2 s −1 to 100 μmol m −2 s −1 ), and recover approximately 60%–90% of the photoinhibition of PSI under light intensities ranging from 142 to 2292 μmol photons m −2 s −1 . Impairing photorespiration resulted in approximately 34% decrease of PSII activity and 43% decrease of PSI activity at the light-saturation point, indicating that PSI was more sensitive to inhibition generated by impairing photorespiration than PSII (Table 2 ). The introduced isoprene pathway was able to recover approximately 75% of the photoinhibition of PSII and nearly 98% of the photoinhibition of PSI under 588 μmol photons m −2 s −1 , the light-saturation point of the WT. This indicates that PSI recovers more easily from the photoinhibition resulted from the impairing of photorespiration than PSII does. Table 2 Measurement of chlorophyll fluorescence kinetics for photorespiration-impaired strains and WT under 588 μmol photons m −2 s −1 Strain rETR(II) Y (II) rETR(I) Y (I) Y (ND) Y (NA) WT 23.63 ± 0.80 0.096 ± 0.004 56.83 ± 3.35 0.37 ± 0.01 0.59 ± 0.01 0.03 ± 0.01 ΔglcD1/D2 15.6 ± 1.35 0.063 ± 0.006 32.6 ± 4.67 0.28 ± 0.05 0.71 ± 0.07 0.01 ± 0.01 IspSΔglcD1/D2 20.03 ± 0.95 0.081 ± 0.004 47.63 ± 5.02 0.33 ± 0.01 0.65 ± 0.02 0.02 ± 0.01 MEP*-IspSΔglcD1/D2 21.63 ± 1.25 0.088 ± 0.005 56.4 ± 7.4 0.35 ± 0.01 0.63 ± 0.02 0.02 ± 0.02 Y (II), effective quantum efficiency of PSII rETR(II), relative electron transport rate of PSII Y (I), effective quantum efficiency of PSI rETR(I), relative electron transport rate of PSI Y (ND), the quantum yield of non-photochemical energy dissipation due to donor-side limitation Y (NA), the quantum yield of non-photochemical energy dissipation due to acceptor-side limitation Data represent the means from three independent measurements ± SD. For each experiment, three technical replicates were performed To further understand the fate of the quantum yield of PSI, the non-photochemical dissipation of the PSI quantum yield: Y (ND) (the quantum yield of non-photochemical energy dissipation due to donor-side limitation) and Y (NA) (the quantum yield of non-photochemical energy dissipation due to acceptor-side limitation) were analyzed (Zhou et al. 2016 ) under the high light intensity of 588 μmol photons m −2 s −1 (Table 2 ). Y (ND) of the photorespiration-impaired strain ΔglcD1/D2 increased 20%, while Y (NA) of ΔglcD1/D2 decreased 6.7% compared to the Y (ND) and Y (NA) of the WT, respectively. This result indicates that the donor-side limitation is the reason for the photoinhibition of PSI, rather than the acceptor side. This further indicates that the excess photochemical energy from the photosystems, rather than the oxygen-induced inhibition of carbon fixation, causes the HCR phenotype when photorespiration is impaired."
} | 7,163 |
22229925 | PMC3262071 | pmc | 1,442 | {
"abstract": "Summary Quorum sensing is a mechanism of cell–cell communication that bacteria use to control collective behaviours including bioluminescence, biofilm formation and virulence factor production. In the Vibrio harveyi and Vibrio cholerae quorum-sensing circuits, multiple non-coding small regulatory RNAs called the quorum-regulated small RNAs (Qrr sRNAs) function to establish the global quorum-sensing gene expression pattern by modulating translation of multiple mRNAs encoding quorum-sensing regulatory factors. Here we show that the Qrr sRNAs post-transcriptionally activate production of the low cell density master regulator AphA through base pairing to aphA mRNA, and this is crucial for the accumulation of appropriate levels of AphA protein at low cell density. We find that the Qrr sRNAs use unique pairing regions to discriminate between their different targets. Qrr1 is not as effective as Qrr2–5 in activating aphA because Qrr1 lacks one of two required pairing regions. However, Qrr1 is equally effective as the other Qrr sRNAs at controlling targets like luxR and luxO because it harbours all of the required pairing regions for these targets. Sequence comparisons reveal that Vibrionaceae species possessing only qrr 1 do not have the aphA gene under Qrr sRNA control. Our findings suggest co-evolving relationships between particular Qrr sRNAs and particular mRNA targets.",
"introduction": "Introduction Quorum sensing is the chemical communication process bacteria use to regulate gene expression in response to changes in cell population density. Quorum sensing relies on the production, secretion and subsequent detection of extracellular signalling molecules called autoinducers (AIs). Quorum sensing ensures that bacteria behave as individuals at low cell density and exhibit group behaviours at high cell density. Quorum-sensing-controlled behaviours include bioluminescence, biofilm formation and virulence factor production ( Davies et al ., 1998 ; Zhu et al ., 2002 ; Hammer and Bassler, 2003 ; Ng and Bassler, 2009 ). Multiple non-coding small regulatory RNAs lie at the centres of the Vibrio harveyi and Vibrio cholerae quorum-sensing circuits and are the focus of this study ( Lenz et al ., 2004 ; Tu and Bassler, 2007 ). Non-coding small RNAs (sRNAs) are widely used regulators in bacteria and eukaryotes. In bacteria, they control traits including nutrient uptake, stress response, viral immunity, and in the present context, quorum sensing ( Waters and Storz, 2009 ). Bacterial sRNAs are classified according to their regulatory mechanism. There are protein activity modulating sRNAs, cis -encoded base pairing sRNAs, trans -encoded base pairing sRNAs, and the recently discovered CRISPR sRNAs ( Waters and Storz, 2009 ). The quorum-regulated sRNAs called the Qrr sRNAs in the V. harveyi and V. cholerae quorum-sensing systems belong to the set of trans -acting sRNAs that function through Hfq-assisted base pairing with target mRNAs to control mRNA translation or stability ( Caron et al ., 2010 ). This class of sRNAs can repress mRNA translation by pairing with the ribosome binding site and occluding ribosome access, typically resulting in mRNA degradation ( Aiba, 2007 ). Alternative mechanisms exist in which sRNAs pair within mRNA coding regions or in intergenic regions of polycistronic transcripts, which leads to RNase E- or RNase III-dependent endonucleolytic cleavage ( Desnoyers et al ., 2009 ; Pfeiffer et al ., 2009 ; Papenfort et al ., 2010 ). sRNAs can also act as activators by pairing with and altering the secondary structures of regions in the 5′ UTR of mRNAs to reveal ribosome binding sites, typically promoting mRNA stabilization and translation ( Frohlich and Vogel, 2009 ). Activation can also occur through sRNA generation of accessible ribosome binding sites via endonucleolytic cleavage or formation of a nuclease barrier at the 5′ end of the target mRNA ( Obana et al ., 2010 ; Ramirez-Pena et al ., 2010 ). In V. harveyi quorum sensing, at low cell density, in the absence of AIs, the quorum-sensing response regulator protein LuxO is phosphorylated ( Freeman and Bassler, 1999 ). Phospho-LuxO activates the expression of five genes ( qrr 1–5) encoding the five Qrr sRNAs ( Tu and Bassler, 2007 ). The Qrr sRNAs activate translation of the low cell density master regulator AphA, which controls ∼ 300 low cell density target genes ( Rutherford et al ., 2011 ). The Qrr sRNAs simultaneously repress translation of the high cell density master regulator LuxR ( Fig. 1 , left) ( Tu and Bassler, 2007 ). At high cell density, when AIs are present, LuxO is dephosphorylated and it is inactive, so production of the Qrr sRNAs ceases. In the absence of the Qrr sRNAs, AphA is not produced, but LuxR translation occurs. LuxR controls ∼ 700 high cell density target genes ( Fig. 1 , right) (J.C. van Kessel, unpublished). The quorum-sensing circuit of the closely related pathogenic bacterium V. cholerae resembles that of V. harveyi , but V. cholerae only has Qrr1–4 and the V. cholerae LuxR homologue is called HapR ( Lenz et al ., 2004 ). In V. harveyi and V. cholerae , in addition to controlling the two quorum-sensing master regulators, AphA and LuxR/HapR, the Qrr sRNAs control other targets and they participate in several feedback loops. These Qrr sRNA-mediated feedback loops fine-tune the quorum-sensing output by providing robust responses to cell population density changes, promoting high fidelity signal transmission, and controlling the input–output dynamic range ( Svenningsen et al ., 2008 ; 2009 ; Tu et al ., 2008; 2010 ; Ng and Bassler, 2009 ; Teng et al ., 2011 ). Fig 1 Model for Qrr sRNA regulation of aphA , luxR / hapR and luxO . At low cell density, phospho-LuxO activates expression of the qrr genes encoding the Qrr sRNAs. The Qrr sRNAs promote translation of the low cell density master regulator AphA and inhibit translation of the high cell density master regulator LuxR/HapR. At high cell density, Qrr sRNA production ceases because dephosphorylated LuxO is inactive. AphA translation stops and LuxR/HapR translation occurs. LuxO production is repressed by the Qrr sRNAs in a negative feedback loop. AphA and LuxR repress each other at the transcriptional level. In this study, we characterize the production pattern of the newly identified quorum-sensing low cell density master regulator AphA in both V. harveyi and V. cholerae . We show that the Qrr sRNAs activate AphA production through direct base pairing to the aphA mRNA 5′ UTR, and this regulatory step is crucial for proper AphA protein accumulation at low cell density. We also find that the Qrr sRNAs use a unique set of pairing regions to activate aphA compared with the regions they use to control other target mRNAs such as luxR and luxO . Qrr1 is less effective than the other Qrr sRNAs in activating aphA because it lacks one of the critical pairing regions. However, Qrr1 is fully functional in its control of mRNA targets that do not require this particular pairing region. Sequence analysis reveals that Vibrionaceae species can possess 1, 4 or 5 Qrr sRNAs. Our evidence indicates that the Qrr- aphA mRNA interaction does not occur in Vibrionaceae species possessing only Qrr1. Rather, only vibrios containing multiple Qrr sRNAs control aphA by this mechanism. We propose that harbouring multiple Qrr sRNAs enables the Qrr sRNAs to diversify and evolve distinct target preferences, and in this case, to ensure optimized quorum-sensing gene expression ( Tu et al ., 2008 ).",
"discussion": "Discussion A set of highly conserved Qrr sRNAs function at the core of the V. harveyi and V. cholerae quorum-sensing circuits. The Qrr sRNAs are expressed when cells are in low cell density mode and they act to repress the production of the high cell density master regulator LuxR/HapR ( Lenz et al ., 2004 ; Tu and Bassler, 2007 ; Bardill et al ., 2011 ). Recently, the Qrr sRNAs were also shown to activate the production of the low cell density master regulator AphA ( Rutherford et al ., 2011 ). As AphA and LuxR/HapR control hundreds of target genes at low cell density and high cell density, respectively, and they mutually repress each other at the transcriptional level, the amount of the Qrr sRNAs present at any time during growth specifies the exact quorum-sensing-controlled gene expression pattern ( Lin et al ., 2007 ; Pompeani et al ., 2008 ; Rutherford et al ., 2011 ). Here we show that the Qrr sRNAs activate aphA through direct base pairing to its mRNA 5′ UTR. Activation is critical for high level production of AphA protein at low cell density, especially in V. harveyi , which exhibits a dramatic increase in AphA compared with that present at high cell density. Based on secondary structure predictions, the ∼ 200 nt long 5′ UTR of aphA mRNA is capable of forming an inhibitory structure masking its ribosome binding site, which presumably leads to translational inhibition. At low cell density, pairing of the Qrr sRNAs to the aphA mRNA 5′ UTR could disrupt this inhibitory structure and expose the ribosome binding site enabling AphA protein translation. Similar ‘anti-antisense’ mechanisms have been described for several other Hfq-chaperone-dependent trans -acting sRNAs including DsrA/RprA/ArcZ- rpoS , RyhB- shiA and GlmZ- glmS in E. coli , Qrr- vca0939 in V. cholerae and recently, PhrS -pqsR in Pseudomonas aeruginosa ( Majdalani et al ., 1998 ; 2001 ; Hammer and Bassler, 2007 ; Prevost et al ., 2007 ; Urban and Vogel, 2008 ; Mandin and Gottesman, 2010 ; Sonnleitner et al ., 2011 ). We engineered 10 mutations (point mutations and deletions) in the aphA 5′ UTR in an attempt to disrupt the putative inhibitory structure and thereby increase basal AphA-GFP levels. None of these mutants exhibited increased GFP production (Fig. S6) indicating that multiple mutations in different regions of the aphA 5′ UTR are likely required to disrupt the inhibitory secondary structure. What is the benefit of Qrr sRNA activation of aphA ? Presumably during the transition from high cell density to low cell density, such as when vibrios exit a host or disperse from a biofilm, the immediate production of the Qrr sRNAs could promote rapid accumulation of AphA by both stabilizing and activating translation of aphA mRNA. This is especially noteworthy given that, in V. harveyi , AphA is undetectable at high cell density. Thus, a rapid and large fold change in AphA occurs at the high to low cell density transition. Presumably, going from ‘no’ AphA to a significant concentration of AphA enables a similar rapid and dramatic change in gene expression of AphA targets. We therefore propose that post-transcriptional rather than transcriptional activation of aphA could be crucial when an instantaneous switch in behavioural modes is required. Indeed, other such regulatory loops involving the Qrr sRNAs exist that affect quorum-sensing dynamics. LuxR/HapR activates qrr expression, which also increases the rapidity of the transition out of high cell density mode ( Svenningsen et al ., 2008 ; Tu et al ., 2008 ). The Qrr sRNAs repress luxO , which delays the transition from low cell density to high cell density mode ( Tu et al ., 2010 ). Finally, the Qrr sRNAs repress luxMN encoding an AI-receptor pair, which adjusts the sensitivity of the quorum-sensing circuit to different AIs ( Teng et al ., 2011 ). Together, these loops exquisitely fine-tune the quorum-sensing transitions presumably to optimize survival in a changing environment. Moreover, we note that the Qrr sRNAs are used repeatedly in these various feedback loops, suggesting an economical solution to control quorum-sensing network dynamics. As the universe of known bacterial sRNAs increases, two important themes are emerging: one is a scenario in which multiple sRNAs regulate the same target, for example, the sRNAs, DsrA, RprA and ArcZ all control the common target rpoS , which defines the gene expression pattern under different stress conditions ( Majdalani et al ., 1998 ; 2001 ; Mandin and Gottesman, 2010 ). The second scenario is one in which the same sRNA regulates multiple targets. For example, RyhB sRNA represses sodB, iscS, cysE and fur and it activates shiA , which together provide growth benefits under iron limiting conditions ( Masse and Gottesman, 2002 ; Prevost et al ., 2007 ; Vecerek et al ., 2007 ; Desnoyers et al ., 2009 ; Salvail et al ., 2010 ), SgrS represses ptsG and manX to relieve sugar-phosphate stress ( Vanderpool and Gottesman, 2004 ; Rice and Vanderpool, 2011 ), Spot42 controls genes in central and secondary metabolism ( Moller et al ., 2002a , b ; Beisel and Storz, 2011 ), and RybB and MicA regulate genes encoding outer membrane proteins that counter cell envelope stress ( Rasmussen et al ., 2005 ; Udekwu et al ., 2005 ; Johansen et al ., 2006 ; Papenfort et al ., 2006 ; Bossi and Figueroa-Bossi, 2007 ; Coornaert et al ., 2010 ; Gogol et al ., 2011 ). These many-to-one and one-to-many regulatory mechanisms give sRNAs overarching power in controlling regulatory networks. We frequently find multiple inputs are wired into sRNA production to ensure strict restriction of their levels, presumably to keep sRNA levels in check. The V. harveyi and V. cholerae Qrr sRNAs function by both scenarios: particular mRNA targets are regulated by multiple Qrr sRNAs and each Qrr sRNA controls multiple target mRNAs. Qrr sRNAs levels are precisely controlled through the feedback mechanisms described in the preceding paragraph. Furthermore, the level of each Qrr sRNA is affected by the other Qrr sRNAs due to dosage compensation ( Svenningsen et al ., 2009 ). Thus, coupling tight control of Qrr sRNA production to a large set of functions provides an orchestrated quorum-sensing response. Additional genes could be controlled by the Qrr sRNAs potentially providing links between quorum sensing and other regulatory networks. Clearly, the Qrr sRNAs share overlapping functions; however, specificity is nonetheless ensured by several different means. First, in spite of their highly conserved sequences, there are particular regions of each Qrr sRNA that can be used to control distinct targets. As shown here, Qrr1 lacks one of the two pairing regions required for aphA activation, suggesting that Qrr1 prefers the targets luxR and luxO . Only about half of the nucleotides in the Qrr sRNAs are identical, suggesting that additional regions could exist to control other targets. It should in principle be possible to further separate regulation of luxR and luxO based on pairing differences. Indeed, mutating UGA ( Figs 3A and 4A , mut iii) in Qrr4 has a more dramatic effect on luxR repression than on luxO repression (Fig. S5). At present, we only know a few Qrr targets, so this idea remains to be further explored as new Qrr targets are identified. Our findings are consistent with those for the sRNAs FnrS, GcvB and Spot42, which show that different stretches are used to control particular target mRNAs ( Durand and Storz, 2010 ; Beisel and Storz, 2011 ; Sharma et al ., 2011 ). Second, even when the pairing regions are conserved, differential regulation of target mRNAs could be achieved based on different expression levels and stabilities of the Qrr sRNAs. The contribution from each Qrr sRNA to regulation of each target mRNA will also be influenced by the efficacy of pairing and the stability of each Qrr-mRNA pair, which, in turn, depend on the avidity of their interactions with the Hfq chaperone and their secondary structures under particular physiological conditions ( Vogel and Luisi, 2011 ). Third, differences in qrr promoter sequences suggest that each qrr is controlled by specific regulators. We know that phospho-LuxO regulates all the qrr genes; however, what additional environmental or intracellular cues affect the expression of one or a subset of the Qrr sRNAs remain undefined. A key finding of this work is that in Vibrionaceae species possessing multiple qrr genes, Qrr1 lacks the region required for aphA activation. Species containing only qrr 1 presumably reflect the ancestral state of this lineage. We suggest that duplication of the ancestral qrr 1 gene in the lineage led to extant species containing multiple qrr genes. Region I in the Qrr sRNAs was co-opted for regulation of a new target, namely aphA . Subsequently, region I was lost from Qrr1, and the other Qrr sRNAs were relegated the role of controlling aphA . Because Qrr2–5 ( V. harveyi ) or Qrr2–4 ( V. cholerae ) contain redundant copies of region I, this region was most likely lost from Qrr1 as a consequence of neutral evolutionary drift. Loss of region I from Qrr1 in these species could be a neutral alteration to the quorum-sensing regulatory circuit. However, we suggest that there may be a selective advantage in possessing Qrr sRNAs devoted to particular regulatory roles, allowing finer tuning of the quorum-sensing circuit. If so, in species containing multiple Qrr sRNAs, Qrr1 could evolve the function of specific tuning of luxR and luxO expression. The present work pinpoints a special role for Qrr1 in regulation of aphA ; however, the other Qrr sRNAs could likewise have exclusive functions. Qrr5 is particularly interesting to us because it only exists in a subset of vibrios including V. harveyi, Vibrio parahaemolyticus and Vibrio vulnificus but not V. cholerae and Vibrio splendidus , which possess only Qrr1–4 ( Lenz et al ., 2004 ; Tu and Bassler, 2007 ; Miyashiro et al ., 2010 ). Our previous studies show that, in V. harveyi , qrr 5 is constitutively repressed under normal growth conditions ( Tu and Bassler, 2007 ). However, Qrr5 is fully functional to repress luxR , luxO , and to activate aphA when expressed in E. coli ( Tu and Bassler, 2007 ). Thus, it will be fascinating to learn under what conditions Qrr5 is produced in V. harveyi , and the functions of its specific target genes. In light of the above results, we predict that Qrr5 specific targets are conserved in vibrio species containing qrr 5 but not in other vibrios."
} | 4,563 |
20657704 | null | s2 | 1,443 | {
"abstract": "All characterized major ampullate silks from orb-web weaving spiders are composites of primarily two different proteins: MaSp1 and MaSp2. The conserved association of MaSp1 and MaSp2 in these spider species, the highly conserved amino acid motifs, and variable ratios of MaSp1 to MaSp2 demonstrate the importance of both MaSp1 and MaSp2 to the strength and elasticity of the fiber. Computer simulated mechanical tests predicted differing roles for MaSp1 and MaSp2 in the mechanical properties of the fibers. Recombinant MaSp1 and MaSp2 proteins were blended and spun into fibers mimicking the computer-simulated conditions. Mechanical testing verified the differing roles of MaSp1 and MaSp2."
} | 172 |
39151012 | PMC11328896 | pmc | 1,444 | {
"abstract": "Proteins self-assemble to function in living cells. They may execute essential tasks in the form of monomers, complexes, or supramolecular cages via oligomerization, achieving a sophisticated balance between structural topology and functional dynamics. The modularity and programmability make DNA origami unique in mimicking these key features. Here, we demonstrate three-dimensional reconfigurable DNA origami pincers (DOPs) that multitask on giant unilamellar vesicles (GUVs). By programmably adjusting their pinching angle, the DOPs can dynamically control the degree of GUV remodeling. When oligomerized on the GUV to form origami cages, the DOP units interact with one another and undergo reorganization, resulting in the capture, compartmentalization, and detachment of lipid fragments. This oligomerization process is accompanied with membrane disruptions, enabling the passage of cargo across the membrane. We envisage that interfacing synthetic cells with engineered, multifunctional DNA nanostructures may help to confer customized cellular properties, unleashing the potential of both fields.",
"introduction": "INTRODUCTION Proteins participate in almost every process of cellular life. They maintain the cell shape, catalyze chemical reactions, regulate cargo transport, and coordinate cell signaling pathways ( 1 ). Intriguingly, protein subunits are often oligomerized into cage-like superstructures, which are strongly guided by symmetry ( 2 ). These so-called protein cages with diversified morphologies support critical cellular functions, providing the predominate means of building complexity in living systems ( 3 ). Although the working mechanisms of most proteins are beyond what scientists can now replicate, the enthusiasm to construct their synthetic equivalents with reduced complexity is everlasting ( 4 , 5 ). Such efforts may have great significance for the development of new biomaterials ( 6 ), nanomedicine ( 7 ), drug delivery ( 8 ), and vaccine ( 9 ) and for the fundamental understanding of the origin of life ( 10 ). DNA nanotechnology has been used to create engineered structures that mimic the key features of biological systems ( 11 – 18 ). The success stems from the fact that DNA is a unique genetic and construction material. It affords versatile design and engineering capabilities, as well as enables highly specific and programmed tasks on the molecular level ( 19 – 24 ). Notably, DNA structures can be designed to work under user-defined interactions and environmental settings, usually impossible to achieve in complex biological systems ( 25 – 28 ). These capabilities have made DNA nanotechnology an important player in synthetic biology, especially providing a mechanistic framework to interact with synthetic cells ( 29 – 32 ). For instance, Franquelim et al. ( 33 ) constructed curved DNA origami inspired by BAR domain proteins to sculpt lipid membranes. Journot et al. ( 34 ) realized DNA triskelion networks that were reminiscent of clathrin-coated pits. Grome et al. ( 35 ) created dynamin mimics using DNA origami spirals to induce membrane tubulation. Furthermore, Birkholz et al. ( 36 ) demonstrated multifunctional DNA nanopores, which could puncture and remodel lipid membranes. In addition, tension-loaded DNA clamps were used to drive membrane tabulation and budding, providing insights into achieving spatiotemporal control over membrane dynamics ( 37 ). Here, we demonstrate three-dimensional (3D) reconfigurable DNA origami pincers (DOPs) that multitask on giant unilamellar vesicles (GUVs) for dynamic GUV remodeling, capturing, compartmentalization, and subsequent detachment of lipid fragments from GUVs, as well as transport of molecules across the membrane by inducing transient pores ( Fig. 1 ). In the monomer state, the membrane-bound DOPs can dynamically control the degree of the GUV morphological changes via programmably adjusting their pinching angle. When oligomerized on the GUV membrane, multiple DOP units interact with one another and self-assemble into origami cages with defined topology. During oligomerization, the reorganization of the DOP units disrupts the GUV membrane. Lipid fragments captured by the DOP units are compartmented by the origami cages and subsequently detached from the GUV. The oligomerization process induces the formation of transient membrane pores, enabling the passage of cargo across the membrane. Fig. 1. Schematic of 3D DOPs that multitask on GUVs. The red structures represent DNA origami pincers (DOPs), and the green vesicles represent giant unilamellar vesicles (GUVs). The reconfigurable DOPs can remodel the GUV morphology in a programmable manner by changing the pinching angle θ. Oligomerization of the DOPs on the GUV leads to the capture, compartmentalization, and detachment of lipid fragments, as well as the formation of transient membrane pores for cargo transport across the membrane. Note that the schematic provides a hypothetical representation, in which the DOPs and GUVs are not depicted to scale.",
"discussion": "DISCUSSION In this work, we have demonstrated that the 3D reconfigurable DOP monomers can work together to regulate the morphological changes of GUVs in a programmable manner. Furthermore, the DOP units can interconnect and self-assemble into origami cages to achieve both defined topology and functional dynamics on GUVs. The DOP oligomerization on the GUV induces transient membrane pores, which allow for the influx of cargo. We envision that, by interfacing cells with engineered dynamic DNA nanostructures, it may foster tantalizing opportunities to interrogate interesting biological questions in the future ( 45 ). For instance, by building synthetic equivalents of proteins, one can grasp the basic relation between the structure and function of proteins in cells, as well as understand when, where, and how proteins occur to function, at least to some extent. In addition, given the ease of engineering from the bottom-up, one can harness the simplicity of DNA origami to create complexity by building artificial systems with nonnatural functionalities. For instance, taking its ability to position bioactive moieties at the nanoscale, the exterior surface, the interior surface, and the interface between the origami units can be versatility modified with customer-defined properties, so that the self-assembled high-order architectures (e.g., interconnection of units into cages and interconnection of cages into cage chains or lattices) may synergistically combine the functions of several protein families, which work independently in biological cells. Closely related to our work, Dietz and Pinner ( 46 ) have recently used DNA origami shells on vesicles to induce membrane budding and mediate scission, opening avenues for the exploration of membrane mechanics. Despite the limited cognition of biological systems as of today, the efforts along this line will add scientific breath for a deeper understanding of cellular life."
} | 1,740 |
31018560 | PMC6523470 | pmc | 1,445 | {
"abstract": "Polymer actuators are important components in lab-on-a-chip and micromechanical systems because of the inherent properties that result from their large and fast mechanical responses induced by molecular-level deformations (e.g., isomerization). They typically exhibit bending movements via asymmetric contraction or expansion with respect to changes in environmental conditions. To enhance the mechanical properties of actuators, a strain gradient should be introduced by regulating the molecular alignment; however, the miniaturization of polymer actuators for microscale systems has raised concerns regarding the complexity of such molecular control. Herein, a novel method for the fabrication of micro-actuators using a simple molecular self-alignment method is presented. Amphiphilic molecules that consist of azobenzene mesogens were located between the hydrophilic and hydrophobic surfaces, which resulted in a splayed alignment. Thereafter, molecular isomerization on the surface induced a large strain gradient and bending movement of the actuator under ultraviolet-light irradiation. Moreover, the microelectromechanical systems allowed for the variation of the actuator size below the micron scale. The mechanical properties of the fabricated actuators such as the bending direction, maximum angle, and response time were evaluated with respect to their thicknesses and lengths. The derivatives of the polymer actuator microstructure may contribute to the development of novel applications in the micro-robotics field.",
"conclusion": "4. Conclusions In this study, micron-sized light-driven actuators were successfully fabricated via thermo-initiated polymerization with molds obtained by the MEMS process. The polymerized molecules were self-aligned as parallel or perpendicular to the surface of the actuators. This alignment was induced by the interaction between the polarity of the amphiphilic molecules and the hydrophilic/hydrophobic properties of the mold surfaces. The splayed (homogeneous and homeotropic) alignment was confirmed by SEM imaging; thus, the opposite bending direction was observed according to the irradiated surface. The actuator bent toward the light when the homogeneous surface was irradiated, and in the opposite direction in the case of the homeotropic surface. This indicates that the homogeneous surface contracts and the homeotropic surface expands under UV irradiation. The bending behavior of the fabricated actuators varied with respect to their lengths and thicknesses. In particular, the longer and thinner actuators exhibited larger bending angles. However, the maximum bending angle was reached quickly, irrespective of the actuator thickness. Thereafter, the actuators gradually bent toward their original positions under continuous irradiation due to the reduction in the strain gradient. The proposed method for the fabrication of micro-actuators with self-alignment and their fast and large photomechanical response could contribute to the development of applications such as wireless microrobots based on UV light irradiation without requiring conventional robot components such as batteries and gears.",
"introduction": "1. Introduction The miniaturization of polymer actuators, which can be controlled through specific motions such as moving and grasping via external stimuli, is a critical factor in the control of micron-sized objects for diagnostics, drug delivery, surgery, and cellular manipulation [ 1 , 2 , 3 ]. Actuators can operate in response to various stimuli such as ion concentration gradients, heat, electrical fields, and light [ 4 , 5 ]. Among them, light offers several advantages for a wide range of applications, as it allows for the operation of the actuator in air and solutions using wireless, prompt, and target stimuli. Moreover, the variation in its wavelength and intensity can result in different actuator movements. Photochromic molecules such as azobenzene mesogens respond to light stimuli via photomechanical reactions within their chromophores [ 6 ]. The conformational changes of azobenzene mesogens are induced by trans–cis isomerization under ultraviolet (UV) and visible light irradiation, which consequently leads to a change in the molecular volume. Their incorporation and specific ordering in a polymer film can create gradients of molecular volume in its bulk [ 7 ]. Even if the variation of the molecular volume is modest, it can generate out-of-plane deformations via accumulation throughout the bulk. Meanwhile, the photoinduced deformation of polymer actuators containing azobenzene mesogens can also be obtained by the photoinduced phase transition between LC and the isotropic state [ 8 ]. Regardless of the mechanism of deformation, it can successively lead to contraction or expansion depending on the alignment of the molecules that contain azobenzene mesogens [ 9 , 10 ]. Although polymer films that contain azobenzene molecules have recently been investigated in relation to the use of their deformation properties in various applications such as light-driven plastic motors, inchworm walkers, and robot arms, all of these applications are at the sub-millimeter scale [ 11 , 12 , 13 ]. To create a specific order of azobenzene mesogens, previous studies have focused on film fabrication with a rubbed polyimide layer. However, this method is unsuitable for microscale products. This limitation can be overcome by using inkjet printing technology, and reducing the film dimensions to 100 μm in width and 500 μm in length [ 14 ]. However, the viscosity of the melted polymers, the monomer de-wetting on the surface, and the nozzle size limit the miniaturization of actuators. Another study fabricated a fiber-type actuator with an approximate diameter of 300 μm by post-cross-linking a series of copolymers [ 15 ]. However, the bending direction was determined as one-way, either toward or away from a light source, due to the cylindrically symmetric structure of the fibers. In this paper, a novel method was proposed for the fabrication of micro-actuators that contain azobenzene polymer films. This was achieved by using molecular self-alignment through the interaction force between their polarity and the surface hydrophilicity, in addition to the molding process using microelectromechanical systems (MEMSs). A reduction in the actuator size to several tens of micrometers was achieved. In addition, for products with a photoinduced bending ability in opposite directions depending on the incident light direction, the molecules were self-aligned by modulating the hydrophilicity of the substrates. Finally, the maximum bending angle was evaluated with respect to the lengths and thicknesses of the fabricated micro-actuators."
} | 1,675 |
25674453 | PMC4320217 | pmc | 1,446 | {
"abstract": "Army ants perform the altruism behavior that an ant sacrifices its own well-being for the benefit of another ants. They build bridges using their own bodies along the path from a food to the nest. We developed the army ant inspired social evolutionary system by using Swarm library. The system has 2 kinds of ant agents, ‘Major ant’ and ‘Minor ant’. They communicate with each other via pheromones. Army ant can recognize them as the signals from the other ants. The pheromones evaporate with the certain ratio and diffused into the space of neighbors stochastically. If the optimal bridge is found, the path through the bridge is the shortest route from the food to the nest. We define the probability for an ant to leave a bridge as to the number of neighboring ants. The constructing method of the optimal route has been proposed. In this paper, the behaviors of ant under the environment with two or more feeding spots were observed. Some experimental results show the behaviors of great interest with respect to altruism of ants. The knowledge discovery of social evolutionary process from some computer simulation results is described in this paper. Electronic supplementary material The online version of this article (doi:10.1186/2193-1801-3-712) contains supplementary material, which is available to authorized users.",
"introduction": "Introduction In animal societies, self-organization is the theory of how minimal complexity in the individual can generate greater complexity at the population. The rules specifying the interactions among the components in the system are implemented by using only local information without global information. In the study of social evolution, army ant performs altruism as one behavior of complexities, where each individual reduces its own fitness but increases the fitness of other individuals in the population. Such behaviors seem to be involved acts of self-sacrifice in order to aid the others. In evolutionary biology, such a behavior is called reciprocal altruism. The concept was initially developed to explain the evolution of cooperation as mutually altruistic acts (Trivers 1971 ). The basic idea is close to the strategy of “equivalent relation” in the study of strategic decision making. Army ants are characterized by their two different phases of activities, a nomadic phase and a stationary phase. During the nomadic phase, army ants move during the day to capture insects, spiders, and so on. The stationary phase starts when the larvae pupate for a few weeks. Moreover, army ants build a living nest with their bodies instead of building a nest like other ants. Each ant will hold on to the other legs and form a linked chain or a ball structure. This behavior is known as a bivouac. This allows the bridging of an empty space. In order to address the self-assembled structure as a particular type of aggregation, Deneubourg et al. defined the probability of an ant entering or leaving chain in (Deneubourg et al. 2002 ). Moreover, they showed that the gregarious behavior facilitates cooperation by Blattella germanica in shelters during the resting period. The probability to leave the shelter was defined. Ishiwata et al. ( 2011 ) developed the simulation system for the foraging behavior and the altruism of army ants by using Swarm library, Swarm-2.2 (Lancaster et al. 2002 ). (The original website www.swarm.org is in the process of being rebuilt.) The probabilities to form the chain defined in (Lioni et al. 2001 ) was used in their simulation experiments. The number of neighboring active ants is considered as the condition for altruistic behavior. Their simulation results show a mimic altruistic behavior. By inspiring Ishiwata’s study, Ichimura et al. developed the multi-agent simulation system to execute more realistic altruistic behavior where two or more kinds of agents realize the division of roles in army ants (Ichimura and Douzono 2012 ). According to the environment in (Ichimura and Douzono 2012 ), the simulation results reported that the optimal path from the food to the nest cannot be always found, because two or more chains in the environment were formed. Although more emergence of altruistic behaviors was observed, but the capabilities of forming chain was dispersed. As a result, the performance of foraging decreases and some ants took a circuitous route. On the contrary, Ichimura et al. defined the evaporation rate of pheromone dues to normal distribution probability and the probability to leave from the chain when the ants in its neighbor region depart gradually in (Ichimura and Douzono 2012 ). The altruism simulation results are reported to find more optimal paths from food to the nest. In this paper, we observed the behaviors of ant agents under the multi feeding spots in the same environment of (Ichimura and Douzono 2012 ). Some experiments with different ratio of feed size were investigated. In general, ant agents take an action to be concentrated in the largest feeding spot. The shortest path from the spot to the nest is constructed and the ants bring feed to the nest. Then, the feeding spots will be disappeared in the order of larger spot. However, it has turned out that there is a certain tendency without regard to the size of feed. The altruism behavior does not work well and the bridge will be broken, if enough ant agents are not gathered into the ditch. As a result, the food at the spots remains to the end of simulation. Moreover, the group of agents was automatically constructed during search. The experimental results with the different number of agents show the building group for the effective search. Such behaviors are also seen in the collaborative social networks. Especially, a research framework for studying social systems uses agent based modeling and simulation. Madey et al. ( 2002 , 2003 ) describe the simulation results for the collaborative social network composed of open source software (OSS) developers and projects. The obtained knowledge in this paper will be useful for the collaborative social networks. The remainder of this paper is organized as follows. The section Simulation environment describes the simulation environment with Swam library. The section Agent behaviors defines the behaviors of agents such as search phase, homing phase (return to the nest), and altruism phase. The section Proposed method describes the proposed method related to pheromone and the leaving probability from chain. Experimental results for simulations are described in the section Experimental results for altruism and the section Experimental results for the formation of group. In the section Conclusive discussion, we give some discussions to conclude this paper.",
"discussion": "Conclusive discussion We developed the army ant inspired social evolutionary system which can perform the altruism. There are 2 kinds of ant agents communicated with each other via pheromones. Moreover, the pheromones evaporate with the certain ratio and diffused into the space of neighbors stochastically. In order to avoid the over-concentration in the chain, the probability of leaving from a chain is introduced. The system with the facilities can find the optimal place of bridge. The path through the bridge is the shortest from foods to the nest. In this paper, the behaviors of ant under the environment with multi feeding spots and the adequate number of agents were observed. The altruism behavior in the few agents to the size of food spot is hard to keep its situation. Such observations of behaviors in the computer simulation strongly will contribute to the shift to knowledge and power from the individual to the collective. We will explore how agent-based modeling and simulation can be used as a research technique to study collaborative social networks. The altruism behaviors described in the paper will be useful to discover the power of the synergy effect in social networks."
} | 1,977 |
39894846 | PMC11788442 | pmc | 1,448 | {
"abstract": "Bacterial biofilms are complex cell communities within a self-produced extracellular matrix, crucial in various fields but challenging to analyze in 3D. We developed a “biofilm-in-capillary” growth method compatible with full-rotation soft X-ray tomography, enabling high-resolution 3D imaging of bacterial cells and their matrix during biofilm formation. This approach offers 50 nm isotropic spatial resolution, rapid imaging, and quantitative native analysis of biofilm structure. Using Bacillus subtilis biofilms, we detected coherent alignment and chaining of wild-type cells towards the oxygen-rich capillary tip. In contrast, the Δ tasA genetic knock-out showed a loss of cellular orientation and changes in the extracellular matrix. Adding TasA protein to the Δ tasA strain restored matrix density and led to cell assembly compaction, but without the chaining observed in wild-type biofilms. This scalable and transferable approach opens new avenues for examining biofilm structure and function across various species, including mixed biofilms, and response to genetic and environmental factors.",
"introduction": "Introduction Many microorganisms develop surface-attached biofilms with a characteristic protective matrix consisting of adhesive macromolecules such as exopolysaccharides (EPS), DNA, and proteinaceous filaments or fibrils 1 , 2 . This heterogeneous gel-like matrix is complemented by small and medium-sized compounds, among them nutrients, signaling molecules, and surfactants 1 – 3 . Although biofilms in natural environments are inhabited by a variety of bacteria and other organisms, structural investigations on model biofilms, e.g., of Bacillus subtilis , help to understand basic principles of biofilm construction, function and development. B. subtilis can produce different types of biofilms depending on culture conditions, such as submerged biofilm within a liquid or pellicles that are formed on surfaces of liquids 1 , 4 , 5 . The extracellular matrix (ECM) of the B. subtilis biofilm contains polysaccharides 6 and the major proteinaceous biofilm component TasA, which can form filaments and fibrils 7 – 9 . The filaments are formed by a strand complementation mechanism 8 , 9 , initiated by TapA, which can also anchor the TasA filaments 10 , 11 . The hydrophobin-like lipoprotein A (BslA) adds additional protection to the surface of the ECM 1 . Different methodological attempts have been made to analyze the entire biofilm architecture using mass spectrometry 12 , magnetic resonance imaging 13 , 14 , scanning transmission X-ray microscopy 15 , 16 , small and wide-angle X-ray scattering 17 , or different electron microscopic techniques 18 . Making use of an extended antibody staining concept, high-resolution light microscopy has been used to reveal the architecture of living V.cholerae 19 biofilm in three-dimensional super-resolution microscopy. A combination of super-resolution PALM with a single objective light sheet and precision genome editing was used in studies of E.coli biofilm structure 20 . Proteus mirabilis biofilm structure and subcellular DNA organization were investigated by 4-fold expansion microscopy after digesting oligosaccharide and protein components by an enzyme cocktail 21 . As biofilms are heterogeneous assemblies containing up to 97% (w/w) water 22 fixation, drying, freezing, and dehydration steps compromise morphology and 3D structure, representing a challenge for these 3D imaging techniques. Here, we employ soft X-ray tomography (SXT) as a label-free imaging modality with a spatial resolution of 25–60 nm to investigate the role of the essential biofilm protein TasA. To understand how tasA deletion affects the biofilm 3D architecture, we developed a biofilm-in-capillary workflow using B. subtilis WT and Δ tasA strains as examples. By 3D imaging, we investigate (i) the collective patterning of bacterial cells in biofilms, (ii) changes in ECM distribution, and iii) phenotypical changes of individual bacteria in suspension. During the process of data acquisition, a set of X-ray projection images (shadows) are collected at different rotation angles around a cylindrical sample represented by a very narrow capillary (typical tip diameter 10–12 µm). Each data acquisition is followed by 3D reconstruction which generates a volume of a specimen with a resolution that may approach 25–60 nm. SXT operates in the “water window” of the electromagnetic spectrum, exploiting the range between carbon and oxygen absorption edges (4.4 nm to 2.3 nm wavelength) for natural contrast of carbon-rich materials and transparency of oxygen-rich aqueous media 23 , 24 . In this energy range, photoelectric absorption is the most dominant process, such that the concentration of chemical species relates to the attenuation of X-rays by the Beer-Lambert law, \\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}$$I(z)={I}_{0}{e}^{-{\\rm{\\mu }}z}$$\\end{document} I ( z ) = I 0 e − μ z , 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}$$\\mu$$\\end{document} μ is the linear absorption coefficient (or shortly LAC) of the material with thickness \\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}$$z$$\\end{document} z . Scattering in the “water window” energy range is negligible hence the LAC is approximated as \\documentclass[12pt]{minimal}\n\t\t\t\t\\usepackage{amsmath}\n\t\t\t\t\\usepackage{wasysym} \n\t\t\t\t\\usepackage{amsfonts} \n\t\t\t\t\\usepackage{amssymb} \n\t\t\t\t\\usepackage{amsbsy}\n\t\t\t\t\\usepackage{mathrsfs}\n\t\t\t\t\\usepackage{upgreek}\n\t\t\t\t\\setlength{\\oddsidemargin}{-69pt}\n\t\t\t\t\\begin{document}$${\\rm{\\mu }}(E)\\approx \\frac{{\\rho }_{m}{N}_{A}}{A}$$\\end{document} μ ( E ) ≈ ρ m N A A , 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}$${\\rho }_{m}$$\\end{document} ρ m , \\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}$${N}_{A}$$\\end{document} N A , and \\documentclass[12pt]{minimal}\n\t\t\t\t\\usepackage{amsmath}\n\t\t\t\t\\usepackage{wasysym} \n\t\t\t\t\\usepackage{amsfonts} \n\t\t\t\t\\usepackage{amssymb} \n\t\t\t\t\\usepackage{amsbsy}\n\t\t\t\t\\usepackage{mathrsfs}\n\t\t\t\t\\usepackage{upgreek}\n\t\t\t\t\\setlength{\\oddsidemargin}{-69pt}\n\t\t\t\t\\begin{document}$$A$$\\end{document} A are the mass density, Avogadro’s number and the atomic mass number, respectively 25 . As a constant X-ray energy of 530 eV is used, the LAC measurement is a function of molecular composition (atomic number and mass) and concentration (mass density). This unique quantitative nature of SXT has been employed to distinguish cellular organelles 26 – 28 , to detect variations in DNA packing 29 , and to study states of viral replication 30 . As an advantage, SXT does not require labeling, fixation, or staining, enabling the unperturbed investigation of hydrated cells. As full-rotation SXT relies on using capillaries, full-rotation SXT has not yet been applied to imaging cells within colonies or tissues. In this work, we develop a “biofilm-in-capillary” growth method compatible with full-rotation SXT of up to 200 μm-thick biofilms with a spatial resolution of 50 nm. Using a machine learning approach, we incorporate single-cell and ECM segmentation for systematic analysis of individual cell phenotypes, their spatial organization, and density of ECM. Employing SXT, we were able to detect and analyze the loss of cell orientation after deletion of tasA . Upon rescue with TasA protein, we could detect partial restoration of the ECM structure. We explore correlated changes in ECM structure and cell phenotypes and discuss the role of TasA in biofilm formation. Altogether, we show that the developed biofilm-in-capillary full-rotation SXT workflow is an efficient method for the visualization and analysis of biofilms at the subcellular level under varying conditions.",
"discussion": "Discussion A “biofilm-in-capillary” workflow enables analysis of bacterial biofilms and quantitative measurements on both the subcellular and macroscopic size scales in a physiological state using full-rotation SXT. The workflow was validated on GFP-expressing B. subtilis biofilms for visualization of structural phenotypes and biochemical density of single cells and ECM within biofilms formed. The bacterium Bacillus subtilis is one of the best-studied model organisms for investigating biofilm formation. In addition to exopolysaccharides, B. subtilis biofilms contain TasA as a major proteinaceous matrix component which is required for matrix formation, protection against oxidative stress, interaction with the membrane and it can also act as a developmental signal stimulating a subset of biofilm cells to revert to a motile phenotype 40 , 46 . We, therefore, have focused on 3D imaging of WT B. subtilis biofilm and compared it to cultures of Δ tasA cells, restoring also the complex architecture of biofilms by addition of TasA protein. By use of full-rotation tomography and automatic segmentation based on machine learning, we were able to detect subtle differences in cellular phenotypes and ECM at statistically significant sample sizes. We could show that deletion of the tasA gene affects bacterial cells grown in suspension differently in comparison to the same bacteria during biofilm formation upon TasA supplementation. We observed significant changes in cellular elongation and chemical density as measured by soft X-ray absorption when B. subtilis cells are grown in biofilms. At the macroscopic level deletion of tasA leads to loss of cellular orientation. The extracellular addition of TasA protein to Δ tasA cultures partially restores the biochemical density and volume of the ECM; on the other hand, the cell organization and their phenotype are not rescued, suggesting complex functions in cellular metabolism and motility upon biofilm formation. Interestingly, based on high-spatial-resolution and sensitivity to chemical density, we were able to detect accumulation of lipids within Δ tasA cells upon TasA supplementation. While we showed that biofilm-in-capillary workflow enables imaging of B. subtilis biofilms at high-spatial resolutions, it would be interesting to apply our approach to other types of biofilms, including those consisting of multiple species 47 , and interactions with artificial additives to biofilms 48 , for example, nanoparticles 49 . A potential extension of the proposed workflow would be localization of specific proteins by correlative fluorescence microscopy. Such correlation of structural information provided by SXT with functional aspects of specific proteins visualized via fluorescence microscopy has been demonstrated by several groups 50 – 53 . Overall, due to high sensitivity to biochemical content SXT imaging combined with biofilm-in-capillary workflow enables unprecedented visualization and quantitative analysis of cells and ECM distribution within a 3D volume of biofilms. The combination of subcellular resolution and density measurements in 3D will provide better insights into the biofilm formation and microenvironment."
} | 2,999 |
39858667 | PMC11767832 | pmc | 1,449 | {
"abstract": "The triboelectric nanogenerator (TENG) has emerged as a promising technology for efficiently converting ambient mechanical energy into electrical energy. Among various designs, the disk-based rotational TENG has demonstrated significant potential, as it can continuously harvest energy in a sliding mode via a grating mechanism. However, horizontal mechanical energy is more common than rotational energy in many practical applications. Herein, the present study introduces a novel device: the double horizontal linear-to-rotational triboelectric nanogenerator (DHLR-TENG). This innovative approach utilizes a gear system to convert horizontal linear mechanical energy into electrical energy. The experimental results revealed that the DHLR-TENG produces a full cycle of alternating current (AC) when integrated into an electrical circuit. It consistently delivers robust performance with an open-circuit voltage of 544 V, a short-circuit current of 61.16 µA, and a maximum power output of 33.27 mW. Additionally, the device durability, capable of withstanding over 1,000,000 cycles, makes it highly effective for powering small electronic devices, such as charging capacitors and illuminating commercial LEDs. The DHLR-TENG’s versatility and efficiency mark it as a major advancement in energy harvesting, with broad implications for powering portable electronic devices in a wide range of environments.",
"conclusion": "4. Conclusions In conclusion, this study has demonstrated a logical approach to enhancing the longevity of TENG by combining a radially arranged TENG with a transmission mechanism, enabling uninterrupted functioning for extended periods of time. This research demonstrated the disk-based DHLR-TENG, which utilizes a gear system to transform linear mechanical energy into rotational energy. To generate a complete cycle of alternating current, two units of TENG were used. These units can be connected to an electric board to combine their outputs and convert AC into DC signals. When an equal velocity is applied in both the forward and backward directions for each unit, a complete cycle of alternating current is generated with an open-circuit voltage ( V OC ) value of 544 V, a short circuit current (I sc ) value of 61.16 µA, a maximum power of 33.27 mW, and a durability of more than 1,000,000 cycles. Due to its exceptional electrical properties, the DHLR-TENG can efficiently provide a substantial quantity of electrical energy to a wide range of tiny electronic devices such as LEDs. Electrical Measurements The Haijie Jiachuang LH4572 stepper motor was used for linear movement between Unit 1 and Unit 2. This motor is equipped with a synchronous belt, a sliding module, high-speed accuracy, quiet operation, and dustproof capabilities. It has a 57-stepper and a stroke length of 200 mm. A function generator was used to produce an electrical signal with a precisely regulated amplitude. An electrodynamic shaker was then used to translate the electrical signals into consistent linear mechanical motion. Subsequently, the DHLR-TENG used this linear mechanical motion for both forward and backward movement. An oscilloscope (Tektronix (MDO3032), Beaverton, OR, USA) used to test the electrical signal of the TENG, including the output voltage and short-circuit current. The oscilloscope had an inbuilt load resistance of 100 MΩ. Then, a Pico ammeter (Keysight (B2981A), Santa Rosa, USA) was used to measure the short-circuit current, and a scanning electron microscope (ZEISS (GeminiSEM300), Oberkochen, Germany) was used to scan the surface of the PTFE.",
"introduction": "1. Introduction The triboelectric nanogenerator (TENG) has gained significant attention in the last few years as a potential substitute for conventional energy-harvesting mechanisms. The TENG device is capable of converting mechanical energy from the surrounding environment into electrical energy through the processes of contact electrification and electrostatic induction [ 1 , 2 , 3 ]. The TENG has gained significant interest as a next-generation technology that surpasses conventional mechanical energy-harvesting technology, owing to its straightforward design process, environmentally friendly functioning, and extensive range of available materials [ 4 , 5 , 6 ]. The functional mechanism of the TENG relies on the combination of contact electrification and electrostatic induction. Initially, when two items with distinct positions in the triboelectric series come into contact, a charge transfer process induces positive and negative charges on the surfaces of both objects [ 7 , 8 , 9 ]. When the two things are placed at a certain distance from each other, an electric field is generated between them. In order to attain electrical balance, the transfer of electrons occurs through electrodes that are affixed to the rear surfaces of the two objects [ 4 , 10 ]. The TENG variant that operates in a sliding mode has garnered significant attention in academic research because of its potential for achieving superior electrical functionality. This is attributed to the fact that when two objects slide against each other, as opposed to simply making contact, a greater amount of charge is transmitted between their surfaces [ 7 , 11 , 12 ]. A disk-based TENG was developed as an illustrative case to capture rotating mechanical energy [ 11 , 13 , 14 , 15 ], demonstrating exceptional electrical characteristics. The disk-based TENG was operated through the rotational motion of two coaxial circular disks. When the two disks make contact, electrostatic induction takes place due to a rotary–sliding motion, resulting in the transfer of electrons between two electrodes on the surface of stators that have complimentary patterns. In contrast to a contact separation mode TENG that operates by linear force, the disk-based TENG operating in sliding mode and driven by rotational force offers distinct advantages in terms of its electrical characteristics [ 2 , 10 , 16 , 17 , 18 ]. For instance, the continuous and steady generation of output power is a characteristic of the TENG under consideration. In contrast, the contact separation TENG exhibits a brief but intense burst of output power, which may pose challenges in terms of power management circuit design. Nevertheless, it should be noted that disk-based TENG is restricted in its operation to a sole rotating force. The availability of rotational mechanical energy is comparatively lower than that of linear mechanical energy [ 2 ]. A limited number of researchers have documented the existence of disk-based TENGs that operate in a sliding mode, utilizing linear mechanical force derived solely from fluid-like sources such as wind and water [ 4 , 19 , 20 , 21 , 22 ]. Furthermore, it has been demonstrated that a fan plays a crucial role in the conversion of linear force to rotating force [ 2 , 3 , 7 ]. In this study, we utilize the force conversion mechanism to enhance the application potential of the TENG. This study introduces a new disk-based TENG that effectively captures ample linear mechanical energy from machine movements while also having the capability to harness energy from human movement. A pair of double disks made of identical materials are prepared for the purpose of providing a complete alternative current (AC) output. A double horizontal-to-rotational triboelectric nanogenerator (DHLR-TENG) is successfully built as an experimental device, with the ability to transform horizontal force into rotational force. A gear system that is appropriately built facilitates the conversion of horizontal mechanical force into rotating mechanical force. In this system, where copper (Cu) slides on PTFE, it has been observed that velocity-strengthening and -weakening phenomena occur due to different sliding velocities in the system. Meanwhile, this work reveals the remarkable durability of the TENG system. Achieving an impressive reliably when operating for more than 1,000,000 cycles, the system highlights significant advancements in optimizing performance and energy conversion and ensures long-term reliability and stability. The generation of a complete alternating current (AC) cycle in this study ensures maximum energy-harvesting efficiency and stability in the rectified direct current (DC) output, addressing fluctuations commonly observed in incomplete AC systems and enhancing the system’s practicality for diverse applications. This extensive durability makes the TENG system promising for practical applications, offering both effective energy conversion and sustained operation over extended periods, making it a viable solution for various energy-harvesting needs. As a result, the DHLR-TENG has the capability to supply electrical energy to compact electronic components such as capacitors and light-emitting diodes (LEDs).",
"discussion": "3. Results and Discussion 3.1. Device Structure Figure 1 a depicts a schematic representation of the design of the DHLR-TENG, which operates via the use of linear mechanical force. The DHLR-TENG consists of two different parts, namely a force-converting part and an energy-generating part, both of which are cube-shaped in structure; Supplementary Figure S12 shows pictorial images of the DHLR-TENG. In the force-conversion process, a button with a column-shaped design is subjected to linear mechanical force. This force is transmitted through a stepper motor and subsequently converted into rotational mechanical force through the use of a gear system. The stepper motor moves forward and backward on Unit 1 and Unit 2, respectively, facilitating this conversion. The transmission of rotating mechanical force to the energy-generating component occurs directly in the alternating states of Unit 1 (U1) and Unit 2 (U2). Upon the removal of the applied linear force on U1, the center of the rack gears reverts back to its initial position via the use of the motor stepper. Figure 1 b depicts the structure design and working principal of a freestanding TENG (FR-TENG) with a two-dimensional schematic of the FR-TENG, consisting of a stator and a rotator. 3.2. Design and Performance Analysis of DHLR-TENG: Mechanisms for Energy Conversion and Output Optimization To showcase the considerable output performance, the voltage and current values are typically determined by measuring the total distance travelled during the displacement of the rack gear in 8 cm increments, considered here to be 8 cm forward and backward motion, under the open-circuit and short-circuit conditions, respectively. As depicted in Figure 2 a,b, the aforementioned high voltage/current outputs possess the capability to energize red LEDs effectively, hence enabling their application in diverse domains like DHLR-TENG for converting linear mechanical energy into electrical energy effectively to power numerous small electronic devices, such as LEDs. Nevertheless, the actual implementation of TENG for various applications is impeded by the significant impedance mismatch between the TENG and the power management unit, resulting in the generation of only a minimal power output. Therefore, it is imperative to develop a comprehensive and effective optimization approach. The sectors are patterned onto the surface of the stator. Figure 1 c depicts the three-dimensional configuration of the FR-TENG, comprising a stator and a rotator with negative triboelectric materials, PTFE. The triboelectric materials consist of the metal electrode E R in the rotator component and the insulator film in the stator. These materials are specifically designed to establish connections with the electrical terminals situated on the side of the stator. The stator was fabricated using a laser-cutting machine developed for the purpose of generating output for the TENG. The E R electrode in the rotator is in a state of electrical buoyancy. On the other hand, the retrograde linear force resulting from the restorative force of the motor stepper cannot be transmitted to the energy-generating component as a result of the single-directional bearing within the gear system of U1. Conversely, U2 commences the generation of electricity at the terminal point of U1, as depicted in Figure 1 d,e. The motor stepper exhibits forward movement for U1 and backward movement. In U1, the linear force imparted to the one-way bearing is converted into a rotating force inside each unit. When the stepper motor is in motion in the forward direction, the one-way bearing located at U1 fails to engage, resulting in the absence of force conversion. The lack of a one-way bearing prevents the generation of a consistent electrical output as well as inhibits the rack gear’s ability to return to its initial position and accept future linear forces in both the forward and backward directions. Supplementary Figure S4 displays the comprehensive specifications of the gear system. The calculation of the number of revolutions of a disk is obtained by multiplying the number of rotations of gear 1 by the rotation ratio of the gear system. Gear 1, having a diameter of 32 mm, undergoes an estimated rotation of 0.867 revolutions when the motor stepper is linearly displaced by a force of 8 cm. The rotational ratio of the whole gear system is roughly 6.92. This means that when gear 1 completes one full revolution, the rotator of the DHLR-TENG completes approximately 6.92 rotations. As an example, upon the application of a linear force, measuring 1 cm, to the button, the rotator disk undergoes an estimated revolution of 0.865 turns. In this study, all measurements were conducted by applying a linear force of 8 cm to depress the button, resulting in an estimated rotation of 6.92 turns for the rotator in the gear configuration. Energy transfer optimization, and calculations of the speeds and ratios are discussed in Supplementary Note S4 and presented in Supplementary Table S1 . Within the energy-generating component, a single disk, referred to as the rotator, undergoes rotation due to the transmission of rotational mechanical force from the energy-converting component. Simultaneously, the other disk, known as the stator, remains stationary as it is affixed to the cube-shaped shell. The Supplementary Materials include Video S1 , which provides real-time visual representations of the DHLR-TENG procedure. Figure 1 f,g depict the surfaces of both the rotator and the stator. The substrate used in this study was composed of glass epoxy material, with a diameter of 150 mm. The surface of the substrate exhibited protruding patterns, which comprised copper (Cu) material. A total of 38 sector units made of copper were arranged in a radial pattern on the stator’s surface, while 19 of these sectors were arranged on the rotator’s surface. The center angle of every sector unit was measured to be 9.47 degrees, while the outermost length was found to be 55 mm in the magnified surface pictures shown in Figure 1 f,g, respectively. The separation lengths of the rotator and stator were 11.15 mm and 1 mm, respectively. The Cu pattern on the stator was segmented into two distinct electrodes, namely electrode 1 and electrode 2. The stator was partitioned into 19 sectors for electrode 1 and 19 sectors for electrode 2, with a finger-type configuration. The disk of the TENG was composed of a rotator and a stator, which were fabricated using a laser-cutting machine. Figure 1 illustrates the whole operational mechanism of the DHLR-TENG used for energy generation. In this study, a polytetrafluoroethylene (PTFE) film was affixed to the stator as a triboelectric layer. The purpose of this layer was to facilitate the generation of electrical energy by rotating and sliding with the projecting Cu on the rotator. Following a pre-charging procedure, the PTFE layer on the stator acquires a negative charge, while the Cu surface on the rotator acquires a positive charge, as determined by the triboelectric series. The consistent rotation of the rotator generates an oscillating flow of electric current, known as alternating current (AC), between electrode 1 and electrode 2 for each unit. In addition, as stated before, U1 and U2 were constructed in a similar manner, with identical materials, sizes, and structures. In Figure 2 a,b, the relationship V = IR (Ohm’s Law) explains the connection between voltage (V), current (I), and resistance (R) in the Cu-PTFE system and is critical to understanding the observed time-dependent behavior. The voltage generated across the interface is directly proportional to the current flowing through the circuit, assuming a constant resistance. However, during sliding, the voltage and current are not static and varies over time due to the initial and final velocities of each disk. This variation led to fluctuations in both voltage and current. The effect of load resistance on the voltage and current output under consistent mechanical energy in this experiment has been investigated in Supplementary Note S1 and Supplementary Figures S8 and S9 . 3.3. Electricity Generation Process The triboelectric materials in this system consist of a metal electrode (E R ) in the rotator and an insulating layer attached to the surface of the stator, which contains two separate electrodes (E 1 and E 2 ). Both the rotator and stator components were fabricated using laser-cutting technology to ensure precise matching and alignment. These electrodes are connected to the electrical terminals, facilitating the TENG output, as shown in Figure 1 b. Initially, in Step I, the E R electrode overlaps with E 1 , causing the triboelectric materials to acquire equal but opposite charges due to electrostatic forces in the freestanding structure. As the rotator begins turning rightward in Step II, E R starts to partially overlap with both E 1 and E 2 , initiating triboelectrification through direct contact. This results in negative charges accumulating on the PTFE surface and positive charges on the metal, with the positive charge density on the rotator being twice that of the stator due to unequal contact areas. The rightward rotation continues in Step III, causing further movement of the E R electrode between E 1 and E 2 , leading to an electron flow from E 1 to E 2 through the external circuit, driven by electrostatic imbalance. Finally, in Step IV, the E R electrode overlaps once again with E 1 , causing the electron flow to reverse from E 2 back to E 1 . This cyclical motion generates an oscillating electrical current or voltage signal, which fluctuates in sync with the rotational movement, resulting in a continuous output of electrical energy from the triboelectric nanogenerator (FR-TENG). In the open-circuit condition, electrons cannot flow between the electrodes. The open-circuit voltage ( V O C ) is defined as the electric potential difference between the two electrodes, expressed as V O C = V E 1 − V E 2 . In Step I, E 1 reaches its maximum potential while E 2 reaches its minimum potential, resulting in the maximum V O C . As the rotator begins to spin, this voltage gradually decreases. Once the rotator passes the midpoint, V O C with an opposite polarity starts to form and continues to build until the rotator reaches Step III. Beyond this point, the V O C changes direction due to the periodic structure. Based on the assumption that the dielectric layer’s thickness is much smaller than the width shown in Figure 2 b, an analytical model can be created. In this model, any overlap between the rotator and the electrodes is treated as a parallel plate capacitor, disregarding edge effects. Using Gauss’s Theorem, the V O C for each unit ( U 1 and U 2) can be analytically calculated, with detailed information provided in Supplementary Figure S10 and Supplementary Note S2 .\n (1) S t e p I : V O C ( i n i t i a l ) = V E 1 − V E 2 = 2 d · σ ε 0 ε r \n (2) \n S t e p I I : V O C θ = V E 1 − V E 2 = d · σ ε 0 ε r θ θ 0 − θ − θ 0 − θ θ \n ( θ approaches neither 0 nor θ 0 )\n (3) S t e p I I I : V O C ( f i n a l ) = V E 1 − V E 2 = − 2 d · σ ε 0 ε r \nwhere d is the thickness of the PTFE layer, σ is the triboelectric charge density on top of the PTFE layer, ε 0 is the dielectric constant of vaccum, ε r is the relative dielectric constant of PTFE, α is the angle at which the rotator rotates away from the initial state, and α 0 is the central angle of a single rotator unit. Equation (2) can only be used to illustrate the changing trend of the V O C (see Supplementary Note S2 ). The theoretical peak-to-peak value of the V O C p − p for each U needs to be calculated by subtracting Equation (3) from Equation (1): (4) V O C , p − p = V E 1 − V E 2 = 4 d · σ ε 0 ε r Additionally, to calculate the open-circuit voltage peak to peak in this system ( V s o c p − p ), substitute Equations (S24) and (S25) into Equation (S23) (see Supplementary Note S2 ): (5) V s o c p − p ( t ) = V U 1 ( t ) , 0 ≤ t p < t p 2 V U 2 ( t ) , t p 2 ≤ t p < t p \nwhere V U 1 ( t ) and V U 2 ( t ) output the voltage peak to peak for Unit 1 and Unit 2, respectively, and t p is the time period modulated by 4 s (so the behavior repeats every 4 s). This diagram displays the voltage required to select V (p−p) for U 1 and U 2, as well as the impact of different time intervals on U 1 and U 2. The maximum voltage of V U 1 is determined to be 160 V, while the minimum voltage is observed at −152 V, resulting in a V U 1 , (p−p) = 312 V. Conversely, the upper limit of voltage, denoted as V U 2 , is established at 160 V, while the lower limit is seen at −168 V, resulting in a value of V U 1 , (p−p) = 328 V. V U 1 and V U 2 are shown in Supplementary Figure S1 . The combined maximum voltage, taking into consideration V U 1 and V U 2 , is 280 V, while the minimum voltage is −264 V. This results in a peak-to-peak voltage value of 544 V for the entire system. According to Figure 2 , the energy generation time for each V U is 2 s. The outcomes of this procedure necessitate a duration of 4 s for both cycles, namely V U 1 and V U 2 . When this process persists and repeats, it leads to the generation of a complete cycle of alternate current when the outputs of V U 1 and V U 2 are coupled. Furthermore, the voltage generation process occurs within a single vibration of the motor stepper, which corresponds to a single frequency of movement. The open-circuit voltage peak to peak ( Vs OC,p−p ) and the short-circuit current peak to peak (I SC,p−p ) of the DHLR-TENG were measured under the nine conditions at a vibration frequency of 3 Hz. The voltage and current signal for 1 Hz are shown in Supplementary Figure S3 . The maximum electrical outputs of the DHLR-TENG, with a V OC value of 544 V, an I SC value of 61.16 μA, and a power of 33.27 μW, were achieved under Condition 6, as shown in Figure 2 a,b and c, respectively. The bar chart in Figure 2 d illustrates the relationship between nine distinct operating conditions and their corresponding voltage outputs, with each condition characterized by a specific velocity value. These results emphasize the versatility of the TENG in energy harvesting, demonstrating its capability to function effectively across a broad range of mechanical inputs, from low-velocity vibrations to high-speed motions. This adaptability makes the TENG a viable solution for capturing and converting energy from diverse environmental sources, ensuring consistent energy output even under fluctuating conditions. The conditions, labeled 1 through 9, represent velocities ranging from 0 cm/s to 8 cm/s. Specifically, Condition 1 exhibits the lowest velocity of 0 cm/s, resulting in a voltage of 0 V, while Condition 2 presents a velocity of 1.14 cm/s and a corresponding voltage of 54 V. Condition 3 follows with a velocity of 2 cm/s and a voltage of 88 V. Condition 4 shows a velocity of 2.66 cm/s, producing a voltage of 208 V. Condition 5, with a velocity of 3.5 cm/s, generates a voltage of 320 V. Condition 6, operating at a medium velocity of 4 cm/s, yields the highest voltage of 544 V. Condition 7 presents a velocity of 5.34 cm/s, with an associated voltage of 378 V. Condition 8 shows a velocity of 6.4 cm/s and a voltage of 136 V, while Condition 9, with the highest velocity of 8 cm/s, produces a voltage of 130 V. Upon observing the bar chart, a discernible trend emerges between velocity and voltage. Initially, from Condition 1 to Condition 5, there is a notable rise in both velocity and voltage, suggesting a positive correlation between the two variables. This trend continues with Condition 6, where both velocity and voltage exhibit further increments. However, from Conditions 7 to 9, despite an increase in velocity, the voltage value experiences a slight decrease compared to Condition 6. Finally, in Condition 9, although the velocity reaches its peak at 8 cm/s, there is a noticeable decrease in voltage compared to the preceding condition. Overall, the bar chart provides a visual representation of the relationship between velocity and voltage across different conditions, starting from 0 cm/s (stationary) to 8 cm/s. While there appears to be a general trend of increasing voltage with higher velocities, the relationship is not strictly linear. The output voltage increases until the velocity reaches 4 cm/s, showing a correlation with rotator speed. However, from Conditions 7 to 9, despite an increase in velocity, the voltage decreases. Equation (5), used to estimate the surface charge density of the PTFE film in the DHLR-TENG, calculates a charge density of 11.33 µC/m 2 at a V s O C , p − p value of 544 V, as shown in Figure 3 a. The interaction between Cu-PTFE during sliding, illustrated in Figure 3 a, exhibits distinct frictional behaviors: velocity strengthening and velocity weakening. Velocity strengthening, represented by the red line, occurs when the frictional resistance between Cu-PTFE increases with sliding velocity. This results in a stable rise in charge density, peaking at 11.33 µC/m 2 under Condition 6, before gradually declining as the system stabilizes. In contrast, velocity weakening, depicted by the blue line, is characterized by a reduction in frictional resistance as the sliding velocity increases. This behavior is reflected in an initial rise in charge density that peaks earlier, at 7.08 µC/m 2 under Condition 5, followed by a rapid decline. These contrasting behaviors underscore the dynamic frictional properties of the Cu-PTFE interface, where velocity strengthening fosters stability, while velocity weakening leads to instability at higher sliding velocities. This graph highlights the relationship between surface charge density and open-circuit voltage across different velocities, indicating that higher velocities generally result in higher charge densities. Notably, the charge density increases with sliding velocity, reaching a maximum at 4 cm/s, a phenomenon attributable to velocity strengthening. These results demonstrate that while the velocity increases linearly, the output voltage does not follow a similar trend. This nonlinearity can be attributed to the mechanical effects associated with higher rotational speeds in the TENG sliding system. At increased speeds, the centrifugal forces and vertical vibrations of the rotor become more pronounced. These effects are exacerbated by imperfections in the fabrication process, which lead to an increase in the gap between the rotor and stator. As a result, the maximum output charge generated by the system declines with increasing rotational speed. This phenomenon explains the observed decrease in output voltage, despite the linear increase in velocity. The findings emphasize the importance of precision in the fabrication process to minimize variations in the gap between the rotor and stator, particularly at higher rotational speeds. Addressing these mechanical limitations could enhance the overall efficiency of the TENG system for high-speed applications. Figure 3 b shows the short-circuit charge (Q SC ) of the DHLR-TENG transferred by one cycle of vibration. To compare an I SC value directly with the corresponding Q SC , the Q SC value was calculated using the time interval integral of I SC . The trend of Q SC according to each condition is identical to those of V OC and I SC . In Condition 6, the maximum value of Q SC was approximately 187.4 nC for one cycle of vibration. Furthermore, power density is calculated by dividing the maximum power output by the area over which the power is distributed, as shown in Figure 3 c. The highest value, determined under Condition 6, is 0.11 mW/cm 2 . On the other hand, according to the geometries of the DHLR-TENG, the volume of the DHLR-TENG is determined to be 193.865 cm 3 . Hence, the power density per volume is 171.616 W/(m 3 ·Hz). These results highlight the DHLR-TENG’s exceptional performance among energy-harvesting TENGs. A detailed comparison is provided in Supplementary Table S2 . Moreover, Figure 3 d illustrates the relationship between V OC and the distance ( x ) between the rotator and the PTFE surface, revealing an inverse correlation between these variables. The separation distance ( x ) can be adjusted to regulate the horizontal force necessary for movement, as illustrated in Figure 1 a. Distance ( x ) refers to the gap between the rotator and the surface of the stator. Strong sliding is facilitated by a high frictional force with a small x , which is the result of high pressure. A low frictional force is generated by low pressure, which triggers feeble sliding due to a large x . The nine conditions of x were used to measure the electrical outputs of the DHLR-TENG. The smallest x was found in Condition 1, while the greatest x was found in Condition 9. x increases linearly from Condition 1 to 9. The value of x is approximately 0 mm to 1.2 mm when Conditions 1 to 9 are met. As a result, the sliding of the two disks becomes more robust as the value of x decreases; however, the rack gear requires a greater amount of mechanical force to be driven. It is anticipated that an optimal condition for x will result in the highest possible power output. The DHLR-TENG’s V OC was measured at a vibration frequency of 3 Hz under the nine conditions. In Condition 4, the DHLR-TENG attained its highest electrical output, with a V OC value of 544 V, as shown in Figure 3 d. A large x value resulted in ineffectual gliding beyond Condition 4. The DHLR-TENG’s electrical output was reduced as a result of a low x value in Condition 4. 3.4. Durability The durability of the TENG in this study has been thoroughly tested, showcasing exceptional stability and efficiency even after 1,000,000 cycles of continuous operation. A key metric for evaluating the stability of TENGs is their ability to sustain a stable electrical output over extended periods. As shown in Figure 4 a,b, the voltage and current waveforms after 1,000,000 cycles exhibit only minor changes, with the voltage decreasing from 544 V to approximately 498 V and the current stabilizing at around 56.02 µA. This slight reduction, representing about a 9% change in output, highlights the robustness of the system and its ability to efficiently convert mechanical energy into electrical energy over extended periods with minimal degradation. A scanning electron microscope (SEM) analysis of the PTFE surface before and after 1,000,000 cycles, as shown in Figure 4 c,d, reveals noticeable changes in surface morphology. These changes suggest that mechanical wear, while minimal, can affect the output capability of the triboelectric layer over time. Although the degradation of output performance caused by device abrasion remains a challenge, the structural design implemented in this study has significantly improved the output stability of the DHLR-TENG. The DHLR-TENG system demonstrates excellent long-term reliability, with its output remaining relatively stable despite minor decreases, making it a promising solution for sustained energy-harvesting applications. Environmental factors, including temperature and humidity, are known to significantly influence the performance of TENGs. Elevated temperatures can amplify the electron thermionic emission effect, resulting in charge loss, while high humidity introduces water molecules that interact with the electrode surface, reducing surface charge density and output performance [ 24 , 25 ]. Such factors, along with other environmental conditions, play a critical role in determining the stability, efficiency, and long-term reliability of TENG systems. The consistent performance observed, with minimal degradation even after 1,000,000 cycles, is a notable achievement that reinforces the TENG’s potential for robust and long-lasting energy-harvesting solutions. Combined with insights from prior studies on environmental influences and the SEM analysis of surface wear, the results highlight the TENG’s suitability for long-term use in various energy-harvesting applications, making it a viable and sustainable solution for powering small electronic devices and sensors. 3.5. Application The free-standing triboelectric nanogenerator (FR-TENG) has a notable practical use in providing power to small electronic devices such as light-emitting diodes (LEDs) through a continuous direct current (DC) output. The operational principle of the TENG is centered on the transformation of mechanical energy into electrical energy by the initiation of frictional interaction between two separate materials. The TENG effectively harnesses and converts mechanical energy, resulting in a uniform output voltage in the form of direct current. This direct current is capable of providing uninterrupted illumination to an LED, ensuring a consistent and uninterrupted light source. The presented practical demonstration of the TENG technology showcases its considerable potential as an autonomous energy source, specifically well suited for sustaining low-power electronic devices. TENGs provide a sustainable and eco-friendly option for powering devices in various environments. They offer a forward-thinking solution for energy generation and usage, enabling more efficient and environmentally sensitive electronic applications. The output voltage possesses the capability to illuminate some LEDs consistently and continuously, as shown in the Supplementary Materials, including Video S2, and Supplementary Figure S7 and Figure 5 b. This is attributed to the alternative current generated by this study, which is subsequently utilized to illuminate some LEDs. In order to demonstrate the capabilities of the DHLR-TENG as an energy provider, a capacitor and commercially available LEDs were linked to the DHLR-TENG using a full wave rectifier, as seen in Figure 5 a. In addition, the DHLR-TENG powered the red commercial LEDs, causing them to illuminate simultaneously, as seen in Figure 5 b. The DHLR-TENG is very efficient at converting mechanical energy into electrical energy. The capacitor linked to the DHLR-TENG saw rapid charging when the DHLR-TENG was activated, as seen in Figure 5 c. The voltage across the capacitors, with a capacitance of (100, 47, 4.7, 2) µF, on the 4.7 µF surpassed 110 V after 240 s. The DHLR-TENG has been shown to efficiently transfer linear mechanical energy into electric energy, making it suitable for powering tiny electronic devices."
} | 8,860 |
35349220 | PMC9328741 | pmc | 1,451 | {
"abstract": "Summary Bio‐based 5‐hydroxymethylfurfural (HMF) serves as an important platform for several chemicals, among which 2,5‐furan dicarboxylic acid (FDCA) has attracted considerable interest as a monomer for the production of polyethylene furanoate (PEF), a potential alternative for fossil‐based polyethylene terephthalate (PET). This study is based on the HMF oxidizing activity shown by Mycobacterium sp. MS 1601 cells and investigation of the enzyme catalysing the oxidation. The Mycobacterium whole cells oxidized the HMF to FDCA (60% yield) and hydroxymethyl furan carboxylic acid (HMFCA). A gene encoding a novel bacterial aryl alcohol oxidase, hereinafter Mycsp AAO, was identified in the genome and was cloned and expressed in Escherichia coli Bl21 (DE3). The purified Mycsp AAO displayed activity against several alcohols and aldehydes; 3,5 dimethoxy benzyl alcohol (veratryl alcohol) was the best substrate among those tested followed by HMF. 5‐Hydroxymethylfurfural was converted to 5‐formyl‐2‐furoic acid (FFCA) via diformyl furan (DFF) with optimal activity at pH 8 and 30–40°C. FDCA formation was observed during long reaction time with low HMF concentration. Mutagenesis of several amino acids shaping the active site and evaluation of the variants showed Y444F to have around 3‐fold higher k cat /K m and ~1.7‐fold lower K m with HMF.",
"introduction": "Introduction Oxidation of 5‐hydroxymethylfurfural (HMF), a dehydration product of C6 sugars, has been a reaction of interest to provide oxidized derivatives, especially 2,5‐furan dicarboxylic acid (FDCA) for use as building block for the polyester polyethylene furanoate (PEF), a biobased alternative to the widely used polyethylene terephthalate (PET). Replacement of one tonne of PET with PEF is expected to reduce the greenhouse gas emission by 30–50% (Davidson et al ., 2021 ). Polyethylene furanoate has also the advantage of possessing better barrier, thermal and mechanical features (Sousa et al ., 2015 , 2016 ). Both HMF and FDCA are identified by the US Department of Energy to be among the important biobased platform chemicals with no negative effect on environment and human health (Werpy et al ., 2004 ; Bozell and Petersen, 2010 ; Sousa et al ., 2015 ; Chen et al ., 2016 ; Troiano et al ., 2020 ). Besides replacing terephthalic acid, FDCA can also replace adipic acid (Sousa et al ., 2015 , 2016 ; Zhang and Deng, 2015 ; Yuan et al ., 2018 , 2020 ) and can serve as a platform to produce other important chemicals such as succinic acid, 2,5‐diformyl furan (DFF), 2,5‐bis(hydroxymethyl)furan (BHMF) and 2,5‐(bishydroxymethyl) tetrahydrofuran (Werpy et al ., 2004 ; Zhang et al ., 2015 ). 2,5‐Furan dicarboxylic acid production from HMF involves three oxidation steps and can proceed via two different pathways (Scheme 1 ). In the first pathway (A), the alcohol group of HMF (1) is oxidized to an aldehyde giving 2,5‐diformylfuran (DFF) (2 * ) , followed by further oxidation of the two aldehyde groups in DFF to FDCA (5) through 5‐formyl‐2‐furoic acid (FFCA) (4) , while the second pathway (B) involves oxidation of the aldehyde group of HMF to 5‐hydroxymethyl‐2‐furoic acid (HMFCA) (3) , followed by further oxidation of the hydroxyl group of HMFCA to FDCA via FFCA (Dijkman, 2015 ; Dijkman et al ., 2015 ; Troiano et al ., 2020 ; Yuan et al ., 2020 ). Several reports on the oxidation of HMF via chemocatalysis, (photo‐)electrochemical catalysis or biocatalysis are available (Ribeiro and Schuchardt, 2003 ; Chadderdon et al ., 2014 ; Zhang and Deng, 2015 ; Carro et al ., 2018a ; Sajid et al ., 2018 ; Sayed et al ., 2020 ; Troiano et al ., 2020 ; Yuan et al ., 2020 ; Kawde et al ., 2021 ). Biocatalytic oxidation of HMF and other aliphatic and aromatic alcohols, diols and polyols using enzymes or microbial cells offers a facile and selective route of reaction under mild conditions. Many microorganisms including bacteria and fungi have shown the ability to metabolize HMF to different products including FDCA (Yuan et al ., 2020 ). In some microorganisms, the HMF oxidizing activity has been attributed to HMF oxidase (HMFO), which is expected to form FDCA as the final product (Dijkman, 2015 ; Troiano et al ., 2020 ; Vinambres et al ., 2020 ). Besides HMFO, other reports describe the use of a cascade of enzymes for achieving oxidation of HMF to FDCA. For example, a cascade of galactose oxidase M3‐5 variant, that oxidizes HMF to DFF, and Escherichia coli periplasmic aldehyde oxidase ABC, oxidizing DFF to FFCA, and consequently to FDCA has been reported (McKenna et al ., 2017 ). This pathway was further developed by including horseradish peroxidase and lipase B from Candida antarctica (McKenna et al ., 2015 , 2017 ). The potential drawbacks of an enzyme cascade system could be the differences in the optimum reaction conditions for the enzymes involved and the cost of enzymes production (Troiano et al ., 2020 ). Scheme 1 Biocatalytic reaction pathways (A and B) for production of FDCA from HMF by three‐step oxidation [O]. The pathway B involves a reduction [H] step prior to oxidation. Glucose–methanol–choline oxidoreductase (GMC) family includes flavoprotein oxidoreductases that are classified into several subfamilies acting on a wide range of substrates from short alcohols such as methanol and choline to more complex substrates such as glucose, aromatic alcohols, fatty alcohols among others (Sützl et al ., 2019 ; Aleksenko et al ., 2020 ; Savino and Fraaije, 2021 ). According to the Pfam database, GMC oxidoreductases include two domains; the first is Pfam00732 with a typical GxGxxG/A motif, an indicator for the Rossman fold, and responsible for binding the adenosine diphosphate (ADP) moiety of the flavin adenine dinucleotide (FAD) prosthetic group, while the second is the less conserved C‐terminal domain, which usually contributes to the substrate specificity of the enzyme. A characteristic of these oxidoreductases is a conserved histidine in the active site that initiates the oxidation via proton subtraction from the substrate and assists in FAD re‐oxidation by the molecular oxygen. Among the subfamilies of GMC oxidoreductases are aryl alcohol oxidases (AAOs) that have been reported to oxidize HMF (Serrano et al ., 2020 ; Lappe et al ., 2021 ). Aryl‐alcohol oxidases (AAOs; EC 1.1.3.7) catalyse the oxidative dehydrogenation of primary alcohols from aliphatic unsaturated or aromatic alcohols to aldehydes using oxygen as a final electron acceptor resulting in production of hydrogen peroxide (H 2 O 2 ) (Urlacher and Koschorreck, 2021 ). Aryl alcohol oxidases have also been classified in the CAZy (the carbohydrate‐active enzymes) database into subfamily AA3_2, where AA3 refers to Auxiliary Activity (AA) superfamily that comprises four subfamilies with different GMC oxidoreductases (Sützl et al ., 2018 ). These enzymes are known to play an important role in biomass degradation by way of supplying hydrogen peroxide to the ligninolytic peroxidases. \n Mycobacterium sp. MS1601 is a promising source of enzymes for oxidation reactions, which is evident from the abundance of gene sequences encoding oxidases and dehydrogenases in its genome (Sayed et al ., 2017a ). The organism has shown unique activity for selective oxidation of branched polyols such as trimethylolpropane (TMP) to 2,2‐bis(hydroxymethyl)butyric acid (BHMB) with high selectivity and yield (Sayed et al ., 2016 , 2017a ) and was also recently shown to oxidize 1,6‐hexanediol to 6‐hydroxyhexanoic acid and adipic acid (Pyo et al ., 2020 ). This report presents a study demonstrating the ability of Mycobacterium sp. MS1601 to oxidize HMF to FDCA and other oxidized intermediates, identification of a novel aryl alcohol oxidase ( Mycsp AAO) in the bacteria that efficiently catalyses oxidation of HMF, heterologous production of the enzyme in Escherichia coli and determination of its substrate scope. Furthermore, through sequence and structure analysis based on homology modelling, an important mutation was generated that resulted in improved HMF oxidation and also revealed FDCA production.",
"discussion": "Discussion The genome sequence of Mycobacterium sp. MS1601 shows the presence of many gene sequences that are annotated as oxidative enzymes (296 dehydrogenases; 168 oxidoreductases and 53 oxidases) (Sayed et al ., 2017b ). The organism has so far been used as an efficient biocatalyst for the highly selective oxidation of polyhydric alcohols (Sayed et al ., 2016 , 2017b ), and in this study for the oxidation of HMF. Interestingly, oxidation of HMF by Mycobacterium sp. MS1601 seems to follow different routes depending on the carbon source used for growing the cells. While the cells grown on glycerol, oxidized HMF to FDCA via FFCA (pathway A in Scheme 1 ) (Fig. 1A ), a typical pathway used by HMFO from Methylovorus sp. MP688 (Dijkman and Fraaije, 2014 ), oxidation by the cells grown on glucose or sorbitol followed an alternative route in Scheme 1 yielding BHMF and HMFCA that was not oxidized further (Fig. 1B and C ). In the latter case, the aldehyde group of HMF is either oxidized directly to HMFCA or after first being reduced to BHMF. Activation of the enzyme(s) responsible for FDCA production with glycerol was in agreement with the previous report involving selective oxidation of trimethylolpropanol (TMP) to its corresponding carboxylic acid by Mycobacterium sp. MS1601 (Sayed et al ., 2017b ). Based on the similarity of the HMF oxidation profile by the glycerol‐activated cells to that by Metsp HMFO, the sequence of the Methylovorus sp. enzyme was used for probing the Mycobacterium sp. MS1601 genome for homologous enzymes. The putative gene sequence with > 30% identity with Metsp HMFO was chosen for this study. Since the identity to Metsp HMFO was low, further sequence analysis was done for the chosen gene from Mycobacterium sp. MS1601 using Pfams and Uniport databases, confirming that the chosen sequence encodes an enzyme belonging to GMC family, to which Metsp HMFO also belongs (Dijkman and Fraaije, 2014 ). However, the first 250 hits from Uniport analysis with high identity were for other GMC members that are not yet experimentally characterized. This observation confirmed the novelty of the chosen protein from Mycobacterium sp. MS1601. The codon‐optimized gene of the above‐mentioned gene from Mycobacterium sp. MS1601was expressed in a soluble form in E. coli BL21 (DE3) at 15°C. The purified Mycsp AAO was shown to be a wide spectrum GMC oxidase enzyme that utilizes oxygen as the final electron receptor and prefers compounds with an aromatic ring as 3,4‐DMBA>HMF>DFF>BHMF>MBA>4‐HBA> furfural>BA compared with aliphatic substrates (Fig. 3A ). The high specific activity of Mycsp AAO against 3,4‐DMBA, MBA and BA is in agreement with the activity of bacterial aryl alcohol oxidase from Sphingobacterium sp. ATM (Tamboli et al ., 2011 ). Our findings indicate that the new GMC member from Mycobacterium sp. MS1601 belongs to the subfamily of aryl alcohol oxidases, also displaying high activity against HMF and its derivatives including DFF and BHMF (Fig. 3A ). Reactions with DFF and furfural as substrates suggest that the enzyme is active also on the aldehyde groups; the oxidation of DFF and BHMF resulted in the formation of FFCA. Aryl alcohol oxidases in rare cases have been shown to catalyse the oxidation of aromatic aldehyde substrates to the corresponding acids (Urlacher and Koschorreck, 2021 ), which is proposed to proceed via the gem ‐diols formed by aldehyde hydration and hence following the mechanism similar to alcohol oxidation (Ferreira et al ., 2010 ). Mycsp AAO displayed very low activity against FFCA and HMFCA as observed even for other AAOs (Carro et al ., 2015 ; Serrano et al ., 2019 ; Viña‐Gonzalez et al ., 2020 ; Lappe et al ., 2021 ), which may be due to interference by the carboxylic group on the binding of the substrate to the enzyme active site. More insights were gained from comparison of Mycsp AAO to the closest fungal aryl alcohol oxidase PDB 3FIM . A characteristic loop covering the active site has been described in the case of PDB 3FIM , and interestingly a similar loop has been identified in the Mycsp AAO model. Most importantly, the previously reported F397 in 3FIM has an equivalent residue F343 in Mycsp AAO (Fig. 5 ). This phenylalanine residue is proposed to attach to the substrates and to be involved in a significant loop displacement necessary to place the substrate inside the active site (Carro et al ., 2018b ; Serrano et al ., 2020 ). Moreover, the analysis of Mycsp AAO 3D model showed a narrow active site due to the presence of two tyrosine residues (Y443 and Y444) located very close to the catalytic H445. These two tyrosine residues may be expected to cause interference for the bulkier substrates to be accepted by the enzyme. \n Mycsp AAO‐Y444F exhibits favourable kinetic parameters against HMF, 4HBA, and furfural compared with the wild‐type enzyme (Table 1 and Fig. S11 ). With HMF, the improved activity of the enzyme variant was ascribed to the significantly higher k cat/ K m and lower K m values (Table 1 ). The higher activity of Mycsp AAO‐Y444F can be further related to the observations recorded for the fungal AAO (PDB 3FIM ). F444 in the enzyme variant is exactly equivalent to the F501 in the fungal AAO (3FIM) structures, which is proposed to be a major player in the substrate activity and specificity. This phenylalanine has also been described to control the access of molecular oxygen to the FAD molecule in the active site (Hernandez‐Ortega et al ., 2011 ; Ferreira et al ., 2015 ; Carro et al ., 2018b ) Hence, Mycsp AAO‐Y444F denotes a back‐to‐ancestral mutation since phenylalanine residue is conserved in many GMC oxidoreductases and specifically in most of the AAOs that have been characterized up to date (Chakraborty et al ., 2014 ; Galperin et al ., 2016 ; Jankowski et al ., 2020 ). Both Mycsp AAO‐WT and Mycsp AAO‐Y444F variants, when tested against a lower concentration of HMF (10 mM in contrast to 31.7 mM) revealed the formation of FDCA over a long incubation time, the Y444F variant giving the higher activity (Fig. 7 ). These observations are in agreement with that reported previously for HMFOs and AAOs (Cavener, 1992 ; Dijkman et al ., 2014 ; Serrano et al ., 2019 ; Vinambres et al ., 2020 ). Very recently, Sánchez‐Ruiz et al . ( 2021 ) reported that FFCA, especially at higher concentration (over 15 mM) has an irreversible inhibition against HMFOs (Sánchez‐Ruiz et al ., 2021 ), which could explain FDCA formation from 10 mM HMF (Fig. 7 ), and not during the oxidation of 31.7 mM HMF with Mycsp AAO‐WT (Fig. 3 ). This finding was further confirmed from screening Mycsp AAO‐WT, ‐Y444F and L298A variants against concentrations of HMF (Fig. S10 ), that showed that increasing the concentration of HMF over 10 mM reduced the initial activity of all the variants although the Y444F variant showed a significantly higher activity. In summary, this report indicates that Mycobacterium sp. MS1601 is an interesting source of enzymes for oxidation of HMF. While the whole cells showed the formation of products with different levels of oxidation including FDCA, the analysis of the selected gene sequence from the genome of Mycobacterium sp. MS1601 indicates that the HMF oxidizing enzyme represents a novel bacterial AAO that oxidizes HMF preferably to FFCA but even to FDCA at lower substrate concentration and longer reaction time. The site‐directed mutagenesis studies on Mycsp AAO indicated that the Y444 in the enzyme’s active site could be replaced by the conserved phenylalanine (F) for improved enzymatic activity. Further improvement through introducing additional rational mutations in Mycsp AAO are required to increase the capacity of Mycsp AAO to oxidize HMF completely to FDCA in a shorter reaction time and higher substrate concentration. On the other hand, being host to several oxidative enzymes, it is more likely that Mycobacterium sp. MS1601 cells utilize an aldehyde dehydrogenase/oxidase activity for oxidation of FFCA in vivo ."
} | 4,074 |
28836729 | PMC5763382 | pmc | 1,453 | {
"abstract": "Summary For the anaerobic remineralization of organic matter in marine sediments, sulfate reduction coupled to fermentation plays a key role. Here, we enriched sulfate‐reducing/fermentative communities from intertidal sediments under defined conditions in continuous culture. We transiently exposed the cultures to oxygen or nitrate twice daily and investigated the community response. Chemical measurements, provisional genomes and transcriptomic profiles revealed trophic networks of microbial populations. Sulfate reducers coexisted with facultative nitrate reducers or aerobes enabling the community to adjust to nitrate or oxygen pulses. Exposure to oxygen and nitrate impacted the community structure, but did not suppress fermentation or sulfate reduction as community functions, highlighting their stability under dynamic conditions. The most abundant sulfate reducer in all cultures, related to Desulfotignum balticum , appeared to have coupled both acetate‐ and hydrogen oxidation to sulfate reduction. We describe a novel representative of the widespread uncultured candidate phylum Fermentibacteria (formerly candidate division Hyd24‐12). For this strictly anaerobic, obligate fermentative bacterium, we propose the name ‘ U Sabulitectum silens ’ and identify it as a partner of sulfate reducers in marine sediments. Overall, we provide insights into the function of fermentative, as well as sulfate‐reducing microbial communities and their adaptation to a dynamic environment.",
"conclusion": "Conclusion The transient exposure to oxygen or nitrate changed the microbial community structure, and impacted the magnitude of net sulfide production as a community function, yet had a minor effect on microbial community composition. This shows that the communities of Janssand intertidal sediments contained organisms that were well adjusted for each of these scenarios, diverting the flow of carbon and energy through the trophic network based on the available electron acceptors. The treatment with oxygen or nitrate did not cause the community to shift to an alternative stable state (Shade et al ., 2012 ). Community stability during the exposure to oxygen or nitrate was enabled by the increased expression of genes involved in oxidative and general stress protection. The stable coexistence of several fermenters and sulfate reducers with nitrate reducers or aerobic respirers supports the recent finding that microbial communities are assembled based on rules that go beyond those of the classical redox tower (Chen et al ., 2017 ).",
"introduction": "Introduction Around 30% of the total oceanic phytoplankton‐derived primary production occurs along the continental margins (Walsh, 1991 ) and up to 50% of this organic matter reaches the surface of shallow coastal sediments. This organic matter can be re‐mineralized by the microorganisms in the surface sediment using a broad suite of electron acceptors, such as oxygen, nitrate, metal oxides and sulfate (Henrichs and Reeburgh, 1987 ; Canfield et al ., 1993 ; Janssen et al ., 2005 ). It has been estimated that about 50% of the total organic carbon mineralization in shallow sediments (Jørgensen, 1982 ) and salt marsh sediments (Howes et al ., 1984 ) and up to 35% of the total mineralization in intertidal flats (Billerbeck et al ., 2006 ) is coupled to sulfate reduction. Yet, despite the global importance of sulfate reduction, the ecophysiology of the involved microorganisms and their environmental controls are poorly constrained. The sulfate‐reducing microbial populations in the surface sediments of intertidal flats are exposed to pulses of oxygen approximately twice daily, because of tidal cycling. In addition, the communities may be regularly exposed to pulses of nitrogen from riverine sources (van Beusekom, 2005 ; Boyer et al ., 2006 ). It is thus very likely that sulfate reducers and also other key anaerobic functional types, such as fermenters, are adapted to these ecosystem dynamics and survive exposure to oxygen and nitrate. Generally, the availability of oxygen leads to a lower relative importance of sulfate reduction, because electron acceptors tend to be consumed in a thermodynamically determined order (the redox cascade). According to this order, oxygen is used first, followed by nitrate, manganese and iron oxides and finally sulfate (Froelich et al ., 1979 ). Hence, sulfate is thought to be the predominant electron acceptor only in the anoxic layers after other electron acceptors are depleted. Sulfate‐reducing bacteria are often strict anaerobes and couple the oxidation of molecular hydrogen or organic compounds to the complete reduction of sulfate to sulfide (Muyzer and Stams, 2008 ; Rabus et al ., 2013 ). Nevertheless, sulfate‐reducing bacteria were detected throughout the whole sediment of an intertidal flat, including the aerobic and denitrifying zones (Llobet‐Brossa et al ., 2002 ; Mußmann et al ., 2005 ; Gittel et al ., 2008 ). In addition, it was found that intertidal flats are a sink for riverine and atmospheric nitrogen (Gao et al ., 2012 ), with the microbial conversion of nitrate to ammonium or dinitrogen (Marchant et al ., 2014 ) and the internal storage of nitrate in benthic diatoms (Stief et al ., 2013 ) being widespread and important processes. Also, nitrite is common in intertidal flats and it was found that some sulfate reducers, like Desulfovibrio desulfuricans , are able to grow on hydrogen coupled to ammonification of nitrate or nitrite (Dalsgaard and Bak, 1994 ). Although, much progress has been made in understanding the key processes and populations in intertidal sediments, for example, elucidating the environmental controls of nitrate respiration (Kraft et al ., 2014 ) and the impact of chemical gradients on community structure (Chen et al ., 2017 ), the trophic network defining combined fermentation and sulfate reduction remains largely unknown. A major challenge in microbial ecology in general is to understand the dynamics of an ecosystem and its impact on the microbial communities (Widder et al ., 2016 ). This can be addressed, for example, by investigating the resistance and resilience of microbial communities to perturbations (Shade et al ., 2012 ; Lee et al ., 2017 ), or by investigating the response of microbial communities to recurring events (Ward et al ., 2017 ). Simple model systems are a promising tool to disentangle community dynamics and constrain cause and effect (Widder et al ., 2016 ). The effects of oscillating redox conditions on microbial communities and on the remineralization of organic matter have been studied, for example, in marine sediments (Aller, 1994 ; Sun et al ., 2002 ), wetland soils (Pett‐Ridge and Firestone, 2005 ; DeAngelis et al ., 2010 ) and rhizospheres (Nikolausz et al ., 2008 ), showing that redox conditions impact the efficiency of organic matter degradation, and that aerobic and anaerobic microbial populations coexist in these ecosystems. Here, we investigated the effect of the tidal cycle, that is, diurnal redox oscillations, on fermentation coupled to sulfate reduction as a community function. We setup defined continuous cultures, which created a homogeneous microbial habitat that selected for communities of sulfate‐reducing and fermentative bacteria. We inoculated the cultures with biomass from tidal flat sediments that were exposed to a tidal cycle. The effect of diurnal exposure to oxygen or nitrate on the microbial activity and community structure was examined by combined chemical, metagenomic and transcriptomic analyses. Using this setup, we gained insights into fermentation coupled to sulfate reduction and the involved trophic networks, as well as into the ecophysiology of an uncultured candidate phylum.",
"discussion": "Results and discussion Physiology of the continuous cultures Six cultures were inoculated with cell suspensions obtained from intertidal sediments of the Janssand tidal flat. The cultures were continuously supplied with sulfate as electron acceptor and a mixture of glucose, seven different amino acids and acetate as electron donors. This mixture was chosen to stimulate the growth of a wide range of organisms, and represents compounds of decaying biomass, which is the main organic carbon source in marine sediments. After 2 days, sulfide was detected in all cultures and increased during the first 150 days, to a concentration of 2–6 mM (Fig. 1 ). All six cultures were incubated for 20 days under identical sulfate‐only conditions to establish anaerobic communities that carry out fermentation‐coupled sulfate reduction. From day 21 onward, four of the cultures were treated with oxygen or nitrate pulses, while two cultures remained untreated. Oxygen was supplied to two replicate cultures (Oxy‐1 and Oxy‐2) for 30 min twice daily, by sparging the cultures with air. Two replicate cultures (Nit‐1 and Nit‐2) were supplied with a nitrate solution for 7 min twice daily. The final two replicate cultures (Con‐1 and Con‐2) did not receive any additional electron acceptor and served as an untreated control. The biomass in each of the cultures remained stable during the entire experiment (OD 600 : ∼0.15; Supporting Information Fig. S2). The sulfide concentrations in the cultures, in combination with the nature of the provided carbon sources, indicated that we selected for a syntrophic community of fermenters and sulfate reducers. It was expected that the fermenting bacteria convert glucose and amino acids to short‐chain fatty acids, lactate, alcohols or hydrogen. These could then be used as carbon sources and/or electron donors by the sulfate‐reducing bacteria (Rabus et al ., 2013 ). Figure 1 Sulfide concentrations in the replicate cultures treated with oxygen (A), treated with nitrate (B) and in the untreated control with sulfate as sole electron acceptor (C). Duplicate measurements of each culture are shown as triangles and circles, the red line depicts the mean of four measurements, the red ribbon represents standard error of the mean. The start of the treatments is indicated by dashed lines, sampling time points for metagenomics and metatranscriptomics are indicated by dotted lines. We characterized the cultures in detail on day 311 (Oxy‐1 and Oxy‐2), day 327 (Nit‐1 and Nit‐2) and day 300 (Con‐1 and Con‐2). During the air supply, the oxygen concentration was stable at around 1.3% air saturation (3.1 µM), while the sulfide concentration decreased by 0.7 ± 0.4 mM (Supporting Information Fig. S3A). In the cultures supplied with nitrate, sulfide concentrations did not decrease (Supporting Information Fig. S3B). The addition of 0.2 millimol nitrate during each oxidant addition cycle yielded a maximal nitrate concentration of 0.5 mM; the added nitrate was metabolized within ∼200 min after termination of the supply (Fig. S3D). In both treatments, we observed the transient production of elemental sulfur. In the oxygen treatment, we measured sulfur concentrations of up to 0.8 mM immediately after the start of aeration, decreasing to ∼0.1 mM within 2–4 h (Supporting Information Fig. S3E). In the nitrate cultures, sulfur was increasing from ∼0.1 mM to up to 0.4 mM within 2 h, and decreased to ∼0.1 mM within 4 h after the start of the treatment (Supporting Information Fig. S3F). Using 15 N‐labelled nitrate we found no production of 15 N‐labelled N 2 , which indicated that ammonia may have been the end‐product of nitrate reduction. Ammonia production could not be assessed directly because of the high background ammonia concentration that resulted from ammonification of the supplied amino acids. Over the 1 year incubation, transient oxygen supply yielded the lowest average sulfide concentrations (2.3 ± 0.3 mM; Fig. 1 A), followed by the cultures that received nitrate (4.2 ± 0.6 mM, Fig. 1 B) and the untreated control cultures (6.3 ± 0.7 mM, Fig. 1 C). Fluctuations in sulfide concentration were highest in the nitrate treatment and lowest in the cultures that were not exposed to oxygen or nitrate. Yet, the cyclic exposure to oxygen or nitrate did not suppress sulfide production (Supporting Information Fig. S3), and thus sulfate reduction, as a community function. Aerobic respiration and ammonification coincided with a decreased magnitude and stability of the sulfide concentration, likely due to microbial re‐oxidation of sulfide, or due to competition between sulfate reducers, aerobes or nitrate ammonifiers. Microbial communities and their response to cyclical exposure to oxygen and nitrate After around 300 days of cultivation, we sequenced the metagenomes of the six continuous cultures. We hypothesized that cyclical exposure to oxygen or nitrate alters resource access to create ecological niches that resemble those present in permeable intertidal sediments. Each treatment would thus select for a different microbial community. Indeed, the community structure was different between the treatments (Fig. 2 and Supporting Information Fig. S4). Yet, all treatments and cultures had a similar microbial community composition (Fig. 2 ). The nitrate treatment caused the strongest microbial community response and favoured fermentative organisms that were less abundant in other treatments, such as Defluviitaleaceae (bin K) and certain Spirochaeta (bin M/bin N) (Figs 2 and 4 A and Supporting Information Fig. S5). Nitrate is energetically favourable over sulfate as an electron acceptor and does not react abiotically with sulfide under the cultivation conditions. Dissimilatory nitrate reduction to ammonium could explain the lower sulfide concentrations as compared to the control, as well as the high relative sequence abundances of organisms affiliating with Desulfuromusa . Moreover, each of the replicate cultures Nit‐1 and Nit‐2 selected for communities of different structure, although they experienced the same selective pressure. The cultures exposed to oxygen were also different from each other, based on the relative abundances of Clostridia (bin H, bin J) and Psychromonas (bin B). Overall, the oxygen treatment seemed to have a lesser impact on the communities than the treatment with nitrate, possibly because oxygen was relatively quickly removed via abiotic reaction with sulfide. In turn, the removal of sulfide may have stimulated the sulfate reducers, causing relative sequence abundances that were higher than in the control treatment. The communities in the untreated replicate controls Con‐1 and Con‐2 had a nearly identical community structure after 300 days of cultivation (Fig. 2 and Supporting Information Fig. S4). Despite the different communities in the nitrate and oxygen treated cultures, fermentation coupled to sulfate reduction was not greatly affected as a community function, as inferred from gene expression (Fig. 3 , Dataset 1) and the production of sulfide (Fig. 1 ). This functional similarity may be explained by the presence of fermentative populations that are phylogenetically different, but perform similar metabolisms (Allison and Martiny, 2008 ). Figure 2 Estimated relative abundances of bins in cultures treated with oxygen (Oxy‐1, Oxy‐2), with nitrate (Nit‐1, Nit‐2) and in untreated controls (Con‐1, Con‐2). The bins were classified to genus level. Populations that affiliated with genera lacking a cultured representative are marked with #. For these bins, we report the closest taxonomically assigned, phylogenetic level, for example, bin K affiliated with an uncultured genus in the family Defluviitaleaceae . Taxonomic assignment was based on the SILVA small ribosomal subunit reference database (SSURef, v123). Relative abundances were obtained by mapping metagenomic sequence reads to the assembled contigs of each bin. The phylogeny of most bins is provided (bin O: Fig. 6 , bin A‐N: Supporting Information Fig. S6‐S9). Figure 3 Metabolic capabilities of the bins (A, B, D‐I, K‐P) based on key genes. Circle size represents relative transcriptional activity averaged over all ten transcriptomes. The absence of a circle shows that gene transcription was not detected in any of the treatments and cultures. Note: For reasons of visualization, all relative transcriptional activities above 3.0 are shown as ≥3.0. Metabolic capabilities were not assessed for the bins C, J, Q and R due to scarce metagenomic data (Table 1 ). Figure 4 Schematic of the trophic network of key populations (A, B, D‐I, K‐P), and transcriptional changes of stress response genes. A. Most abundant obligate fermentative heterotrophs (Fermenters), sulfate‐reducing bacteria (Respirers I) and associated respiratory heterotrophs (Respirers II) in the three different conditions. The network is based on metagenomic and/–transcriptomic data. All 14 shown bins were present in all cultures. Circle size represents estimated relative abundance. Only one organism (bin D) was abundant in all cultures. Arrows depict key pathways that occur in all (grey), two (blue) or one condition (red). Bins C, J, Q and R were not included due to scarce metagenomic data (Table 1 ). B. Change of gene transcription caused by the treatment with oxygen or nitrate. Transcriptional changes of genes involved in redox‐ and general stress response were among the highest of all detected genes. Values are log2‐transformed ratios of gene transcription in replicate cultures after and before the treatment, that is, a value of 1 means that gene transcription was twice as high after the treatment than before the treatment; a value of −2 means a fourfold decrease in transcription. The guilds are represented by the 11 most abundant bins for clarity. To study the communities in detail, we focused on metagenomic bins with relatively long contigs, relatively equal coverage distribution and a consistent taxonomic signature (Table 1 ). These bins can be considered provisional genome sequences that represented the genetic repertoire of the most abundant populations (Fig. 2 ). All enriched populations were present in the inoculum, yet with low relative sequence abundances (Table 1 ), as the in situ community experienced different conditions and was thus dominated by different clades (Lenk et al ., 2011 ). The 16S rRNA gene sequences of 15 of the bins were used for a detailed analysis of their taxonomic and phylogenetic affiliation (Supporting Information Figs S6–S9). Proteobacteria , Firmicutes and Spirochaeta were the phyla with the overall highest relative abundance (Fig. 2 ) and dominated the metagenomes of each treatment. Of all 18 bins, only bin D affiliating with sulfate‐reducing Desulfotignum had a high relative abundance across all treatments and cultures, while the other 17 bins were either always relatively low abundant or thrived only under certain conditions (Fig. 2 ). Most organisms were predicted to have a fermentative (bin G–P) or sulfate‐reducing (bin D and F) metabolism (Fig. 3 ). Bin A–C, E, Q and R belonged to heterotrophs that were selected in the oxygen or nitrate treatments and hence a respiratory metabolism is most likely. For the bins C ( Thioalkalispira ), Q and R ( Bacteroidetes ) meaningful metabolic inferences were impossible because the binned metagenomic data were too scarce (Table 1 ), bin J was not further analysed due to a high percentage of contamination (40%). Table 1 Properties of the 18 bins obtained from metagenomes of the six continuous cultures (Oxy‐1/2, Nit‐1/2, Con‐1/2). Bin A B C D E F G H I Affiliation Rhodo bacterales Altero monadales Chromati ales Desulfo bacterales Desulfuro monadales Desulfo vibrionales Firmi cutes Clostridi ales Clostridi ales Size (Mb) 3.85 3.08 1.55 4.77 4.10 4.11 4.27 4.45 4.59 Number of contigs 3272 1897 1650 594 237 885 2310 351 562 N50 contig length (kb) 1.5 2.5 1.1 126 58.9 14.6 2.8 27.6 107 GC content (%) 60.1 50.6 47.6 51.9 50.2 53.9 40.7 37.6 39.9 Number of CSCGs 131 115 71 132 131 138 121 112 119 Number of tRNAs 41 34 16 43 50 60 62 31 51 Completeness (%) 71.1 73.2 33.5 84.9 89.3 86 72.9 74.2 79.4 Contamination (%) 25 7 3 15 11 8 9 8 20 \n In situ relative abundance, Mean ± S.D. (%) 0.05 ± 0.026 0.035 ± 0.022 0.013 ± 0.009 0.041 ± 0.024 0.048 ± 0.028 0.062 ± 0.036 0.002 ± 0.001 0.02 ± 0.012 0.031 ± 0.02 \n Bin \n \n J \n \n K \n \n L \n \n M \n \n N \n \n O \n \n P \n \n Q \n \n R \n Affiliation Clostridi ales Clostridi ales Spiro chaetales Spiro chaetales Spiro chaetales Fermenti bacteria Anaerolineales Bacteroi detes Bacteroi detes Size (Mb) 6.69 3.86 3.36 3.57 3.78 2.92 2.51 1.68 2.30 Number of contigs 3459 248 40 317 691 96 1147 1661 2657 N50 contig length (kb) 3.3 118 144 160 10.1 222 6.0 1.2 0.99 GC content (%) 34.5 45.4 53.5 36.8 35.7 56.6 52.8 43.5 42.1 Number of CSCGs 182 126 129 88 93 98 128 69 100 Number of tRNAs 41 39 47 47 36 39 36 11 11 Completeness (%) 86.1 90.7 88 82.1 75.3 76.9 86.7 40.5 42.4 Contamination (%) 40 18 10 15 17 5 9 2 21 \n In situ relative abundance, Mean ± S.D. (%) 0.019 ± 0.01 0.023 ± 0.011 0.021 ± 0.015 0.021 ± 0.015 0.016 ± 0.011 0.019 ± 0.01 0.015 ± 0.009 0.005 ± 0.004 0.012 ± 0.008 CSCG: Conserved single‐copy gene. S.D.: Standard deviation of the mean. To infer the metabolic activity of key organisms in the cultures and in response to the applied treatment, we sequenced ten metatranscriptomes after 300 (control), 311 (oxygen treatment) and 327 (nitrate treatment) days of incubation. We sampled 1 h before the treatment and directly after the treatment ended (Supporting Information Fig. S3C, D). This sampling strategy enabled us to sketch a trophic network in the cultures (Fig. 4 A) and look at differences in their gene expression caused by the treatments. The transcriptional activity mirrored relative abundance, such that populations that were abundant in a treatment, were also most active. The differences in overall gene transcription before and after the treatment were not very pronounced. The relative transcription of most genes involved in anabolism, catabolism and energy metabolism showed minor changes, suggesting that after 300 days the key organisms were very well adapted to the provided cyclic environment. Consistent and large differences caused by the treatment were mainly detected for genes involved in oxidative (Fig. 4 B) and general (Fig. 4 C) stress protection. The genes that were transcribed by the populations indicated that each population had a slightly different strategy to cope with stress. Sulfate reducers Most of the sulfate reduction was likely performed by Deltaproteobacteria affiliating with Desulfotignum balticum (bin D, Supporting Information Fig. S6) and Desulfovibrio profundus (bin F, Supporting Information Fig. S6). Both organisms constitutively expressed bd‐type terminal oxidases to respire oxygen and protect oxygen‐sensitive enzymes (Ramel et al ., 2013 ). Genes encoding for sulfate adenylyltransferase ( sat ), adenylyl‐sulfate reductase ( aps ) and dissimilatory sulfite reductase ( dsr ) were also transcribed by both organisms (Fig. 3 , Dataset 1). Desulfotignum dominated all conditions based on relative abundance, yet Desulfovibrio seemed to have a higher relative transcription of dsr genes than Desulfotignum in the nitrate‐supplied and untreated cultures (Supporting Information Fig. S5). Desulfotignum transcribed NiFe(Se)‐hydrogenases (e.g., hyb ) and c‐type cytochromes, which are needed to use hydrogen as an electron donor (Heidelberg et al ., 2004 ). Desulfotignum also constitutively expressed carbon monoxide dehydrogenase and acetyl‐CoA synthase ( cooS/acsA ), the key genes in the acetyl‐CoA pathway for acetate oxidation or carbon dioxide fixation (Fig. 3 ). Both, autotrophic growth and heterotrophic growth using genes of the acetyl‐CoA pathway has been previously described for D. balticum (Kuever et al ., 2001 ). Together, the high transcriptional activity of these genes (Fig. 3 , Dataset 1) indicated that the organisms most likely activated acetate to acetyl‐CoA, which can be used as a carbon or energy source (Schauder et al ., 1988 ), while simultaneously gaining energy by H 2 oxidation and sulfate reduction (Kuever et al ., 2001 ). However, it cannot be ruled out that the organisms grew chemolithoautotrophically using the same genes in reverse (Wood‐Ljungdahl pathway), despite the excess of organic carbon sources; a possible scenario that merits further investigation. The Desulfovibrio population (bin F) also transcribed Ni/Fe hydrogenases ( hyb/hyd ) and appeared to consume hydrogen. It also transcribed genes for formate‐hydrogen lyase ( hycE ) and formate oxidation (Fig. 3 ), consistent with the physiology of many Desulfovibrio species (Barton and Fauque, 2009 ). Desulfovibrio and Desulfotignum were found to co‐occur in extreme environments such as petroleum reservoirs (Li et al ., 2017 ) and marine deep subsurface sediments (Fichtel et al ., 2012 ), their niches possibly separated by the use of different alcohols and short‐chain fatty acids (Fichtel et al ., 2012 ). The coexistence of Desulfotignum and Desulfovibrio populations in each treatment of the experiment, revealed two stable ecological niches for sulfate reducers in our cultures. Obligate fermenters All Clostridiales (bin G‐K, Supporting Information Fig. S7), Spirochaetales (bin L‐N, Supporting Information Fig. S7) and Anaerolineales (bin P) were strictly fermentative, based on their gene content and transcriptional activity. They transcribed thioredoxins, peroxiredoxins and rubredoxins to protect their enzymes against oxidative stress during oxygen or nitrate treatments (Fig. 4 B). The organisms transcribed hydrogen‐producing hydrogenases and their associated electron transfer apparatus, but lacked a respiratory chain. All fermenters transcribed genes for electron transport complexes ( rnf ), which apparently enabled them to harness a proton/sodium motive force to reduce ferredoxins by oxidizing NADH. Glucose and amino acids supplied with the medium were the main substrates, as shown by highly transcribed sugar and amino acid transporters (Dataset 1). Defluviitaleaceae (bin K) are known sugar degraders (Ma et al ., 2017 ), yet they have so far not been linked to fermentation in marine sediments. All Firmicutes (bin G–K) transcribed V‐type and F‐type ATP synthases. It was shown that F‐type ATP synthases act as sodium pumps in certain Clostridia (Ferguson et al ., 2006 ), so it is unclear whether these organisms harnessed a proton motive force to generate ATP. The three Spirochaetes only encoded a vacuolar type ATP synthase and are thus likely dependent on substrate level phosphorylation during fermentation. Transcription of acyl phosphatase and formate acetyltransferase (pyruvate‐formate lyase) suggested that acetate and formate were end‐products of fermentation, in addition to hydrogen. All three end‐products seemed to be used by the two reducers, suggesting a syntrophic relationship between fermenters and sulfate reducers. The uncultured Spirochaeta bin L (Supporting Information Fig. S8) also transcribed genes to metabolize a large number of carbohydrates. The transcriptional activity indicated that this organism is able to import diverse sugars, into the cell and shuttle them into glycolysis or the pentose phosphate way (Supporting Information Fig. S10). Based on the transcription of key metabolic genes, the organisms affiliating with Clostridiales (bin G‐I) seemed to have very similar physiologies, which was also the case for the organisms affiliating with Spirochaeta (bin L‐N) (Fig. 3 ). However, each population appeared to use slightly different glycosyl hydrolases (Supporting Information Table S1), and sets of genes involved in fermentation and energy conversion. For instance, the Clostridiales transcribed pyruvate synthase ( porC ), lactate dehydrogenase ( ldh ) and nitrite reductase ( nasD ), which the Spirochaeta did not transcribe. In turn, the Spirochaetes seemed to have a much higher expression of citrate synthase ( citA ) and isocitrate dehydrogenase ( icd ), key genes involved in the citric acid cycle (Fig. 3 ). The Firmicutes bin G exhibited high numbers of transcripts for amino acid importers and amino acid metabolism (e.g., glutamate dehydrogenase), whereas the Firmicutes bin K exhibited mainly transcripts of sugar importers and glycolysis. These differences in gene transcription indicate substrate preferences that may explain the observed coexistence of Anaerolineales , Clostridiales and Spirochaetales in our cultures and in anoxic marine ecosystems, for example, cold seeps (Dowell et al ., 2016 ; Ruff et al ., 2016 ), and hint towards metabolic complementation within the fermentative network (Xia et al ., 2014 ). Facultative aerobes and nitrate respirers Populations affiliating with Alphaproteobacteria and Gammaproteobacteria (Supporting Information Fig. S9, bins A–C) were detected in the transient oxygen and nitrate cultures and were minor constituents in the sulfate‐only cultures (Fig. 2 ). The transient exposure to oxygen and nitrate apparently selected for these organisms, which were capable of respiration. Genes encoding respiratory complexes I‐IV and genes of the citric acid cycle were present and actively transcribed in the Rhodobacterales (bin A) and Alteromonadales (bin B). Compared to the fermenters, the respiratory organisms showed low transcriptional activity of sugar and amino acid transporters. Thus, it is likely that the respiratory organisms mainly used fermentation products, such as acetate, as electron donors. Hydrogen did not seem to be a major energy source for these organisms, as transcriptional activity of hydrogenases was not detected. In contrast, the Rhodobacterales actively transcribed all sox genes that are needed for sulfide and sulfur oxidation. Both organisms transcribed genes involved in polyhydroxybutyrate (PHB) and polyphosphate metabolisms. This indicates that PHB may have accumulated under anoxic conditions driven by polyphosphate hydrolysis, and was oxidized under oxic conditions, a well‐known strategy for biological phosphorus removal (Wu et al ., 2010 ). Indeed, in the Rhodobacterales , polyphosphate kinase and poly‐beta‐hydroxybutyrate polymerase were downregulated during the period of air supply (Dataset 1). The population related to Desulfuromusa bakii (bin E), did not have or transcribe dsr genes and was apparently not performing sulfate reduction. This organism was only selected in cultures with transient nitrate supply and showed a strong global transcriptional response to nitrate availability. In response to nitrate, it transcribed genes for citric acid cycle enzymes, complex I, nitrate‐induced formate dehydrogenase ( fdn ), periplasmic nitrate reductase ( nap ) and pentaheme nitrite reductase ( nrf ). It likely performed nitrate ammonification with substrates such as amino acids, acetate and formate. Desulfuromusa bakii and related bacteria are known as sulfur‐reducing, and often facultatively fermentative bacteria (Liesack and Finster, 1994 ). Hence, in the absence of nitrate the organisms selected here may also have performed fermentation of amino acids and/or dicarboxylates. Ecophysiology of U Sabulitectum silens \n We also detected an organism (bin O) that affiliated with the candidate phylum Fermentibacteria (formerly candidate division Hyd24–12) (Kirkegaard et al ., 2016 ). These organisms were present in all cultures, but were only abundant in the untreated cultures that were not exposed to oxygen or nitrate (Fig. 2 ). The contigs of this bin were very long (up to 538 kb; N50: 222 kb), the provisional genome had a size of 2.9 Mb and was inferred to be 77% complete (Supporting Information Table S2). Annotation of the genes encoded on the contigs of bin O suggested that the organisms have a typical gram‐negative cell envelope with a complete peptidoglycan biosynthesis pathway and an active outer membrane transport system ( tonB / exbBD ). Glycolysis and the non‐oxidative pentose phosphate pathway were complete (Fig. 5 ). The presence of largely complete operons coding for genes involved in lipid biosynthesis, cofactor biosynthesis, amino acid metabolism and nucleotide metabolism indicated that these bacteria are likely not dependent on others for the generation of the major cellular building blocks. The organism transcribed an H + /Na + ‐translocating V‐type ATP synthase as well as numerous protein complexes that translocate sodium ions across the cell membrane, such as an electron transport complex protein ( rnf ), a NADH‐oxidoreductase ( ndh ) and a Na + ‐translocating decarboxylase ( oad/gcd ). This combination of proteins indicated that the organism was able to synthesize ATP using a sodium motive force (Mulkidjanian et al ., 2008 ). However, the organism lacked a complete citric acid cycle and a respiratory chain. Single genes for flagellar biosynthesis and twitching motility were transcribed, yet the pathways for motility were incomplete (Fig. 5 ). Bin O lacked many of the mechanisms for oxidative and general stress protection (Fig 4 B), which may explain its low abundance in the oxygen and nitrate treated cultures. The metagenome and metatranscriptome indicated that the organism is a non‐motile, strictly anaerobic, obligate fermenter. In accordance with the recently proposed taxonomy for uncultured microorganisms (Konstantinidis et al ., in press), we propose to name the organism U Sabulitectum silens (gen. et sp. nov.; sabulum (lat.) – sand; tectus (lat.) – covered, roofed; silens (lat.) – still, silent). The candidate phylum Fermentibacteria belongs to the Fibrobacteres‐Chlorobi‐Bacteroidetes superphylum (Supporting Information Fig. S11) and comprises one class‐, one order‐, four family‐ and at least nine genus‐level clades (Fig. 6 ). The four family‐level clades were previously indicated (Kirkegaard et al ., 2016 ). Figure 5 Metabolic map of U Sabulitectum silens (bin O) showing central pathways that the organism transcribed in the sulfate‐only treatment (Con‐5, Con‐6). Transcribed genes are shown as blue arrows, genes of annotated pathways that were not detected as red arrows. Enzymes are abbreviated with letters, the full list as well as further metabolic pathways are provided in Supporting Information Table S3. Dashed blue circles depict additional pathways that were detected. Figure 6 Phylogeny of candidate phylum Fermentibacteria , showing the affiliation of all publicly available, non‐redundant 16S rRNA gene sequences, including the provisional species U Sabulitectum silens and Ca . Fermentibacter daniensis (black). The phylum comprises one class (at a threshold sequence identity of 78.5%), one order (at 82%), four families (at 86.5%) and at least nine genera (at 94.5%). The origin of the sequences is color‐coded (red: methane seeps; red bold: anaerobic methanotrophic enrichment cultures; orange: hypersaline mats; pink: springs; light blue: dolphin, dark blue: anaerobic digesters: grey: other) and indicates niche‐differentiation among Fermentibacteria . An extensive list of ecosystems harbouring Fermentibacteria is provided in the Supplementary Results. Phylogeny is based on the SILVA small subunit ribosomal database SSURef 123.1 (released 03/2016). The scale bar shows estimated sequence divergence. Fermentibacteria sequence alignments and phylogeny are provided as an ARB database (Dataset 3). The parameters that were used to compile the sequence database are described in the Supporting Information. The nearest relative of U Sabulitectum silens is the recently described Ca . Fermentibacter daniensis, an anaerobic fermenter that is possibly involved in the sulfur‐cycle (Kirkegaard et al ., 2016 ). In contrast to Ca . F. daniensis, U S. silens did not seem to possess or transcribe genes for sulfhydrogenases, despite the presence of sulfur in the cultures. Overall, both organisms appear to have similar lifestyles based on their transcriptional activity, despite their phylogenetic distance, suggesting that this lifestyle might be common among the phylum Fermentibacteria . Thus, it is not surprising that, so far, the phylum comprises sequences that almost exclusively originated from anoxic, organic and/or methane‐rich ecosystems (Fig. 6 ), including sulfidic cave biofilms (Macalady et al ., 2006 ), sulfur‐rich springs (Elshahed et al ., 2007 ), methane seeps (Ruff et al ., 2015 ; McKay et al ., 2016 ; Trembath‐Reichert et al ., 2016 ), mud volcanoes (Pachiadaki et al ., 2011 ; Chang et al ., 2012 ), methane hydrates (Mills et al ., 2005 ), marine sediments (Schauer et al ., 2011 ), coral reef sands (Schöttner et al ., 2011 ), microbial mats (Harris et al ., 2013 ; Schneider et al ., 2013 ), marine sponges (Simister et al ., 2012 ) and anaerobic digesters (Nelson et al ., 2012 ; Kirkegaard et al ., 2016 ). The physiology that U Sabulitectum silens exhibited in our cultures (Figs 3 , 4 B and 5) suggests that Fermentibacteria are strict anaerobes that produce hydrogen and acetate from the fermentation of amino acids and sugars, in these ecosystems."
} | 9,295 |
25503502 | PMC4264028 | pmc | 1,454 | {
"abstract": "Recent surge in the development of superhydrophobic/superoleophobic surfaces has been motivated by surfaces like fish scales that have hierarchical structures, which are believed to promote water or oil repellency. In this work, we show that the under-water oil repellency of fish scales is entirely due to the mucus layer formation as part of its defense mechanism, which produces unprecedented contact angle close to 180°. We have identified the distinct chemical signatures that are responsible for such large contact angle, thereby making fish scale behave highly superoleophobic inside the water medium. In absence of the mucus layer, it is found that the contact angle decreases quite dramatically to around 150°, making it less oleophobic, the degree of such oleophobicity can then be contributed to its inherent hierarchical structures. Hence, through this systematic study, for the first time we have conclusively shown the role of the fish's mucus layer to generate superoleophobicity and negate the common notion that hierarchical structure is the only reason for such intrinsic behavior of the fish scales.",
"conclusion": "Conclusion Here we have performed systematic static contact angle characterization of fish scales with and without mucus layer. The model fluid used is silicon oil and we have measured the contact angle of silicon oil on fish scales submerged in water using needle-free drop deposition technique coupled with proper illumination to decipher the drop contact area on the fish scale. This measurement technique allowed us to correctly quantify the contact angle for fish scales with mucus layer and we have observed an unprecedented contact angle of around 180° for under-water fish scales. Also, through careful characterization of the mucus layer, we found that certain functional groups in mucus layer (hydroxyl (-OH), amine (-NH2) and carbonyl (CO)) are responsible for such extreme superoleophobicity of the fish scales. By removing the mucus layer, we found that the degree of superoleophobicity of the fish scales reduces significantly to a static contact angle of around 150°, which can then be attributed to the hierarchical micro/nano-structures on the fish scales. Hence, the present study conclusively proves that under-water superoleophobicity of fish scales is due to its inherent mucus layer formation, generated through its own defense mechanism (for e.g., foreign oil droplet coming in contact with its scales), rather than the common notion of the presence of hierarchical structures of the fish scales. This study further opens up new paradigm towards proper understanding of under-liquid wettability of such marine life, which is often been biomimicked through extensive micro/nanofabrication process to generate superoleophobic/superhydrophobic/omniphobic surfaces.",
"discussion": "Results and Discussion Figure 1 provides the contact angle measurements for the fish scales with and without mucus layer. It is found that in air medium, the contact angle with oil drop changes from 30.6 ± 7.7° in case of scales with mucus to 24.1 ± 0.7° for scales without mucus. However, in air medium with water drop, the contact angle is found to be 34.1 ± 2.7° with mucus, which changes to 85.7 ± 4.7° for scales without mucus (please refer to Table S1 in Supplementary Information for the magnitude of contact angles for different cases). Such increase in contact angle for water in air medium has also been reported for Oncorthynchus mykiss fish 3 10 . In case of water medium, with oil drop, it is found that the fish scale with mucus layer behaves as a superoleophobic surface with unprecedented contact angle ≈ 180° (measured value is 178.4° ± 1.3). Such superoleophobicity has not been determined accurately in earlier studies 4 5 11 for fish scales. Through our careful mucus removal process, it is found that the same fish scales without mucus shows a reduced superoleophobic behavior, with a measured contact angle of 148.9 ± 5.5°. The superoleophobicity of fish scales without mucus layer can be considered as the contribution purely due to the geometrical asperities. However, our study clearly demonstrates that such hierarchical structures cannot alone contribute to unprecedented superoleophobic behavior of fish scales under-water. The inherent mucus layer, which is produced due to external stress like a pollutant oil drop in marine ecosystem, is responsible for this added superoleophobicity of fish scales under-water. The role of surface roughness on the wetting nature of fish scales can theoretically be presented using the Cassie-Baxter and Wenzel theories where the prior information regarding roughness parameters (height, pitch, width, etc. of the individual micro feature) and equilibrium contact angle on a flat smooth surface of the same material are needed. Hence, in the case of fish scales the knowledge of the geometrical micro features of fish scale and equilibrium contact angle on the smooth fish scale is an utmost requirement. The microscopic images (SEM as presented in this study) can provide geometrical features but on the other hand obtaining perfectly flat and smooth fish scale for equilibrium contact angle measurement is impractical. Hence, in most of the cases the fish scales are mimicked on the glass, silicon or the PDMS, using cumbersome micro-nano fabrication techniques, and the role of surface roughness is studied, using Cassie-Baxter and Wenzel theories 3 6 11 . Further the role of surrounding medium and wettability of substrate can be understood using mathematical argument as presents here. Let us consider a substrate in two different media i.e., in air and in water. Let us denote the interfacial surface tensions and contact angles for different combinations as γ ij , γ ik and θ ij , respectively, where i and j represent different phases i.e., air (a), water (w) or oil (o), and k represents the solid substrate. Based on Young's equation, one can write 13 : Therefore, one can write, For two different substrates i.e., with and without mucus, one can obtain two different equations using Eq. (2). Further, the difference in the wettability of these two substrates can be presented as Here, superscript ‘n' and ‘nm' denote two substrates with mucus and without mucus, respectively. It is to be noted that change in the wettability due to presence and absence of mucus i.e., is a not only a function of contact angles in air and water media ( ) but it also depends on magnitudes of different interfacial tension (γ wa , γ ow and γ oa ). Therefore, it can be inferred that apart from the equilibrium contact angles on such surfaces, the interfacial tension also dictates the decrease in the oleophobicity. To further delineate the effect of mucus layer on the wetting characteristics of the fish scales, we have scraped the mucus layer from the whole fish using the backside of the sterile surgical knife and collected the mucus sample on a glass side. The mucus sample was sandwiched between the two glass slides for uniform spreading of the sample and later taken apart and dipped into the glass couvette filled with DI water to measure the contact angle of oil drop on mucus layer alone. It is found that mucus layer on glass surface produces a similar unprecedented contact angle ≈ 180°, which further collaborates our earlier findings that mucus layer is solely responsible for such unprecedented superoleophobic behavior of fish scales (refer Supplementary Video S1 ). However, the mucus layer on a glass substrate looks wavy and rough as these are soft gel like material which has confirmed to its own minimum energy state, when deposited on a flat glass substrate. The key here is to emphasise that such conformity of the mucus layer on a glass surface has no bearing on the underlying roughness of the glass substrate, which in this case is almost flat. Therefore, through such exercise (as demonstrated in Video S1 ), we are able to demonstrate the role of chemical signature, which dictates such oil repellency behavior. An Attenuated Total Reflection Fourier Transform Infrared (ATR-FTIR) spectroscope was used to obtain the chemical signature of the fish scales, with and without mucus. The chemical signature of mucus layer obtained from FTIR data, which is presented in Fig. 2 , shows the presence of functional groups in the range from 1600 cm −1 to 3400 cm −1 , which are hydroxyl (-OH), amine (-NH2) and carbonyl (CO) 14 . Also, the mucus layer exhibits three distinct peaks that correspond to molecules, representing alcohols and phenols, which disappears for the mucus free sample. These functional groups along with several other components of fish mucus may lead to the formation of an oil repelling layer between the fish scale and the polluting oil through its interaction with the surrounding water medium. Due to this, the foreign oil drop virtually does not feel the presence of the underlying micro/nano-structured surface and probably creates an unfavorable spreading parameter 13 , which thereby produces such unprecedented superoleophobicity."
} | 2,264 |
28458657 | PMC5394123 | pmc | 1,455 | {
"abstract": "Black Band Disease (BBD), the destructive microbial consortium dominated by the cyanobacterium Roseofilum reptotaenium , affects corals worldwide. While the taxonomic composition of BBD consortia has been well-characterized, substantially less is known about its functional repertoire. We sequenced the metagenomes of Caribbean and Pacific black band mats and cultured Roseofilum and obtained five metagenome-assembled genomes (MAGs) of Roseofilum , nine of Proteobacteria, and 12 of Bacteroidetes. Genomic content analysis suggests that Roseofilum is a source of organic carbon and nitrogen, as well as natural products that may influence interactions between microbes. Proteobacteria and Bacteroidetes members of the disease consortium are suited to the degradation of amino acids, proteins, and carbohydrates. The accumulation of sulfide underneath the black band mat, in part due to a lack of sulfur oxidizers, contributes to the lethality of the disease. The presence of sulfide:quinone oxidoreductase genes in all five Roseofilum MAGs and in the MAGs of several heterotrophs demonstrates that resistance to sulfide is an important characteristic for members of the BBD consortium.",
"conclusion": "Conclusions The functional repertoire of BBD microbial communities highlights the role of the cyanobacterium R. reptotaenium as the engineer of the disease consortium through the production of organic carbon and nitrogen, as well as novel secondary metabolites. The Roseofilum MAGs recovered here from five different coral species and two different oceans contain a suite of both ribosome-dependent and non-ribosomal peptide and polyketide biosynthetic clusters, at copy numbers similar to or greater than other filamentous cyanobacteria (Shih et al., 2015 ). These secondary metabolites may be the key to understanding microbe-microbe interactions in the BBD microbial community, such as previously demonstrated for the role of lyngbic acid in inhibiting QS in Vibrio species (Meyer et al., 2016 ). The genomic content of Bacteroidetes and Proteobacteria in BBD consortia suggests that heterotrophs fill the role of degrading and recycling organic matter within the mat, fueling the continued growth of Roseofilum . This dependence of Roseofilum on heterotrophic partners is consistent with the fact that filamentous cyanobacteria cannot be easily cultured axenically. However, the relationship between Roseofilum and its heterotrophic partners is not exclusively beneficial. For example, despite its potential for sulfide detoxification, Roseofilum migrates away from the build-up of sulfide within in situ BBD mats (Glas et al., 2012 ). In aerated cultures where Desulfovibrio is absent (and therefore not creating sulfide), Roseofilum can easily be overwhelmed by heterotrophic growth if the culture medium is too rich and must be transferred regularly to maintain active growth ( personal observation ). A similar accumulation of heterotrophic cells over time was documented in two different unicyanobacterial cultures of mats from Hot Lake, Washington (Cole et al., 2014 ). The conservation of taxonomic groups in the minimal saltwater culture strongly suggests that once the disease consortium is established, the BBD microbial mat does not require the coral host tissue to thrive. This suggests that the pathogenesis of Roseofilum and other disease consortium members is contextual and has important implications for the mitigation of BBD in coral reef systems, primarily that if the conditions initiating the establishment of the microbial mat can be prevented, the disease can be thwarted.",
"introduction": "Introduction Black Band Disease (BBD) is a globally distributed coral disease that devastates dozens of species of corals, including large, reef-building scleractinians (Sato et al., 2016 ). It is easily recognized by the appearance of a dense dark purple or black band, which is the visible accumulation of phycoerythrin-rich filamentous cyanobacteria. In the Caribbean, known as a “hotbed” of coral disease (Weil et al., 2006 ), BBD is more prevalent in warmer, shallower waters (Kuta and Richardson, 2002 ), and manipulative experiments have demonstrated that BBD progresses faster at higher light levels and temperature (Kuehl et al., 2011 ; Sato et al., 2011 ). The cyanobacterium recently renamed Roseofilum reptotaenium has been implicated as the causative agent of BBD within the polymicrobial disease consortium (Casamatta et al., 2012 ). Strains of Roseofilum have been cultivated in the laboratory, but like other filamentous cyanobacteria, Roseofilum cannot be fully isolated, only grown in non-axenic, unicyanobacterial cultures (Richardson et al., 2014 ). This conserved interdependence of Roseofilum and heterotrophic bacteria may be linked to metabolic requirements and may explain why this cyanobacterium is found within apparently healthy coral microbiomes (Meyer et al., 2016 ). In many ways, BBD consortia resemble microbial mats found in tropical lagoons (Echenique-Subiabre et al., 2015 ), mangroves (Guidi-Rotani et al., 2014 ), and modern marine stromatolites (Ruvindy et al., 2016 ) that are characterized by steep physicochemical gradients and the presence of Cyanobacteria, Bacteroidetes, and Proteobacteria (especially Alpha-, Delta-, and Gammaproteobacteria; Bolhuis et al., 2014 ). With the large size of filamentous cyanobacteria relative to other bacterial cells, their production of exopolysaccharides that consolidate the mat, and their critical roles in converting inorganic carbon, nitrogen, and sulfur to organic forms, Cyanobacteria are the pioneering microorganisms of microbial mats (Bolhuis et al., 2014 ). Likewise, Roseofilum is the engineer of the black band consortium, creating a narrow polymicrobial band on the surface of reef-building corals (Miller and Richardson, 2011 ; Casamatta et al., 2012 ; Richardson et al., 2014 ). While the composition of BBD has been studied for decades (Miller and Richardson, 2011 ), the recent application of high-throughput sequencing to characterize the microbial community structure in both the black band layer and adjacent healthy coral epimicrobiome has uncovered two important factors in understanding this disease. First, BBD is highly localized, and in as little as 10 cm from the leading edge of the black band mat, the microbial community in the surface mucus layer is indistinguishable from that of healthy corals (Meyer et al., 2016 ). Second, Roseofilum is a rare but ubiquitous member of healthy Caribbean coral microbiota, implying that growth of Roseofilum is constrained in healthy tissue until undefined restrictions are removed (Meyer et al., 2016 ). Together, this suggests that while Roseofilum is capable of engineering the highly altered black band layer, its influence is spatially limited and its pathogenesis is contextual. To uncover the mechanisms through which Roseofilum proliferates and engineers a new environment on the surface of corals during BBD, we examined the metagenomic potential of members of black band consortia in situ and in vitro .",
"discussion": "Results and discussion Microbial community structure from 16S rRNA gene sequencing Bacterial community structure was determined for 24 BBD consortia, 30 healthy coral surface microbiomes, and one Roseofilum culture by amplification and Illumina sequencing of the V6 hypervariable region of 16S rRNA genes (Table S1 ). Consistent with previous results from Caribbean corals only (Meyer et al., 2016 ), the reanalysis of the bacterial community structure with both Caribbean and Pacific corals demonstrates the prevalence of Roseofilum , Rhodobacteraceae, Bacteroidales, and Desulfovibrio in black band consortia in contrast to the high proportions of Gammaproteobacteria (such as Halomonas and Moritella ) and Actinobacteria (such as Renibacterium ) in healthy epimicrobiomes (Figure 1 ). Comparison of the composition of 24 in situ BBD samples and a unicyanobacterial, but not axenic Roseofilum culture revealed that three proteobacterial genera were depleted in the culture. The relative abundance of Vibrio in the culture was two orders of magnitude lower (0.001% of reads in the Roseofilum culture vs. ~0.1% in BBD), Arcobacter was three orders of magnitude lower (0.0004% vs. ~0.1%), and Desulfovibrio was three orders of magnitude lower (0.0002% vs. ~0.01%). Figure 1 Community shifts in Black Band Disease (BBD) consortia and healthy coral epimicrobiomes . Relative abundance of the 10 most common bacterial genera in 16S rRNA gene libraries from 54 coral surfaces. Each column in the heatmap represents an individual coral surface microbiome sample. The remarkable degree of conservation of taxa in the Roseofilum culture compared to in situ BBD metagenomes suggests a strong interdependence of members of the disease consortium that is unrelated to the coral host. Specifically, no major BBD taxonomic groups were entirely missing in the Roseofilum culture and the proportions of Bacteroidetes, Deltaproteobacteria, and Gammaproteobacteria were relatively unchanged between the Roseofilum culture and in situ BBD. However, three proteobacterial genera ( Vibrio, Arcobacter , and Desulfovibrio ) were detected in lower relative abundance in the culture compared to in situ BBD, based on 16S rDNA amplicon libraries. Although these genera may not be the most abundant organisms within the BBD consortium, each has the potential to impact the progression of the disease, as both Vibrio and Arcobacter are genera containing pathogens and both Arcobacter and Desulfovibrio are involved in sulfur cycling. Recent metatranscriptomic analysis of BBD consortia in the Red Sea demonstrated that the most highly expressed functional gene among BBD vibrios was a thiamin transporter (Arotsker et al., 2016 ), suggesting that Roseofilum , as a producer of thiamin, has the capacity to encourage the growth of vibrios that lyse coral tissue (Arotsker et al., 2009 ) during the colonization of adjacent healthy tissue. However, interactions between Roseofilum and vibrios are likely complex, as we previously demonstrated mutual exclusion between Roseofilum and Vibrio spp., that may be mediated through production of lyngbic acid by Roseofilum that is capable of disrupting the CAI-1 vibrio quorum-sensing pathway (Meyer et al., 2016 ). While previous work revealed a signficant co-occurrence of Roseofilum and Desulfovibrio in BBD consortia (Meyer et al., 2016 ), Desulfovibrio was not co-enriched with Roseofilum in the culture. This indicates that the co-occurrence of these two genera is related to opportunistic interactions on the coral surface, rather than a specific interactive relationship. Furthermore, it suggests the two groups contribute individually to disease progression. Assembled metagenomes and metagenome-assembled genomes from black band disease consortia Four metagenomes of the disease consortia were generated from the black band community on Pseudodiploria strigosa and Orbicella annularis corals from Belize, Goniopora fruticosa from Guam, and a pooled sample from M. cavernosa and O. faveolata from Florida. A fifth metagenome was generated from a unicyanobacterial, but not axenic, culture of R. reptotaenium isolated from a black band layer on M . cavernosa collected in Florida. Each metagenomic dataset contained 16,853,081–44,199,247 read pairs after quality filtering (Table S2 ). Annotated metagenomic assemblies contained 69,548–612,757 protein-coding genes (Table S2 ). The coverage of metagenomic sequencing was assessed with non-pareil using the unassembled quality-filtered sequencing reads and ranged from 87 to 99%, indicating that this sequencing depth was adequate to capture most of the unique sequences in the extracted community DNA, even if not all of this diversity was captured in the metagenomic assemblies. To assess the quality of the metagenomic assemblies, the unassembled sequencing reads were mapped to the metagenomic assemblies, with an overall alignment rate that ranged from 41% (BLZD) to 92% (Cyano), where a higher alignment rate indicates that the metagenomic assembly encompassed more of the diversity present in the sequencing reads (Table S2 ). Between 2 and 24 unique genomes were detected within each metagenomic assembly (Table S2 ) by Anvi'o (Eren et al., 2015 ), using four different databases of single copy genes as in Delmont and Eren ( 2016 ). The number of protein-coding genes in each metagenome ranged from 69,548 to 612,757 (Table S2 ). Of the annotated genes in the assembled metagenomes, 12 to 47% could be assigned to bacterial phyla (16,674–79,576 bacterial genes per metagenome). Less than 1% of annotated genes in any metagenomic assembly could be assigned to archaeal phyla. The remaining genes were either unassigned or identified as eukaryotic. A total of 53 bacterial phyla and 6 archaeal phyla were identified in the metagenomic assemblies (Table S3 ). The four most abundant bacterial phyla (Cyanobacteria, Bacteroidetes, Proteobacteria, and Firmicutes) constituted 80–95% of the annotated bacterial genes in the four environmental metagenomic assemblies (Figure S1 ). In the Roseofilum culture, these four phyla constituted 68% of annotated genes with taxonomic assignment, and Planctomycetes were more abundant than Firmicutes (Figure S1 ). The proportions of functional genes assigned to Cyanobacteria and Bacteroidetes in the metagenomic assemblies matched the proportion of these phyla detected by 16S rRNA gene surveys, while the proportion of Proteobacteria was higher in 16S gene surveys than in metagenomic assemblies (Figure S1 ). Binning of the assembled metagenomic datasets produced a total of 26 MAGs, with estimated genome coverage ranging from 9 × to 776 × (Table 1 ). The annotated MAGs are publicly available in IMG and under the accession numbers listed in Table 1 . MAGs had varying levels of estimated genome completeness and redundancy (Table 1 ). With the exception of the LKpool Roseofilum , which was assembled from pooled coral hosts, the MAGs of Roseofilum were relatively complete and contained low redundancy (4% or less of the single-copy genes in the Roseofilum MAGs were present in more than one copy). Only one cyanobacterial MAG and one full-length cyanobacterial 16S rRNA gene were retrieved from each assembled metagenome. To corroborate the low diversity of cyanobacterial genomes, contigs identified as containing cyanobacterial genes based on IMG annotation from each of the five metagenomic assemblies were pooled in silico and analyzed for the number of genomes present. Assembled contigs annotated as Cyanobacteria from the Belize Orbicella metagenome (BLZ4) contained up to three copies of single copy genes, while cyanobacterial contigs from the other metagenomes contained one to two copies of single copy genes, indicating a low diversity of Cyanobacteria in the black band layer. The cyanobacterial 16S rRNA gene sequence was identical in metagenomic assemblies from the three Caribbean samples and the unicyanobacterial culture, while minor variation (98.9% similarity) was detected in the cyanobacterial 16S rRNA gene sequence from Guam. The 16S rRNA gene sequences recovered from the Caribbean cyanobacterial MAGs are identical to the reported sequence in R. reptotaenium (EU743965) with the exception of one nucleotide at the 5′ end of the published sequence for the type species. The two-way average nucleotide identity of the four Caribbean Roseofilum MAGs was >99% similarity, while the Guam Roseofilum MAG showed around 94% similarity to the Caribbean Roseofilum MAGs. In addition to the five Roseofilum MAGs, MAGs of 12 Bacteroidetes, and 9 Proteobacteria were recovered by binning, including an unclassified gammaproteobacterium, Bacteroidales, and Desulfovibrio , which have been detected as common in BBD consortia through 16S amplicon surveys (Figure 1 ; Miller and Richardson, 2011 ; Meyer et al., 2016 ). Functional characteristics of black band disease metagenomes and metagenome-assembled genomes The five Roseofilum MAGs had similar functional profiles based on KEGG Ontology (Figure S2 ), as well as similar gene synteny (Figure S3 ) in contrast to the more taxonomically and functionally diverse Bacteroidetes and Proteobacteria MAGs (Figure S2 ). Pan-genomic clustering of protein-coding genes in Roseofilum and Geitlernema MAGs (Figure 2 ) reveals a conserved core of Roseofilum functional genes that is distinct from the Geitlerinema MAG based on sequence variation, as expected for taxa assigned to different genera. However, the overall functional profile of all eight MAGs, based on SEED subsystems, is well-conserved (with the exception of the less-complete LKpool Roseofilum ; Figure 2 ). Clustering of Roseofilum MAGs based on the relative abundance of proteins showed a clear geographic signal, with Caribbean MAGs clustering together and Pacific MAGs (from three different research groups) clustering together (Figure 2 ). A total of 36,649 genes in 8,690 protein clusters were identified in the eight MAGs. The Roseofilum core indicated in Figure 2 contained 27,153 genes in 3,470 protein clusters, while the Geitlerinema specific bin contained 2,702 genes. Clusters in the Roseofilum core were defined as being present in at least five of the seven Roseofilum MAGs. The remaining, less-conserved regions contained a total of 6,794 genes in 2,634 protein clusters. Of these less-conserved protein clusters, 40% (1055 clusters) contained only one gene and 18% (485 clusters) contained only two genes. Figure 2 Roseofilum \n pan-genome . Protein clusters from seven Roseofilum and one Geitlerinema metagenome-assembled genomes (MAGs) are displayed in the central dendrogram and genes present in each MAG are indicated on concentric rings. The outer-most ring displays Single Copy Genes present in the Roseofilum core genome. The next two rings display the number of genes and number of MAGs, respectively, for each protein cluster. Upper right: Relative abundance of genes in each MAG assigned to Level 1 SEED subsystems annotation. The % completeness and % redundancy of each MAG is based on the presence/absence of single-copy genes. Clustering of MAGs (right) was based on the relative abundance of proteins and reflected the location: Green for Caribbean samples and Teal for Pacific samples. “Geit” is the cultured Geitlerinema sp. BBD 1991 MAG isolated from Orbicella annularis in Florida (Den Uyl et al., 2016 ), “LKpool” is the Roseofilum MAG from a pooled sample from Montastraea cavernosa and O. faveolata from Florida, “Cyano” is the cultured Roseofilum MAG isolated from M . cavernosa collected in Florida, “BLZ4” is the Roseofilum MAG from Orbicella annularis in Belize, “BLZD” is the Roseofilum MAG from Pseudodiploria strigosa in Belize, “Cya2” is the Roseofilum MAG from Montipora hispida on the central Great Barrier Reef (Sato et al., 2016 ), “AO1” is the cultured Roseofilum MAG isolated from Pavona duerdeni on the central Great Barrier Reef (Buerger et al., 2016 ), and “Guam” is the Roseofilum MAG from Goniopora fruticosa in Guam. To determine the potential roles and interactions between different members of the polymicrobial disease consortium, we tested for differences in the relative abundances of protein-coding genes assigned to SEED subsystems (Overbeek et al., 2014 ) level 3 functions. Examination of differentially represented functions among the three predominant phyla (Cyanobacteria, Bacteroidetes, Proteobacteria) revealed 87 subsystems with statistically different abundances (Table S3 ). Of these differentially represented functions, 13 had an effect size >0.6, meaning that the difference in the mean relative abundance was >60% and encompassed mostly functions that were present solely in the Roseofilum MAGs and not in the MAGs of heterotrophs (Figure 3 ). Genes associated with carbon fixation (Subsystem CO 2 uptake, carboxysome), micronutrient acquisition (Subsystem ECF class transporters), and phototrophy (Subsystems Cytochrome B6-F complex, Photosystem II, and Phycobilisome) were more abundant in the Roseofilum MAGs than the MAGs of heterotrophs (Figure 3 ). In addition, functions related to carbon respiration (Subsystems Cyanobacterial bypass in the TCA, Maltose and maltodextrin utilization), and the redox-regulated molecular chaperone Hsp33 (Subsystem Steptococcus pyogenes recombinatorial Zone) were primarily unique to the Roseofilum MAGs (Figure 3 ). Figure 3 Functional differences in metagenome-assembled genomes from Black Band Disease consortia . Differences in the relative abundance of Level 3 SEED subsystems annotations among metagenome-assembled genomes from three phyla. Functional annotations with the same lettering color belong to the same Level 1 SEED subsystems category. Only features passing the filtering criteria are shown ( q < 0.05, effect size > 0.6, multiple test correction with Storey FDR). Differentially abundant functions between the three phyla also suggest a strong internal cycling of organic nitrogen within the black band layer in which Bacteroidetes and Proteobacteria degrade amino acids to produce urea and urea is degraded by Roseofilum . Both Bacteroidetes and Proteobacteria MAGs contained genes for the degradation of branched chain amino acids (leucine, isoleucine, valine) that were not present in the five Roseofilum MAGs, and the Roseofilum MAGs had more genes associated with the degradation of urea than Bacteroidetes and Proteobacteria MAGs (Table S4 ). In addition, genes for the production and degradation of cyanophycin, a polypeptide-like nitrogen storage compound, were present in the five Roseofilum MAGs and in the Cyano_bin9_Gammaproteobacteria MAG (Table S5 ). Genes for the assimilation of ammonia and nitrate were common in the MAGs of both Roseofilum and heterotrophs, demonstrating that each member of the consortium utilizes multiple pathways to assimilate nitrogen in the nitrogen-limited coral reef ecosystem. Genes for nitrogen fixation ( nifHDK ) were detected in the five Roseofilum MAGs as well as in two Proteobacteria MAGs (BLZ4_bin3_Alteromonadales, BLZ4_bin7_Rhodospirillales; Table S5 ). The diversity of methods utilized by Roseofilum for the acquisition of nitrogen mirrors those found in the Geitlerinema strain isolated from BBD (Den Uyl et al., 2016 ). The multiple, diverse Bacteroidetes MAGs recovered here from both Caribbean and Pacific BBD underscore the previously underappreciated predominance of Bacteroidetes in BBD. While Bacteroidetes have been detected in 16S rRNA gene surveys of BBD and have been hypothesized to play a role in the pathogenesis of BBD (Frias-Lopez et al., 2004 ), the abundance of Bacteroidetes MAGs and high relative abundance of genes assigned to Bacteroidetes in the metagenomic assemblies (Table S3 ) suggests this group has a substantial role in the BBD consortia. Marine Bacteroidetes are known for their ability to degrade complex polymers (Fernández-Gómez et al., 2013 ) and may thus play a critical role in the progression of BBD. Whether these Bacteroidetes are simply opportunistic in the degradation of coral tissue under a BBD mat or whether they contribute to the invasion of adjacent healthy tissue is yet to be determined. Recent work has demonstrated that benthic cyanobacterial mats in Caribbean coral reef ecosystems are stimulated by the localized degradation of organic matter in sediments (Brocke et al., 2015 ), further suggesting that microbial mat-forming filamentous cyanobacteria like Roseofilum require the release of nutrients by heterotrophs to proliferate. Here, we found that both Proteobacteria and Bacteroidetes in the five metagenomes carry glycoside hydrolase families for the degradation of starch/glycogen, oligosaccharides, and chitin (data not shown). Future metatranscriptomic studies may reveal which groups are most actively degrading complex carbon in the black band layer. Akin to other cyanobacterial mats, a gradient of decreasing oxygen and increasing sulfide with depth in the black band layer has been detected (Glas et al., 2012 ), and the accumulation of sulfide may have an important role in BBD as sulfate reduction appears to be necessary for the initiation of the disease but not for the progression of already established BBD (Richardson et al., 2009 ; Brownell and Richardson, 2014 ). Previous investigations of key sulfur-cycling genes in BBD have identified dissimilatory sulfate reductase genes in Desulfovibrio (Bourne et al., 2011 ) and sulfur oxidation genes in Rhodobacteraceae (Bourne et al., 2013 ). However, the low levels and lack of diversity in sulfur oxidation genes associated with BBD imply that sulfide is not broken down, but rather accumulates beneath the mat (Bourne et al., 2013 ). These results are consistent with the current metagenomic study. Dissimilatory sulfate reduction genes ( dsrA ) were detected only in the two Desulfovibrio MAGs and not in the other major groups found in the black band consortium (Table S5 ). No sulfur oxidation genes of the sox pathway were detected in any of the five BBD metagenomes nor in the 26 MAGs, indicating that this pathway is likely rare in the black band layer, allowing the accumulation of sulfide during anaerobic phases of the diel cycle (Glas et al., 2012 ). While few members of the disease consortium may be capable of oxidizing sulfide, it is possible that sulfide is detoxified through the action of sulfide:quinone oxidoreductase ( sqr ) genes. Each of the Roseofilum MAGs contains two unique copies of sqr , with the exception of the less complete Florida MAG (LKpool) that has only one assembled copy (Figure 4 ). The Roseofilum MAGs contain both Type VI and Type II sqr genes, which are adapted to high sulfide concentrations and low sulfide affinity, respectively (Marcia et al., 2010 ). The detoxification of sulfide appears to be a common feature in genomes of filamentous cyanobacteria that form mats, as this lifestyle promotes the growth of sulfate reducers and the sulfide they produce inhibits photosynthesis through the irreversible blockage of Photosystem I (Voorhies et al., 2012 ). The MAG of the Geitlerinema strain cultured from BBD (Den Uyl et al., 2016 ) also contains one copy of a Type VI sqr , that is about 58% similar to the Type VI sqr gene in the Caribbean Roseofilum MAGs (Figure 4 ). Neither Roseofilum nor Geitlerinema from black band contained the third type of sqr found in Cyanobacteria, Type I that has a high sulfide affinity and is used in anoxygenic photosynthesis in the filamentous cyanobacterium Geitlerinema sp. PCC 9228 (Grim and Dick, 2016 ). Single copies of sqr were also found in the unclassified gammaproteobacterium from the Roseofilum culture (IMG gene id 2628135161), in Bacteroidetes MAGs from Florida (2628091167), Belize (2628117824), and Guam (2628109795), and in the Rhodospirillales MAG from Belize (2628122119), indicating that the detoxification of sulfide is a common feature in members of the BBD polymicrobial consortium. Figure 4 Maximum likelihood tree of sulfide:quinone oxidoreductase ( sqr ) genes from filamentous cyanobacteria . The five metagenome-assembled genomes of Roseofilum contained both Type VI and Type II sqr genes and are compared to sqr genes from other Oscillatoriales genomes. Branch labels are preceded by the IMG gene id and colored boxes indicate the habitat from which the cyanobacteria were isolated and whether or not it forms mats, based on publicly available metadata. Metabolism of DMSP (dimethylsulfoniopropionate) has been implicated as an important process in the marine sulfur cycle, particularly in coral reefs, where DMSP functions as an osmolyte and is rapidly degraded by bacteria in the coral surface microbiome (Raina et al., 2009 ). No genes in the five metagenomes were annotated as DMSP lyases, which may reflect either the use of alternative pathways for DMSP degradation or the replacement of native coral commensals that would normally degrade DMSP with bacteria that are opportunistic generalists, lacking specialized adaptations to the coral epimicrobiome niche. In the light of extensive coral tissue death beneath the cyanobacterial mat and the lack of DMSP production genes in Roseofilum , it is unlikely that significant quantities of DMSP would be available in the black band layer, though this has not been directly measured. Production of secondary metabolites by black band disease consortia Cyanobacteria are known for their ability to produce diverse natural products using non-ribosomal peptide synthetase (NRPS) and polyketide synthase (PKS) pathways, including toxins and siderophores (Leão et al., 2012 ; Shih et al., 2015 ). At least four unique NRPS/PKS biosynthetic clusters were detected per Roseofilum MAG, along with biosynthetic clusters for terpenes, bacteriocins, and cyanobactins (Figure 5 ). The abundance of biosynthetic clusters in the Roseofilum MAGs is consistent with the distribution of biosynthetic clusters in cyanobacterial genomes, as uncovered in a recent large-scale comparative genomics study (Shih et al., 2015 ). Natural products from marine cyanobacteria have been extensively explored for bioactive properties in the development of anti-inflammatory and anti-cancer drugs (Nunnery et al., 2010 ), though the intended purpose of these compounds from the perspective of cyanobacteria is potentially as grazing deterrents (Nagle and Paul, 1999 ), for UV protection (Sorrels et al., 2009 ) and allelopathy (Leão et al., 2012 ), or for interference in bacterial communication (quorum sensing inhibition; Dobretsov et al., 2010 ; Meyer et al., 2016 ). The Roseofilum MAGs contained the highest number of biosynthetic clusters, followed by Bacteroidetes MAGs, and the fewest clusters were detected in Proteobacteria MAGs. Two of the Proteobacteria MAGs ( Desulfovibrio from the Belize Orbicella and Vibrio from the Guam Goniopora ) had no detectable biosynthetic clusters for secondary metabolites. The role of secondary metabolites in the establishment and persistence of BBD consortia is an area of ongoing active research. Figure 5 Frequency of antibiotic and secondary metabolite biosynthetic clusters in binned genomes from Black Band Disease consortia . Biosynthetic clusters were identified with AntiSMASH. Each column represents one metagenome-assembled genome and bubbles represent the relative abundance of biosynthetic clusters. NRPS, Non-ribosomal peptide synthase; PKS, polyketide synthase; Other KS, PKS other than type 1, 2, 3, or trans-AT."
} | 7,722 |
36156954 | PMC9507689 | pmc | 1,457 | {
"abstract": "The combination of deep neural networks and reinforcement learning had received more and more attention in recent years, and the attention of reinforcement learning of single agent was slowly getting transferred to multiagent. Regret minimization was a new concept in the theory of gaming. In some game issues that Nash equilibrium was not the optimal solution, the regret minimization had better performance. Herein, we introduce the regret minimization into multiagent reinforcement learning and propose a multiagent regret minimum algorithm. This chapter first introduces the Nash Q-learning algorithm and uses the overall framework of Nash Q-learning to minimize regrets into the multiagent reinforcement learning and then verify the effectiveness of the algorithm in the experiment.",
"conclusion": "3. Conclusion In summary, we had used regret minimization the study of multiagent and put forward the MARMQ algorithm, and compared it with Nash Q-learning. The results showed that MARMQ's performance in Travelers Game, Centipede Game, and Grid-World was better than Nash Q-learning. It showed that when MARMQ uses regret minimization strategy of action, it was better than the Nash strategy that was more conducive to the maximization of its own interests. Nash Q-learning was the maximization of their own interests as the starting point. This would often fall into the predicament of the prisoner, and MARMQ was seeking to minimize the regret, which relieved this trap to a certain extent. We also proposed another form of training in MARMQ, i.e., independent training. In independent training, each agent could no longer use the Q table of other agent to minimize regret but to learn the Q table of each agent. Such independent training had undoubtedly increased time costs, but the performance in the traveler's game was very good. All algorithms and trainings could pave the way to the direction of future research.",
"introduction": "1. Introduction Strengthening learning allows agent to continuously undergo self-learning through the way of interacting with the environment and ultimately meets the expected goals. It is a type of try and error (TE) learning. TE learning first appeared in the study of cats in 1898, and he conducted TE experiments on cats [ 1 ]. Watkins and Dayan proposed the famous Q-learning algorithm in 1989 [ 2 ], which combines time series difference with the Markov decision process (MDP) and Bellman equation. It is the most classic algorithm of reinforcement learning. In 2015, Mnih proposed the deep Q-learning network [ 3 ] (DQN) by combining the deep neural network and Q-learning and achieved superhuman performance in the game. Since then, reinforcement learning has fully entered the deep reinforcement learning stage. The achievements of deep strengthening learning are remarkable. DQN could achieve better results in 49 ATARI games than professional players. Google's DeepMind team created the Go robot AlphaGo in 2016, which uses the DQN network and could learn itself. AlphaGo won the game with the top Go master of human beings, and Li Shishi's games won all of them [ 4 ]. In-depth reinforcement learning has also been applied to solve problems such as MUJUCO [ 5 ] and 3D maze. In-depth strengthening learning has become the most potential research direction in the complexity of the real world and has also made great contributions in the field of artificial agent. It is difficult to migrate to multiagent in the success of a single agent. Multiagent reinforcement learning (MARL) and single agent reinforcement learning from the state of the most different one is determined by multiple agents rather than a single agent. As the environment becomes uncertain, each agent must face the optimal strategy movement problem; that is, the optimal strategy is constantly changing with the changes in other agent parties [ 6 , 7 ]. Because of this, most of the single agent reinforcement learning algorithms are not effective in the multiagent environment, and as the state action space of each agent grows in the index level, the problem of Gaowei curse have become more serious in the multiagent environment. Hence, MARL proposes a series of new technologies and methods to achieve the purpose of accelerating the entire learning process through sharing of knowledge and learning between agent [ 8 , 9 ]. Another difference between multiagent and single agent to enhance learning is that there are multiple agents, and there could be cooperation or competition (game) relationships between agents, and this relationship is determined by the reward function [ 10 ]. If there is a purely cooperative relationship between the agents, their reward functions are the same, and the learning goals are to maximize the common income. If there is a purely competitive relationship between the agents, their reward function is zero-sum. If there is neither a complete cooperation nor a complete competition, their relationship is called mixed cooperation and competition [ 11 ]. Main challenges of multiagent reinforcement learning are the instability of the environment, some observations, rewards allocation, and computing complexity. The calculation complexity is a problem that all multiagent learning algorithms have to face. The calculation complexity contains other challenges. The combination of regrets in online learning and the combination of multiagent strengthening learning is a new direction in the development of multiagent in recent years. Online learning is a type of machine learning. Unfortunately, the difference between the rewards related to the action and the return he actual actions is big [ 12 ], the regret minimum related algorithm in the online learning could be used to solve the problem of expansion game under nonperfect information, such as the virtual regret minimization (Counterfactual Regret Minimization, CFR [ 13 ]) and its latest variant CFR+ [ 14 ]. CFR could be achieved in the game of double zero and nonperfect information game to Nash balance. Brown and Sandholm could defeat top professional players in one-to-one Texas Poker with the help of Sandholm. Noam Brown's agent Pluribus [ 15 ] could achieve the same excellent results in the six Texas Poker. Noam Brown and others combined CFR with deep neural networks to propose a deeper virtual regret minimum algorithm (DCFR [ 16 ]). In the field of multiagent reinforcement learning, the deep role [ 17 ] algorithm proposed by combining CFR and self-play could identify partners and opponents in the game. A regret-based minimization algorithm (ARM [ 18 ]) was proposed by Jin et al in order to combine the concept of advantageous functions in enhanced learning with CFR. Steinber [ 19 ] also combines CFR and enhanced learning. All these items should be redirected to MARL based on the Nash Q-learning framework. Nash Q-learning [ 20 ] use Nash balance to solve the problem of zero-sum game under the multiagent body. NASH Q-learning is an early work of multiagents to strengthen learning. In the face of the instability of multiagent reinforcement learning, it is not possible to use neural networks for powerful search, but to search each agent first, whenever, their decisions are based on Nash balance. The author also proves the convergence of the algorithm under the theoretical level. This article draws on this idea to give the agent a priori knowledge so that their decisions are minimized based on regret and proposes multiagent regret minimum algorithms. Multiagent regrets the minimum algorithm compared to Nash Q-learning, which could not only deal with zero-sum game problems but also handle nonzero-sum games. In Nash Q-learning, the Q function has made two changes in two aspects. The first is to overcome the instability. Another is the input of the Q function of the current state and the joint action. We herein propose a multiagent regret minimization and strengthen the learning (MARMQ) algorithm. MARMQ was above the Nash Q-learning framework. The selection part of the counterpart was improved and Lemke–Howson algorithm was replaced with iterative regret minimum algorithms not looking for the Nash balance point but to minimize the motion of each agent. At the same time, when the Q value was updated, the Q value of the minimum movement in the next state was used as part of the update and ultimately enabled the agent to learn the regret minimum strategy."
} | 2,098 |
26162885 | PMC4561719 | pmc | 1,458 | {
"abstract": "Methanogens are anaerobic archaea that grow by producing methane, a gas that is both an efficient renewable fuel and a potent greenhouse gas. We observed that overexpression of the cytoplasmic heterodisulfide reductase enzyme HdrABC increased the rate of methane production from methanol by 30% without affecting the growth rate relative to the parent strain. Hdr enzymes are essential in all known methane-producing archaea. They function as the terminal oxidases in the methanogen electron transport system by reducing the coenzyme M (2-mercaptoethane sulfonate) and coenzyme B (7-mercaptoheptanoylthreonine sulfonate) heterodisulfide, CoM-S-S-CoB, to regenerate the thiol-coenzymes for reuse. In Methanosarcina acetivorans , HdrABC expression caused an increased rate of methanogenesis and a decrease in metabolic efficiency on methylotrophic substrates. When acetate was the sole carbon and energy source, neither deletion nor overexpression of HdrABC had an effect on growth or methane production rates. These results suggest that in cells grown on methylated substrates, the cell compensates for energy losses due to expression of HdrABC with an increased rate of substrate turnover and that HdrABC lacks the appropriate electron donor in acetate-grown cells.",
"introduction": "INTRODUCTION Methane is a combustible gas fuel that can be used to produce heat and electricity. Methanogens are anaerobic archaea that naturally produce methane in soil, ocean sediment, and the gastrointestinal tracts of animals ( 1 ). In anaerobic environments, methanogens catalyze the terminal step in the carbon cycle by reducing C 1 compounds (such as CO 2 , CO, formate, methanol, methylsulfides, and methylamines) and acetate to methane gas, which diffuses up to aerobic zones, where it can be oxidized by methanotrophic bacteria ( 2 ). Because of their intrinsic ability to produce methane, methanogens are harnessed in anaerobic digesters to turn municipal, agricultural, and industrial waste products into renewable fuel (biogas), heat, and electricity ( 3 ). The model organism Methanosarcina acetivorans is one of a small group of methanogens capable of producing methane both from acetate and from methylated substrates, such as methanol, methylamines, and methylsulfides ( 4 ). In the methylotrophic methanogenesis pathway, electrons obtained from oxidizing one molecule of methanol to carbon dioxide are used to reduce three molecules of methanol to methane ( 5 – 7 ). The cell conserves energy by coupling generation of a transmembrane ion gradient (ΔΨ) to electron transport between reduced electron carriers and the terminal electron acceptor, a coenzyme M (2-mercaptoethane sulfonate) (CoM)-coenzyme B (7-mercaptoheptanoylthreonine sulfonate) (CoB) heterodisulfide molecule (CoM-S-S-CoB) that is produced in the last step of methanogenesis ( 6 ). The CoM-S-S-CoB heterodisulfide must be reduced to the CoM-SH and CoB-SH thiols to be reused for subsequent rounds of methanogenesis. Reduction of CoM-S-S-CoB is performed by one of two heterodisulfide reductases: two-subunit membrane-bound HdrED (energy-conserving CoM-S-S-CoB reductase) and three-subunit cytosolic HdrABC (heterodisulfide reductase) ( 8 ). Genes encoding HdrED ( MA0687-MA0688 ) are essential for growth on trimethylamine, methanol, acetate, and methanol plus acetate ( 8 ). HdrED oxidizes the membrane electron carrier methanophenazine (MP) and reduces the CoM-S-S-CoB heterodisulfide. In the process, HdrED conserves energy using a q-loop mechanism to contribute to the transmembrane ion gradient by translocating protons across the cell membrane ( 9 ). The flow of protons back into the cell via ATP synthase produces ATP, which is then used for other biosynthetic reactions ( 10 ). The hdrA1C1B1 ( MA3126-MA3128 ) operon (here referred to as hdrABC ) is expressed on methylotrophic substrates but is nonessential. In contrast to HdrED, the HdrABC enzyme is cytoplasmic and does not conserve energy. M. acetivorans also expresses genes encoding HdrD2 ( MA0526 ), HdrA2 and polyferredoxin ( MA2867-MA2868 ), and HdrC2B2 ( MA4236-MA4237 ), all of which are constitutively expressed ( 8 ). The physiological roles of the hdrD2 , hdrA2 , and hdrC2B2 genes are not yet understood. In hydrogenotrophic methanogens, such as Methanococcus , HdrABC accepts electrons from hydrogen via a bound hydrogenase, Vhu. An electron bifurcation mechanism is used to drive reduction of CO 2 to formyl-methanofuran by the formyl-methanofuran dehydrogenase Fmd, coupling the reaction with the thermodynamically favorable reduction of CoM-S-S-CoB ( 11 , 12 ). Though it does not itself conserve energy, HdrABC is essential for methanogenesis by Methanococcus because it provides the low-potential electrons necessary for the first step of the hydrogenotrophic methanogenesis pathway. However, because functional hydrogenases are not expressed in M. acetivorans and some other Methanosarcinales species, Vhu or another hydrogenase cannot be the electron donor for HdrABC in these organisms. In light of this fact, it is possible that HdrABC may use one of three general mechanisms for catalysis: an electron confurcation mechanism (defined as when one electron each from two separate donors is transferred to one two-electron acceptor), where electrons from both reduced ferredoxin (FdxH 2 ) and reduced F 420 (8-hydroxy-5-deazaflavin [F 420 H 2 ]) are used to reduce two molecules of CoM-S-S-CoB ( Fig. 1A ); a direct-reduction mechanism, where either FdxH 2 or F 420 H 2 could reduce CoM-S-S-CoB ( Fig. 1B ); or an electron bifurcation mechanism that couples oxidation of FdxH 2 to reduction of both CoM-S-S-CoB and another electron carrier, such as F 420 ( Fig. 1C ). Based on the lack of hydrogenase activity in M. acetivorans and because the reaction would be energetically favorable, HdrABC was thought to use FdxH 2 as a substrate to directly reduce CoM-S-S-CoB ( Fig. 1B ) ( 8 ). FIG 1 Putative models for the role of HdrABC during methylotrophic growth. (A) In an electron confurcation mechanism, HdrABC uses electrons from both FdxH 2 and F 420 H 2 in a 1:1 ratio to reduce two molecules of CoM-S-S-CoB and bypasses Rnf, Fpo, and HdrED energy-conserving enzyme complexes. (B) In a direct-reduction mechanism, HdrABC bypasses either Rnf or Fpo ion translocation steps by using FdxH 2 or F 420 H 2 as an electron donor. (C) HdrABC uses a bifurcation mechanism to reduce CoM-S-S-CoB while interchanging electrons between FdxH 2 and F 420 H 2 electron carriers. The model in panel B most closely matches the experimental data. k a is the rate of HdrABC enzyme activity; k CH4 is the rate of methane production; t g is the generation time. Whether HdrABC uses both or either ferredoxin and F 420 as an electron carrier has predictable consequences for the cell. Ferredoxin is also thought to be the electron donor for the sodium-pumping ferredoxin-methanophenazine oxidoreductase, Rnf (named for homology to the Rhodobacter nitrogen fixation enzyme complex) ( 13 – 15 ). On methylotrophic growth in high-salt (HS) medium, Rnf pumps 0.04 sodium ion across the cell membrane per 2 electrons ( 16 ). F 420 H 2 is the electron donor for the proton-pumping F 420 H 2 -methanophenazine Fpo complex. Fpo pumps two protons across the cell membrane per two electrons ( 17 ). The reduced methanophenazine (MPH 2 ) produced by both Rnf and Fpo is oxidized by HdrED to translocate protons across the cell membrane. Therefore, if HdrABC confurcates electrons from both FdxH 2 and F 420 H 2 in a 1:1 stoichiometry, HdrABC will compete for electrons with both Rnf and Fpo complexes, decreasing electron flux through HdrED and resulting in lower ion motive force generated per mole substrate consumed ( Fig. 1A ). If HdrABC uses a direct mechanism using either FdxH 2 or F 420 H 2 to reduce CoM-S-S-CoB, then it will compete with either Rnf or Fpo for substrate and lower the electron flux through HdrED. However, because Rnf and Fpo have a 50-fold difference in transmembrane ion translocation activity, the effect of HdrABC enzyme activity on the ion motive force generated would depend on whether HdrABC competes with Rnf or Fpo ( Fig. 1B ). Finally, if HdrABC can bifurcate electrons by using FdxH 2 to reduce both CoM-S-S-CoB and F 420 , then HdrABC may compete with Rnf for substrate but increase the electron flux through the more energy-efficient Fpo ( Fig. 1C ). We tested these hypotheses by adding a second copy of the methylotrophic HdrABC operon to the M. acetivorans chromosome and by using computational metabolic-flux modeling. Instead of observing a growth defect compared to the parental strain, as we had anticipated, we discovered that the mutant cells had growth kinetics and biomass identical to those of the parent while increasing the methane production rate by 30%. This suggests that an increased substrate uptake rate can compensate for decreased metabolic efficiency and that hdrA1B1C1 in M. acetivorans uses a direct mechanism ( Fig. 1B ) rather than an electron bifurcation mechanism to reduce CoM-S-S-CoB.",
"discussion": "DISCUSSION M. acetivorans has evolved multiple mechanisms for survival under a wide array of changing growth conditions. It can adjust to take advantage of changes in carbon and energy sources by altering gene expression of multiple substrate-specific methyltransferases to utilize methanol, methylamines, and methylsulfide ( 5 , 34 , 36 – 44 ). It can also adjust to changes in salt concentration by synthesizing glycine betaine as an osmoregulator; by upregulating phosphate and sodium transporters; by using a sodium proton antiporter, MrpA, to optimize the transmembrane proton gradient to maintain optimal ATPase function; and by using a promiscuous H + /Na + ATP synthase ( 31 , 45 , 46 ). Our observations suggest Hdr enzymes have also evolved as a consequence of selective pressure to respond to fluctuations in substrate availability. Phylogenomic analyses show that there are several classes of Hdr enzymes in methanogens. Obligate hydrogenotrophic methanogens generally have two sets of HdrABC enzymes (one encoding selenocysteines), while Methanosarcinales has two kinds of Hdr enzyme (HdrABC and HdrED), as well as multiple copies of Hdr subunit genes. Most Methanosarcinales genomes contain hdrABC genes that share similarity with methylotrophic and obligate acetoclastic Methanosarcinales \n hdrABC genes, while the other hdrA2 and hdrC2B2 genes are more similar to the obligate hydrogenotrophic methanogen hdrABC genes ( 8 ). Therefore, within the Methanosarcinales lineage, it appears as if there was a strong selective pressure to evolve increased HdrABC levels to uncouple methanogenesis and growth. Our physiological data suggest hdrABC genes allow methylotrophic methanogens to rapidly take up substrate and grow, albeit at submaximal efficiency. This scenario is consistent with M. acetivorans as a k -strategist at high substrate concentrations that turns over substrate faster but less efficiently. (Microbial k -strategists are characterized by an inability to adjust their growth rate in order to compete for high concentrations of substrate. Instead, k -strategists compete effectively at low substrate concentrations by having increased substrate affinity and uptake.) By converting substrate to methane faster and rapidly dropping the concentration of available substrate in the environment, over time, Methanosarcinales may outcompete other microbes for substrate as long as it is still able to generate ATP above a threshold maintenance level. Multiple versions and copies of Hdr enzymes could have evolved to form specialized protein-protein interactions. Methanococcus HdrABC can use an electron bifurcation mechanism and forms specific protein-protein interactions with the hydrogenase Vhu; formyl-methanofuran dehydrogenase, Fmd; and formate dehydrogenase, Fdh ( 11 , 12 ). HdrABC enzymes in M. acetivorans cannot interact with Vhu or another hydrogenase because none are expressed. However, Fmd or another protein partner(s) is possible. HdrED may also participate in protein-protein interactions in the cell. The HdrD subunit of HdrED interacts with acetyl-CoA decarbonylase/synthase (ACDS) and methylene-tetrahydromethanopterin reductase (Mer) ( 47 ). Further experiments are necessary to determine which proteins and enzymes interact with each other, how these relationships change as Methanosarcina switches between carbon and energy sources, and the degree of relevance to growth of the organism. Hdr enzymes may also tailor electron donor specificity as cells switch from one methanogenesis pathway to another, for instance, between methylotrophic and acetoclastic pathways. Though HdrABC and Rnf compete for the same electron donor in methanol-grown cells, Rnf is upregulated 4- to 10-fold when cells are grown on acetate and genes necessary for the oxidative branch of the methanogenesis pathway are poorly expressed ( 15 , 33 , 34 , 36 ). Transcriptional-fusion experiments suggest HdrA2, polyferredoxin, and HdrC2B2 genes in M. acetivorans are upregulated on acetate, and acetate-grown cells are known to have ferredoxin-heterodisulfide reductase activity (Fho), which is coupled with Rnf ( 16 ). If overexpressed HdrABC cannot kinetically compete with Fho, it may explain why overexpression of the methylotrophic HdrABC is not able to increase the rate of methanogenesis from acetate ( 8 ). If the methylotrophic HdrABC enzyme interacts directly with Fmd, or if it requires a ferredoxin that is not expressed on acetate, HdrABC would not be able to compete with Rnf for electrons when cells are grown on acetate. This idea is consistent with the fact that the acetate-induced HdrA2 gene is expressed as part of an operon that also contains a polyferredoxin. Regardless of substrate specificity details, our results suggest manipulating Hdr gene expression increases the rate of substrate turnover by bypassing rate-limiting redox reactions, further enhancing the k -strategist metabolism of M. acetivorans ."
} | 3,526 |
32233442 | PMC7227016 | pmc | 1,459 | {
"abstract": "The\nability to actuate liquids remains a fundamental challenge\nin smart microsystems, such as those for soft robotics, where devices\noften need to conform to either natural or three-dimensional solid\nshapes, in various orientations. Here, we propose a hierarchical nanotexturing\nof piezoelectric films as active microfluidic actuators, exploiting\na unique combination of both topographical and chemical properties\non flexible surfaces, while also introducing design concepts of shear\nhydrophobicity and tensile hydrophilicity. In doing so, we create\nnanostructured surfaces that are, at the same time, both slippery\n(low in-plane pinning) and sticky (high normal-to-plane liquid adhesion).\nBy enabling fluid transportation on such arbitrarily shaped surfaces,\nwe demonstrate efficient fluid motions on inclined, vertical, inverted,\nor even flexible geometries in three dimensions. Such surfaces can\nalso be deformed and then reformed into their original shapes, thereby\npaving the way for advanced microfluidic applications.",
"conclusion": "Conclusions In\nsummary, we designed nanostructured\nsurfaces using hierarchical textures to optimize shear hydrophobicity\nand tensile hydrophilicity to tune droplet transport and adhesion\non arbitrarily oriented devices. We demonstrated that the CYTOP/ZnO/Al\nsheets enabled a significantly better actuation performance when compared\nto conventional ZnO/Si acoustic wave devices. These CYTOP/ZnO/Al sheets\nwere uniquely capable of efficiently driving droplets across a wide\nrange of inclination angles (0° < α ≤ 180°).\nOur designed devices can be flexibly deformed and then maintain their\nshape, while enabling efficient liquid pumping. We envisage that the\nsurface design principles and demonstrated devices/applications will\nbe pivotal for liquid sampling for smart diagnostic systems within\nmobile and disposable settings."
} | 461 |
29774049 | PMC5946492 | pmc | 1,460 | {
"abstract": "Background Methanogenic biodegradation of aromatic compounds depends on syntrophic metabolism. However, metabolic enzymes and pathways of uncultured microorganisms and their ecological interactions with methanogenic consortia are unknown because of their resistance to isolation and limited genomic information. Results Genome-resolved metagenomics approaches were used to reconstruct and dissect 23 prokaryotic genomes from 37 and 20 °C methanogenic phenol-degrading reactors. Comparative genomic evidence suggests that temperature difference leads to the colonization of two distinct cooperative sub-communities that can respire sulfate/sulfite/sulfur or nitrate/nitrite compounds and compete for uptake of methanogenic substrates (e.g., acetate and hydrogen). This competition may differentiate methanogenesis. The uncultured ε - Proteobacterium G1, whose close relatives have broad ecological niches including the deep-sea vents, aquifers, sediment, limestone caves, spring, and anaerobic digesters, is implicated as a Sulfurovum -like facultative anaerobic diazotroph with metabolic versatility and remarkable environmental adaptability. We provide first genomic evidence for butyrate, alcohol, and carbohydrate utilization by a Chloroflexi T78 clade bacterium, and phenol carboxylation and assimilatory sulfite reduction in a Cryptanaerobacter bacterium. Conclusion Genome-resolved metagenomics enriches our view on the differentiation of microbial community composition, metabolic pathways, and ecological interactions in temperature-differentiated methanogenic phenol-degrading bioreactors. These findings suggest optimization strategies for methanogenesis on phenol, such as temperature control, protection from light, feed desulfurization, and hydrogen sulfide removal from bioreactors. Moreover, decoding genome-borne properties (e.g., antibiotic, arsenic, and heavy metal resistance) of uncultured bacteria help to bring up alternative schemes to isolate them. Electronic supplementary material The online version of this article (10.1186/s13068-018-1136-6) contains supplementary material, which is available to authorized users.",
"conclusion": "Conclusions The metabolic roles of uncultured microorganisms prevalent in previous phenol-degrading methanogenic bioreactors are predicted by genome-resolved metagenomics. Comparative genomics enriches our view on the microbial syntrophic and competitive interactions in phenol-degrading methanogenic consortia. Revealing a relationship between the formation of distinct but cooperative sulfate/sulfite/sulfur or nitrate/nitrite-reducing sub-communities and deteriorated methanogenic activity justifies biological manipulation to maximize methanogenesis in full-scale anaerobic digesters. While genome-resolved metagenomics shows its power in mining uncultured bacteria, other complementary approaches, such as activity- and cultivation-based ones, are needed to further validate their genome-resolved physico-chemical properties and metabolic pathways.",
"discussion": "Results and discussion Metagenome-assembled genomes The genomic DNA from two phenol-degrading methanogenic bioreactors (operated for 193 days at mesophilic (37 °C, MP) and ambient temperatures (20 °C, AP) were sequenced, yielding 11.0 and 8.6 Gbp of clean paired-end reads (100 bps in length), respectively (Additional file 1 : Table S1). The two MP data sets with library insert sizes of 180 and 800 bps were co-assembled into a total of 172 Mbp scaffolds with lengths ranging from 1 to 423 Kbp. Independent mapping of the raw reads of MP and AP samples to MP scaffolds and the visualization of coverage profiles in a two-dimensional coverage plot (Fig. 1 ) displays differential coverage of scaffolds, indicating that the identical microbial populations occurred with different abundances in the MP and AP bioreactors, which is also reflected by their different bacterial community composition revealed by read-based 16S rRNA gene analysis [ 3 ]. Grouping all scaffolds by their coverages clustered the scaffolds into 23 putative genome bins (labeled from G1 to G23, Fig. 1 ). These bins were further discriminated using a PCA of TNFs to remove potential contamination from other species in the identical coverage-defined bin, yielding refined genomes with sizes between 1.78 and 4.91 Mbp, GC content from 37.0 to 68.6%, and estimated genome completeness over 85% for most genomes (Fig. 1 ). Phylogenetic analysis reveals the novelty of the reconstructed genomes Taxonomic assignment of full-length or near-full-length 16S rRNA gene sequences of the 19 curated genomes suggests that they originate from a wide range of phylogenetic distances (Fig. 2 ), including 16 members from 11 bacterial orders and 3 methanogens from three archaeal orders. Four genomes without near-complete (> 1000 bp) 16S rRNA genes, i.e., G9, G16, G17, and G18, were assigned according to the lowest common ancestors (LCAs) of their genomic essential single-copy marker genes (ESCGs), namely, to the families Synergistaceae (G9, 100% LCAs) and Anaerolineaceae (G16, G17, and G18 with 92, 85, and 86% LCAs). Fig. 2 Maximum-likelihood 16S rRNA gene phylogenetic tree showing the placement of reconstructed genomes. The 16S rRNA genes of 4 out of total 23 genomes, namely, G9, G16, G17 and G18, are not successfully retrieved. The closest representative to each sequence is shown in parentheses (Accession Number). The accumulative bar chart indicates the relative abundance of each 16S rRNA gene sequence in the metagenomes of SEED, AP (20 °C) and MP (37 °C) sludge \n The phylogenic analysis demonstrates that most of the reconstructed genomes belong to uncultured microorganisms previously known only by their 16S rRNA genes. The genome sequences were obtained for an uncultured ε -proteobacterium dominant in the AP (20 °C) reactor (G1, Additional file 1 : Figure S2a, Table S3), a Chloroflexi T78 clade representative abundant in the MP (37 °C) reactor (G3, Additional file 1 : Figure S2b), and less abundant uncultured microbes with different degrees of genetic novelty, including members of one unclassified OP8 class (G4, Additional file 1 : Figure S2c), two unclassified Bacteroidetes families [WCHB1-69 (G11) and PL-11B8 wastewater-sludge group (G19)], two unclassified Synergistaceae genera (G13), two Syntrophobacterales genera (G20, G21), two euryarchaeotal genera [ Methanobacterium (G8) and Methanolinea (G10)], and one Mycobacterium species (G15) (Fig. 2 ; Additional file 1 : Table S3). Moreover, the 16S rRNA genes of G6 (1517 bp, 99.2%), G7 (1534 bp, 99.5%), and G14 (1526 bp, 99.9%) are nearly identical to Brachymonas denitrificans [ 32 ], Advenella faeciporci [ 33 ], and Cryptanaerobacter phenolicus [ 34 ], which are cultured species with isolates but still without a complete reference genome. The draft genomes obtained provide genomic insights into metabolic pathways of these bacterial populations in methanogenic environments, as discussed below. G1: an uncultured versatile ε -proteobacterial S/H 2 /N-metabolizer with remarkable environmental adaptability Uncultured ε -Proteobacterium G1 is highly enriched in AP (20 °C) digester (Figs. 2 ; 6 b). This microorganism and its uncultured relatives, i.e., 16S rRNA gene clones from methanogenic digesters [ 4 , 35 ] or benzene-degrading enrichments [ 36 – 38 ], are embedded in the cluster of deep-sea hydrothermal vent sulfur-oxidizing chemolithoautotrophic ε - Proteobacteria (i.e., Sulfurovum , Nitratifractor , and Sulfurimonas ) and most resembles Sulfurovum (Fig. 2 ). This phylogenetic relatedness is also evident in a PCA biplot depicting the similarities in KEGG metabolic profiles for all sequenced ε -Proteobacteria, in which the Sulfurovum -like G1 is clustered with hydrothermal vent strains (E90 and E98), whereas clearly segregated from the pathogenic Helicobacter (E31–E38) and Campylobacter species (E8–E30) (Additional file 1 : Figure S3). However, G1 is probably a novel ε - Proteobacteria sublineage, because it shows only 93.2–94.4% 16S rRNA gene similarities to Sulfurovum spp. (i.e., NBC37-1, S. lithotrophicum , and recently nominated S. aggregans ) and has a much smaller genome (1.78 Mb; > 99% completeness) than NBC37-1 (2.56 Mb). Genome-based physiological predictions demonstrate that G1 shares many genetic commonalities with Sulfurovum sp. NBC37-1 and other S/H 2 -oxidizing metabolizers, including Nitratifractor , Sulfuricurvum , Nitratiruptor , Arcobacter, and Sulfuromonas (Additional file 1 : Figure S2a, Table S4). Sulfur metabolism G1, like its autotrophic relatives (e.g., NBC37-1) [ 39 – 41 ], encodes all of the genes involved in the reductive tricarboxylic acid cycle (rTCA) for carbon (CO 2 ) fixation and gene clusters for oxidative phosphorylation, hydrogen utilization (described below), polysulfide respiration, sulfide oxidation, and membrane-bound respiratory nitrate reduction (Fig. 3 ), providing the genetic basis for their versatile respiration. G1 has cytoplasmic and periplasmic sulfide:quinone oxidoreductases (Sqr) that can catalyze sulfide (HS − ) oxidation to elemental sulfur (S 0 ), a process perceived as contributing to filamentous sulfur formation in hydrothermal vents [ 40 ]. Moreover, G1 encodes genes involved in sulfur metabolism and homologous to those of NBC37-1, including (I) a polysulfide reductase cassette (PsrABC) that catalyzes polysulfide respiration coupled to hydrogen oxidation, (II) a sulfate adenylyltransferase (Sat) that catalyzes adenosine phosphosulfate oxidation to sulfate, and (III) a thiosulfate sulfurtransferase that catalyzes thiosulfate oxidation to sulfite. These genes indicate the similar capacity of the organism in utilizing sulfur compounds as both electron donors and acceptors (Additional file 1 : Table S7-G1). Fig. 3 Central metabolism and solute transport in methanogenic digester ε -Proteobacterium G1. The closest cultured relative of G1, i.e., deep-sea vent Sulfurovum sp. NBC37-1, was used for comparison. Metabolic pathways for which enzymes are not encoded in either G1 or NBC37-1 but in other deep-sea vent ε - Proteobacteria are filled in grey. IM inner membrane, OM outer membrane, Fd ferredoxin, Cyt cytochrome, H 2 ase hydrogenase, Sqr sulfide–quinone oxidoreductases, Fcc flavocytochrome c sulfide dehydrogenase, Nif nitrogen-fixing proteins, Nar membrane-bound nitrate reductase, Nap periplasmic nitrate reductase, cdNir cytochrome cd1nitrite reductase, Nor nitric oxide reductase, Nos nitrous oxide reductase, Mdh malate dehydrogenase, Sdh succinate dehydrogenase (The figure was modified from Nakagawa et al. [ 40 ]) \n However, unlike NBC37-1, G1 has no sulfur-compound oxidation (Sox) system, which is typical in all other isolated or sequenced S/H 2 -oxidizing ε -proteobacterial chemoautotrophs except for Sulfurospirillum deleyianum , Nautilia profundicola , and Thiovulum sp. (Additional file 1 : Table S4). Intriguingly, nitrogen fixation genes are encoded by both Sox-lacking G1 and N. profundicola with compact genomes (1.78 and 1.68 Mb, Additional file 1 : Table S4-Yellow) and SoxCDYZ-lacking Sulfuricurvum and Nitratifractor (Additional file 1 : Table S4-Green), whereas such genes are absent from those SoxCDYZ-carrying species of Sulfurovum , Sulfurimonas and Arcobacter (Additional file 1 : Table S4-Purple). Combined, G1 might be a key player in driving sulfur cycling in the methanogenic reactors. Hydrogen utilization as an energy source Hydrogen-oxidizing sulfur respiration pathways using a hydrogenase and Psr have been noted in the energy metabolism of characterized deep-sea vent sulfur-metabolizing ε - Proteobacteria [ 42 ]. G1 also encodes four Ni–Fe hydrogenase subunits (2597532332-35, including two H 2 -uptake type and two H 2 -sensing type) and three Psr subunits ABC (2597532414-16) most similar to those in strain NBC37-1. The presence of these genes suggests the capacity to oxidizing hydrogen when using oxidized sulfur compounds as an electron acceptor. The presence of gene cassettes coding respiratory nitrate reductases (2597531175-77) and aerobic cbb3-type cytochrome c oxidase (2597531014-16) indicates that this bacterium is likely also capable of hydrogen oxidation by respiring nitrate or oxygen (Fig. 3 ). Moreover, G1 contains complete H 2 -uptake hydrogenase clusters (e.g., HydABCD, HyaCD and HypFBCDEA, Additional file 1 : Table S7-G1) which are well conserved in ε - Proteobacteria [ 40 , 43 ]. By contrast, this bacterium, similar to the pathogenic Helicobacter species, lacks H 2 -evolving type of Ni–Fe hydrogenases typical in deep-sea vent ε -proteobacterial genomes (e.g., strains NBC37-1 and SB155-2) [ 40 ]. This type of hydrogenase is associated with the hydrogen release in formate (via formate dehydrogenase H) or carbon monoxide oxidation, or energy conservation during methanogenesis [ 44 ]. In addition, the presence of H 2 -sensing Ni–Fe hydrogenases in G1 (methanogenic digesters) and deep-sea vents strains, including NBC37-1 and SB155-2, is likely necessitated by the relative low concentrations of H 2 in a methanogenic digester (typically < 0.1 μM [ 45 ]) or deep-sea vents (< 10 μM [ 40 ]). In summary, G1 is probably a substrate competitor of hydrogenotrophic methanogens (i.e., Methanobacterium and Methanolinea ) in both methanogenic reactors, considering its full metabolic capacity in H 2 -oxidizing sulfur-compound respiration. Environmental adaptation and ecological niches G1 encodes plentiful enzymes to support microaerobic growth and cope with oxygen or oxidative stress (Additional file 1 : Table S7-G1), including enzyme complexes I–V and cbb3-type cytochrome c oxidase that carries out a complete oxidative phosphorylation pathway (Fig. 3 ), antioxidant enzymes including alkyl hydroperoxide reductases and peroxidases, and iron cofactored superoxide dismutase. Moreover, like strain NBC37-1, G1 contains a large number of two-component regulatory system genes to sense and respond to changes in environmental cues, as well as a variety of transport enzyme systems to flexibly respond to environmental minerals or multidrugs, including (I) detoxification mechanisms of heavy metals including copper, cadmium, zinc, arsenate, and (II) resistance enzyme systems of acriflavin and vancomycin (Fig. 3 and Additional file 1 : Table S7-G1). G1 possesses at least four couples of coding genes of chromosomal toxin–antitoxin (TA) systems (e.g., RelE–YefM, CopG–RelE, and YefM–YoeB). The exact roles of TA systems in cells are unclear, but their prevalence in bacterial genomes could be associated with stress resistance, population growth regulation, biofilm formation, or even niche-specific colonization [ 46 – 48 ]. In addition, G1 contains two regions of CRISPRs and CRISPR-associated protein-coding genes (IMG gene IDs: 2597532104-07; 2597531940-43), which may severe as defense systems against the invasion of exogenous genetic materials (e.g., phage infection). Combined, our results suggest that ε - Proteobacterium G1 can be a facultative anaerobic metabolizer of hydrogen and sulfur compounds. These genetic potentials related to versatile metabolic capacities and remarkable environmental adaptability may provide solid foundations for the widespread distribution of G1-resembling 16S rRNA clones or populations (BLASTN similarity > 95%, bits core > 2300) in various artificial systems and environmental niches, such as benzene-degrading sulfate-reducing bioreactors [ 36 – 38 ], acetate-amended aquifers [ 49 ], antibiotic-receiving river sediments [ 50 ], sulfidic cave biofilms and springs [ 51 – 53 ], and limestone sinkholes [ 54 ]. Notably, recent protein stable isotope probing experiments with labeled acetate unveils some uncultured Epsilonproteobacteria as highly efficient dominant acetate scavengers in a sulfate-reducing microbial community mineralizing benzene [ 55 ]. Assume that G1 could also utilize acetate, it can compete with Methanosaeta , which are the dominant methanogens present in both bioreactors. T78 clade: phylogeny, occurrence, and metabolic potentials Phylogenetic and taxonomic analysis of the full-length 16S rRNA gene sequences (1489 bp) of G3, which is markedly enriched in the MP reactor (Fig. 6 b) shows that this T78 clade species is most closely related to members of Anaerolineaceae , including Bellilinea sp. clone De3218 (95.2%) and Longilinea sp. clone 48IIISN (95.0%), followed by Leptolinea BUT1_OTUB3 (88.9%), Levilinea clone SBYH_799 (88.9%), and Anaerolinea thermophila UNI-1 (87.2%) (Fig. 2 ; Additional file 1 : Figure S2b). The T78 clade is a Chloroflexi cosmopolitan in freshwater lakes and springs [ 56 , 57 ], sediments [ 58 – 60 ], and anaerobic digestion and biogas systems [ 61 – 63 ]. Genomic analysis shows that G3 encodes complete KEGG pathway enzymes for the beta oxidation of butyrate via crotonoyl-CoA to acetyl-CoA (Additional file 1 : Figure S5a, Table S7-G3). G3 also has electron transfer flavoproteins (2600084386-87), NAD(P)-dependent iron-only hydrogenases (2600085382-85), and NAD(P) transhydrogenases (2600085557-60), which may be responsible for electron transfer and energy conservation via proton reduction (i.e., hydrogen production) coupled to proton translocation. Overall, the presence of these genes implicates that G3 is likely capable of butyrate oxidation. In addition, G3 has nearly a dozen ADH and AdDH enzymes for alcohol dehydrogenation and a complete list of enzyme-coding genes for KEGG glycolysis/gluconeogenesis and pentose phosphate pathways, suggesting that this T78 clade representative can metabolize alcohols and carbohydrates (Additional file 1 : Table S7-G3). Syntrophorhabdus : key phenol-degrading syntrophs Biodegradation mechanisms of aromatic compounds have been well documented in isolates of nitrate-reducing bacteria (NRB) [ 64 , 65 ], SRB [ 66 – 68 ] and iron-reducing bacteria (IRB) [ 57 ]. Here, we reconstructed two near-complete genomes (G2 and G5) of genus Syntrophorhabdus , which is known to syntrophically degrade phenol under methanogenic conditions [ 69 ]. Phylogenetic distance of 16S rRNA gene, in silico DNA–DNA hybridization value (DDH), average amino acid identity (AAI), and average nucleotide identity (ANI) (Additional file 1 : Table S5) congruously reveal that G5 is from a novel Syntrophorhabdus species (Fig. 2 ; Additional file 1 : Table S5). A gene-by-gene manual comparison suggests this organism shares numerous gene cassettes with S . aromaticivorans (G2 and strain UI) for (I) the syntrophic biodegradation of phenol, 4-OHB, benzoyl-CoA (using non-ATP-dependent benzoyl-CoA reductase, Additional file 1 : Table S7) and benzoate (Fig. 4 a and Additional file 1 : Figure S4) and (II) a novel reverse electron transport mechanism [ 11 ] in Rnf-lacking syntrophic metabolizers of aromatic compounds (Additional file 1 : Text S1). Fig. 4 Anaerobic phenol and benzoate pathways ( a ) and the Pps-Ppc operons ( b ). a Phenol is anaerobically degraded to benzoyl-CoA, which is then degraded via ‘Dch-Had-Oah’ pathway and beta oxidation to butyrate and acetate. b The pathways can be conducted by anaerobes under methanogenic, nitrate-reducing, sulfate-reducing and iron-reducing conditions. The percentage under each enzyme-coding gene indicates its similarity (%) to the gene counterpart in Syntrophorhabdus aromaticivorans strain UI. OHB 4-hydroxybenzoate, 4-BCL 4-hydroxybenzoate-CoA ligase, 4-Hbcr 4-hydroxybenzoyl-CoA reductase, BCL Benzoate-CoA ligase, BCR Benzoyl-CoA reductase, Dch Cyclohexa-1,5-dienecarbonyl-CoA hydratase, PaaK putative phenylacetate-CoA ligase, Had 6-hydroxycylohex-1-en-1-carbonyl-CoA dehydrogenase, PaaK putative phenylacetate-CoA ligase, Oah 6-oxo-cyclohex-1-ene-carbonyl-CoA hydrolase, PaaK putative phenylacetate-CoA ligase, ACSL long-chain acyl-CoA synthetase \n A further comparison of the phenylphosphate synthase and carboxylase gene clusters (Pps–Ppc operons) in Syntrophorhabdus with those encoded by NRB, SRB, and IRB reveals that they share essential gene subunits, i.e., PpsAB and PpcBY in the Pps–Ppc operons (Fig. 4 b). The PpcX (UbiD-like) and PpcY (UbiX-like) subunits encoded in the downstream of Pps–Ppc operons of syntrophs and NRB were distantly homogenous to 4-hydroxybenzoate (4-OHB) decarboxylases (BLASTP similarity: 28 and 47%; bit score: 170 and 172), which are enzymes responsible for phenol decarboxylation to 4-OHB [ 70 ]. It has been demonstrated that PpcX was transcribed to directly carboxylate phenol to 4-OHB in iron-reducing archaea Ferroglobus placidus [ 71 ], while PpcY (originally referred as ORF8 in T. aromatica ) was transcribed alongside PpsAB and PpcB genes in Geobacter metallireducens during anaerobic growth on phenol [ 72 ]. Based on these observations, it is speculated that PpcX and PpcY are putative 4-OHB decarboxylases responsible for a previously unrecognized phenol decarboxylation pathway in Syntrophorhabdus . Cryptanaerobacter : unrecognized genetic contents and metabolic pathways The first and only isolate of Cryptanaerobacter , i.e., strain LR7.2, was described as an anaerobe that presumably utilized phenol or 4-OHB as an energy source and electron acceptor for growth in pure culture with essential complex supplements, seemingly converting these compounds into benzoate via unknown “unusual anaerobic respiration” [ 34 ]. However, the authors could neither identify electron donors and carbon sources, nor explain the stimulated growth of strain LR7.2 by sulfite. For more than a decade, studies have been investigating whether Cryptanaerobacter , like its Pelotomaculum relatives [ 73 ], syntrophically oxidizes organic substrates, or like its Desulfitobacterium relatives [ 74 ] performs sulfite/sulfonates reduction. In this study, we present genomic evidence that G14 performs (I) phenol/4OHB degradation and assimilatory sulfite reduction and (II) syntrophic propionate oxidation (Fig. 5 ) in methanogenic environments. Fig. 5 KEGG biodegradation pathways encoded in the genome of Cryptanaerobacter sp. G14. Each reaction is labeled with the abbreviated IMG gene IDs (e.g., ‘2600290001’ as ‘0001’) for the enzyme-coding genes. ADH alcohol dehydrogenases, AdDH aldehyde dehydrogenases, MQ menaquinone, Fd ferredoxin \n Gene-by-gene analysis of the first Cryptanaerobacter genome shows that unlike phenol-degrading Syntrophorhabdus , NRB, SRB, and IRB that utilize a phosphorylation–carboxylation pathway to convert phenol to 4-OHB, G14 encodes homologs of 4-OHB decarboxylase subunits BCD (2600292530-32, Additional file 1 : Table S6) that convert phenol into 4-OHB via a carboxylation pathway (Fig. 4 a). Moreover, G14 has the genes coding putative 4-hydroxybenzoate-CoA ligases (4-BCL, 2600292885), putative 4-hydroxybenzoyl-CoA reductase (4-HBCR, 2600294143-44), and benzoate-coa ligase (BCL, 2600294302) to further convert 4-hydroxybenzoate (4-OHB) to benzoate (Additional file 1 : Table S6). Notably, G14 encodes an anaerobic sulfite reductase (asrABC; 2600293676-78), NADPH-dependent flavin oxidoreductases (e.g., 2600293674), and putative NADPH-dependent hemoprotein (2600293675), which are engaged in the NADPH-dependent assimilatory sulfite reduction (ASR) to sulfide via the following reaction: \\documentclass[12pt]{minimal}\n\t\t\t\t\\usepackage{amsmath}\n\t\t\t\t\\usepackage{wasysym} \n\t\t\t\t\\usepackage{amsfonts} \n\t\t\t\t\\usepackage{amssymb} \n\t\t\t\t\\usepackage{amsbsy}\n\t\t\t\t\\usepackage{mathrsfs}\n\t\t\t\t\\usepackage{upgreek}\n\t\t\t\t\\setlength{\\oddsidemargin}{-69pt}\n\t\t\t\t\\begin{document}$${\\text{SO}}_{3}^{2 - } + 3{\\text{NADPH}} + 3{\\text{H}}^{ + } \\leftrightarrow {\\text{S}}^{2 - } + 3{\\text{NADP}}^{ + } + 3{\\text{H}}_{2} {\\text{O}} .$$\\end{document} SO 3 2 - + 3 NADPH + 3 H + ↔ S 2 - + 3 NADP + + 3 H 2 O . \n Therefore, G14 may perform phenol decarboxylation and assimilatory sulfite reduction to derive energy for growth. These metabolic capacities of G14 make the previously-observed sulfite stimulated growth of C. phenolicus plausible [ 34 ]. Notably, G14 encodes three complete sets of ABC-type transport systems for sulfonate (Additional file 1 : Table S6), a substrate utilizable by its sulfite-reducing Desulfitobacterium relatives as a terminal electron acceptor (TEA) for growth [ 74 ]. Nevertheless, to ascertain whether G14 assimilates sulfonate is impossible merely by reference-based genome annotation, because the enzymes responsible for sulfonate assimilation in Firmicutes remain unknown. Assuming that the ABC-type sulfonate transport system, sulfite reductases, and hypothetical proteins of G14 enable its uptake of sulfonate, phenol/4-OHB may also be degraded via sulfonate respiration. Future studies are needed to test whether sulfonate supports growth of Cryptanaerobacter species and which enzymes will be expressed by Desulfitobacterium isolates to utilize (aliphatic) sulfonate as TEA for growth with the release of sulfonate sulfur as sulfide. On the other hand, G14 encodes a complete methylmalonyl-CoA (MMC) pathway for the syntrophic conversion of propionate into succinyl-CoA and succinate. These two compounds are then metabolized via oxaloacetate to pyruvate and eventually transformed to acetyl-CoA and acetate through a series of enzymatic reactions (Fig. 5 and Additional file 1 : Table S6). The MMC pathway is a common mechanism for propionate oxidation in many mesophilic syntrophs, and the eight catabolic genes encoding the MMC pathway in G14 are most similar (66–89%, averaged 81%, Additional file 1 : Table S6) to those of its closest cultured relative, Pelotomaculum thermopropionicum SI (Fig. 2 , 93.8% 16S similarity; AAI: 53.2%), a syntrophic propionate-oxidizing thermophilic anaerobe [ 75 ]. Like strain SI, G14 encodes at least five alcohol dehydrogenases (ADH) and one acetaldehyde dehydrogenase (AdDH) for alcohol (e.g., propanol) oxidation (via aldehyde) to carboxylic acids (Fig. 5 and Additional file 1 : Table S6), such as propionate. The propionate could be further oxidized by G14 in syntrophy (with a hydrogen-scavenging partner) using an ion-translocating ferredoxin oxidoreductase genes (IFO)-associated cassette (2600291645-53, Additional file 1 : Table S6) that encodes a heterodisulfide reductase complex, hydrogenase subunits, and putative ion-translocating Fd:NADH oxidoreductase subunits for energy conservation, as is the case for many other Rnf-lacking syntrophic metabolizers [ 11 ]. Moreover, genome analyses reveal two iron-only hydrogenases and one Ni-Fe hydrogenase in G14 (Additional file 1 : Table S6), presumably engaging in discharging reducing equivalents (i.e., electrons) after propionate oxidation via three reductive steps, namely, menaquinone reduction by a succinate dehydrogenase (2600294405-07), NAD + reduction by a malate dehydrogenase (2600292336), and ferredoxin reduction by a pyruvate/ferredoxin oxidoreductase (2600292338) (Fig. 5 ). In addition, G14 has two gene clusters containing formate dehydrogenases. However, whether G14 can produce formate to exhaust reducing equivalents or pool electrons requires further investigation. Combined, the genomic evidence reveals that Cryptanaerobacter sp. G14 may play dual flexible and significant roles in methanogenic environments, both as a sulfite-respiring, aromatic compound degrader (i.e., phenol and 4-OHB) and a syntrophy specialist that uptakes volatile fatty acids (e.g., propionate) and alcohols (e.g., propanol, ethanol) generated by upstream fermentative or acidogenic bacteria and syntrophically degrades them to acetate, hydrogen, and carbon dioxide. These compounds are then available as substrates for downstream methanogens. Nitrate/nitrite denitrifiers in 37 °C phenol-degrading consortia KEGG pathway annotation of Burkholderiales genomes, i.e., G6 and G7, identifies gene cassettes and metabolic pathways (Additional file 1 : Table S7-G6 and G7) related to (I) nitrate and/or nitrite denitrification (e.g., Additional file 1 : Figure S5b for G6), (II) RnfABCDGE type electron transfer, (III) oxygen/oxidative tolerance, (IV) propionate biodegradation (via acryloyl-CoA pathway) (Additional file 1 : Figure S5c-III for G6), and (V) metabolism of pyruvate and lactate. These genetic potentials are well supported by the fully validated capacities of their closest cultured isolates, namely, B. denitrificans and A. faeciporci (Fig. 2 ), to use these organic substrates as carbon and energy sources [ 32 , 33 ]. In agreement with their oxygen/oxidative tolerance, G7 encodes two gene cassettes for oxic/anoxic biodegradation of 4-OHB (to pyruvate, Additional file 1 : Figure S5d-I) and catechol (to acetyl-CoA using catABC and pcaDIJ operons, Additional file 1 : Figure S5d-II and Table S7-G7), while G6 encodes four enzymes that catalyze oxic/anoxic oxidation of benzoate and phenol (to catechol; Additional file 1 : Table S7-G6). Moreover, protein-coding genes related to amino acid metabolism are the most abundant in both G6 (169 KEGG orthology, i.e., KO) and G7 (201 KO), followed by those genes involved in utilization of carbohydrates (G6: 134 KO; G7: 158 KO). These genome-encoded metabolic potentials of G6 and G7 are well in agreement with the validated capacities of their closest relatives to utilize a variety of amino acids as energy and carbon sources for anaerobic growth [ 32 , 33 ]. Notably, besides an acryloyl-CoA pathway, G6 also encodes all enzymes essential for propionate degradation via a MMC pathway (Additional file 1 : Figure S5c-I) and a reductive carboxylation pathway (Additional file 1 : Figure S5c-II and Table S7-G6). Sulfate and sulfite metabolizers in the 20 °C phenol-degrading consortia Three genomes, namely, G15, G21, and G22, are more enriched in the AP reactor (Fig. 6 b), display full genetic capacities to utilize sulfur compounds as electron acceptors. Fig. 6 Substrate competition between methanogenic consortia and sulfur/nitrogen compounds metabolizers. a MP: 37 °C; AP: 20 °C. Green arrow lines implicate syntrophic phenol-degrading methanogenic pathway. Blue arrow lines indicate pathway competition for methanogenic substrates and other organic molecules. APS adenosine 5′-phosphosulfate, GSB green sulfur bacteria Chlorobiaceae . b Percent in heatmap cells denotes relative abundance (%) of each genome in a metagenome \n G15 encodes complete enzymes (i.e., CysND, CysC, CysH, and sir) for assimilatory sulfate reduction via sulfite to sulfide (Additional file 1 : Table S7-G15). Consistent with most of its well-understood Mycobacterium relatives [ 76 , 77 ], this uncultured facultative anaerobic organism encodes biodegradation enzymes of xenobiotic polycyclic aromatic hydrocarbons (e.g., naphthalene, dichloropropene, and phenanthrene), carbohydrates (e.g., starch, sucrose, and glucose), and fatty acids (i.e., hexadecanoate, butyrate, and propionate). Smithella sp. G21 encodes enzymes for syntrophic oxidation of propionate (i.e., MMC oxidation pathway and acryloyl-CoA pathway) and butyrate (via crotonoyl-CoA), an iron-only hydrogenase, formate dehydrogenases, alcohol dehydrogenases, dissimilatory sulfite reductases (DsrABD, 2603687864-66) and associated electron transfer proteins (DsrKJO), and polysulfide reductases (Psr) (Additional file 1 : Table S7-G21), revealing its potential in VFAs and alcohol metabolism, hydrogen production, formate oxidation, and the uptake of sulfur compounds as electron acceptors. G22 resembles Desulfovibrio aminophilus DSM 12254 (99.9% 16S similarity, Fig. 2 ; 70% DDH), an amino acid-degrading and sulfate-reducing bacterium (isolated from an anaerobic dairy wastewater lagoon) that also utilizes formate, H 2 /CO 2 , and ethanol as electron donors [ 78 ]. Differentiated methanogenic phenol-degrading metabolic pathways Tracking habitat origins of close relatives of uncultured microorganisms in our MP and AP digesters by their reconstructed 16S rRNA genes reveals that these microbes are widespread in methanogenic bioreactors or enrichments receiving wastewater-borne aromatic compounds and sulfate [ 1 , 3 , 4 , 35 – 37 ]. The temperature difference between MP (37 °C) and AP (20 °C) bioreactors acclimates two divergent methanogenic communities that significantly differ in community composition (Figs. 1 and 6 b), and methane-producing rate (200 vs. 283 CH 4 -COD/g-VSS/day) and phenol-degrading rate (274.0 and 363.6 g-phenol/g-VSS/day) [ 3 ]. Besides potential temperature dependence of biodegradation rate, genome-resolved evidence suggests that the higher methane production and faster phenol degradation at 37 than 20 °C are associated with different microbial syntrophic and competitive relationships (Fig. 6 a) besides potential differences in enzymatic activities, as described below. In both MP and AP reactors where effluent monitoring on Day 113, 130, 146, 151, and 167 suggests the presence of benzoate, acetic acid, ethanol, and butanol (Fig. 6 a), phenol is syntrophically degraded to benzoyl-CoA by Syntrophorhabdus spp. (G2 and G5) in the presence of hydrogen-scavenging microorganisms, such as the hydrogenotrophic archaeal Methanobacterium (G8) and Methanolinea (G10). Phenol can be also converted to 4-OHB and benzoyl-CoA by Cryptanaerobacter sp. G14. Then, benzoyl-CoA is degraded by Syntrophorhabdus spp. and other syntrophs, such as Syntrophus spp., to fatty acids (e.g., butyrate/propionate/acetate) and/or alcohols (e.g., butanol/propanol/ethanol). These byproducts can be further selectively oxidized to methanogenic substrates (i.e., acetate, formate, or hydrogen) by syntrophs, including 37 °C-enriched T78 clade bacterium (G3) and Syntrophus aciditrophicus (G12), and/or 20 °C-enriched (G14), Syntrophus (G20) and Smithella (G21). After that, acetate is used by M. concilii (G23) for direct acetoclastic methanogenesis, whereas hydrogen and formate are probably utilized by hydrogenotrophic Methanobacterium (G8) and Methanolinea (G10) to reduce carbon dioxide to methane. However, the temperature difference induces the great shift of community composition in the seed sludge (Fig. 6 b) and the development of two distinct but cooperative sub-communities that, respectively, metabolize sulfate/sulfite/sulfur and nitrate/nitrite in AP and MP reactors, leading to potential substrate competition with methanogens and syntrophic bacteria (Fig. 6 a). In the AP reactor, sulfate is converted to sulfide through dissimilatory reduction by SRB Desulfovibrio (e.g., G22) and assimilatory reduction by Mycobacterium spp. (G15). This accompanies competitive uptakes of methanogenic substrates with methanogens (i.e., hydrogen, formate, acetate) or of small organic substrates (e.g., ethanol or amino acids) with syntrophs. Sulfite can be assimilated by Cryptanaerobacter (G14) to stimulate its growth and 4-OHB transformation activity. Sulfide, a notorious and ubiquitous product in anaerobic digestion processes, is oxidized by phototrophic green sulfur bacteria (GSB) Chlorobiaceae , which occurred only in the AP reactor and accounted for 3.6 and 4.5% of the amplicon and metagenome 16S rRNA gene sequences, respectively [ 3 ], to regenerate sulfite and sulfate, yielding element sulfur as an intermediate product. The element sulfur and polysulfide are reducible by uncultured ε -proteobacterium G1 and Smithella spp. G21 with the oxidation of hydrogen and formate, respectively. Therefore, SRB ( Desulfovibrio ), sulfur-reducing bacteria (G1), and GSB ( Chlorobiaceae ) can form a cooperative metabolic network in which sulfur compounds are recycled and exchanged among the partners. This leads to continuous competitive depletion of methanogenic substrates, thus deteriorating methane production by methanogens. In contrast, the co-occurrence of substrate competitors, i.e., nitrate/nitrite-denitrifying Brachymonas sp. G6, nitrate-reducing G1, and nitrite-denitrifying Advenella sp. G7 in the MP digester could also be detrimental to methanogenesis, although their total relative abundance is much lower than the sulfate/sulfite/sulfur-respiring sub-community in the AP reactor (Fig. 6 b). For example, our genomic evidence highlights that G6 can utilize organic acids (e.g., acetate, butyrate, benzoate, lactate, pyruvate), alcohols (e.g., ethanol), and some amino acids as carbon and energy sources, while G7 can assimilate acetate, propionate, lactate and pyruvate for growth. Therefore, they can compete with acetoclastic methanogens for acetate (e.g., G23) and with syntrophs for these organic acids and alcohols. The sustainable levels of nitrate (9.4–23.6 mg/L) in the MP bioreactor are probably attributed by the transient introduction of dissolved oxygen during bioreactor feeding."
} | 9,227 |
24616716 | PMC3935151 | pmc | 1,461 | {
"abstract": "Past studies of hydrogen cycling in hypersaline microbial mats have shown an active nighttime cycle, with production largely from Cyanobacteria and consumption from sulfate-reducing bacteria (SRB). However, the mechanisms and magnitude of hydrogen cycling have not been extensively studied. Two mats types near Guerrero Negro, Mexico—permanently submerged Microcoleus microbial mat (GN-S), and intertidal Lyngbya microbial mat (GN-I)—were used in microcosm diel manipulation experiments with 3-(3,4-dichlorophenyl)-1,1-dimethylurea (DCMU), molybdate, ammonium addition, and physical disruption to understand the processes responsible for hydrogen cycling between mat microbes. Across microcosms, H 2 production occurred under dark anoxic conditions with simultaneous production of a suite of organic acids. H 2 production was not significantly affected by inhibition of nitrogen fixation, but rather appears to result from constitutive fermentation of photosynthetic storage products by oxygenic phototrophs. Comparison to accumulated glycogen and to CO 2 flux indicated that, in the GN-I mat, fermentation released almost all of the carbon fixed via photosynthesis during the preceding day, primarily as organic acids. Across mats, although oxygenic and anoxygenic phototrophs were detected, cyanobacterial [NiFe]-hydrogenase transcripts predominated. Molybdate inhibition experiments indicated that SRBs from a wide distribution of DsrA phylotypes were responsible for H 2 consumption. Incubation with 13 C-acetate and NanoSIMS (secondary ion mass-spectrometry) indicated higher uptake in both Chloroflexi and SRBs relative to other filamentous bacteria. These manipulations and diel incubations confirm that Cyanobacteria were the main fermenters in Guerrero Negro mats and that the net flux of nighttime fermentation byproducts (not only hydrogen) was largely regulated by the interplay between Cyanobacteria , SRBs, and Chloroflexi .",
"conclusion": "Conclusion In this study, a suite of methods identified that a variety of Cyanobacteria were the dominant fermentive organism responsible for hydrogen production during nighttime constitutive metabolism. Furthermore, hydrogen production was driven by daytime carbon storage, and total hydrogen produced was a fraction of the total fermentation potential, with the majority of fermentation products being organic acids (especially acetate). This work also identified uptake of acetate during nighttime by both sulfate reducing bacteria and filamentous Chloroflexi provided an important linkage to Cyanobacteria . Taken together, these results indicate the nighttime fermentation of stored light energy can explain the close association of the filamentous Chloroflexi and of the SRB with cyanobacterial filaments. Conflict of interest statement The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.",
"introduction": "Introduction Hypersaline microbial mats, living analogs of early life on Earth (Des Marais, 2003 ), are compact and structured laminations of highly diverse microbial communities that undergo significant redox changes over the diel (day-night) cycle, alternating between oxic and anoxic states. The metabolic diversity of microbial mats is reflected in a diverse potential for H 2 metabolism (Hoehler, 2005 ), and nitrogen fixation (Omoregie et al., 2004a , b ), and previous work has documented significant efflux of H 2 from hypersaline microbial mats (Hoehler et al., 2001 ; Burow et al., 2012 ) under dark anoxic conditions. Hence, these systems are of interest not only in an ecological frame of reference, but also for bioenergy science. In previous work we showed that nighttime production of hydrogen gas from hypersaline mats of Elkhorn Slough, CA, USA originated within the photic layer of the mats and was primarily attributable to the fermentation activity of Cyanobacteria , especially the dominant filamentous cyanobacterium Microcoleus chthonoplastes , and was largely insensitive to nitrogen fixation (Burow et al., 2012 ). In the Elkhorn Slough mats, hydrogen consumers [sulfate-reducing bacteria (SRB)] were present in close physical association with hydrogen producers, and significantly reduced hydrogen efflux (Burow et al., 2013 , in press ). Understanding the ecological and environmental factors that control net H 2 production is thus critical to understanding the role of H 2 cycling in mat structure and ecology. In the present work we have characterized and quantified fermentative activity and consumption of fermentation products in two mat types from the previously documented site at Guerrero Negro, B.C.S., Mexico, that exhibit a range of net H 2 production rates. “GN-S” are well-developed subtidal mats located in pond 4 near pond 5 of the salt works and constructed primarily by the cyanobacterium Microcoleus chthonoplastes that have been described extensively in previous reports (Spear et al., 2003 ; Ley et al., 2006 ; Feazel et al., 2008 ; and Robertson et al., 2009 ). “GN-I” are intertidal mat communities constructed largely by Lyngbya sp. The difference in dominant cyanobacterium and extent of development of accessory populations are reflected in differing chemical behavior of the two mat types, including in H 2 efflux. Previous work on the GN-I mats documented that the integrated H 2 production rate is equivalent to 16% of net daytime carbon fixation (on a per-electron basis), and individual bubbles at the mat surface may contain up to 10% hydrogen in the pre-dawn hours (Hoehler et al., 2001 ; Hoehler, 2005 ). Moreover, the amount of H 2 efflux has been found to vary over more than four orders of magnitude, as a function of environmental forcing (Bebout et al., 2002 , 2004 ; Hoehler, 2005 ; Burow et al., 2012 ). In comparison, the submerged mats experience relatively stable conditions and release less H 2 . The extensive diversity, dominance patterns, and dynamic microbial response over space and time found within microbial mats are well suited for the development and deployment of advanced sequencing and isotope probing techniques to identify the complex biological interactions as well as energy and nutrient cycles of these highly diverse systems. When combined with traditional biogeochemical techniques, these methods can provide insights into energy and nutrient cycling in and through these systems. Using a comparative approach, the relative role of fermentation in the carbon and hydrogen cycles of these two mat types, and the ecology surrounding these cycles, was examined via a holistic set of molecular, isotopic, and biogeochemical methods. Specifically, we employed pyrotag libraries, functional gene sequencing of [NiFe]-hydrogenase ( hoxH ) and dissimilatory sulfite reductase ( dsrA ), stable isotope probing of labeled 13 C-bicarbonate and 13 C-acetate, Catalyzed Reporter Deposition Fluorescence In Situ Hybridization (CARD-FISH) probing of Chloroflexi and SRB clades, in combination with measurements of hydrogen, hydrogen sulfide, and organic acids to examine microbial mats manipulated with inhibitors that disrupt the sulfur, nitrogen, and carbon cycles. The results from these experiments indicated that in GN-I mats constitutive fermentation served to liberate roughly 80% of the photosynthetically fixed electrons into the bulk pool and thereby formed a basis for close trophic coupling between Cyanobacteria , filamentous anoxygenic phototrophs, and SRBs in hypersaline mats of both Guerrero Negro and of Elkhorn Slough.",
"discussion": "Discussion Previous work (Skyring et al., 1989 ; Burow et al., 2012 ) suggested that nighttime production of reduced gases results from photoautotrophy and storage of reduced carbon by Cyanobacteria with subsequent fermentation of stored photosynthate following the onset of dark, anoxic conditions (Hoehler et al., 2001 ; Des Marais, 2003 ). In Elkhorn Slough mats, Cyanobacteria were indicated as the dominant fermenter (Burow et al., 2012 ) and sulfate reducers as a key consumer of H 2 (Burow et al., in press ). This study builds on previous reports of H 2 cycling in microbial mats in three important regards. First, bulk chemical cycling and the underlying ecology are shown to be common features of geographically diverse mats, as well as mats constructed by distinctly different Cyanobacteria . Second, organic acid cycling is characterized and quantified, and shown to represent a significant component of overall carbon and electron flow in the studied mats. Last, multiple methods are utilized to demonstrate that exchange of fermentation products serves to directly link Cyanobacteria with sulfate reducers and anoxygenic phototrophs. Fermentation is quantitatively important in mats If daytime oxygenic photosynthetic fixation of carbon drives the subsequent nighttime fermentation to hydrogen and organic acids, a series of changes in metabolites should be observable as: (1) accumulation during the day and release at night of inorganic carbon based on measurements of the flux of dissolved inorganic carbon (DIC) across the mat-water interface, (2) accumulation during the day and depletion at night of small stored carbon polymers within the mat (e.g., glycogen), and (3) a rise in organic acid flux within the mat at night. These were all observed in the studies presented here. As shown in Table 2 and Figure 6 , inorganic carbon was incorporated into mats during daytime. Figure 7 indicated glycogen as the primary fixed carbon storage molecule in hypersaline microbial mats, with GN-S mats accumulating much more glycogen. Approximately ~81% of all carbon fixed during the day in GN-I mats was subsequently fermented at night, with most of the net accumulation of fermentation products occurring as organic acids rather than as hydrogen. The small amount of hydrogen released by fermentation activity relative to organic acids was unsuspected given that the hydrogen concentrations in the mat go through a four order of magnitude change in concentration throughout a diel cycle and particularly given that net hydrogen fluxes in GN-I mat are 10 times greater than in the GN-S mat. A stoichiometric fermentation of glucose to acetic acid, carbon dioxide, and hydrogen would produce nearly 2:1 electron ratios of acetate and hydrogen, but the ratios measured in this study were typically closer to 100:1 (Table 2 ) and suggests that organic acids provide the most quantitatively observable flux of reductant and energy available to the broader community of microbes within the mat under dark/anoxic conditions. Constitutive fermentation by Cyanobacteria was responsible for a majority of hydrogen and organic acid production The results of this study indicated similar diel patterns in organic acid and hydrogen production in hypersaline microbial mats from different locations and of different types. Because fermentation of photosynthate represents a loss of reducing power and delivers the lowest energy yield among potential catabolic processes, it could be viewed as a process to be minimized (that is, to be employed only if required by the demands of nighttime metabolism). One potentially large demand for such fermentation would be the energy required to fuel dinitrogen fixation and, for this reason, we examined whether inhibition of dinitrogen fixation (by addition of ammonium as a source of combined nitrogen) diminished the yield of fermentation products. As shown in Figure 2A , and consistent with previous observations in Elkhorn Slough, ammonium addition does not appear to affect fermentation. This suggests that fermentation may be a constitutive aspect of metabolism in these mats, rather than being regulated in response to energetic or other demands of metabolism. The glycogen data and relative ratios of organic acid flux indicate that fermentation of glycogen to acetate was the dominant nighttime pathway, with a smaller component of mixed acid fermentation. Photosynthetically-fixed carbon was indicated to be the feedstock for fermentation. In GN-I mats, enhanced levels of bicarbonate incorporation (Figure 3 ) were reflected in significantly higher net efflux of fermentation products than in GN-S mats (Figure 2 ). Indeed, when DCMU was used to limit oxygenic photosynthesis, net fermentation dropped more than in GN-S mats, indicating oxygenic photosynthesis in Cyanobacteria was the most likely cause behind net fermentation productivity of these mats. At the same time, NanoSIMS measurements of bicarbonate incorporation (Figure 4 ) show increased accumulation of label in GN-I Microcoleus filaments over GN-S filaments. These findings point to the accumulation of glycogen being more common in GN-S mats, but increased fixation and increased catabolic metabolism being more common in GN-I mats and fits the observation that GN-I mats are adapted to exist in a dynamic turbulent intertidal zone and GN-S mats are adapted to quiescence (Bebout et al., 1994 ). Pyrotag assays (Figure 1 ) corroborated previous reports of the abundance (approximately one fourth of DNA sequences from pyrotag libraries) of Chloroflexi in GN-S mats (Ley et al., 2006 ) as well as a similar level of Chloroflexi pyrotags seen in GN-I mats. Moreover, previous studies have shown that anoxygenic phototrophy was found to account for 10–40% of carbon fixation in GN-S mats (Finke et al., 2013 ), assumed to be by phototrophic sulfide oxidation. It has also been demonstrated, in the case of hot springs microbial mats that filamentous photoautotrophic Chloroflexi can have a role in fermentation, including hydrogenase transcript expression at night (Klatt et al., 2013 ), or a role indirectly driving fermentation (Otaki et al., 2012 ). Yet, in this study anoxygenic phototrophy appears to play only a minor role in nighttime fermentation in Guerrero Negro mats. Expression ratio (cDNA:DNA pyrotags) data demonstrated that Cyanobacteria (specifically genus Microcoleus in GN-S mat and both Microcoleus and genus Lyngbya in GN-I mat) maintain a consistent level of ribosomal expression between day and night and at a level much higher than any other phylogenetic group detected. The only hydrogenases attributable to phototrophs that were expressed at night (Figure 5 ) were associated with Cyanobacteria ; no type 3b [NiFe]-hydrogenases from any anoxygenic phototrophic Chloroflexi group were recovered. However, given that marine filamentous anoxygenic phototrophs are diverse and mostly uncharacterized in Guerrero Negro mats (Nübel et al., 2001 ; Ley et al., 2006 ) the phylogeny of novel Chloroflexi hydrogenases present in these systems is an avenue of future study. But overall, given the dominance of several types of Cyanobacteria in pyrotags and hydrogenase transcripts, Cyanobacteria were likely the metabolically dominant phototrophic fermenters in mats. Interestingly, the HoxH tree (Figure 5 ) does suggest that a Cyanobacteria other than Microcoleus chthonoplastes PCC 7420 was the main hydrogen producer in GN-S mats and that different species of Cyanobacteria may differ in their capacity for hydrogen production. Under natural conditions, both the GN mats and Elkhorn Slough mats (Burow et al., 2012 ) were characterized by net fluxes of hydrogen and organic acids out of the mats at night due to fermentation activity. However, in mats incubated with DCMU for the previous photoperiod, net hydrogen fluxes were reduced relative to the unamended treatments, whereas the flux of organic acids out of the mats was not significantly different (Figures 2C , D ), This differential effect of DCMU on hydrogen vs. organic acid flux was not previously observed in Elkhorn Slough samples (Burow et al., 2012 ). Overall, though the DCMU in mats has been shown to inhibit photosystem II in Cyanobacteria and the establishment of anoxic conditions, the response of whole mat communities to this photosystem shutdown in daylight was variable across different mat types and is still poorly understood, especially with respect to the daytime sulfide cycling. Mechanisms in both Cyanobacteria and in Cyanobacteria -associated members, such as phototrophic sulfide oxidation, may be acting in concert to alter both the cycling of hydrogen as well as the original production of hydrogen. Hydrogen and organic acid availability leads to uptake by Cyanobacteria -associated microbes in hypersaline microbial mats Release of cyanobacterial fermentation products within the closely packed matrix of the mat offers a flux of potential substrate to a range of terminal metabolizers. We hypothesized that SRB were the primary consumers of hydrogen and organic acids under dark, anoxic conditions, due to the abundance of sulfate. Like in hot spring microbial mats (Otaki et al., 2012 ) they were suspected to require close physical proximity for hydrogen uptake. In GN-I mats, inhibition with molybdate significantly increased accumulation of hydrogen at quantitatively similar levels to physical disruption (Figure 8B ). This was consistent with previous findings in the hypersaline microbial mats of Elkhorn Slough (Burow et al., 2012 ), and suggests that a physical association between Cyanobacteria and SRB underlies most of the observable consumption of fermentation products within these mats (Burow et al., in press ). Molybdate also enhanced accumulation of hydrogen in the GN-S mats, but physical disruption in those mats did not result in significantly greater net hydrogen flux relative to controls (Figure 8D ). Thus, while SRB appear to be the dominant sink for fermentation products under dark conditions in GN-S mats, physical associations appear to be less important than in the GN-I and Elkhorn Slough mats, though there is evidence that physical proximity could still be necessary (Fike et al., 2008 ). However, the failure of physical disruption techniques to separate mat members apart in GN-S mats cannot be discounted, nor can the possibility of unique motile SRBs in GN-S mats be discounted. We show here that disruption of the GN-I microbial mat, with a presumed separation of diverse members of the mat community from Cyanobacteria , led to a proportionally greater increase in the flux of organic acids, relative to hydrogen. This was consistent with the idea that organic acid consumption was also dependent (and possibly even more dependent than hydrogen) on tight physical association between producing and consuming organisms. Preliminary efforts were made to identify organisms that may be consuming acetate, via NanoSIMS analysis of samples incubated with 13 C-labeled acetate under dark/anoxic conditions. NanoSIMS analysis (Figure 9 ) confirmed that filamentous members of Chloroflexi and Desulfobacteraceae were significant consumers of acetate at night and may be important members of these close spatial associations, though no known filamentous SRB could be identified in DsrA phylogenetic analysis. Pyrotags of Proteobacteria show that purple non-sulfur bacteria were also quantitatively important, particularly in the GN-I mats (Figure 1B ). The present study did not specifically investigate acetate uptake by these organisms. However, the diverse and robust nature of the metabolism within this group (with various members being able to conduct autotrophic, heterotrophic, photosynthetic, chemotrophic, aerobic, and anaerobic metabolisms) suggests that they should also be examined for significant nighttime consumption of fermentation products in GN-I mats."
} | 4,910 |
38966358 | PMC11220614 | pmc | 1,463 | {
"abstract": "Biohybrid systems for solar fuel production integrate artificial light-harvesting materials with biological catalysts such as microbes. In this perspective, we discuss the rational design of the abiotic–biotic interface in biohybrid systems by reviewing microbes and synthetic light-harvesting materials, as well as presenting various approaches to coupling these two components together. To maximise performance and scalability of such semi-artificial systems, we emphasise that the interfacial design requires consideration of two important aspects: attachment and electron transfer. It is our perspective that rational design of this photosensitiser–microbe interface is required for scalable solar fuel production. The design and assembly of a biohybrid with a well-defined electron transfer pathway allows mechanistic characterisation and optimisation for maximum efficiency. Introduction of additional catalysts to the system can close the redox cycle, omitting the need for sacrificial electron donors. Studies that electronically couple light-harvesters to well-defined biological entities, such as emerging photosensitiser–enzyme hybrids, provide valuable knowledge for the strategic design of whole-cell biohybrids. Exploring the interactions between light-harvesters and redox proteins can guide coupling strategies when translated into larger, more complex microbial systems.",
"introduction": "Introduction Generating fuels from renewable energy, such as sunlight, is of paramount importance in our urgent transition away from fossil fuels. In its most basic form, three components are required for photoredox catalysis: a catalyst for chemical reduction, a catalyst for chemical oxidation, and a light absorber to harvest solar energy. These components can be synthetic, natural, or a combination of both. Semi-artificial photosynthesis has been previously defined as interfacing biological catalysts, which provide unparalleled catalytic specificity for the production of desirable chemicals, with synthetic light-harvesting materials to create biohybrid systems. 1 Here, we focus specifically on the integration of microbial catalysts directly with synthetic nanoparticle light-harvesters. We refer the reader to previous reviews on other systems of semi-artificial photocatalysis, including photoelectrochemical cells (PECs), which are not discussed here. 2,3 While the use of microbes as biological catalysts in solution with light-harvesting nanoparticles has been investigated for decades, 4 a landmark study in 2016 initiated a renaissance of research into these biotechnologies by demonstrating that nanoparticles could be synthesised and then localised at the surface of the microbe. 5 A long-lived physical interface between microbe and photosensitiser eliminates diffusion as a limiting factor. Synthetic materials can have advantages over natural photosystems in terms of resistance to photodamage; some subunits of natural photosystems require repair as often as every 30 minutes. 6 Further, microbial approaches exploit the capability of live cells to self-replicate and repair, reducing the input of materials such as purified proteins that require costly and time-consuming production. The ability to genetically program cells also provides opportunities to engineer them as factories for a specific chemical product. 7,8 While light-harvesters and microbes have been previously reviewed in depth, and whole-cell biohybrids discussed and analysed, 2,9–16 the challenge of creating a defined electron transfer interface between photosensitiser and microbe has only been highlighted. 17 The nature of this abiotic–biotic interface in existing biohybrids is often unclear and difficult to characterise retrospectively. 18 Thus, the question we aim to address here is: how can we establish controllable and characterisable pathways for electron transfer, from photosensitiser to cytoplasmic enzymes, that can be built-in by design? What research should we turn to for creating these biohybrid blueprints, and where are remaining gaps of knowledge that prevent progression? In this perspective, we propose that the rational design of these systems should exploit microbes with natural mechanisms of electron uptake, and involve engineered attachment of the photosensitiser to the biological catalyst. We support this view with a detailed discussion of research that we hope can benefit the ongoing optimisation of whole-cell biohybrids in the future."
} | 1,115 |
26793184 | PMC4707226 | pmc | 1,464 | {
"abstract": "Previous studies of microbial communities in deep-sea hydrothermal ferric deposits have demonstrated that members of Zetaproteobacteria play significant ecological roles in biogeochemical iron-cycling. However, the ecophysiological characteristics and interaction between other microbial members in the habitat still remain largely unknown. In this study, we investigated microbial communities in a core sample obtained from shallow hydrothermal iron-oxyhydroxide deposits at Nagahama Bay of Satsuma Iwo-Jima, Japan. Scanning electron microscopic observation showed numerous helical stalk structures, suggesting the occurrence of iron-oxidizing bacteria. Analysis of 16S rRNA gene sequences indicated the co-occurrence of iron-oxidizing Zetaproteobacteria and iron-reducing bacteria such as the genera Deferrisoma and Desulfobulbus with strong correlations on the sequence abundance. CARD-FISH indicated that the numbers of Zetaproteobacteria were not always consistent to the frequency of stalk structures. In the stalk-abundant layers with relatively small numbers of Zetaproteobacteria cells, accumulation of polyphosphate was observed inside Zetaproteobacteria cells, whereas no polyphosphate grains were observed in the topmost layers with fewer stalks and abundant Zetaproteobacteria cells. These results suggest that Zetaproteobacteria store intracellular polyphosphates during active iron oxidation that contributes to the mineralogical growth and biogeochemical iron cycling.",
"introduction": "Introduction Members of the class Zetaproteobacteria were first found at the Loihi Seamount, Hawaii (Moyer et al., 1995 ). Mariprofundus ferrooxydans is the only isolate characterized as a marine iron-oxidizing bacterium within the Zetaproteobacteria (Emerson et al., 2007 ). Previous studies of 16S rRNA genes in deep-sea hydrothermal fields showed that members of the Zetaproteobacteria are widely distributed in iron-oxyhydroxide deposits on the plate spreading centers, hot-spot seamounts, and island arcs (Davis et al., 2009 ; Kato et al., 2009 ; Forget et al., 2010 ; Edwards et al., 2011 ; McAllister et al., 2011 ; Fleming et al., 2013 ). These observations strongly suggest that microbial communities involving Zetaproteobacteria play significant ecological roles in biogeochemical iron and other elemental cycles. In the deep-sea hydrothermal iron-oxyhydroxide deposits, it has been demonstrated that the dissolved oxygen is present but generally lower than that of the surface seawater, e.g., less than 50 μM of oxygen was observed in iron-rich mats around the Loihi Seamount (Glazer and Rouxel, 2009 ), suggesting that members of the Zetaproteobacteria preferentially inhabit and grow by oxidizing ferrous iron to ferric iron at the sub-oxic redox condition. Consistently, a kinetic model study using a pure culture supported the notion that the habitable zone of iron-oxidizing microorganisms is severely and sensitively constrained by in situ oxygen concentration, and the maximum value for the geochemical niche approaches ~50 μM (Druschel et al., 2008 ). In addition, a genomic study of Mariprofundus ferroxydans PV-1 (Singer et al., 2011 ), which was isolated from hydrothermal venting at Loihi Seamount, revealed that it has the complete TCA cycle, the ability to fix CO 2 , and genes encoding aerotaxis as well as antioxidant functionalities. Although strain PV-1 does not always represent metabolic pathways and functions of all iron-oxidizing Zetaproteobacteria , the genomic information suggest that this strain is capable of sensing and responding to the redox state of the iron-oxyhydroxide deposits. Comparative genomic analyses of single-amplified genomes also indicated the niche specialization of Zetaproteobacteria , most likely controlled by the oxygen tolerance (Field et al., 2015 ). However, ecophysiology and growth/survival strategy of Zetaproteobacteria correlated with other members in the iron-oxidizing microbial ecosystem are still largely unknown. Satsuma Iwo-Jima is a small volcanic island located at ~40 km south of Kyushu Island, Japan. The volcanic activity provides a shallow hydrothermal field in the Nagahama Bay, where the formation of iron-oxyhydroxide deposits, including chimney-like structures was widely observed on the seafloor (Figure 1 ). Pilot geological and sedimentological studies of this environment showed that the depositional rates are exceptionally high, ranging from 2.8 to 4.9 cm per year (Kiyokawa and Ueshiba, 2015 ). Light microscopic observation of these deposits showed twisted stalk structures, suggesting the occurrence of iron-oxidizing microbial communities that mediate the formation process of iron-oxyhydroxide deposits. Figure 1 Regional (A) and local (B) maps of Satsuma Iwo-Jima . (C) An overview photo of the Nagahama Bay. The seawater is brownish-red due to the presence of iron oxyhydroxides. The yellow star indicates the sampling point in this study. In this study, we investigated microbial communities in the shallow hydrothermal iron-oxyhydroxide deposits (water depth: ~3 m) in the Nagahama Bay. To understand the distribution and ecophysiological characteristics of Zetaproteobacteria cells in this iron-rich habitat, we obtained a 50 cm-long core sample and studied microbial communities using scanning electron microscopy (SEM), image-based cell count, and catalyzed reporter deposition-fluorescence in situ hybridization (CARD-FISH) techniques as well as diversity and correlation analyses of 16S rRNA gene-tagged sequences.",
"discussion": "Discussion In the 50 cm-long core sample obtained from shallow hydrothermal iron deposits in the Nagahama Bay, members of the phylum Chloroflexi are the most predominant microbes throughout the core column. All the Chloroflexi sequences were most likely non-phototrophic, mesophilic to moderatory thermophilic anaerobes and no sequences related to phototrophic members are present, e.g., members of the class Anaerolineae are known to be strictly anaerobic heterotrophs that have been often detected from various organic-rich sedimentary environments, e.g., terrestrial soils, hydrothermal sediments, and marine subsurface sediments (Inagaki et al., 2006 ; Yamada et al., 2006 ). In the iron deposits that we examined, these Chloroflexi sequences present 8–20% of the total reads in all samples, indicating that, like other sedimentary microbial ecosystems, these members may play ecological roles in biogeochemical carbon cycling. It may also be possible that Anaerolineae ferment organic matter and produce fermented secondary products for iron-reducing bacteria (also iron-oxidizing members, if these are mixotrophs). In addition, the phylum Chloroflexi has been reported to be dominant in iron reducing enrichment culture (Hori et al., 2015 ) and Ardenticatena maritima that can reduce ferric iron was isolated from iron-rich coastal hydrothermal field (Kawaichi et al., 2013 ). Therefore, we cannot exclude the possibility that some of the Chloroflexi detected here might also associate with ferric iron reduction. In fact, we observed potential sulfate reducers such as members of the Deltaproteobacteria throughout the cored iron deposit we examined. If some of those members have multiple metabolic functions that can utilize either iron or sulfate as an electron acceptor, the presence of high concentration of iron-oxyhydroxides might be advantageous for iron reduction (Chapelle and Lovley, 1992 ) and hence they might contribute to iron cycling with Zetaproteobacteria . It is not surprising in this iron-rich shallow sedimentary microbial ecosystem that many 16S rRNA sequences in all the examined depths are closely related to known sulfate-reducing bacteria (i.e., Deltaproteobacteria ). Because the prevalence of sulfate-reducing conditions down to at least 40 cm below the seafloor and hence, even through the iron deposits are soft and highly permeable for discharging hydrothermal fluids and/or the overlying seawater, the redox state within the iron deposits are heterologous but generally sub-oxic to anaerobic conditions, providing habitable zones for both Chloroflexi and sulfate reducers with iron oxidizers. Beta-diversity analysis of microbial communities indicated the zonation pattern of three ecological niches in the iron deposits: the top (10 cm), middle (17–33cm), and bottom layers (~40 cm), suggesting the occurrence of redox-sensitive microbial habitats associated with iron and other elemental cycles. In fact, microscopic observations revealed that twisted stalk structures are most abundant in the middle layer. In addition, based on the visual observation, upward hydrothermal fluid flows seem to form complicated webs of the fluid passage in the middle zone (Kuratomi et al., 2015 ), which may create the patchily heterologous redox states that biogeochemical iron and other elemental cycles preferentially occur. The twisted stalk structure is a typical morphological feature formed by iron-oxidizing Zetaproteobacteria under relatively low oxygen concentrations ranged from 3 to 23 μM (Krepski et al., 2013 ). Because we did not find any other known iron-oxidizers such as Marinobacter and Hyphomonas in the shallow hydrothermal iron deposits that we examined, the detected members of Zetaproteobacteria should play a primary role in the formation process of thick iron-oxyhydroxide deposits in the Nagahama Bay. Our previous observations indicated that the vertical formation rate of the iron deposit mound is ~12 mm y −1 (Kuratomi et al., 2015 ), which is similar to the formation rate of ~19 mm y −1 estimated by using a pure culture of M. ferroxydans (Chan et al., 2011 ). It is also conceivable that abiotic autocatalysis on the iron-oxyhydroxide stalks may accelerate the iron deposition (Rentz et al., 2007 ). On the other hand, we did not observed any stalk- or sheath-like structures in the topmost and bottom layers, implying that abiotic iron oxidation is dominant or other biotic oxidation process occurs without formation of the unique structures. The community correlation network analysis based on Spearman's rank of SSU rRNA sequences at genus-level classification also exhibited the co-occurrence of iron-oxidizing Zetaproteobacteria and other anaerobic respirers in the iron-oxyhydroxide deposits. For example, Deferrisoma and Desulfobulbus showed a strong correlation to Zetaproteobacteria (Figure 8 ). Deferrisoma is mesophilic heterotrophs that can grow optimally at 50°C and use ferric iron or elemental sulfur as electron acceptors (Slobodkina et al., 2012 ), and Desulfobulbus is characterized as sulfate and iron reducers (Holmes et al., 2004 ). It has been reported that some sulfate-reducing bacteria are tolerant to relatively high concentrations of oxygen and have aerobic respiration systems (Muyzer and Stams, 2008 ). Although it still remains unknown if these members primarily utilize ferric iron for the anaerobic energy respiration instead of sulfur compounds or even both flexible to the in situ redox state, the strong correlations between iron oxidizers and reducers indicate the occurrence of iron cycling in this environment. Another interesting finding in this study would be a discrepancy between the frequency of twisted stalk structures and the number of Zetaproteobacteria cells. Both deep sequencing analysis of 16S rRNA and CARD-FISH consistently showed that Zetaproteobacteria were most abundant in the top layer (i.e., CH1_7), although fewer stalk structures were observed (Figure 4 ). Although the taxonomic compositions of Zetaproteobacteria are varied in sediment depths, OTU1 was found to dominate all sedimentary horizons where we observed stalk-like structures (see Supplementary Figure 1 ). We might not be able to eliminate a possibility that minor Zetaproteobacteria or other iron-oxidizers may contribute to iron oxidation and mineral growth at top and bottom layers, however, we infer that OTU1-relatives play a significant ecological role in this habitat. In marked contrast, the occurrence of polyphosphate-accumulated cells was observed only in the middle layer, whereas no polyphosphate accumulations were observed in the top and bottom layers. It has been reported that a diverse array of microbes are capable of polyphosphate accumulation in natural and artificial environments (Hupfer et al., 2007 ), suggesting that intracellular polyphosphate grains act as a reservoir of energy and phosphate. In fact, M. ferroxydans PV-1 has genes encoding polyphosphate conversion and stores it via ATP under the aerobic conditions of excess carbon and energy substrates. Although the pure culture of M. ferroxydans PV-1 cannot grow using any organic carbon, some single amplified genomes obtained from the Loihi Seamount implicate the metabolic potential for heterotrophy or mixotrophy of Zetaproteobacteria (Field et al., 2015 ). Taken together, we interpret that Zeteproteobacteria in the middle layer oxidize ferrous iron, produce stalks, and store polyphosphate grains inside the cell. In the top and bottom layers, they may be different metabolic and/or physiological state(s), i.e., gaining energy from iron-oxidation and/or heterotrophy without forming stalk-like structures. This might be an adaption mechanism of iron-oxidizing Zetaproteobacteria ; however, mechanisms for their redox-sensing and energetic response to the environmental change still remain largely elusive. In conclusion, we show that Zetaproteobacteria play significant ecological roles in iron cycling with other iron reducers and construct a unique microbial ecosystem in shallow hydrothermal iron-oxyhydroxide deposits. In addition, we first demonstrated that Zetaproteobacteira in the natural ecosystem store polyphosphate grains inside cells for energy storage and a phophate source with iron-oxidation and stalk formation. We also observed that Zetaproteobacteria without pholyphosphate grains with fewer stalk formations have largest population, indicating a possibility that Zetaproteobacteria changed their metabolism depending on the environments. These results provide some new insights into ecophysiology of iron-oxidizing Zetaproteobacteria and the microbial ecosystem in marine hydrothermal environments."
} | 3,586 |
35521414 | PMC9064385 | pmc | 1,466 | {
"abstract": "Biomimetic polymeric materials, adopting the basic molecular design principles of biological materials, have been extensively studied in recent years but it is still challenging to combine assorted mechanical characteristics in a single material. Here, we present a simple and effective strategy to prepare mechanically robust yet resilient biomimetic polymer networks by utilizing dual noncovalent and covalent cross-linkings. Tailoring the dual cross-links consisting of thiourea noncovalent interactions and epoxy–amine covalent linkages in the biomimetic polymer networks enables a rare combination of excellent elastic modulus (1.1 GPa), yield stress (39 MPa), extensibility (320%), as well as complete strain and performance recovery after deformation at room temperature. The biomimetic polymer networks also exhibit highly adaptive mechanical properties in response to multiple-stimuli including strain rate, temperature, light, and solvent."
} | 237 |
39503013 | PMC11534796 | pmc | 1,467 | {
"abstract": "Introduction Nonlinear and non-stationary processes are prevalent in various natural and physical phenomena, where system dynamics can change qualitatively due to bifurcation phenomena. Machine learning methods have advanced our ability to learn and predict such systems from observed time series data. However, predicting the behavior of systems with temporal parameter variations without knowledge of true parameter values remains a significant challenge. Methods This study uses reservoir computing framework to address this problem by unsupervised extraction of slowly varying system parameters from time series data. We propose a model architecture consisting of a slow reservoir with long timescale internal dynamics and a fast reservoir with short timescale dynamics. The slow reservoir extracts the temporal variation of system parameters, which are then used to predict unknown bifurcations in the fast dynamics. Results Through experiments on chaotic dynamical systems, our proposed model successfully extracted slowly varying system parameters and predicted bifurcations that were not included in the training data. The model demonstrated robust predictive performance, showing that the reservoir computing framework can handle nonlinear, non-stationary systems without prior knowledge of the system's true parameters. Discussion Our approach shows potential for applications in fields such as neuroscience, material science, and weather prediction, where slow dynamics influencing qualitative changes are often unobservable.",
"introduction": "1 Introduction Nonlinear, non-stationary processes are abundant in various natural and physical phenomena. For instance, the dynamics of neurons are known to be strongly dependent on the state of the brain, determined by varying levels of attention, arousal, anesthesia, and sleep depth, as well as on different behavioral patterns such as movement (Steriade et al., 1993 ; Buzski, 2002 ; Tokuda et al., 2019 ; Vohryzek et al., 2020 ). Similarly, the response of physical systems can qualitatively change due to bifurcation phenomena as sample properties or experimental conditions vary (Bnard, 1901 ; Ertl, 1991 ; Itoh and Kimoto, 1996 ; Raab et al., 2023 ). Various mathematical frameworks have been proposed to model non-stationary dynamics (Waddington, 1961 ; Kaneko and Tsuda, 2003 ; Rabinovich et al., 2001 ; Katori et al., 2011 ; Patel et al., 2021 ). One plausible and simple depiction is that system parameters vary over time or in different contexts (Patel et al., 2021 ). Consider either a discrete nonlinear dynamical system: \n (1) \n x ( n + 1 ) = f ( x ( n ) ; λ ) , \n or a continuous dynamical system: \n (2) \n d x d t = f ( x ; λ ) , \n where x ∈ℝ n represents the dynamical variable expressing fast dynamics, and λ is a parameter of function f whose value can potentially lead to bifurcation in the dynamics of x . In the context of modeling static nonlinear systems with a fixed value of λ, recent advancements in machine learning have enabled the rules governing the underlying system to be extracted and learned from observed time series data with much higher accuracy than before. In particular, by learning from time series data, reservoir computing has facilitated the creation of autonomous dynamical systems within the model that can generate time series resembling those of the target system, achieving high accuracy even in challenging problems such as learning chaotic systems. Furthermore, recent studies have demonstrated the prediction of unobserved bifurcations that are not present in the learning data (Kong et al., 2021 ; Patel et al., 2021 ; Kim et al., 2021 ; Itoh, 2023 ). In their settings, they have succeeded in predicting unknown bifurcations that occur when the parameter λ takes values other than those used when generating the observed data. For example, Patel et al. addressed the bifurcation parameter λ of a chaotic dynamical system not as static value but as a variable changing very slowly over time, and learned the time series generated by this system. After learning the one-step-ahead prediction task, they added a feedback loop to the reservoir, creating a closed-loop model that can generate time series as an autonomous dynamical system. They showed that, although learning the time series of x using the conventional reservoir computing framework alone does not predict unobserved bifurcations, successful learning can be achieved by separately providing the reservoir with the true value of the parameter at each moment as an additional input. When the parameter values inputted during the prediction phase were different from those during learning, the model was able to predict bifurcations not included in the training data. Kim et al. demonstrated that the emergence of a Lorenz attractor not present in the training data could be predicted by first inputting time series generated from the Lorenz equations along with the true bifurcation parameter values into the reservoir, then forming a closed-loop model to create an autonomous dynamical system, and finally changing the input parameter values. These studies indicate that predicting unknown bifurcation phenomena is possible by additionally inputting the value of the bifurcation parameter into the reservoir. This suggests that the reservoir computing framework is capable of learning not just specific dynamical systems but families of dynamical systems, suggesting the potential to predict the emergence of system states qualitatively different from those observed in real data. However, these prior studies assume that the true value of the parameter is known, which is not the case in most real-world scenarios, including in brain data observation. Therefore, the question arises whether the behavior of non-stationary systems with temporal parameter variations can be predicted solely from observed time series data. Various methods, including recurrence plots (Marwan and Kraemer, 2023 ), supervised learning (Zhai et al., 2024 ), slow feature analysis (Wiskott and Sejnowski, 2002 ; Antonelo and Schrauwen, 2012 ), and hierarchical structures (Yonemura and Katori, 2021 ; Katori, 2019 ; Gallicchio et al., 2017 ; Tamura et al., 2019 ), have been reported for extracting the slowly moving components of system dynamics. In this study, we leverage the reservoir computing framework to address this problem. Our central idea is based on the following consideration: in a typical scenario, a reservoir receives a signal derived from a nonlinear dynamical system, such as one variable of the state vector x — e.g., x 1 —, in one step and predicts its value in the next time step. Previous studies have indicated that establishing generalized synchronization between the reservoir and the original system generating the input signal is crucial for achieving accurate predictions (Rulkov et al., 1995 ; Carroll, 2020 ; Lu et al., 2018 ; Lu and Bassett, 2020 ), where generalized synchronization refers to the condition that the listening reservoir's state, u ( t ), is a continuous function, Ψ( x ), of the state of the original system, x . Especially, if the function Ψ( x ) is invertible, the reservoir's state u ( t ) has all the information about x . It is reasonable to predict the value of another element — e.g., x 2 —, from partial observation of the system — e.g., only x 1 —, if generalized synchronization is established between the original system and the reservoir (Lu et al., 2017 ). Now, considering the parameter λ varies slowly over time as expressed in Equation 2 , the following system can be formulated: \n (3) \n { d x d t = f x ( x ; λ ) d λ d t = f λ ( x , λ ) \n Let X be a concatenation of x and λ, defined as X = t ( t x , λ), then this system can be represented as a single ordinary differential equation (ODE): \n (4) \n d X d t = F ( X ) . \n We assume that the signal is generated from the trajectory of this concatenated system's attractor. When the signal originating from x is input into the reservoir and invertible generalized synchronization between the reservoir state u and X is achieved, the reservoir's state has full information about λ. While above discussion is speculative, previous studies have shown that by adjusting the reservoir's timescale and structure, the reservoir can successfully extract the slow dynamics of the signal source system (Manneschi et al., 2021 ; Jaeger, 2008 ; Gallicchio et al., 2017 ; Tanaka et al., 2022 ; Yonemura and Katori, 2021 ). The extraction of such slow or static system states within the reservoir computing framework, where internal couplings are not altered during learning, suggests that unsupervised extraction of such information is possible using reservoirs. We first aim to verify whether it is possible to extract the true variation of parameter λ's by simply extracting the slowly varying variables within the reservoir (Section 3.1 Experiment 1). Patel et al. ( 2021 ) have demonstrated that predicting the time series of the concatenated system X cannot be achieved by a simple single reservoir. The challenges addressed in this paper are twofold: (1) estimating the unobservable slowly varying parameter values (Section 3.1 Experiment 1), and (2) predicting unknown bifurcations in the fast dynamics under the variation of such parameters (section 3.2 Experiment 2). While the second challenge has been tackled by Patel et al. and Kim et al. in scenarios where the true parameter value is known, in this study, we explore the possibility of learning from observational data generated by nonlinear systems and predicting unknown bifurcations without the knowledge of true parameter values. We allow the bifurcation parameter values to change over time but assume these changes occur on a significantly longer timescale compared to the system's fast dynamics. Previous studies suggest that extracting the slowly changing parameter values from time series observations in an unsupervised manner may allow us to substitute the true parameter value with an estimated one. The architecture of the model proposed in this study comprises two types of reservoirs stacked in layers: a slow reservoir with long timescale internal dynamics and a fast reservoir with short timescale dynamics. Assuming a nonlinear system with a very slowly changing bifurcation parameter value as the signal source, we input observational data obtained from the fast dynamics into these reservoirs. We found that when the variables that change slowly are extracted from the internal state of the slow reservoir, they trace the temporal variation of the system's parameter. Although the variables extracted from the slow reservoir differ in amplitude scale from the true parameter values, we show that adding these variables and the observational time series to the fast reservoir allows for the prediction of unknown bifurcations, resembling the true parameter values provided in prior studies.",
"discussion": "4 Discussion We have demonstrated that unsupervised extraction of the very slowly changing parameters of the dynamical system generating the signals is possible by simply feeding the observation to a reservoir with a long-time scale and selecting the internal nodes of the reservoir with slowly varying states. Furthermore, we have shown this reservoir's capability to predict bifurcations not present in the training data, such as the death of chaotic oscillations, by inputting the extracted slow features and observation signal into another reservoir. Kim et al. ( 2021 ) and Patel et al. ( 2021 ) demonstrated the prediction of unobserved bifurcations not present in the training data using a reservoir computing framework. Their work illustrated the remarkable capability of reservoir computing to learn the parameter dependencies within dynamical system flows and to reproduce unknown bifurcations. However, they treated the parameters as known, which is not the case in real-world applications, where the values of these parameters often cannot be observed. In this study, we introduce two reservoirs: a slow feature predictor that forecasts the movement of these slow features, and a fast reservoir that predicts the values of the observed time series. By inputting the slow features resulting from the unsupervised extraction, we establish a closed-loop model that operates as a fully autonomous dynamical system during the predicting phase. This demonstrates the ability to forecast the emergence of unknown bifurcations without any direct observation of the parameter value. Nonlinear, non-stationary processes are abundant in various natural and physical phenomena. Additionally, numerous scenarios probably exist where slow dynamics inducing qualitative dynamics changes remain unobservable. The potential applications of this approach are vast, spanning fields such as neuroscience (including electrophysiological measurements, electroencephalography, functional Magnetic Resonance Imaging, and disease progression with tipping points), material science (including surface science), and weather prediction and control. The main limitation of the current work is the lack of understanding of the mechanism underlying the phenomenon wherein the behavior of slowly moving nodes, selected heuristically from within the slow reservoir, is similar to variations in the original system's parameters. As mentioned in the Introduction, the observations made could be explained if the reservoir can achieve generalized synchronization with the target system (Carroll, 2020 ; Rulkov et al., 1995 ), including the slow parameter dynamics. For the results shown in Figure 3 , we conducted the same numerical simulation using a reservoir with linear dynamics by replacing the activation function in Equation 8 with the identity map. The results show that a linear reservoir with a slow time constant does not yield parameter estimation, even with supervised training, where the internal state of the reservoir is fitted to the true parameter value ( Supplementary Figure S2 ). This suggests that the nonlinearity of the reservoir is crucial for the current results. However, generalized synchronization would not fully explain the current results. For example, fast-moving nodes also exist within the slow reservoir. It is not trivial that in the internal state of the reservoir, u s ∈ℝ N s , the directions of fast and slow fluctuations align along axes, u 1 , u 2 , ⋯ , u N s (i.e., different nodes). In fact, it would not be surprising if fast and slow fluctuations were superimposed at all single nodes. Therefore, further investigation is required to elucidate the logical reason why simply selecting slow-moving nodes worked well as a heuristic. Conversely, employing a more sophisticated method to separate the directions of fast and slow fluctuations might lead to better performance (Antonelo and Schrauwen, 2012 ). Previous works have extensively explored the behavior of complex systems around tipping points (Dakos et al., 2008 ; Veraart et al., 2011 ; Liu et al., 2013 ). For instance, the Dynamical Network Biomarker (DNB) method captures the increase in temporal fluctuations and the intensified correlation associated with critical slowing down Liu et al. ( 2013 ). Unlike the approach in the current study, which involves learning the flow of the dynamical system in a relatively low-dimensional, deterministic, and strongly nonlinear phase space, the DNB method utilizes the generic behavior near bifurcation points in very high-dimensional systems based on linearization around a fixed point. Given their distinct advantages, combining these methods in the future might improve the prediction and control of non-stationary, nonlinear systems. Kim et al. ( 2021 ) the emergence of chaotic attractors in the Lorenz system by extrapolating the parameter space and learning in regions without chaotic attractors, where only two stable fixed points exist. In our research, we have extracted the slowly changing parameters of the target system by receiving its generated time series through the reservoir. However, applying our method to predict the emergence of a chaotic strange attractor by learning observation from the Lorenz system with stable fixed points is currently challenging because our method relies on observing long time series to extract slow features, whereas the target system does not produce a long time series with oscillations if the trajectory converges to a fixed point. A new framework would be necessary to estimate the parameter changes in a system with stable fixed points, e.g., by introducing external perturbations to the target system and receiving its response through the reservoir."
} | 4,179 |
27136455 | PMC4852939 | pmc | 1,473 | {
"abstract": "Plant associations with root microbes represent some of the most important symbioses on earth. While often critically promoting plant fitness, nitrogen-fixing rhizobia and arbuscular mycorrhizal fungi (AMF) also demand significant carbohydrate allocation in exchange for key nutrients. Though plants may often compensate for carbon loss, constraints may arise under light limitation when plants cannot extensively increase photosynthesis. Under such conditions, costs for maintaining symbioses may outweigh benefits, turning mutualist microbes into parasites, resulting in reduced plant growth and reproduction. In natural systems plants commonly grow with different symbionts simultaneously which again may interact with each other. This might add complexity to the responses of such multipartite relationships. We experimented with lima bean ( Phaseolus lunatus ), which efficiently forms associations with both types of root symbionts. We applied full light and low-light to each of four treatments of microbial inoculation. After an incubation period of 14 weeks, we quantified vegetative aboveground and belowground biomass and number and viability of seeds to determine effects of combined inoculant and light treatment on plant fitness. Under light-limited conditions, vegetative and reproductive traits were inhibited in AMF and rhizobia inoculated lima bean plants relative to controls (un-colonized plants). Strikingly, reductions in seed production were most critical in combined treatments with rhizobia x AMF. Our findings suggest microbial root symbionts create additive costs resulting in decreased plant fitness under light-limited conditions.",
"introduction": "Introduction Rhizobia and arbuscular mycorrhizal fungi (AMF) represent two major groups of plant-associated microbial mutualists [ 1 – 3 ]. Legume-associated rhizobia are nitrogen-fixing bacteria that play a key role for local and global nitrogen cycles, dramatically influencing the productivity and species composition of natural and agricultural ecosystems [ 2 – 4 ]. AMF colonize roots of the host plant, form extensive networks, and participate in the acquisition of nutrients (namely phosphorus) and water [ 1 ]. The association between plants and AMF is likely the most ubiquitous of all mutualisms, having been observed in 400 million year-old fossils and persisting in more than 80% of extant land plants [ 5 ]. Like rhizobia, AMF are considered keystone species in terrestrial ecosystems [ 1 ] due to their critical impact on plant growth and species composition [ 6 ]. Root colonization with rhizobia and AMF generally have positive effects on plant growth [ 2 , 7 , 8 ], including increases in vegetative and reproductive traits [ 9 , 10 ]. In legumes, which are frequently colonized by both types of microbial symbionts simultaneously, plant growth is usually enhanced by the dual symbiosis [ 11 ]. However, while most studies have been carried out under optimal conditions, in several common bean ( Phaseolus vulgaris ) varieties, under drought stress both dual and single colonization with different rhizobia strains and AM fungi can result in negative effects for the plant host [ 12 ]. Under such water limited conditions Franzini and co-workers [ 12 ] showed that AMF inhibited rhizobial nodule development and N 2 fixation, and thus caused diminution of plant growth. While costs of simultaneous colonization by rhizobia and AMF for their legume hosts have been reported under water limited conditions, less information exists on the effects of another key plant resource: light. Both rhizospheric associations, rhizobia and AMF, incur significant costs as consumers of plant photosynthates as the combined demand of symbionts can reach up to 28% of carbon fixed by the plant [ 13 ]. Plants can compensate for this cost through sink stimulation of photosynthesis, which is considered to be an adaptation to take advantage of nutrient supply enhancement by the symbiont without compromising the total amount of photosynthates available for plant functioning [ 13 ]. While sink stimulation is generally an efficient strategy to compensate for costs of carbohydrate allocation, the question arises of how plants respond to microbial inoculation when photosynthesis cannot easily be increased [ 14 ], for example under light-limited conditions [ 15 , 16 ]. We hypothesize that under such conditions the costs for maintaining the symbioses may outweigh the benefits, ultimately turning the mutualist microbes into parasites, resulting in reduced growth and reproduction of colonized plants. In nature, light availability is often a variable resource due to competition among plant species and, depending on cultivation method, also shows strong variation in agricultural systems [ 17 ]. In a few previous studies it was shown that the effects of light limitation on plant growth did not differ between unfertilized plants growing with symbionts and plants growing without symbionts but with additional fertilization [ 18 , 19 ], thereby neglecting the role of nutrient supply by the symbionts as the ultimate benefit of these symbioses for the plant. In the present study we used lima bean (Phaseolus lunatus), a model plant in chemical ecology and an important food plant [ 20 – 24 ] to better understand the concerted effects of AMF/rhizobial colonization and light availability on vegetative and reproductive traits. Plants were inoculated with rhizobia (R) and AMF and blocks were exposed to two different levels of light availability. Treatments included: no symbiont (R-AMF-), AMF only (R-AMF+), rhizobia only (R+AMF-), and both symbionts (R+AMF+). While effects of light on plant resource allocation patterns (including carbon, nitrogen and phosphate) to either rhizobia [ 15 ] or AMF [ 25 – 30 ], have been studied using both empirical and modelling approaches, research on the effects of light upon interactions with multiple bacterial and fungal rhizospheric symbionts simultaneously is limited. However, as plants are frequently colonized by multiple microbial symbionts considering interactive effects among microbes, which may range from competition to cooperation of mutualists, is of high importance [ 31 , 32 ]. This study aims to answer the following specific questions: i) What are the separate and interacting effects of rhizobia and arbuscular mycorrhizal fungi on growth and reproduction of lima bean, and ii) what are the outcomes of these belowground mutualistic interactions regarding plant growth and reproduction when light availability is limited. To our knowledge, our study is among the first to analyze the interactive effects of AMF, rhizobia and light availability, addressing potential costs of these microbes when photosynthesis is limited.",
"discussion": "Discussion In our study we quantitatively analyzed effects of light availability on the tripartite symbiosis of mycorrhizal fungi (AMF), rhizobia, and lima bean. Our results showed enhancing effects on growth and reproduction by belowground symbionts under full light, but reduced plant development and reproduction in inoculated plants under shaded conditions. Moreover, nodulation of plants was reduced by AMF resulting in interacting effects of both types of symbionts on growth and reproduction. Interacting effects of rhizobia and AMF on plant growth are sometimes suggested to be rather uncommon [ 39 , 40 ], however, we could demonstrate an antagonistically interacting effect on aboveground biomass. This antagonistic effect was also obvious for the reproduction of lima bean but only under conditions of light limitation. This implies that light availability mediates the outcome of a dual infection with different root symbionts. Carbon requirements by root-associated microbial mutualists are thought to be easily compensated through sink stimulation (up-regulation) of photosynthesis [ 13 , 26 ]. Our study shows that even though sink stimulation may be an efficient strategy to compensate for photosynthate losses to microbial symbionts under full light, the situation can be fundamentally different when light is limited. Under such conditions, mutualistic microbes may act as parasites that exploit resources and reduce host fitness [ 15 , 41 , 42 ]. As spatial and temporal variation in light availability are ubiquitous in nature [ 17 ], as are plant associations with multiple rhizospheric microbes, our findings suggest that light variation represents a widely overlooked key factor determining the outcome of plant-microbe interactions. Inquiry into the mutualism/parasitism continuum offers insight into the delicate coevolution of plants and their belowground symbionts. Kiers and Denison [ 43 ] have shown that rhizobial mutualisms are maintained by host sanctioning of “cheating” genotypes while AMF mutualisms are maintained by reciprocal rewards between genotypes of hosts and AMF that allocate a greater amount of nutrient or photosynthate to the more generous partner. Though plants have relatively fine control over these mechanisms via selective partitioning within root systems when multiple microbial genotypes are present [ 43 ], an overall change in an abiotic condition such as light is likely to have a disruptive effect on the symbiosis. Whether either sanctioning or reciprocal rewards are at play, mutualistic cooperation also favors mutualist fitness, thus we predict a destabilization of mutualisms under unfavorable abiotic conditions that induce a symbiotic lifestyle switch. Our study shows that light availability is a key factor in determining the threshold between mutualist and parasite in such interactions. Legumes, rhizobia, and mycorrhizal fungi usually form a tripartite symbiosis, and legume performance may therefore be affected by potential interactions between both symbionts. In our study, nodulation was reduced for plants grown with AMF. Nutrient provisions by one microbial symbiont may reduce the provisional benefits of the other. Multiple studies have addressed reduced benefits of rhizospheric symbioses in the presence of non-limiting resources [ 19 , 44 , 45 ]. The reduced nodulation we observed in the presence of AMF might also be explained by competition for colonization sites, even if evidence for this is controversial [ 43 ]. AMF are sometimes described as the dominant symbiont due to the prioritization of phosphorous over N provision to both host and rhizobia during AMF colonization [ 46 ]. However, we found that this effect tended to be more obvious under full light, whereas under shaded conditions nodulation was generally weak regardless of the status of AMF infection, suggesting competition for photosynthates. This interaction was also reflected by an antagonistic effect of both symbionts on biomass and seed production under light limitation, which is in dramatic contrast to synergistic effects under full light reported in the present study and others [ 46 ]. Differences in abiotic conditions (e.g. light availability) may therefore be interpreted as an important mediator of such mutualistic relationships. Many questions remain regarding whether all legumes may exhibit similar decreases in fitness parameters due to AMF and rhizobial colonization under light-limited conditions. A better understanding of these interactions is of high relevance. As of 2003, grain and forage legumes represented 27% of all primary agricultural production [ 20 ]. In natural ecosystems, ranging from forests to grasslands, microbial associations with legumes have been shown to be responsible for the provision of the majority of available nitrogen [ 2 , 20 ]. Considering the ubiquity of the highlighted symbiotic associations and the economic importance of legumes, conditional symbiotic lifestyle switching (mutualist to parasite) has significant ramifications regarding the stability of mutualisms and the productivity of agro-ecological systems. It further remains to be tested whether the effects we report here hold true under natural conditions. While we used a natural rhizobia strain derived from a wild lima bean population, the AMF inoculum we used in the present study represents a commercial product. Thus, it likely contains AMF species and strains that form beneficial interactions with a broad range of different host plant species but may not interact with lima bean plants in nature. Also, using an inoculum composed of various fungi leaves the question which of the strains actually forms beneficial mycorrhiza and causes in the observed effects. Alternatively, several AMF species may co-colonize plant roots simultaneously and thus can introduce uncontrolled variation into the experimental system due to variable synergistic effects or competition among fungi but also between fungi and rhizobia. However, as we observed consistent effects of AMF and rhizobial inoculation, as well as for plants inoculated with both root symbionts, our statements on the effect of light limitation on this experimental tripartite interaction seem justified. Whether or not such effects determine the outcome of diverse tripartite plant-AMF-rhizobia interactions in natural ecosystems remains to be tested."
} | 3,283 |
26403924 | PMC4703631 | pmc | 1,474 | {
"abstract": "Many anaerobic conversions proceed close to thermodynamic equilibrium and the microbial groups involved need to share their low energy budget to survive at the thermodynamic boundary of life. This study aimed to investigate the kinetic and thermodynamic control mechanisms of the electron transfer during syntrophic butyrate conversion in non-defined methanogenic communities. Despite the rather low energy content of butyrate, results demonstrate unequal energy sharing between the butyrate-utilizing species (17 %), the hydrogenotrophic methanogens (9–10 %), and the acetoclastic methanogens (73–74 %). As a key finding, the energy disproportion resulted in different growth strategies of the syntrophic partners. Compared to the butyrate-utilizing partner, the hydrogenotrophic methanogens compensated their lower biomass yield per mole of electrons transferred with a 2-fold higher biomass-specific electron transfer rate. Apart from these thermodynamic control mechanisms, experiments revealed a ten times lower hydrogen inhibition constant on butyrate conversion than proposed by the Anaerobic Digestion Model No. 1, suggesting a much stronger inhibitory effect of hydrogen on anaerobic butyrate conversion. At hydrogen partial pressures exceeding 40 Pa and at bicarbonate limited conditions, a shift from methanogenesis to reduced product formation was observed which indicates an important role of the hydrogen partial pressure in redirecting electron fluxes towards reduced products such as butanol. The findings of this study demonstrate that a careful consideration of thermodynamics and kinetics is required to advance our current understanding of flux regulation in energy-limited syntrophic ecosystems. Electronic supplementary material The online version of this article (doi:10.1007/s00253-015-6971-9) contains supplementary material, which is available to authorized users.",
"introduction": "Introduction Anaerobic conversion of fatty acids, such as butyrate, involves a close interaction of different microbial groups. Butyrate-oxidizing bacteria convert 1 mol of butyrate to 2 mol of acetate and hydrogen. This reaction is energetically feasible only by product removal mediated by acetoclastic and hydrogenotrophic methanogens. Such mutually dependent microbial consortia are referred to as syntrophic communities (Kleerebezem and Stams 2000 ; Schink 1997 ; Stams 1994 ). Both, the hydrogen and acetate transferred between these syntrophic partners, serve as electron carriers with carbon dioxide and methane being the final products. The control of the electron transfer in methanogenic ecosystems is not yet fully understood. Only few studies have focused on flux regulation in syntrophic communities that are active at the thermodynamic boundary of life and share the little amount of energy available. Two methanogenic coculture studies investigated the bioenergetics of either butyrate or ethanol degradation; however, they have been performed in batch mode (Dwyer et al. 1988 ; Seitz et al. 1988 ). A lack of adaptation of the syntrophic partner organisms to batch reactor conditions may cause lag periods, reduced activity, or the uncoupling of syntrophic growth which leads to unreliable starting conditions. These bottlenecks can be overcome by continuous reactor operation. Seitz et al. ( 1990a , b ) give examples for continuous syntrophic coculture studies on ethanol. Analyzing the thermodynamic system state during syntrophic ethanol conversion, Seitz et al. ( 1990b ) found an unequal distribution of the total Gibbs energy change among the hydrogen-producing acetogen (23 %) and the hydrogenotrophic methanogen (77 %). Smith and McCarty ( 1989a , b ) performed ethanol perturbations of propionate and ethanol-fed enrichments to study the kinetic and thermodynamic control of reduced product formation such as propanol and long-chained fatty acids. However, energy sharing between the different microbial groups was not further investigated. This study aims to elucidate the kinetic and thermodynamic control mechanisms of the electron transfer during syntrophic butyrate conversion in non-defined methanogenic communities. For this purpose, butyrate and ethanol-fed continuously stirred tank reactors (CSTRs) were perturbed with increased ethanol concentrations either at bicarbonate-limiting or non-limiting conditions. The relation of the functional groups participating in syntrophic butyrate and ethanol conversion are shown in Fig. 1 Fig. 1 Syntrophic interactions during anaerobic conversion of butyrate and ethanol in non-defined methanogenic associations. ΔG \n 01 Gibbs energy change under standard conditions and pH 7.0 (kJ/mol donor), But butyrate, EtOH ethanol, Ac acetate",
"discussion": "Discussion In this study, the combination of continuous cultivation, liquid measurements, online off-gas measurements, and model description successfully contributed to the identification of thermodynamic and kinetic control parameters during anaerobic butyrate conversion in ethanol and butyrate-fed methanogenic enrichments. An overview of the kinetic parameters determined and a comparison to literature is given in Online Resource Table S1 . The growth yields estimated according to the Gibbs energy dissipation method were in the range of reported values. The Y XC4ox/But of S . cellicola strain 19J-3, which showed 98 % gene similarity with the butyrate-utilizing species found in this study, has not yet been reported. However, a total biomass yield on butyrate has been reported for its closest relative, Syntrophospora bryantii DSM 3014T (94.6 % gene similarity), in coculture with different hydrogenotrophic methanogens (0.041–0.115 mol- X /mol-But, assuming 55 % protein content per gram dry weight) (Dong et al. 1994 ; Wu et al. 2006 ). The estimated total biomass yield of the enriched butyrate-utilizer and hydrogenotrophic methanogens on butyrate (0.088 mol- X mol-But −1 ) falls into the reported range. To identify the microbial composition and to confirm the assumed catabolic reactions shown in Fig. 1 , DGGE analysis was performed on samples taken prior to the perturbation experiments C1 and C2. As expected, a similar microbial composition was found in both experiments. All reactions shown in Fig. 1 , except for acetoclastic methanogenesis, were identified based on the comparison of the gene similarities between the enriched species and the closest cultivated relative. Since a significant fraction of Methanosaeta -like species was observed using phase-contrast microscopy (Online Resource Fig. S2 ), acetoclastic methanogenesis was assumed as the acetate consuming reaction in the model. Syntrophic acetate-oxidizing bacteria can consume acetate in cooperation with hydrogenotrophic methanogens and are known to occur at conditions inhibitory to acetoclastic methanogens, e.g., high ammonium concentrations (>5.0 g L −1 NH 4 + -N) and high VFA levels (Schnürer et al. 1999 ; Westerholm et al. 2011 ). Such inhibitory conditions have not been observed in course of experiments C1 and C2. In addition, the retention time applied in this study (10–13 days) is rather short compared to the doubling times of acetate-oxidizing syntrophs, e.g., 20–25 days for Clostridium ultunense in coculture with a hydrogenotrophic methanogen under mesophilic conditions (Hattori 2008 ; Schnürer et al. 1997 ). These facts make a significant contribution of syntrophic acetate-oxidizers unlikely, although the reaction catalyzed by this microbial group, anaerobic acetate conversion, was exergonic throughout the experiments. The reverse pathway of anaerobic acetate oxidation, referred to as homoacetogenesis (reduction of CO 2 by H 2 ) was thermodynamically unfavorable throughout the two experiments. Therefore, syntrophic acetate conversion and homoacetogenesis have been neglected in the model description. Kinetic control of electron transfer This study showed a clear influence of the hydrogen partial pressure on the biomass-specific flux of ethanol and butyrate conversion. A significant decrease of q But was observed in perturbation experiment C2 even though anaerobic butyrate conversion remained exergonic. In the absence of significant product accumulation, the hydrogen partial pressure was the single parameter impacting reaction kinetics. The q But clearly decreased as the hydrogen partial pressure increased during perturbation, and q But increased again at the end of the perturbation when the initially low hydrogen partial pressure was restored. These conditions allowed to calculate a K i ,H2,C4ox of 0.074 ± 0.013 μM dissolved hydrogen (9 ± 2 Pa H 2 in the gas phase) which is about ten times lower than the K i ,H2,C4ox proposed in the Anaerobic Digestion Model No. 1 (ADM1) (see Online Resource Table S1 ). ADM1 is a generalized anaerobic digestion model established by the IWA task group to provide a common platform for process description and further development (Batstone et al. 2002 ). The lower K i ,H2,C4ox obtained in this study supports that anaerobic butyrate conversion is already significantly inhibited at lower hydrogen partial pressures, as previously theoretically elaborated by Kleerebezem and Stams ( 2000 ). Furthermore, an inhibitory effect of the hydrogen partial pressure on ethanol degradation was observed in both perturbation experiments. Eichler and Schink ( 1984 ) reported on hydrogen inhibition of anaerobic ethanol conversion in a pure culture of Acetobacterium carbinolicum strain WoProp1 grown on ethanol. Based on growth curves obtained either under H 2 /CO 2 or N 2 /CO 2 atmosphere (80 %/20 %), a K i ,H2,EtOH equal to 1.408 ± 0.253 μM of dissolved H 2 was derived. The K i ,H2,EtOHox determined in this study (0.515 ± 0.022 μM dissolved H 2 or 63 ± 3 Pa H 2 in the gas phase, R 2 = 0.983) was about three times lower, indicating a much stronger inhibitory effect of hydrogen on ethanol conversion. Moreover, in both experiments, the increasing hydrogen partial pressure was associated with decreasing q CH4,Acm which suggests hydrogen inhibition on acetoclastic methanogenesis. However, based on the present experiments, it was not possible to either confirm or refute the effect of hydrogen on acetoclastic methanogenesis. A tight coupling between hydrogen-producing and hydrogen-consuming organisms is essential to syntrophic methanogenic conversions. In this regard, the K S,H2 is an important kinetic parameter because a low K S,H2 permits efficient hydrogen uptake even at low hydrogen concentrations and reduces the inhibitory effect of hydrogen on the hydrogen-producing partner. The observed q CH4,Hym of hydrogenotrophic methanogenesis was only one third of the q CH4,Hym,max (0.500 mol-CH 4 (mol- X Hym ) −1 h −1 ) reported for Methanobacterium flexile strain GH and M . subterraneum strain A8p (Kotelnikova et al. 1998 ; Zhu et al. 2011 ), the two closest cultured relatives. These observations indicate that the hydrogenotrophic methanogens were operating below maximum capacity due to hydrogen limitation. Given above specific methane production rates, a K S,H2 of 52 ± 10 Pa (0.430 ± 0.082 μM dissolved H 2 ) was deduced, which lies in the reported range for several hydrogenotrophic methanogens (see Online Resource Table S1 ). In line with previous findings in defined methanogenic cocultures on lactate and formate (Junicke et al. 2015a , b ), an overcapacity of hydrogenotrophic methanogens was observed during syntrophic butyrate and ethanol conversion in non-defined methanogenic enrichments, reflecting the robustness of syntrophic bioconversions and enabling stable reactor performance. In a chemostat the biomass-specific growth rate equals the dilution rate. Previous coculture studies on lactate showed that the syntrophic partners follow different strategies to adapt to a common biomass-specific growth rate (Junicke et al. 2015b ). In the present study, the hydrogenotrophic methanogens compensated their low biomass yield per electron-mole of substrate ( Y X /e ) with a 2-fold higher biomass-specific electron transfer rate ( q e ), compared to the butyrate-utilizing partner. These findings provide further support for the previously reported growth strategies in defined methanogenic cocultures on lactate. Thermodynamic control of electron transfer Thermodynamic analysis, combined with system modeling and reaction kinetics provides valuable insights into thermodynamic feasibility of the underlying reactions, pathway reversibility, and energy sharing between syntrophic partners. In perturbation experiment C1, the increase of the hydrogen partial pressure was associated with an increasing actual Gibbs energy change of anaerobic butyrate conversion which became positive 20 h after increasing the influent ethanol concentration (Fig. 4 a). At the same time q But was effectively zero (Fig. 3 a), providing experimental evidence for the thermodynamic control of anaerobic butyrate conversion and the need to implement thermodynamic restrictions in energy-limited anaerobic digestion models, as previously proposed in (Kleerebezem and Stams 2000 ; Kleerebezem and van Loosdrecht 2006 ). In anaerobic digestion models, such as ADM1, thermodynamic constraints are still neglected, thus violating thermodynamic principles. By combining the model-derived q rates with thermodynamic analysis, it is furthermore possible to conclude on the reversibility of biochemical pathways. For example, a strongly negative q But concomitant with a positive Δ G 1 for butyrate conversion would reflect the reversibility of butyrate conversion. In experiment C1, however, q But remained close to zero at positive Δ G 1 for anaerobic butyrate conversion. These findings suggest that the reverse reaction of butyrate conversion did not occur. Pathway reversibility of anaerobic butyrate conversion was previously theoretically investigated by González-Cabaleiro et al. ( 2013 ). It was predicted that the reversibility of butyrate conversion is rather unlikely due to biochemical limitations, which agrees with the results of this study. Since anaerobic bioconversions proceed close to thermodynamic equilibrium, it is of great interest to understand how thermodynamics affect the energy sharing among the syntrophic partners. Seitz et al. ( 1990b ) investigated the energy distribution of defined methanogenic cocultures in ethanol-fed chemostats. They found an unequal distribution of the total Gibbs energy change between the hydrogen-producing acetogen (23 %) and the hydrogenotrophic methanogen (77 %). Unequal energy sharing was also demonstrated during syntrophic lactate conversion in different methanogenic cocultures (Junicke et al. 2015a , b ). However, opposite to the results of Seitz et al. ( 1990b ), the lactate-utilizing species shared a larger fraction of the total energy (79–83 %) compared to the hydrogenotrophic methanogen (17–21 %). This study revealed unequal energy distribution between the butyrate-utilizing species (17 %), the hydrogenotrophic methanogens (9–10 %), and the acetoclastic methanogens (73–74 %) during syntrophic butyrate conversion. As for the coculture study on lactate (Junicke et al. 2015a , b ), a larger energy fraction was devoted to the hydrogen-producing acetogen while the hydrogenotrophic methanogen gained considerably less energy. The lower energy gain results in a low biomass yield which requires a larger q e in order to maintain equal biomass-specific growth rates during syntrophic cooperation. Therefore, the different growth strategies are consistent with and directly follow from the unequal energy distribution between the syntrophic partners. Reduced product formation Formation of reduced products occurs as a side-reaction in the presence of excess electrons. It provides an additional electron sink when (i) the enzymatic capacity of the primary reaction is exceeded, (ii) product inhibition occurs, (iii) the primary reaction becomes thermodynamically unfeasible, or (iv) the electron acceptor of the primary reaction becomes limiting. Since reduced products represent energetically dense chemicals, the conditions of their formation are focus of on-going research (González-Cabaleiro et al. 2013 ; Steinbusch et al. 2008 ; Steinbusch et al. 2011 ). So far, the role of electron transfer in the form of hydrogen remains unclear and it is unknown at which hydrogen partial pressure a switch between methanogenesis and reduced product formation occurs. In experiment C1, butanol formation was observed 20 h following perturbation with increased ethanol concentrations, concomitant with increasing hydrogen partial pressures and decreasing carbon dioxide partial pressures (Fig. 2 ). The decrease of the carbon dioxide partial pressure led to bicarbonate limitation of hydrogenotrophic methanogenesis, and resulted in a further increase of the hydrogen partial pressure. The Δ G 1 of butyrate conversion became positive at 20 h, accompanied by butanol production and ethanol accumulation. Ethanol conversion remained thermodynamically feasible during the perturbation experiment. Smith and McCarty ( 1989a , b ) reported on similar observations in ethanol and propionate-perturbed CSTRs. They showed that the ethanol-oxidizing bacterium catalyzed the reduction of propionate with ethanol to propanol and acetate, and not the propionate-oxidizing bacterium. This was a striking observation since propionate conversion ceased due to elevated hydrogen partial pressures, and it was expected that the propionate-oxidizing bacterium would perform an alternative reaction to gain sufficient energy for growth. The formation of reduced products such as butanol reflects the redirection of electron fluxes towards an alternative electron acceptor when the hydrogenotrophic methanogen is limited. Smith and McCarty ( 1989b ) argued that this mechanism may result in an altered overall stoichiometry, which is marked by lower hydrogen production in order to circumvent kinetic and thermodynamic limitations. Furthermore, they hypothesized that the increased ethanol consumption rate may increase the need for the use of alternative enzyme systems. In the present study, butanol formation was observed at increasing hydrogen partial pressures after perturbation of the butyrate and ethanol-fed CSTR with increased ethanol concentrations. A shift from methanogenesis to reduced product formation was found when hydrogenotrophic methanogenesis was bicarbonate limited and when the hydrogen partial pressure exceeded 40 Pa. These findings imply that the hydrogen partial pressure may be an important control parameter to direct electron fluxes towards the formation of a valuable product such as butanol."
} | 4,689 |
39702246 | PMC11660635 | pmc | 1,476 | {
"abstract": "Inspired by the natural symbiotic relationships between diverse microbial members, researchers recently focused on modifying microbial chassis to create artificial coculture systems using synthetic biology tools. An increasing number of scientists are now exploring these systems as innovative biosynthetic platforms for biomass conversion. While significant advancements have been achieved, challenges remain in maintaining the stability and productivity of these systems. Sustaining an optimal population ratio over a long time period and balancing anabolism and catabolism during cultivation have proven difficult. Key issues, such as competitive or antagonistic relationships between microbial members, as well as metabolic imbalances and maladaptation, are critical factors affecting the stability and productivity of artificial coculture systems. In this article, we critically review current strategies and methods for improving the stability and productivity of these systems, with a focus on recent progress in biomass conversion. We also provide insights into future research directions, laying the groundwork for further development of artificial coculture biosynthetic platforms. Graphical Abstract",
"conclusion": "Conclusions Artificial microbial coculture systems have gained increasing attention due to their unique characteristics and functionalities, often surpassing those of individual microbial populations. In coculture systems, members can engage in one-way, two-way, or even multi-way communication through the exchange of signalling molecules, detection, and mutual responses [ 124 , 125 ]. This mutual coordination ensures the stability of the community’s structure and function [ 126 , 127 ]. By specializing in distinct tasks, each member enables the system to perform complex functions that are beyond the capabilities of a single strain. However, constructing stable and productive coculture systems is far more challenging than cultivating individual strains, primarily due to the difficulty in maintaining optimal population ratios and striking a balancing anabolism and catabolism over long-term cultivation. In this review, we summarized the issues challenging the stability and productivity of coculture systems. We also introduce strategies and tools for optimizing these systems for biomass conversion. Finally, we provided perspectives and future research directions for coculture systems as novel biosynthetic platforms, laying a foundation for the development of more sophisticated coculture systems in the future.",
"introduction": "Introduction The production of bio-based chemicals, derived from various biotechnological or chemical processes using organic streams such as non-food lignocellulosic biomass or municipal wastes, as well as CO 2 , provides environmentally friendly alternatives to fossil fuels and derivatives [ 1 ]. Microbial cell factories employ various strategies to convert renewable resources into fermentable sugars. However, converting these substrates for biosynthesis presents challenges for the monoculture, whether natural or engineered. For instance, strains of the Trichoderma genus, one of the predominant genera used in industrial enzyme production, exhibit relatively low β-glucosidase activity, an enzyme crucial for cellulose degradation that works synergistically with cellobiohydrolase and endoglucanase [ 2 ]. As a result, the monoculture of Trichoderma is unable to produce the full range of enzymes necessary for the complete breakdown of cell wall components. While CO 2 conversion can be performed by cyanobacteria or microalgae, extracting sugars such as sucrose from the culture supernatant is costly, and large-scale sucrose production often leads to contamination issues [ 3 ]. In most cases, multiple modifications to cellular metabolic pathways are necessary to enable the synthesis of desired chemicals from CO 2 [ 4 ]. In nature, most microorganisms interact with microbial communities or complex ecosystems to increase their chances of survival and growth [ 5 ]. Natural microbial symbioses have undergone millions of years of evolutionary selection, resulting in intricate yet stable interactions among various strains. Lichens serve as a prime example of such stable, self-sustaining symbiotic organisms, composed of photosynthetic autotrophs and heterotrophic fungi, and are distributed globally [ 6 ]. Utilizing naturally occurring microbial communities offers a promising approach for degrading complex substrates, though controlling the behaviour of community members can be challenging [ 1 ]. For instance, in cocultures of naturally occurring strains, the excessive accumulation of intermediate products may negatively affect the final yield. Moreover, unlike engineered strains, naturally occurring strains in cocultures lack the ability to precisely fine-tune the metabolic functions of individual bacteria to meet specific needs. Advances in synthetic biology technologies have enabled researchers to mimic natural symbiotic systems by creating artificial coculture systems with different microbial chassis, serving as next-generation biosynthesis platforms. Artificial microbial coculture involve the collaboration of two or more microbial members to establish a reaction network for chemical production [ 1 ]. Using the strategy of coculture, one-pot conversion of renewable resources into fine chemicals can be achieved, offering distinct advantages over cultivating each microorganism independently [ 7 ]. By dividing labor, the members of the coculture system reduce metabolic burdens on individual species, especially in lengthy product pathways [ 8 ]. Photosynthetic microorganisms, which can harness CO 2 and light energy for organic carbon production, can be incorporated into coculture systems to provide carbon sources for heterotrophic species [ 9 , 10 ]. This approach enables the environmentally friendly synthesis of products from CO 2 in systems that combine autotrophic and heterotrophic species. Additionally, many anaerobic bacteria, such as Clostridia , are capable of degrading cellulose through cellulosomes, spurring significant scientific interest in developing coculture systems with cellulose-degrading strains for biofuel production [ 11 , 12 ]. Considering the advantages of artificial coculture systems, additional efforts have been made to establish various coculture systems for the production of valuable products, such as amino acids, peptides, proteins, lipids, biofuels, polyphenols, alkaloids, terpenoids, and so on. In Tables 1 and 2 , we outline the applications of artificial coculture systems comprising two or more species. The stability of an artificial coculture system refers to its ability to sustain consistent microbial interactions and performance over time, while productivity refers to the amount of product generated per unit of time within the coculture. To fully harness artificial microbial coculture systems as innovative platforms for biomass conversion, it is essential to enhance both the stability and productivity of these systems [ 5 , 13 ]. Table 1 The applications of artificial two-species coculture systems as biosynthetic platforms Products Coculture systems Substrate/mid-products Production before optimization Optimization method Coculture production Refs Amino acids L-Lysine E. coli–C. glutamicum Starch/ lysine, glucose, amylase ~ 0.77 g/L with glucose Establishing cross-feeding interactions between member species 0.4 g/L without glucose [ 128 ] L-Pipecolic acid E. coli–C. glutamicum Starch/ lysine, glucose, amylase ~ 4 mM* with glucose Establishing cross-feeding interactions between member species 3.4 ± 0.1 mM without glucose [ 128 ] γ-Amino butyric acid C. lacerate–L. plantarum Glucose, soybean flour, rice bran — Adjusting environmental parameters 15.53 mg/mL [ 129 ] Peptides Nisin Y. lipolytica–L. lactis Sugar beet molasses/ lactic acid 176 mg/L* Adjusting environmental parameters 270 mg/L [ 130 ] Dentigerumycin E Streptomyces sp. JB5 –Bacillus sp. GN1 Malt extract, glucose, yeast extract — Adjusting environmental parameters 34 mg [ 131 ] Fengycin B. subtilis–C. glutamicum Maltodextrin, sucrose, yeast extract / proline 871.86 mg/L* Dividing metabolic pathways and enhancing the transport of essential metabolites 2.31 g/L [ 132 ] Proteins Alpha amylase B. cereus–B. thuringiensis Starch 14.5 ± 0.1 U/ml/min Adjusting environmental parameters 44.0 U/ml/min [ 133 ] Total protein Microalgae– L. starkeyi CO 2 0.3 g/g* Establishing cross-feeding interactions between member species 0.15 g/g (lipids and carbohydrates production was higher) [ 134 ] Lipids Lipid, total fatty acid C. pyrenoidosa–R. glutinis CO 2 , glucose Lipid: 0.74 ± 0.05 g/L, Total fatty acid: 91.2 ± 2.57 mg/L/day*( C. pyrenoidosa )、70.08 ± 1.97 mg/L/day*( R. glutinis ) Adjusting environmental parameters Lipid: 2.48 ± 0.09 g/L, total fatty acid: 175.64 ± 2.32 mg/L/day [ 135 ] Lipid R. glutinis–C. vulgaris Acetate, CO 2 1 g/L Adjusting environmental parameters 2.6 g/L [ 136 ] Biofuels H 2 E. aerogenes–C. butyricum Crude glycerol, apple pomace hydrolysate 19.46 mmol/L Adjusting environmental parameters 26.07 ± 1.57 mmol/L [ 137 ] n-Butanol E. coli – E. coli Cellulose hydrolysate of rice straw/ butyrate 0.093 g/L/h* Dividing metabolic pathways 0.163 g/L/h [ 138 ] Ethanol A. niger–S. cerevisiae Potato waste/ glucose 21.58 g/L* Constructing confined spaces 37.93 g/L [ 71 ] Methane C. cellulovorans– methanogens Sugar beet pulp/ H 2 , CO 2 — Adjusting environmental parameters 34.0 L/kg [ 139 ] Isobutanol T. reesei–E. coli Pretreated corn stover — Improving the biosynthesis of essential intermediate metabolites 1.88 g/L [ 140 ] Isopropanol E. coli–E. coli Cellobiose/ glucose — Utilization of quorum-sensing 16.0 ± 2.2 mM [ 81 ] Ethanol A. niger–S. cerevisiae Tofu waste/ sugar 7.69 g/L Adjusting environmental parameters 11.39 g/L [ 141 ] Fusel alcohol E. coli–E. coli Distillers’ grains with solubles hydrolysates 12 g/L* (from glucose synthetic medium) Adjusting environmental parameters 10.3 g/L [ 142 ] Polyphenols p -Coumaric acid, Caffeic acid S. cerevisiae–S. cerevisiae Carboxymethyl-cellulose/ glucose 0.46 mg/L*, — Establishing cross-feeding interactions between member species and adjusting environmental parameters 71.71 mg/L, 8.33 mg/L [ 143 ] Caffeic acid E. coli–C. glycerinogenes Sugarcane bagasse hydrolysate/ shikimate 133.10 mg/L* Enhancing the transport of essential metabolites 1943.2 mg/L [ 144 ] Salidroside E. coli–E. coli Cellobiose, glucose, xylose/ Glucose, tyrosol 128.2 mg/L Utilization of quorum-sensing 1.18 g/L [ 82 ] Coniferol, chavicol B. subtilis–E. coli Corncob slurry/ferulic acid, coumaric acid — Application of biosensors 55 ± 2.5 mg/L, 72 ± 1.3 mg/L [ 77 ] β-carotene S. elongatus–P. putida CO 2 / sucrose ~ 0.06 g/L Hydrogel encapsulation 1.3 g/L [ 145 ] Others 3-Hydroxypropionic acid S. elongatus–E. coli CO 2 / sucrose ~ 10 mg/L* Adjusting environmental parameters 68.29 mg/L [ 7 ] Medium-chain-length polyhydroxyalkanoate P. putida–E. coli Lignocellulose/ acetic acid, free fatty acids — Reducing competition between species members 1.64 g/L [ 25 ] Polyhydroxybutyrate S. elongatus–H. boliviensis CO 2 / sucrose ~ 0.005 mg/L/day Hydrogel encapsulation 28.3 mg/L/day [ 9 ] Polyhydroxyalkanoate S. elongatus–P. putida CO 2 , 2,4-dinitrotoluene/ sucrose — Hydrogel encapsulation 5 mg/L/day [ 146 ] Lactate T. asperellum–L. paracasei Cellulose/ glucose 1.38 g/L 3D printing 13.82 g/L [ 102 ] —, the production data were not provided in the original reference. *, Data from monoculture production Table 2 The applications of artificial more species coculture systems as biosynthetic platforms Products Coculture systems Carbon source, precursor/ mid-products Production before optimization Optimization method Coculture production Refs Caffeic acid S. cerevisiae–S. cerevisiae–S. cerevisiae Carboxymethyl-cellulose/ glucose, p -Coumaric acid 8.33 mg/L Adjusting environmental parameters 16.91 mg/L [ 143 ] Lipopeptide C. glutamicum–B. amyloliquefaciens–P. pastoris Kitchen waste/ L-proline, amylase ~ 29.77 mg/L* Dividing metabolic pathways 74.13 mg/L [ 147 ] H 2 Enterococcus–Enterococcus – Enterococcus Wheat-straw xylan — Dividing metabolic pathways 79.54 mL/g wheat-straw xylan [ 148 ] Lipid R. opacus–R. jostii–R. jostii Lignin, glucose/ acetyl-CoA, β-ketoadipyl-CoA — Reducing competition between species members 0.08 g/g cell dry weight [ 149 ] Lipopeptides B. amyloliquefaciens–C. glutamicum–C. glutamicum–P. pastoris Kitchen waste, glucose/ proline, serline, amylase ~ 193.47 mg/L* Dividing metabolic pathways and enhancing the transport of essential metabolites 269.17 mg/L [ 150 ] Butyl Butyrate C. acetobutylicum–C. tyrobutyricum-E. coli–T. asperellum Microcrystalline cellulose/ lipase, acetate, butyrate, butanol 13.52 g/L with glucose Dividing metabolic pathways 2.94 g/L without glucose [ 151 ] Butyric acid T. reesei–L. pentosus–C. tyrobutyricum–L. brevis Cellulose, xylose/ acetate, lactate, butyrate, O 2 9.5 g/L Membrane separation 10.2 g/L [ 103 ] Electricity S. elongatus-E. coli–S. oneidensis–G. sulfurreducens CO 2 / sucrose, lactate, acetate ~ 0.6 W·m −2 Dividing metabolic pathways and adjusting environmental parameters 1.7 W·m −2 [ 152 ] Iturin A B. amyloliquefaciens–B. subtilis–B. subtilis Food waste/ amylase, lipase, glucose, fatty acid chain 7.66 mg/L Dividing metabolic pathways 8.12 mg/L [ 153 ] —, the production data were not provided in the original reference. *, data from monoculture production In this article, we review recent advances in utilizing coculture systems as biosynthetic platforms for biomass conversion. We also discussed the factors that influence stability and productivity in coculture systems, along with strategies and tools to address these challenges. Additionally, we proposed further perspectives on optimizing coculture systems through the development of novel tools, analysis of interspecies relationships, and the regulation of metabolic balance. These insights provide a foundation for future efforts aimed at production of bio-based chemicals from renewable resources using coculture systems."
} | 3,547 |
33119710 | PMC7595309 | pmc | 1,477 | {
"conclusion": "Concluding remarks and further perspectives It is now clear that the bacterial QS response is not only regulated by self-made AIs, and the boundary between “self-sensing” and other signaling networks becomes blurry. Combining information about the environment together with QS, especially when the system is governed by a positive feedback loop, is proposed to be critical to coordinate bacterial group behaviors within a heterogeneous environment [ 32 ]. In addition to the examples discussed above, certain microbiota species affect V . cholerae virulence in an AI-2-dependent but LuxP-independent manner [ 33 ]. Upon attack by bacteria, mammalian cells produce a molecule that is structurally distinct but functionally similar to AI-2. In turn, this host-produced AI-2 mimic is detected by the bacterial LuxPQ and LsrB QS receptors [ 34 ]. Yet, the importance of this reciprocal interkingdom communication pathway in pathogenesis and symbiosis is not clearly defined. Moreover, V . cholerae possesses an additional QS circuit detecting an AI called 3,5-dimethyl-pyrazin-2-ol (DPO) with the receptor VqmA, separate from the multi-HK receptor pathway [ 35 ] ( Fig 2 ). Interestingly, both DPO and AI-3 depend on TDH for biosynthesis, and DPO is a structural isomer of one of the compounds in the EHEC AI-3 family [ 10 ]. The VqmA/DPO system has been proposed as a bypass mechanism to optimize QS functions within specific niches, such as within the host. Whether the Vqm system also responds to host-derived signals remains to be studied. What is the driving force for pathogenic and symbiotic bacteria to evolve to integrate both self-made AIs and host-derived chemical cues into their QS circuit? We envision that although the initial interactions between the host-derived signals and the QS receptors could be coincidental as some of these host-derived compounds structurally resemble the cognate self-made signal(s), one intriguing possibility is that these receptors might have evolved to serve as dual-function sensors to use the host-derived signal as a proxy for locating different regions in the animal host, as we have discussed in some of the Vibrio QS systems. Further investigation into the binding capacity and modulatory effects of host metabolites with other QS receptors present in different species could test this idea.",
"introduction": "Introduction Quorum-sensing (QS) systems, which rely on the production and detection of chemical signals called autoinducers (AIs) made by the bacteria themselves, are classically thought to be employed as a means to sense “self,” ensuring that bacteria cooperate and share resources to benefit their kin. Thus, most QS receptors are found to be specific for their cognate AIs. Although stringent signal specificity is considered fundamental to the fidelity of QS, receptors that respond broadly to non-self AIs have been identified. These “promiscuous” QS receptors are thought to function as interspecies signaling systems that are implicated in both competition and cooperation between microbes in polymicrobial communities [ 1 , 2 ]. Additional signal-sensing strategies have evolved for the QS systems in pathogenic and symbiotic bacteria which need to interact intimately with their hosts. Here, we discuss the organization and functions of QS circuits that harbor a dual-sensing function by detecting both endogenously produced AIs as well as chemical cues present inside the host. Co-opting the use of QS circuits to incorporate both microbial and host-derived information into their sensing repertoire allows proper spatial and temporal regulation of the expression of determinants critical for pathogenesis and symbiosis. AI-3/epinephrine sensing by EHEC QseC The QseC histidine kinase (HK) of enterohemorrhagic Escherichia coli O157:H7 (EHEC) is the earliest reported example of a QS receptor detecting both bacteria-made AIs and host-generated signals [ 3 ]. EHEC colonizes the human colon, and virulence is dependent on Shiga toxin, flagella/motility, and a type III secretion system (T3SS). Expression of the genes encoding these factors are all activated by QseC upon detection of various signals [ 4 ]. QseC is required for EHEC motility by modulating the phosphorylation state of its cognate response regulator (RR) QseB. Phosphorylated QseB binds to the regulatory region of flhDC (the master regulator of the flagellar regulon) ( Fig 1 ) [ 5 ]. The dephosphorylation of QseB by QseC is critical to derepress flhDC and maintain motility gene expression, particularly since another HK PmrB also phosphorylates QseB [ 6 , 7 ]. QseC also phosphorylates 2 additional RRs QseF and KdpE, and together, these 2 RRs activate the expression of T3SS genes on a pathogenicity island called locus of enterocyte effacement (LEE) as well as the Shiga toxin gene stx2 ( Fig 1 ) [ 4 , 8 , 9 ]. 10.1371/journal.ppat.1008934.g001 Fig 1 The AI-3/Epi/NE signaling pathway in EHEC. The HK QseC detects Epi and NE made by the host, as well as AI-3 produced by the EHEC enzyme TDH. Binding of these molecules to the HK enables QseC to control the activity of the 3 RRs KdpE, QseB, and QseF, which results in regulation of downstream virulence genes. AI-3, autoinducer-3; EHEC, enterohemorrhagic Escherichia coli O157:H7; Epi, epinephrine; HK, histidine kinase; NE, norepinephrine; RRs, response regulator; TDH, threonine dehydrogenase. Created with BioRender.com. The kinase activity of QseC is modulated by multiple signals ( Fig 1 ). QseC is activated by a self-produced autoinducer AI-3, which consists of a group of molecules belonging to the pyrazinone family, whose biosynthesis depends on threonine dehydrogenase (TDH) [ 10 ]. Synthetic AI-3 compounds added to EHEC cells activate virulence gene expression through QseC with varying potencies and specificities; however, a direct ligand binding interaction between AI-3 and QseC has not been demonstrated [ 10 ]. QseC also separately detects human adrenergic hormones epinephrine (Epi) and norepinephrine (NE) [ 3 , 4 , 11 ]. Epi/NE directly activates the kinase activity of QseC in vitro [ 11 ] and QseC-dependent virulence gene expression in EHEC [ 4 ]. The in vivo role of QseC sensing of Epi/NE in host colonization was studied using Citrobacter rodentium carrying a LEE island similar to that from EHEC [ 12 ]. C . rodentium is deficient for colonizing dopamine β-hydroxylase knockout (Dbh −/− ) mice, which do not produce Epi/NE. Similarly, qseC null mutants were also impaired for colonizing the mouse intestine, highlighting the importance of host signal sensing during host colonization [ 12 ]. Overall, these studies have established that QseC acts a crucial link integrating both host-derived signals (Epi and NE) and self-produced AI molecules (AI-3). It should also be noted that the exact regulatory mechanisms of QseC on target gene regulation are diverse among different E . coli subtypes. For example, in uropathogenic E . coli (UPEC), QseB phosphorylation state can be cross-regulated by another HK PmrB in response to iron to confer polymyxin resistance [ 7 , 13 , 14 ]. While the role of Epi/NE sensing may not be universal among different E . coli strains and subtypes, QseC signaling has been shown to be critical for virulence in many strains of enteric pathogens such as Salmonella , UPEC, and enteroaggregative E . coli (EAEC) [ 4 , 15 , 16 ]. Vibrio QS systems that detect host-generated signals Many Vibrio species including Vibrio cholerae , Vibrio harveyi , and Vibrio fischeri spend part of their life cycle inside animal hosts either as a pathogen or as a symbiont. These species use multiple QS systems to regulate the expression of the genes involved in host colonization [ 17 ]. Emerging evidence suggests that Vibrio species, similar to EHEC, also integrate host-derived chemical cues to modulate their overall QS responses. To illustrate this idea, we first focus on the canonical QS circuit of V . cholerae composed of 4 HK receptors CqsS, LuxPQ, CqsR, and VpsS [ 18 ] ( Fig 2 ). At low cell density (LCD), these 4 HKs function in parallel to phosphorylate RR LuxO through an intermediate phosphotransfer protein LuxU. Phosphorylated LuxO promotes and inhibits the production of master transcriptional regulators AphA and HapR, respectively, resulting in the activation of virulence and biofilm gene expression at LCD, which is critical for V . cholerae host colonization [ 18 ]. At high cell density (HCD), binding of the cognate signals to the receptors leads to kinase activity inhibition, resulting in dephosphorylation of LuxO and expression of HCD QS genes. Some of the AIs detected by these QS receptors are well characterized: CqsS and LuxPQ detect the Vibrio -specific signal CAI-1 ( S -3-hydroxytridecan-4-one) and the “universal” signal AI-2 in its cyclic, borated form ( S -2-methyl-2,3,3,4-tetrahydroxytetrahydrofuran-borate), respectively. AI-2 is made by many bacteria via the enzyme LuxS and is considered an interspecies signal [ 19 ]. The 2 additional V . cholerae QS receptors, CqsR and VpsS, have been demonstrated to respond to self-made chemicals present in spent culture media; however, the identities of these signals remain unknown [ 18 ] ( Fig 2 ). Similar parallel circuit architecture is found in other Vibrio species; however, the receptors used for signal perception can be variable. For example, the LuxN HKs in V . harveyi and Vibrio parahaemolyticus , and AinR HK in V . fischeri , which are all absent in V . cholerae , detect acyl homoserine lactones (AHLs; Fig 2 ) [ 19 ] and are distinct from the cytosolic LuxR AHL receptor in V . fischeri . Here, we will discuss how 2 host-derived signals, ethanolamine and nitric oxide (NO), are detected and integrated into these parallel HK-based QS systems. 10.1371/journal.ppat.1008934.g002 Fig 2 Vibrio signaling pathways for QS and host sensing. Various HK receptors in Vibrio species recognize distinct signals produced by the bacterium (sensing “self”) and/or produced by host cells or neighboring bacteria (sensing “other”). The conservation of the HKs varies among species, and this figure indicates the proteins and signals determined in the literature for each species (V.h, Vibrio harveyi ; V.c., Vibrio cholerae ; V.p., Vibrio parahaemolyticus ; V.f., Vibrio fischeri ). Signals produced by enzymes (if known) are indicated for each system. In the absence of cognate signals, phosphorylation (P) of LuxU and LuxO leads to production of AphA and low production of LuxR (V.h.)/HapR (V.c.)/OpaR (V.p.)/LitR (V.f.) and expression of biofilm, virulence, and type III secretion genes. In the presence of signals, dephosphorylation of LuxU drives production of LuxR/HapR/OpaR/LitR and expression of bioluminescence, proteases, and type VI secretion genes. The NO/H-NOX/HahK pathway in V.f. inhibits syp gene expression and biofilm formation. The VqmA/DPO pathway inhibits biofilm formation in V.c. AI-2, autoinducer-2; CAI-1, cholera autoinducer-1; CP, cytoplasm; DPO, 3,5-dimethyl-pyrazin-2-ol; HAI-1, harveyi autoinducer-1; HK, histidine kinase; H-NOX, heme nitric oxide/oxygen binding; NO, nitric oxide; PP, periplasm; QS, quorum sensing. Created with BioRender.com. Integration of ethanolamine sensing into the QS circuit Ethanolamine is a common intestinal metabolite generated during host and bacteria membrane turnover. In an unbiased chemical screen, ethanolamine was found to specifically interact with the periplasmic ligand-binding domain of CqsR. In V . cholerae mutants expressing only CqsR but not the other 3 QS receptors, ethanolamine induces a premature HCD QS response to inhibit virulence gene expression and limit host colonization [ 20 ]. Yet, V . cholerae mutant defective in producing ethanolamine is still proficient in QS, suggesting ethanolamine functions only as an external cue for CqsR, and additional signals must be endogenously made by V . cholerae and detected by CqsR. While the exact physiological function of ethanolamine sensing by CqsR remains unclear, the ethanolamine concentration is notably higher in the large intestine than that in the small intestine [ 20 ], and therefore, ethanolamine could be used as a proxy for niche identification. Interestingly, previous studies have demonstrated that ethanolamine both positively and negatively affects host colonization and virulence during infection with other enteric pathogens [ 21 , 22 ]. Integration of NO sensing into the QS circuit NO is produced by a variety of animal cells as an antibacterial mechanism. Upon NO sensing, some bacteria express a set of nitrosative response genes to counteract this toxic compound [ 23 ]. Heme NO/O 2 binding (H-NOX) proteins are a broadly conserved family of sensor proteins that bind NO within an Fe(II)-heme domain [ 24 ]. H-NOX modulates the activity of a HK called H-NOX-associated QS kinase (HqsK) encoded in the same operon as H-NOX [ 25 – 27 ]. In V . harveyi and V . parahaemolyticus , HqsK feeds into the parallel QS circuitry made of LuxPQ, CqsS, and LuxN described above ( Fig 2 ). In the absence of NO, HqsK phosphorylates LuxO via LuxU. When NO is present, it binds to H-NOX, and this complex inhibits the kinase activity of HqsK. This decreases the pool of phosphorylated LuxU and LuxO, resulting in a premature HCD QS response (e.g., increase in light production) in V . harveyi [ 25 , 26 ]. In V . fischeri , H-NOX/NO inhibits the HqsK homolog, HahK ( Fig 2 ), resulting in decreased biofilm formation via inhibition of syp transcription [ 28 ] and decreased expression of genes encoding hemin transport [ 29 ]. It is not yet clear if H-NOX influences HKs in the V . fischeri QS circuit, although LuxPQ and the downstream components are conserved in V . fischeri ( Fig 2 ). NosP proteins are another class of NO-sensing proteins that are widely conserved in bacteria and are also encoded in operons with cognate signaling proteins [ 30 ]. In V . cholerae , NosP (also called VpsV) binds NO and inhibits the autokinase activity of VpsS (encoded in the same operon) in vitro [ 30 ] ( Fig 2 ). In this way, V . cholerae NosP bound to NO appears to function analogously to V . harveyi H-NOX to inhibit phosphorylation of LuxU and could potentially feed into the QS pathway. However, the exact physiological role of NO sensing by NosP/VpsV in V . cholerae QS gene regulation is unclear. Because there are no identified NO synthase genes encoded in these Vibrio species, it is hypothesized that NO acts as an interkingdom signaling molecule between bacterium and host. For example, in V . fischeri , during early stages of colonization of the light organ in the bobtail squid Euprymna scolopes , the surface epithelium of the squid secretes mucus that contains NO [ 31 ]. V . fischeri cells first adhere to the mucus and form aggregates, and then the bacteria disperse and migrate through pores to eventually colonize the crypts of the light organ. NO inhibits aggregation and biofilm formation, and other signals such as calcium positively influence biofilm formation [ 28 ]. Thus, the integration of both NO and calcium signaling may balance aggregation and biofilm formation to a level that enables the bacteria to disperse from the aggregates to colonize the light organ. Additionally, the repression of hemin transport by HahK likely prepares the bacteria for the iron limited environment of the host light organ. In tandem, these results indicate that the V . fischeri bacteria rely on NO host signaling to colonize and adjust to the environment in the light organ. Concluding remarks and further perspectives It is now clear that the bacterial QS response is not only regulated by self-made AIs, and the boundary between “self-sensing” and other signaling networks becomes blurry. Combining information about the environment together with QS, especially when the system is governed by a positive feedback loop, is proposed to be critical to coordinate bacterial group behaviors within a heterogeneous environment [ 32 ]. In addition to the examples discussed above, certain microbiota species affect V . cholerae virulence in an AI-2-dependent but LuxP-independent manner [ 33 ]. Upon attack by bacteria, mammalian cells produce a molecule that is structurally distinct but functionally similar to AI-2. In turn, this host-produced AI-2 mimic is detected by the bacterial LuxPQ and LsrB QS receptors [ 34 ]. Yet, the importance of this reciprocal interkingdom communication pathway in pathogenesis and symbiosis is not clearly defined. Moreover, V . cholerae possesses an additional QS circuit detecting an AI called 3,5-dimethyl-pyrazin-2-ol (DPO) with the receptor VqmA, separate from the multi-HK receptor pathway [ 35 ] ( Fig 2 ). Interestingly, both DPO and AI-3 depend on TDH for biosynthesis, and DPO is a structural isomer of one of the compounds in the EHEC AI-3 family [ 10 ]. The VqmA/DPO system has been proposed as a bypass mechanism to optimize QS functions within specific niches, such as within the host. Whether the Vqm system also responds to host-derived signals remains to be studied. What is the driving force for pathogenic and symbiotic bacteria to evolve to integrate both self-made AIs and host-derived chemical cues into their QS circuit? We envision that although the initial interactions between the host-derived signals and the QS receptors could be coincidental as some of these host-derived compounds structurally resemble the cognate self-made signal(s), one intriguing possibility is that these receptors might have evolved to serve as dual-function sensors to use the host-derived signal as a proxy for locating different regions in the animal host, as we have discussed in some of the Vibrio QS systems. Further investigation into the binding capacity and modulatory effects of host metabolites with other QS receptors present in different species could test this idea."
} | 4,498 |
36865097 | PMC9980142 | pmc | 1,478 | {
"abstract": "The Bacillus subtilis extracellular biofilm matrix includes an exopolysaccharide that is critical for the architecture and function of the community. To date, our understanding of the biosynthetic machinery and the molecular composition of the exopolysaccharide of B. subtilis remains unclear and incomplete. This report presents synergistic biochemical and genetic studies built from a foundation of comparative sequence analyses targeted at elucidating the activities of the first two membrane-committed steps in the exopolysaccharide biosynthetic pathway. By taking this approach, we determined the nucleotide sugar donor and lipid-linked acceptor substrates for the first two enzymes in the B. subtilis biofilm exopolysaccharide biosynthetic pathway. EpsL catalyzes the first phosphoglycosyl transferase step using UDP-di- N -acetyl bacillosamine as phospho-sugar donor. EpsD is a GT-B fold glycosyl transferase that facilitates the second step in the pathway that utilizes the product of EpsL as an acceptor substrate and UDP- N -acetyl glucosamine as the sugar donor. Thus, the study defines the first two monosaccharides at the reducing end of the growing exopolysaccharide unit. In doing so we provide the first evidence of the presence of bacillosamine in an exopolysaccharide synthesized by a Gram-positive bacterium.",
"conclusion": "Overarching Conclusion The study of glycoconjugate biosynthesis pathways requires a concerted effort of different approaches as individual bioinformatic, biochemical, and genetic approaches often provide incomplete details. In this study, we establish the sequential characterization of the B. subtilis EPS steps by applying biochemical assays and phenotypic screening to the first two membrane-associated processes in the pathway – EpsL and EpsD. The major advantage of addressing steps in the pathway in their biosynthetic order is that the characterization of each enzyme provides the substrate for investigating the following step. Additionally, as enzyme expression and isolation (either in a cell envelope fraction or in a detergent-solubilized form) is included in the process, it enables the chemoenzymatic synthesis of products for additional analysis and use in related pathways. The established enzyme assays also provide the opportunity for small molecule inhibitor screening, both individually (EpsL or EpsD) or as biosynthetic partners (EpsL and EpsD). Taken together, these studies set a clear course for analysis of the downstream EPS glycosylation pathway and the development of a complete picture of EPS structure.",
"introduction": "INTRODUCTION Biofilms are self-associating microbial systems that contain surface-adherent individuals within an extracellular matrix ( 1 ). The non-pathogenic bacterium, Bacillus subtilis ( Bs ) , has been used extensively for understanding biofilm formation due to its ease of genetic manipulation and its extensive applied uses across diverse sectors of our economy ( 2 ). The B. subtilis biofilm matrix contains multiple specific components: BslA (a hydrophobin-like protein that confers hydrophobicity and structure to the community), fibers of the protein TasA (required for the structural integrity of biofilm), extracellular DNA (eDNA, important at early stages of biofilm formation), poly-γ-glutamic acid (γ-PGA, possible function in water retention), and an exopolysaccharide (EPS) ( 3 ). The EPS is the main carbohydrate component of the B. subtilis matrix and is critical for biofilm architecture and biofilm function ( 4 , 5 ). Despite considerable interest in understanding biofilm biosynthesis and regulation, the individual building blocks for this macromolecular glycoconjugate have not been determined. Biosynthesis of EPS is dependent on enzymes expressed from a 15-gene epsABCDEFGHIJKLMNO ( epsA–O ) operon, which has a similarity with the Campylobacter jejuni pgl operon ( Figure 1A ) ( 6 ). These enzymes have been annotated based on sequence analysis as a phosphoglycosyl transferase (PGT), glycosyl transferases (GTs), uridine diphosphate sugar (UDP-sugar) modifiers, a regulatory enzyme, and a flippase ( 5 , 7 , 8 ). However, most of the membrane-associated enzymes that are involved in the biosynthesis of exopolysaccharide in B. subtilis have not been biochemically characterized. Furthermore, analysis of exopolysaccharide composition has afforded conflicting information. Even studies of the same strain of B. subtilis (namely NCIB 3610) provided different carbohydrate compositions depending on the bacterial growth conditions and/or methods of extraction and purification ( 6 ). For example, when grown in a glutamic acid and glycerol-rich media, an EPS fraction contained glucose, N -acetylgalactosamine (GalNAc), and galactose (Gal) ( 9 , 10 ). The same strain grown in lysogeny broth (LB) media that included magnesium and manganese divalent cations, produced an EPS fraction containing mannose and glucose ( 11 , 12 ). Furthermore, growth in a minimal media supplemented with glucose (MMG) produced an EPS fraction containing poly- N -acetylglucosamine (GlcNAc) ( 5 ). UDP- N,N ’-diacetylbacillosamine (UDP-diNAcBac) is a prokaryote-specific nucleotide sugar donor ( 13 ). The monosaccharide component, diNAcBac, was originally discovered in Bacillus licheniformis ( 14 ). Based on in vitro activity and sequence similarity, EpsC, EpsN, and EpsM are proposed to produce UDP-diNAcBac in B. subtilis ( Figure 1B ). EpsC contains the sequence motifs found in other dehydratases and is a UDP-GlcNAc 4,6-dehydratase that converts UDP- N -acetylglucosamine (UDP-GlcNAc) to UDP-2,6-dideoxy-2-acetamido 4-keto glucose (UDP-4-keto) ( 15 ). It catalyzes the NAD + -dependent elimination of water across C5 and C6, while oxidizing C4 of UDP-GlcNAc. The resulting α,β-unsaturated ketone is reduced by hydride addition at C6, followed by tautomerization and regeneration of NAD + to provide the UDP-4-keto sugar and cofactor for a new catalytic cycle. The penultimate enzyme, EpsN, is a pyridoxal 5’-phosphate (PLP)-dependent aminotransferase that transfers an amine from L-glutamate to the C4 of UDP-4-ketosugar to provide UDP-2,6-dideoxy 2-acetamido 4-amino glucose (UDP-4-amino) ( 16 ). The subsequent enzyme, EpsM, is an acetyltransferase that transfers an acetyl group from acetyl coenzyme A (AcCoA) onto UDP-4-amino sugar to provide UDP-diNAcBac ( 17 ). To further support the assignment of these Eps enzymes, isofunctional homologs in Campylobacter , in particular C. jejuni (PglF, PglE, PglD) ( Figure 1A ), have been biochemically characterized and shown to make UDP-diNAcBac in a similar fashion ( 13 , 18 – 20 ). Consistent with this, EpsCNM from B. subtilis and PglFED from C. jejuni ( Cj ) have 54, 64, and 50% sequence similarity, respectively ( 15 ). Our overarching goal is to elucidate the composition and structure of the B. subtilis biofilm matrix EPS. Given the inconsistencies obtained from direct analysis of the extracted EPS material, we elected to start by determining the identity of the individual monosaccharides at the reducing end of the exopolysaccharide. In this work, we investigate and define the substrate specificity of two enzymes encoded within the eps operon, EpsL and EpsD, annotated as a phosphoglycosyl transferase and glycosyl transferase, respectively, using biochemical and genetic complementation approaches. We present experimental evidence supporting the designation of EpsL as a PGT which installs diNAcBac as the first monosaccharide onto a undecaprenol phosphate (UndP) carrier. We also identify EpsD as the second enzyme, and the first GT, in the pathway that likely installs GlcNAc onto the diNAcBac-appended lipid anchor. Thus, a key polyprenol-diphosphate-linked disaccharide is proposed and can be made available through chemoenzymatic synthesis. Therefore, our work sets the stage for future analysis of downstream glycosyltransferase reactions in the EPS pathway.",
"discussion": "Discussion It is extremely challenging to elucidate the structures of complex glycoconjugates directly from bacterial extracts. A case in point is the major polysaccharide found in the extracellular matrix of B. subtilis biofilms, which has remained undefined, despite considerable experimentation for many years. This is an important area of research as biofilm formation is a prevalent behavior displayed across multiple microbial species and exopolysaccharide production is highly correlated with biofilm formation ( 35 ). In this study, we have applied complementary biochemical and genetic approaches to establish the function of essential enzymes that catalyze key early steps in biofilm biosynthesis from the Bacillus subtilis epsA-O operon. Overall, the sequences of protein encoded by the operon support the expression of enzymes involved in UDP-sugar biosynthesis as well as several GTs and a PGT with unknown substrate specificity and roles in biofilm biosynthesis, however, in the absence of targeted analysis, the eps pathway cannot be defined. EpsL is a functional PGT that utilizes UDP-diNAcBac. Bioinformatic analysis suggested that many of the genes in the epsA-O cluster showed similarity to the pgl gene cluster, which is responsible for the general protein N-glycosylation pathway in C. jejuni ( 32 , 36 ). As the pgl gene cluster had been biochemically characterized and shown to be involved in the biosynthesis of UDP-diNAcBac and a heptasaccharide product containing diNAcBac at the reducing end of the glycan ( 37 ), this similarity provided the foundation for exploration of the function of selected enzymes in the B. subtilis EPS pathway. Previous sequence analysis and in vitro characterization of EpsCNM suggested that these enzymes are responsible for the biosynthesis of UDP-diNAcBac ( 15 – 17 ). Sequence analysis also identified EpsL as a close homolog of the C. jejuni and C. concisus PglCs, which are structurally and biochemically well-characterized PGTs ( Figure 2 ) ( 32 ). The identification of a PGT is noteworthy as these enzymes catalyze phosphosugar transfer from UDP-diNAcBac to a polyprenol phosphate carrier as the first membrane-associated step in many glycoconjugate assembly pathways ( 38 ). Thus, we designed a strategy to implement an in vitro biochemical activity assay using UndP as the acceptor substrate and a series of [ 3 H]-labeled and unlabeled UDP-sugars, including UDP-diNAcBac. Following heterologous expression, solubilization, and purification, EpsL was used to screen enzyme activity in vitro . Complementary assays using either radiolabeled sugars or the UMP-Glo ® assay were applied to confirm that EpsL prefers UDP-diNAcBac as phosphosugar donor and affords the Und-PP-diNAcBac product ( Figure 3B – C ). These in vitro biochemical assay results were supported by genetic analyses using biofilm formation as the phenotypic readout. This revealed that the B. subtilis epsL deletion mutant could be genetically complemented by the pglC coding sequence of C. jejuni ( Figure 3D ). Thus, we conclude that EpsL catalyzes the first step in EPS biosynthesis pathway to form Und-PP-diNAcBac. Moreover, we show the first experimental evidence of the function of a UDP-diNAcBac utilizing PGT in a Gram-positive bacterium and the presence of diNAcBac as the first sugar at the reducing end of EPS in B. subtilis . These findings are significant; diNAcBac was first discovered in Bacillus licheniformis ( 14 ), however, to date the diNAcBac sugar has only been described in N- and O-linked glycoproteins, lipopolysaccharide (LPS), and the capsular polysaccharide (CPS) of diverse Gram-negative bacteria ( 13 ). EpsD is a UDP-GlcNAc-dependent N -acetyl glucosamine transferase in B. subtilis . The successful characterization of the first step in EPS pathway provided the Und-PP-diNAcBac substrate for exploring the next enzyme in the EPS biosynthesis. In this case, although the epsA-O gene cluster revealed five candidate GTs with predicted GT-A or GT-B fold, the assignment of structure to functional specificity could not be definitively predicted. However, the similarity of epsA-O cluster genes with C. jejuni N-glycosylation pathway genes helped us to narrow down the candidates to EpsD and EpsF as possible GTs for the subsequent step in the pathway. Our bioinformatic analysis suggested that both EpsD and EpsF share similarity with PglA of C. jejuni and selected Neisseria sp. ( Figure 4A ) and we additionally knew that both EpsF and EpsD were essential for biofilm formation in B. subtilis ( 5 ). The possibility that EpsF was the next enzyme in the biosynthetic pathway was ruled out by the inability of pglA genes of C. jejuni and Neisseria sp. to rescue the biofilm formation upon expression in epsF deletion mutant of B. subtilis ( Figure S8A ). However, comparable experiments with EpsD provided new insight as genetic complementation with the C. jejuni pglA was able to partially rescue the biofilm-negative phenotype in the epsD deletion mutant of B. subtilis ( Figure 4B ). In contrast, the expression of two pglA variants , which catalyze the addition of Gal in the second step of the Neisseria pgl pathway( 39 ) did not rescue the phenotype in the epsD deletion mutant. Although, the partial complementation of p glA of C. jejuni in epsD deletion mutant did not confirm the preference of EpsD for GalNAc it provided the possibility that the preferred sugar substrate could be the related HexNAc sugar, GlcNAc. This hypothesis was supported by using a biochemical approach where the cell envelope fraction of E. coli expressing EpsD was used to assess activity using Und-PP-diNAcBac and four different commercially available 3 H-labeled UDP-sugars as donor substrates. The in vitro assay results provided further insight into the EpsD sugar substrate selectivity; EpsD showed a clear preference for UDP-GlcNAc over the other UDP-sugars tested with significant conversion UDP-[ 3 H]GlcNAc to Und-PP-diNAcBac-[ 3 H]GlcNAc ( Figure 4D ). This supports the function of EpsD in the second step in the EPS pathway. Interestingly, EpsD was also able to transfer [ 3 H]GalNAc to Und-PP-diNAcBac although with far lower efficiency. This donor substrate promiscuity displayed by EpsD not only explains the partial genetic complementation of epsD deletion mutant of B. subtilis with p glA of C. jejuni but also provides insight into the step downstream. As previously established, PglA transfers GalNAc onto Und-PP-diNAcBac in C. jejuni N-glycans ( 32 , 40 ). Thus the partial complementation observed upon expressing pglA in the B. subtilis epsD deletion mutant suggests that Und-PP-diNAcBac-GalNAc is not a preferred acceptor for the next GT in the B. subtilis EPS biosynthetic pathway, resulting in the observed partial biofilm phenotype. It also suggests possible acceptor substrate promiscuity of the next GT in line. Summarizing new insights into the Bacillus subtilis EPS biosynthetic pathway The characterization of EpsL and EpsD in this study has set the foundation for characterizing the remaining GTs in the EPS biosynthesis pathway, which would ultimately enable us to define the EPS sugar composition and structure. Based on the experimental evidence provided in this study, we propose the current EPS glycosylation pathway ( Figure 5 ). EpsCNM have already been shown to biosynthesize UDP-diNAcBac ( 15 – 17 ). EpsL is a PGT that transfers diNAcBac onto Und-P converting it to Und-PP-diNAcBac. EpsD further extends this glycan by transferring GlcNAc onto the product from EpsL thus converting it to Und-PP-diNAcBac-GlcNAc. These findings also indicate a divergence in the B. subtilis EPS glycosylation pathway after the synthesis of Und-PP-diNAcBac (as diNAcBac-GlcNAc-) compared to C. jejuni (diNAcBac-GalNAc-) and N. gonorrhoeae (diNAcBac-Gal-) pathways. Homologs of EpsL and EpsD are present broadly across the B. subtilis clade. This suggests the presence of similar glycosylation pathways and exopolysaccharides in many Bacillus species and provides an opportunity to explore the diversity of diNAcBac-containing clusters and the associated exopolysaccharides. Overarching Conclusion The study of glycoconjugate biosynthesis pathways requires a concerted effort of different approaches as individual bioinformatic, biochemical, and genetic approaches often provide incomplete details. In this study, we establish the sequential characterization of the B. subtilis EPS steps by applying biochemical assays and phenotypic screening to the first two membrane-associated processes in the pathway – EpsL and EpsD. The major advantage of addressing steps in the pathway in their biosynthetic order is that the characterization of each enzyme provides the substrate for investigating the following step. Additionally, as enzyme expression and isolation (either in a cell envelope fraction or in a detergent-solubilized form) is included in the process, it enables the chemoenzymatic synthesis of products for additional analysis and use in related pathways. The established enzyme assays also provide the opportunity for small molecule inhibitor screening, both individually (EpsL or EpsD) or as biosynthetic partners (EpsL and EpsD). Taken together, these studies set a clear course for analysis of the downstream EPS glycosylation pathway and the development of a complete picture of EPS structure."
} | 4,358 |
34890232 | PMC8664264 | pmc | 1,479 | {
"abstract": "A robot learns to follow a path to exit a maze through sensorimotor learning that is induced by an organic neuromorphic circuit.",
"introduction": "INTRODUCTION In all living organisms, the sensory and motor systems coordinate with each other, forming a unified entity ( 1 – 4 ). In this sensorimotor integration, the processing of senses in the sensory system occurs jointly with motor behaviors while, simultaneously, motor actions are under continuous sensory guidance. For instance, the action-to-sense direction can occur in vision (move of the body or saccadic eye movements to actively visualize the environment) and in olfaction (active sampling with sniffs to perceive a smell) ( 5 ). In the opposite direction of sense to action, sensory stimuli trigger motor actions, e.g., the presence of an object in the visual field initiates and guides movement ( 6 ). Even simple invertebrate organisms such as insects (e.g., drosophila, locust, etc.), whose neuronal circuits are easily traceable, exhibit a repertoire of intelligent behaviors due to sensorimotor integration ( 7 ). These behaviors are either hardwired and predefined (reflex-like) or learned as sensorimotor associations that are context dependent. More complex behaviors and learning build upon low-level reflexes and sensorimotor associations. A simplified mechanistic yet insightful version of sensorimotor integration was proposed by Braitenberg ( 8 ), with vehicles as a metaphor. In these hypothetical vehicles, primitive forms of intelligence that are found in low-level species such as exploratory, avoidance, and escape behavior emerge by coupling sensory signals and motor commands via excitatory/inhibitory and ipsilateral/contralateral connections ( 8 – 12 ). On top of this hardwired coupling, behavioral learning is promoted by adaptable sensory-to-motor connections, thus forming sensorimotor associations that represent a simple and generalized mechanism for behavioral emergence. Although conceptually primitive, these vehicles represent a prominent platform to develop and assess neuromorphic circuits for learning sensorimotor processes and behavioral tasks in robotics, as well as for energy-efficient and distributed data handling/processing ( 13 – 15 ). Neuronal computation can be directly emulated in the analog-digital domain on neuromorphic circuits, thereby providing real-time communication between the analog world, accessed by the sensorimotor system and the digital unit(s) of robotic platforms ( 13 , 16 – 18 ). Nevertheless, these neuromorphic circuits are usually large scale and implemented in custom-made robotic systems ( 13 , 16 – 19 ). For example, the silicon-based SpiNNaker engine, which has been used for sensorimotor learning, consists of 48 chips and 18 processors per chip ( 16 , 17 ). Despite the notable demonstrations of high complexity, an in materio computing perspective may provide elegant and simplified solutions in robotics. Emerging materials and devices, for instance, have novel properties and can unlock circuit functionalities unattainable by conventional electronics, as they are able to emulate directly bioinspired and biorelevant functionalities such as synaptic plasticity, neuronal functions, homeostasis, and self-healing ability, without needing complex circuitry ( 20 ). Moreover, the mechanics of the embodiment (i.e., physical body) are crucial in robotics, for instance, the use of inertia for energy-efficient locomotion and morphological adaptation for locomotion in unstructured or complex environments ( 19 , 21 , 22 ). Small-scale (i.e., simple and consisting of limited components) neuromorphic circuits for sensorimotor control and optimization of the utmost simplicity are thus of great importance for understanding the fundamental relationships between the sensory and motor systems. Only recently, small-scale neuromorphic circuits based on metal oxide neuromorphic devices have been used for local computation and control in robotic systems. Improved balance with low-latency and adaptive behavior in mobile robotics has been achieved with memristor-based adaptive filters and arrays ( 23 , 24 ). Robotic arm control that is tolerant to damages has also been demonstrated with metal oxide transistors ( 25 ). Nevertheless, in the above cases, learning is either offline ( 23 ) or outside the sensorimotor loop ( 24 ), and the implementation requires many conventional silicon components ( 23 – 25 ). Organic electronic materials have recently emerged for neuromorphic electronics because of excellent tuneability, high stability, and low-voltage, low-power operation ( 26 – 31 ). Organic materials are soft, can be solution-processed or printed at relatively low thermal budget, and can be integrated on large-area, rigid, as well as conformal substrates ( 32 , 33 ). The flexible and biocompatible nature and the mixed ion-electronic conduction of semiconducting polymers also allow for enhanced connections with biological and biohybrid systems ( 34 , 35 ). Despite these notable demonstrations, organic neuromorphic circuits have only been assessed so far for their trainability and adaptability in on-bench applications such as small-scale artificial neural networks, logic gates, and sensors ( 35 , 36 ), all systems that perceive the external stimuli without any behavioral context and outcome. However, building and evaluating intelligent systems require a holistic approach with embodiment, with agents that perform actions to explore the environment and perceive in real time the corresponding consequences ( 4 , 14 , 37 ). Creating sensorimotor associations in locally trained organic neuromorphic circuits with ease of fabrication and unconventional form factors (i.e., solution processable, printable, large-area integration, and mechanical conformity) can lead to optimized systems with “on the edge” decentralized/distributed learning and reduced communication latencies (via large-area integration in flexible/stretchable substrates), fault tolerance due to redundancy or self-repairing (via large-area integration and self-healing ability), versatility, and low-power consumption (via low-voltage operation). In this work, we introduce sensorimotor integration and local learning in a target behavioral task that requires mobility, which is enabled by a simple and low-voltage organic neuromorphic circuit. A standalone robot learns to navigate itself in a two-dimensional maze by following a planned path, after training its organic neuromorphic circuit with direct and real-time feedback from the sensorimotor system ( Fig. 1A ). Through online learning within the sensorimotor loop, the organic neuromorphic circuit establishes an association between the robot’s sensory and motor units. This association is necessary for accomplishing the navigation task. Specific motor actions are triggered by visual stimuli that function as navigation cues. This sensorimotor integration, which happens locally and in the analog domain, guides the robot to the exit. The robot, its sensors/actuators, and the neuromorphic circuit are battery-powered and operate autonomously. The work showcases the use of organic neuromorphic electronics as local and decentralized learning circuitry for mobile applications in environments with energy restrictions. Fig. 1. Path-planning robot with an organic neuromorphic circuit for sensorimotor integration. ( A ) An autonomous robot gradually learns to navigate in a maze by following navigation cues to the exit. Processing and learning toward the target task are achieved locally with an organic neuromorphic circuit. ( B ) Detailed schematic of the robotic system. Static, low-level control of the sensorimotor system is carried out by a central unit in the digital domain. The sensorimotor system and the organic neuromorphic circuit operate in the analog domain, and a real-time, sensorimotor loop is established between the control unit (digital domain) and the sensorimotor system/neuromorphic circuit (analog domain). The neuromorphic circuit consists of organic synaptic transistors: a volatile (OECT) and a nonvolatile (MEM) device. While operating within the loop, the neuromorphic circuit receives optomechanical sensory signals (at the gates of the devices G OECT and G MEM ) to perceive (adapt) the (to the) environmental stimuli and sends motor commands ( V M ) to the actuators of the robot for locomotion. With its trainability and adaptivity, the circuit forms sensorimotor associations through training that are necessary for accomplishing the target task. ( C ) The turning behavior of the robot at a maze intersection is influenced nondeterministically and in real time by the output voltage V M of the neuromorphic circuit.",
"discussion": "DISCUSSION Inspired by the biological process of sensorimotor integration, we demonstrated a standalone robot that learns with a simple yet effective neuromorphic circuit. An organic neuromorphic circuit is used as a low-voltage, analog computing core of the sensorimotor loop in robotics made of education-purpose components. While the robot explores the environment, real-time sensorimotor signals are merged in the organic neuromorphic circuit, and, through local/decentralized training on the circuit, a visuomotor association is gradually formed. With this sensorimotor integration, the robot learns to associate navigation cues with behavioral outcome and is able to follow a planned path to the exit of a maze. Once the sensorimotor association is established, the robot is able to navigate inside the maze toward the exit through unknown paths. This demonstration shows how low-voltage and easy-to-tune organic devices can function as adaptive elements capable of forming multimodal associative links for autonomous learning. It highlights the ease of fabrication, integration, and training of organic neuromorphic circuits for decentralized sensorimotor integration and paves the way for sophisticated systems that include a plethora of sensory streams to allow more complex behaviors, advanced learning in circuits, or even in materio sensing, computing, and actuating with high-performing organic materials. By integrating sensory, actuating, learning, and self-repairing primitives in materio, intelligence can be distributed and incorporated in the fabric of agents. The combination of organic neuromorphic electronics with education-purpose robotics will also lead to a versatile platform for physical modeling and rapid prototyping of intelligent, real-world systems."
} | 2,630 |
32490601 | PMC7268258 | pmc | 1,480 | {
"abstract": "Abstract Biological nitrogen fixation emerging from the symbiosis between bacteria and crop plants holds promise to increase the sustainability of agriculture. One of the biggest hurdles for the engineering of nitrogen‐fixing organisms is an incomplete knowledge of metabolic interactions between microbe and plant. In contrast to the previously assumed supply of only succinate, we describe here the CATCH ‐N cycle as a novel metabolic pathway that co‐catabolizes plant‐provided arginine and succinate to drive the energy‐demanding process of symbiotic nitrogen fixation in endosymbiotic rhizobia. Using systems biology, isotope labeling studies and transposon sequencing in conjunction with biochemical characterization, we uncovered highly redundant network components of the CATCH ‐N cycle including transaminases that interlink the co‐catabolism of arginine and succinate. The CATCH ‐N cycle uses N 2 as an additional sink for reductant and therefore delivers up to 25% higher yields of nitrogen than classical arginine catabolism—two alanines and three ammonium ions are secreted for each input of arginine and succinate. We argue that the CATCH ‐N cycle has evolved as part of a synergistic interaction to sustain bacterial metabolism in the microoxic and highly acid environment of symbiosomes. Thus, the CATCH ‐N cycle entangles the metabolism of both partners to promote symbiosis. Our results provide a theoretical framework and metabolic blueprint for the rational design of plants and plant‐associated organisms with new properties to improve nitrogen fixation.",
"introduction": "Introduction Nitrogen is a fundamental element of all living organisms and the primary nutrient that impacts crop yield (Socolow, 1999 ). Despite being highly abundant in the atmosphere, plants can only assimilate nitrogen in reduced forms such as ammonium. More than 125 megatons of nitrogen are fixed annually by the industrial Haber–Bosch process into ammonium and applied to increase agricultural crop production (Graham & Vance, 2000 ). Endosymbiosis between legumes and soil bacteria termed rhizobia is capable to fix nitrogen biologically. On a global scale, anthropogenic nitrogen delivered to the environment surpasses annual supplies by natural biological nitrogen fixation on land (Gruber & Galloway, 2008 ) leading to serious environmental impacts from climate change to the disruption of eco‐systems and pollution of coastal waters. Improving the ability of plants and plant‐associated organisms to fix atmospheric nitrogen has inspired biotechnology for decades (Beatty & Good, 2011 ; Bhardwaj et al , 2014; Gupta et al , 2015), not only for the apparent economic and ecological benefit that comes with the replacement of chemical fertilizers but also more recently for opportunities toward more sustainable agriculture and the potential to reduce greenhouse gas emissions. To catalyze atmospheric nitrogen fixation, rhizobia use a specific enzyme termed nitrogenase. Attempts to transfer and improve nitrogenase genes clusters have to date focused largely on organisms such as Escherichia coli (Dixon & Postgate, 1972 ; Wang et al , 2013 ). More recently, the emerging field of synthetic biology provides an alternative approach to engineer designer nitrogenase gene clusters in bacteria (Temme et al , 2012; Li et al , 2016 ; Burén et al , 2018; Yang et al , 2018). Despite these promising results, engineered organisms based on heterologous expression of nitrogenase genes have not yet come close to the efficiency of natural rhizobia–legume symbiosis systems (Beatty & Good, 2011 ; Good, 2018 ). While the molecular mechanism of the nitrogenase reaction has been resolved with atomistic detail (Hoffman et al , 2009 , 2014 ; Seefeldt et al , 2009 ; Sippel & Einsle, 2017 ), the precise nature of metabolic interactions between plants and bacteria to sustain the energy‐intensive process of nitrogen fixation has remained an open question. The current model of nutrient exchange in rhizobia–legume symbiosis postulates that, in exchange for fixed nitrogen, the plant provides C4‐dicarboxylic acids such as succinate, which is metabolized through the tri‐carboxylic acid (TCA) cycle to generate ATP and reduction equivalents needed for the nitrogenase reaction (Watson et al , 1988 ; Yurgel & Kahn, 2004 ; Clarke et al , 2014 ). However, multiple lines of evidence argue against a simple exchange of succinate for ammonium during symbiosis (Udvardi & Kahn, 1993 ; Kahn et al , 1985 ). The nitrogenase is highly sensitive to oxygen, which irreversibly inactivates the enzyme. While the microoxic environment encountered by rhizobia inside root nodules promotes nitrogenase activity, it also inhibits the catabolism of succinate through the TCA cycle. This is because increases in NADH and NADPH levels inhibit key enzymes of the TCA cycle including citrate synthase, isocitrate dehydrogenase, and 2‐oxoglutarate dehydrogenase, a process termed redox inhibition (Dunn, 1998 ; Prell & Poole, 2006 ). Thus, the TCA cycle probably operates below its full aerobic potential. Furthermore, if the metabolism of symbiotic nitrogen‐fixing bacteria is based exclusively on the provision of succinate, then the bacterial nitrogen requirement must be covered solely by the nitrogenase reaction. However, nitrogen‐fixing root‐nodule bacteria (termed bacteroids) do not self‐assimilate but rather secrete large quantities of ammonium (Bergersen & Turner, 1967 ; Brown & Dilworth, 1975 ; Udvardi & Poole, 2013 ) suggesting that the plant provides the bacteroids with a nitrogen‐containing nutrient to cover their nitrogen needs. Finally, the degradation product of succinate through a fully operational TCA cycle is carbon dioxide. However, it has been reported that nitrogen‐fixing bacteroids also secrete the amino acids alanine and aspartate (Kretovich et al , 1986 ; Waters et al , 1998 ; Allaway et al , 2000 ) suggesting a partially operating TCA cycle to yield alanine or aspartate. Based on metabolic considerations of inefficient TCA cycle operation under microaerobic conditions, the inability of bacteroids to self‐assimilate nitrogen, and evidence for secretion of alanine or aspartate by nitrogen‐fixing bacteroids, we postulated a nitrogen‐containing nutrient that is plant‐provided in addition to dicarboxylic acids. Since the plant must provide the N‐containing compound in sufficient quantities, we reasoned that an amino acid might be a likely candidate. Based on the finding that nitrogen‐fixing bacteroids utilize succinate and secrete the amino acids alanine and aspartate (Kretovich et al , 1986 ; Waters et al , 1998 ; Allaway et al , 2000 ; Day et al , 2001 ), we concluded that the plant‐provided compound must comprise at least two nitrogen atoms to enable two consecutive transamination reactions. The first nitrogen is used for transamination of the ketoacid derived from succinate while the second nitrogen atom is utilized for transamination of the ketoacid derived from the plant‐provided compound. Six out of the twenty natural amino acids (Arg, His, Lys, Gln, Asn, and Trp) contain two or more nitrogen atoms and thus are likely candidates. Thereof, His, Lys, and Gln can be excluded because their degradation involves a compulsory 2‐oxoglutarate dehydrogenase step, which is subjected to redox inhibition and disfavored under microoxic conditions (Salminen & Streeter, 1990 , 1992 ). Furthermore, we also excluded Trp and Asn because their catabolism enters the TCA cycle at the level of pyruvate and oxaloacetate, respectively, which limits energy metabolism within a partially operating TCA cycle. Based on these theoretical considerations, we postulated that the remaining amino acid arginine is a likely candidate for the nitrogen‐containing compound provided upon symbiosis. Here, we report on the CATCH‐N cycle based on the co‐catabolism of plant‐provided arginine and succinate as part of a specific metabolic network to sustain symbiotic nitrogen fixation as a synergistic interaction. Using 13 C and 15 N isotope tracing experiment in Bradyrhizobium diazoefficiens in conjunction with in planta transposon‐sequencing analyses and enzymatic reaction network characterization in Sinorhizobium meliloti , we uncovered the principle of the metabolic inter‐species interaction leading to the nitrogen‐fixing symbiosis between plants and bacteria. Collectively, we demonstrate that the CATCH‐N metabolism is governed by highly redundant functions comprised of at least 10 transporter systems and 23 enzymatic functions. In sum, our systems‐level findings provide the theoretical framework and enzymatic blueprint for the optimization and redesign of improved symbiotic nitrogen‐fixing organisms.",
"discussion": "Discussion Here, we report on the CATCH‐N cycle operating on the provision of the two substrates arginine and succinate by the plant as part of a specific metabolic network that drives the process of symbiotic nitrogen fixation in rhizobia. The CATCH‐N cycle shares aspects with the plant mitochondrial arginine degradation pathway (Polacco et al , 2013 ; Winter et al , 2015 ) used to liberate ammonium upon germination; however, it delivers up to 25% higher yield in nitrogen in the form of two alanines and three ammonium secreted for each co‐fed arginine and succinate. Thus, from the plant's perspective, the CATCH‐N cycle multiplies the nitrogen releasing capacity of arginine. On the level of bacteroids, the CATCH‐N cycle provides an elegant metabolic solution for maintaining an active respiratory chain under the highly acidic and microoxic conditions present within the lysosomal compartment of the symbiosome, which is in agreement with the symbiosome‐as‐phagolysosome hypothesis stated by Mellor and colleagues (Mellor, 1989 ). In this model, nitrogen fixation is essential for the bacteria, which otherwise have no use for the ammonium that the plant finds so valuable (Udvardi & Kahn, 1993 ). It addresses the question whether bacteria can opt‐out of nitrogen fixation, which might allow cheaters to prosper. If neutralizing acid is what nitrogen fixation does for the bacteria, bacteria that do not fix N 2 will be killed as a direct consequence of their lack of cooperation, turning this mutualism into a kind of extortion. If proton consumption is the bacteroid's goal in fixing nitrogen, the plant could control nitrogen fixation by controlling input of carbon through the provision of arginine and dicarboxylate but also through its willingness to accept aspartate or alanine as products from the symbiosome and might control the efficiency of the bacterial neutralization in this way. The existence of different alternatives may be linked to the variation of symbiotic effectiveness. Thus, the CATCH‐N cycle also functions as an effective mechanism to promote the survival of bacteroids within infected plant cells. Equimolar arginine and succinate serve as substrates, and a molar ratio of nitrogen to the oxygen of 1:4 is required to operate the CATCH‐N cycle. Therefore, nitrogen fixation still depends on oxygen as terminal acceptor, while harnessing elementary nitrogen as the second electron acceptor for reducing equivalents generated by the metabolism. Also, the CATCH‐N cycle requires a constant flux of 8 protons into the symbiosome to maintain the pH balance of the reaction. These protons must be translocated by the action of plant ATPases as part of the symbiosome acidification process. Thus, the operation of the CATCH‐N cycle depends on the presence of an active plant metabolism. From the nitrogen balance standpoint, a feedback loop exists between the nitrogenase function of bacteroids and the availability of arginine within the host plant. Ammonium released by bacteroids is rapidly incorporated by plant cells into glutamate, glutamine, and aspartate that all serve as precursors for the biosynthesis of ornithine and subsequently for arginine occurring within chloroplasts. The output of the CATCH‐N cycle results in a net gain of assimilated nitrogen that subsequently amplifies the plant's arginine biosynthesis capacity as part of a positive feedback mechanism. As humanity faces global challenges with population growth and climate change, we need to rethink how tomorrow's agriculture will look like. Thereby, systems‐biology approaches to broaden our understanding of plant–microbe interactions, as well as the design of synthetic nitrogen‐fixing microbes that mimic natural symbiosis with plants, hold significant promise. Our integrated model of the CATCH‐N cycle provides new insights into the principles underlying legume symbiosis and comprises an important stepping stone for the rational biotechnological engineering of artificial nitrogen‐fixing microbes and improved crop plants to ensure food and climate security."
} | 3,213 |
24646570 | null | s2 | 1,481 | {
"abstract": "Microbes have long been adapted for the biosynthetic production of useful compounds. There is increasing demand for the rapid and cheap microbial production of diverse molecules in an industrial setting. Microbes can now be designed and engineered for a particular biosynthetic purpose, thanks to recent developments in genome sequencing, metabolic engineering, and synthetic biology. Advanced tools exist for the genetic manipulation of microbes to create novel metabolic circuits, making new products accessible. Metabolic processes can be optimized to increase yield and balance pathway flux. Progress is being made towards the design and creation of fully synthetic microbes for biosynthetic purposes. Together, these emerging technologies will facilitate the production of designer microbes for biosynthesis."
} | 203 |
34218505 | PMC8518482 | pmc | 1,482 | {
"abstract": "Abstract Morphological and phenological traits are key determinants of the structure of mutualistic networks. Both traits create forbidden links, but phenological traits can also decouple interaction in time. While such difference likely affects the indirect effects among species and consequently network persistence, it remains overlooked. Here, using a dynamic model, we show that networks structured by phenology favour facilitation over competition within guilds of pollinators and plants, thereby increasing network persistence, while the contrary holds for networks structured by morphology. We further show that such buffering of competition by phenological traits mostly beneficiate to specialists, the most vulnerable species otherwise, which propagate the most positive effects within guilds and promote nestedness. Our results indicate that beyond trophic mismatch, phenological shifts such as those induced by climate change are likely to affect indirect effects within mutualistic assemblages, with consequences for biodiversity.",
"introduction": "INTRODUCTION For a century, the mechanisms that promote species coexistence in nature have fascinated the biologists, as the pervasive competitive interactions among species are expected to drive species exclusion and to limit coexistence (Gause, 1934 ; Volterra, 1928 ). This question becomes even more intriguing when dealing with complex systems because theoretical works have shown that the stability of a natural community should decrease with the number of species it contains and with the number of interactions among them (Gardner & Ashby, 1970 ; May, 1972 ). So far, this historical issue has been addressed by studying how the structure of ecological networks, either food webs or mutualistic networks, determine species coexistence and community stability (Bastolla et al., 2009 ; Montoya et al., 2006 ; Neutel et al., 2002 , 2007 ; Okuyama & Holland, 2008 ; Otto et al., 2007 ; Thébault & Fontaine, 2010 ). However, the consequences of the species traits that shape the structure of these networks have been seldom considered in this context, despite the growing empirical evidence that traits, such as species phenology, are key for understanding the temporal dynamics of networks (CaraDonna et al., 2021 ). Recent findings have highlighted that ecological networks are structured by multiple species traits, such as, pollination webs, flower shape and the length of the feeding apparatus of pollinators (Junker et al., 2013 ; Stang et al., 2006 ), flowering and flying phenology (Gonzalez & Loiselle, 2016 ; Junker et al., 2013 ), floral height (Junker et al., 2013 ) or floral scent (Schiestl, 2010 ). Even if all these traits can play a similar structural role on overall network structure, for instance by promoting nestedness (Encinas‐Viso et al., 2012 ; Junker et al., 2013 ; Santamaría & Rodríguez‐Gironés, 2007 ), they do not structure interactions with the same mechanisms, potentially affecting species coexistence. While some species traits, such as morphological traits, decrease competition only by defining forbidden interactions among species with different traits, other kinds of species traits, such as phenological traits, can also decrease competition by decoupling interactions in time. This fundamental difference between the two types of traits implies that the latter trait type can allow species from the same guild with distinctive trait values to interact indirectly, as they can share interaction partners at different times, whereas the former type of trait does not allow species to share interaction partners as soon as they differ in their traits. Such a difference for indirect interactions between morphological and phenological traits is likely to have important consequences as indirect effects are known to play a fundamental ecological and evolutionary role, as shown in food webs (Montoya et al., 2009 ; Salas & Borrett, 2011 ) and mutualistic networks (Guimarães et al., 2017 ; Pires et al., 2020 ). However, whether indirect effects among species depend on the type of traits that shape interaction networks remains unexplored and so do the consequences for species coexistence. Contrasting effects of morphological and phenological traits might be especially important in pollination networks because the coexistence of mutualistic networks is expected to strongly depend on the relative importance of indirect competition and indirect facilitation within guilds, either plant or pollinator. Indeed, Bastolla et al., ( 2009 ) showed that the nestedness of mutualistic networks increases network persistence by minimising competition while preserving facilitation. In the case of interactions structured by morphological traits, the absence of competition between two pollinators, or plants, is expected to be coupled with the absence of indirect facilitation between these pollinators, or plants, because the species involved do not share mutualistic partners (Figure 1 ). In contrast, when interactions are structured by phenological traits, they can be decoupled in time thus removing competition but maintaining facilitation between the two pollinators, as they can still share the same mutualistic partners (Figure 1 ). From the schematic example presented in Figure 1 , we expect that a network mainly structured by phenological traits buffers competition but maintains facilitation within plant and pollinator guilds, contrary to a network structured by morphological traits. We thus hypothesise that in plant–pollinator networks, differences in phenological traits among species might promote greater coexistence than species differences in morphological traits because phenology differences might increase the relative importance of facilitation over competition among plants and pollinators. FIGURE 1 Schematic pollination networks with no structuring trait (left), structured by a morphological trait (center) or by a phenological trait (right). Links between pollinators and plants represent mutualistic interactions (+/+) whereas indirect effects within the pollinator guild are represented by dashed arrows. Gaussians represent the distributions of the values of the morphological trait or the flowering/flight periods for plants and pollinators, and the overlap among them (colored area) represents the interaction strength Here we test this hypothesis and quantify how phenological and morphological traits affect the relative strength of competition and facilitation and the persistence of plant–pollinator networks. To do so, we develop a dynamic model of pollination networks including intra‐guild competition for access to mutualistic partners and measure direct and indirect effects among species over all possible paths in the networks. Our results reveal that niche partitioning due to the phenological and morphological traits, henceforth phenological and morphological forcing, respectively, strongly differ in their consequences on pollination network structure and persistence when there is intra‐guild competition.",
"discussion": "DISCUSSION Our results show that the structuring effect of phenological traits on plant–pollinator interactions dampens the negative effects of competition for mutualistic partners on species persistence, leading to greater diversity and network nestedness than when interactions are structured by morphological traits. As hypothesised, we find that these two types of traits affect indirect effects in two very distinct ways: while the structure imposed by morphological traits decreases both competition and positive indirect effects among species from the same guild, the structure imposed by phenological traits increases competition and positive indirect effects among species from the same guild. Most importantly, once differences in network nestedness and diversity are accounted for, we show that phenological traits lead to a less negative, or more positive, balance between competition and positive indirect effects within guilds at equilibrium than morphological traits. Since indirect effect estimation is based on a linear approximation around the equilibrium state, we cannot estimate indirect effects during the transient dynamics leading to the equilibrium, which prevents us from properly assessing if they are a cause or a consequence of network persistence. However, there is no difference in persistence between networks structured by phenology (i.e. with a phenological forcing) and networks structured by morphology (i.e. with a morphological forcing) when intra‐guild competition is null. This suggests that the positive effect of phenological forcing on persistence results from changes in net effects of competition and facilitation between species from the same guild. Taken together, our results show that the type of species traits shaping interactions in mutualistic networks affects species coexistence, by altering the balance between competition and facilitation among species from the same guild. The benefits of phenological traits mainly occur because they decouple interactions in time, making the balance between facilitation and competition less negative than morphological traits. Our results provide a mechanism that might explain the importance of phenological traits relatively to morphological traits in seasonal pollination networks (CaraDonna et al., 2017 ; Gonzalez & Loiselle, 2016 ; Manincor et al., 2020 ; Sonne et al., 2020 ) and suggest that the seasonal structure is key to the maintenance of diversity in mutualistic communities. These findings can be generalised to other traits than phenology, while they allow to decouple interactions in time or space without leading to resource depletion. Indeed, any trait following this assumption and decoupling interactions in time or space, as for example traits associated with daytime activity and flower opening (e.g. diurnal vs. night) or with flight and flower heights, should similarly maintain indirect facilitation within guilds and promote species coexistence. For instance, differences in flight and flower heights could allow two pollinator species that fly at different heights to avoid competition while still interacting with the same plant population, or even with the same individual plant if an individual plant has flowers at different heights. As plant–pollinator interactions have been shown to differ at small spatial and temporal scales (Albrecht et al., 2012 ; Cusser et al., 2021 ; Knop et al., 2017 ), we expect that the mechanisms highlighted in this study are widespread in pollination networks. Competition is known as an important evolutionary and ecological driver of plant–pollinator interactions (Bartomeus et al., 2021 ; Jones et al., 2012 ; Levin & Anderson, 1970 ) because plants or pollinators strongly compete among them to access their mutualistic partners in pollination networks (Campbell, 1985 ; Henry & Rodet, 2018 ; Pleasants, 1980 ). However, most of the theoretical studies tackling this point modelled competition independently from plant to pollinator interactions (Bastolla et al., 2009 ; Pascual‐García & Bastolla, 2017 ), while few other studies suggest that accounting for the seasonal structure in competition increases network persistence (Encinas‐Viso et al., 2012 ; Rudolf, 2019 ). Our modelling approach allows structuring competition depending on the sharing of mutualistic partners in time. Doing so, we show that competition is a major driver of the persistence of plant–pollinator networks and that the differential effects of phenological and morphological traits depend on the presence of competition, our scenario with no competition being a null expectation or a control. Further, we also show that when competition is present, the structuring effect of phenological traits favours positive indirect effects within guilds, that is facilitation, thereby maintaining diversity. Such effect not only comes from indirect effects among species sharing mutualistic partners, that is paths of length two, but also from indirect effects among species from the same guild over longer paths as our calculation includes all possible paths. Furthermore, our results highlight that the persistence of specialist species is key to understand the structuring effects of phenology at equilibrium. As revealed by Saavedra et al., ( 2011 ), we found that specialists are the species that promote the most the nestedness of networks at equilibrium, as they create heterogeneity in degree distribution (Bascompte et al., 2003 ), but they are also the most vulnerable species. Including a seasonal structure better protects specialist species from extinction, which provides new insights on mechanisms that could maintain those vulnerable species in networks. Consequently, we find that phenological forcing increases nestedness much more than morphological forcing, which is expected to increase the resilience and the robustness of the networks to perturbations (Memmott et al., 2007 ; Thébault & Fontaine, 2010 ). Moreover, we find that specialist species propagate more positive indirect effects to other species relative to their direct competitive effects than generalists. Thus, in addition to promoting positive indirect effects within guilds by decoupling interactions in time, the structuring effect of phenology protects species that have a lower negative balance between positive indirect effects and competitive effects, thereby tilting the balance even more towards facilitation rather than competition. Recent studies showed that climate warming is shifting pollinator flight periods and flowering periods, leading to changes in the seasonal structure of pollinator and plant assemblages, which either increase or decrease phenological overlaps among species depending on the location (Diez et al., 2012 ; Duchenne et al., 2020 ; Theobald et al., 2017 ). Such changes are likely to cause mismatches among interacting species (Memmott et al., 2007 ; Revilla et al., 2015 ) and to decrease the robustness of the network to any other perturbation, thus leading to synergistic effects among perturbations (Revilla et al., 2015 ). Beyond and more insidiously than trophic mismatch, our results highlight that phenological shifts are likely to affect indirect effects such as competition pressures (Rudolf, 2019 ) and facilitation in mutualistic assemblages, with currently unknown consequences for biodiversity. Further, since specialist species are often presented as generally more sensitive to perturbation (Clavel et al., 2011 ) and we showed that they tend to propagate more positive indirect effects than other species, our results suggest that perturbations targeting specialists are likely to increase the propagation of negative indirect effects. However, network reorganisation following perturbations can also happen (Burkle et al., 2013 ) and might buffer the effects of specialist extinctions. Our results are theoretical and focus on ecological dynamics only. Since in diverse communities competition can constrain species’ evolutionary trajectories (Mazancourt et al., 2008 ), it is likely that ecological and evolutionary equilibrium differ. An interesting perspective would be to investigate the consequences of eco‐evolutionary dynamics of morphological and phenological traits on the competition–facilitation balance and related network persistence. In addition, future steps would be to estimate the real benefits of empirical seasonal structures on coexistence. To do so, the challenge is not only to assess the relative importance of phenological and morphological overlaps among species within ecological networks (CaraDonna et al., 2017 ; Sonne et al., 2020 ) but also to solve the complex ‘inverse problem’ to parametrise models using empirical seasonal and morphological structures (Tarantola, 2005 )."
} | 3,967 |
38260425 | PMC10802270 | pmc | 1,483 | {
"abstract": "Stony coral tissue loss disease (SCTLD) has devastated coral reefs off the coast of Florida and continues to spread throughout the Caribbean. Although a number of bacterial taxa have consistently been associated with SCTLD, no pathogen has been definitively implicated in the etiology of SCTLD. Previous studies have predominantly focused on the prokaryotic community through 16S rRNA sequencing of healthy and affected tissues. Here, we provide a different analytical approach by applying a bioinformatics pipeline to publicly available metagenomic sequencing samples of SCTLD lesions and healthy tissues from four stony coral species. To compensate for the lack of coral reference genomes, we used data from apparently healthy coral samples to approximate a host genome and healthy microbiome reference. These reads were then used as a reference to which we matched and removed reads from diseased lesion tissue samples, and the remaining reads associated only with disease lesions were taxonomically classified at the DNA and protein levels. For DNA classifications, we used a pathogen identification protocol originally designed to identify pathogens in human tissue samples, and for protein classifications, we used a fast protein sequence aligner. To assess the utility of our pipeline, a species-level analysis of a candidate genus, Vibrio , was used to demonstrate the pipeline’s effectiveness. Our approach revealed both complementary and unique coral microbiome members compared to a prior metagenome analysis of the same dataset.",
"conclusion": "Conclusions In all, we provide a novel pipeline to understand coral disease, and further investigate the role of bacteria and DNA viruses in SCTLD. Our new pipeline found both incongruities and parallels with the original analysis of these data. For instance, both studies revealed a prevalence of Rhodobacteraceae and Flavobacteriaceae in lesion samples. One of the major differences between the studies was the dominance of Synechococcus and Vibrio in this study, which was likely driven by the use of k-mers to quantify taxonomic abundance (as the protein analysis with read counts aligned more with previous results). In addition to providing novel insights, our pipeline also reduced the computational load for downstream analysis, making large metagenomic analysis more accessible to those with less computational resources. This method is useful not only for coral scientists, but also for fields that study non-model organisms and lack comprehensive genomic resources. In corals, the lack of such resources can impede the progress of metagenomic analysis and the identification of potential pathogens. Specifically, host-associated metagenomes without a reference genome can dominate and confound metagenomic results ( Rosales et al. 2022 ). As genome assemblies are unlikely to be available anytime soon for the over 30 species of coral affected by SCTLD, this pipeline offers a useful tool for the simultaneous examination of viruses, bacteria, and eukaryotic species living on or within the host tissue that may be associated with infection for any coral species or non-model organism.",
"introduction": "Introduction Stony coral tissue loss disease (SCTLD) was discovered off the coast of Miami, FL in 2014 and has since had negative consequences on the function of coral reefs across Florida and the Caribbean ( Walton et al. 2018 ; Alvarez-Filip et al. 2022 ). To date, despite many efforts, no pathogen has been definitively identified as the causative agent of SCTLD. The stony coral (order Scleractinia) microbiome is a complex system of interactions between the host, bacteria, viruses, fungi, archaea, and algal symbionts ( Bourne et al. 2009 ); thus a disturbance in any number of these symbiotic relationships could be involved in SCTLD progression. Multiple studies have explored viruses that may infect stony coral symbionts, notably Symbiodiniaceae , but no causative relationships have been detected ( Work et al. 2021 ; Veglia et al. 2022 ; Beavers et al. 2023 ; Howe-Kerr et al. 2023 ). Bacterial species are particularly under scrutiny for their potential involvement in SCTLD, due to the effectiveness of antibiotics in halting lesion progression in multiple affected coral species ( Aeby et al . 2019 , Neely et al. 2020 ; Shilling et al. 2021 ; Studivan et al. 2023 ). Consequently, SCTLD studies have predominantly focused on understanding changes in the bacterial community between apparently healthy and SCTLD-affected corals. Studies to identify bacterial pathogens have relied primarily on small subunit 16S ribosomal RNA (rRNA) sequencing, followed by computational analysis ( Callahan et al. 2016 ) typically using the Silva database ( Quast et al. 2013 ) to assign and classify Amplicon Sequence Variants (ASVs) into taxa. ASVs found in diseased lesion samples are then compared to samples from apparently healthy colonies to determine which ASVs are associated with the tissue loss lesions ( Meyer et al. 2019 ; Rosales et al. 2020 ; Clark et al. 2021 ). These methods have characterized many notable shifts in coral bacterial communities due to SCTLD and several bacterial taxa associated with SCTLD lesions, including Rhizobiales , Clostridiales , Peptostreptococcales-Tissierellales , Rhodobacteraceae , Flavobacteriaceae , and Vibrionaceae ( Rosales et al. 2023 ). However, because of the difficulty in determining whether an associated bacterial taxon is a harmless commensal, an opportunistic secondary infection, or the primary pathogen, none of the bacterial taxa associated with SCTLD have been identified as the causative agent. An alternative approach to understanding disease dynamics is the use of metagenomic sequencing, in which all of the DNA from a source is sequenced, including not only the host, but also viruses, bacteria, and eukaryotic species living on or within the host tissue. For example, by analyzing metagenomic sequencing data of human tissue samples taken from the site of infection, researchers have identified pathogenic agents in brain infections ( Salzberg et al. 2016 ; Wilson et al. 2019 ), corneal infections ( Eberhart et al. 2017 ), and other diseases ( Kostic et al. 2012 ). The sensitivity of this approach relies on first, sequencing the source DNA deeply enough to capture the pathogen of interest, and second, the existence of genome assemblies closely related to the pathogen in public databases. While the number of complete genomes has grown enormously over the past two decades, databases still contain few or no genomes for non-model organisms, including scleractinian corals. Currently, data from only one SCTLD metagenome study is publicly available. While the authors of that study ( Rosales et al. 2022 ) were able to assemble and annotate genomes for SCTLD-associated bacterial taxa such as Rhodobacterales , Rhizobiales , and Flavobacteriales , the results were focused on only five of the twenty diseased lesion tissue samples, all from the same coral species ( Stephanocoenia intersepta ), because the majority of samples were dominated by host sequences. In metagenomic studies, host sequences can confound results, so they are typically removed by aligning all reads to a host reference genome ( Gihawi et al. 2023 ; Lu et al. 2022 ). Currently, the GenBank database has 53 genome assemblies from scleractinian corals, of which only seven are at the chromosome level ( NCBI 2023 ). Of these 53 genomes, none are from the species of corals previously investigated for SCTLD ( Rosales et al. 2022 ), emphasizing the additional challenges associated with using metagenomics in non-model organisms. Additionally, given the complex symbiotic microbiome (i.e., algal symbiont, viruses, and prokaryotic community) of stony corals ( Bourne et al. 2009 ), the host DNA is only one of the hurdles. In this study, we applied new classification methods to understand this devastating coral disease. We used a method to filter host reads from metagenome data by using data collected from apparently healthy corals of the same species to approximate a species-specific healthy host coral genome and microbiome. We then applied the Kraken software suite for pathogen identification ( Lu et al. 2022 ) using KrakenUniq ( Breitwieser et al. 2018 ) to identify putative pathogens present in diseased samples and not present in healthy ones. Using these methods, we identified a number of taxa that have previously been associated with SCTLD, providing further support for their involvement in SCTLD pathogenesis. Finally, from the pool of bacterial taxa we identified as associated with SCTLD lesions, we selected a candidate genus, Vibrio , with which to explore the utility of our pipeline at a finer taxonomic level.",
"discussion": "Discussion In this study, we used previously published sequencing data from coral affected by SCTLD and developed a novel metagenomic analysis pipeline to explore the microbial communities present in those data. The data consisted of samples from four coral species collected from Florida’s coral reefs during a SCTLD outbreak. To investigate the microbial taxonomy of these samples, the previous study used small subunit rRNA gene assemblies and metagenome-assembled genomes (MAGs). Our investigation differed by using the Kraken software suite and focusing on unique k-mer count data to understand abundances. By predominantly working with k-mers instead of MAGs, we maximized the utility of the read data, as our approach allowed us to capture reads that may have been discarded during MAG assembly and binning. Samples with high intrapopulation diversity can provide challenges in MAG assembly and binning ( Ramos-Barbero et al. 2019 ; Meziti et al. 2021 ), which may have hindered the generation of a MAG bin in the previous analysis. The previous work also did not filter out host sequences because reference genomes do not exist for the sampled coral species. However, here we applied a novel technique to filter these reads by using data derived from apparently healthy (AH) samples as a surrogate for reference genomes. This allowed us to examine unique sequences from DL samples by approximating a species-specific host coral genome and reduce computational load. In addition, in this study, we investigated the SCTLD DNA virome, which has not been previously reported. The families Rhodobacteraceae and Flavobacteriaceae were found to be associated with SCTLD in our protein analysis, consistent with other SCTLD studies (reviewed in Papke et al. 2024 ), including previous examinations of the same sample set ( Rosales et al. 2020 , 2022 ). Rhodobacteraceae is one of the most common bacterial families associated with coral diseases ( Gignoux-Wolfsohn et al. 2017 ) in diverse geographic locations, but no member has been identified as a causative coral disease agent ( Mouchka et al. 2010 ), and members of this group are broadly found across ocean habitats and fulfill diverse ecological functions ( Brinkhoff et al. 2008 ). Flavobacteriaceae has been enriched in White Band Disease in the Scleractinia staghorn coral Acropora cervicornis ( Gignoux-Wolfsohn and Vollmer 2015 ), but has never been identified as a causative agent in coral tissue loss. In SCTLD, Flavobacteriaceae has previously been found enriched in unaffected (non-lesion) tissue from diseased colonies, potentially indicating colony stress and initial dysbiosis due to disease ( Rosales et al. 2023 ). In contrast to the protein analysis, which used read count data, our DNA-level analysis using k-mers did not detect a singular genus belonging to Rhodobacteraceae or Flavobacteriaceae in high proportions among the unique diseased reads, providing support for the idea that members of these two families are likely a diverse set of bacteria species that associate with SCTLD opportunistically. In our DNA-level analysis, Vibrionaceae and Synechococcaceae were among the most abundant families within the unique DL reads and these high abundances showcase how the k-mer based abundance approach can provide a different perspective of highly abundant taxa in particular samples. These families were not observed as highly abundant in the MMSeqs2 protein analysis; however, we were primarily interested in whether new candidates emerged from the MMSeqs2 analysis, not the relative proportions of the candidates. Interestingly, the genera Synechococcus and Vibrio were not detected in the previous analysis of these data ( Rosales et al. 2022 ). In the 16S rRNA metabarcoding study of these same samples, Synechococcus and Vibrio were differentially abundant in M. meandrites samples ( Rosales et al. 2020 ). These results did not show a clear trend as Synechococcus was present across multiple AH samples and Vibrio was not prevalent across DL coral samples. In this study, Synechococcus was the most or second most proportionally abundant genus across all four coral species. Synechococcus belong to phylum Cyanobacteria , which are photosynthetic picoplankton ( Kim et al. 2018 ), and typically not involved in pathogenesis. Even so, Synechococcus have been enriched in other SCTLD studies that compared healthy colonies and healthy tissue on diseased colonies, and it was hypothesized that their increase in abundance is a response to disease stress ( Rosales et al. 2023 ). The high proportion of Synechococcus in this study supports the suggestion that Synechococcus may have some role in microbial community interactions during SCTLD. Vibrio also have been associated with SCTLD (reviewed in Papke et al. 2024 ), including in these samples ( Rosales et al. 2020 ), but in contrast to Synechococcus , Vibrio have been associated with other coral tissue loss diseases, such as white pox disease ( Kemp et al. 2018 ) and yellow band disease ( Cervino et al. 2008 ), and diseases in other marine organisms, such as sponges ( Dincturk et al. 2023 ) and crustaceans ( de Souza Valente and Wan 2021 ). In three coral species analyzed here, Vibrio was one of the two most abundant genera, suggesting it may be associated with SCTLD tissue loss. In the fourth species, M. meandrites, Vibrio was only the sixth most abundant genus. However, M. meandrites had the fewest AH reads to create the database used for filtering the DL reads ( Table 1 ), which may reduce the effectiveness of our novel pipeline that relies on reads from control groups. This was also a surprising result, since as mentioned, Vibrio were abundant in the 16S rRNA M. meandrites analysis of these samples ( Rosales et al. , 2020 ). When using our pipeline to explore a species-level analysis within the candidate genus Vibrio , we found read matches to V. mediterranei/shilonii ( Tarazona et al. 2014 ), V. coralliilyticus , V. harveyi , and V. owensii – all known coral pathogens associated with bleaching and tissue loss ( Kushmaro et al. 1997 ; Ben-Haim et al. 2003 ; Luna et al. 2007 ; Ushijima et al. 2012 ; Munn 2015 ). V. mediterranei/shilonii , which comprised a majority of the classified Vibrio species in D. labyrinthiformis and D. stokesii, was previously found to be responsible for the annual bleaching of the scleractinian coral Oculina patagonica off the Israeli coast from 1993 ( Kushmaro et al. 1996 , 1997 ) to 2003 ( Reshef et al. 2006 ). Additionally, when V. mediterranei / shilonii is experimentally introduced together with V. coralliilyticus to healthy corals, they appear to have synergistic pathogenic effects ( Rubio-Portillo et al. 2014 ). Given these associations, V. mediterranei / shilonii , may be of particular interest in future SCTLD studies as potentially increasing the virulence of SCTLD, as does V. coralliilyticus ( Ushijima et al. 2020 ), which also appeared in our samples. However, the very low similarity between the cultured V. coralliilyticus sequences in Ushijima et al. (2020) and our Vibrio sequences leads us to believe that multiple Vibrio species may be involved in SCTLD lesion development. It is important to note that our picture of the SCTLD microbiome is restricted by the genomes in the databases used. The Vibrio genus has been found to have a large degree of genetic plasticity between species ( Gu et al. 2009 ), so while the classifications to different Vibrio species may truly represent the presence of an array of Vibrio species, it may instead be the result of various reads from a novel Vibrio species matching different Vibrio species based on closest genomic similarity. Therefore, while matches to different Vibrio species and their potential role in SCTLD may offer some insights, a more robust interpretation is to consider the implications of disease association by Vibrio at the genus level. The etiology of SCTLD has been hypothesized to be viral ( Work et al. 2021 ), and gene expression data show that there is an increase in coral viral immune response in corals with SCTLD ( Beavers et al. 2023 ). Researchers have explored the potential role of RNA viruses in SCTLD, but no RNA viruses have been found exclusively in corals with SCTLD ( Veglia et al. 2022 ) and these viruses are likely ubiquitous in corals without any potential relationship to SCTLD ( Howe-Kerr et al. 2023 ). However, the involvement of DNA viruses has not previously been explored. Our data show the majority of DNA viruses in diseased samples represent phages. Not surprisingly, phage sequences correspond with some of the most abundant bacteria identified in this study, such as Rhodobacteraceae, Vibrionaceae, and Synechococcaceae . The Paracoccus phage, which infects Rhodobacteraceae, and the Vibrio phage would be interesting to further explore as potential avenues for disease mitigation. In addition to phages, sequences were found with similarities to the Chrysochromulina ericina virus. However, with only two contigs and little coverage of its genome, we do not believe this virus plays a role in SCTLD. Thus, we did not find any DNA viruses with a definitive association with SCTLD. Future studies may consider viral enrichment protocols prior to sequencing to help better characterize the SCTLD DNA virome. In addition to differences in the bacterial and viral communities, members of Symbiodiniaceae were found to be abundant in DL samples, particularly in M. meandrites . SCTLD disrupts the relationship between the host coral and its Symbiodiniaceae through symbiont necrosis and peripheral nuclear chromatin condensation, among other physiological changes ( Landsberg et al. 2020 ). This may result from an increase in rab7 signaling among the Symbiodiniaceae to degrade dead and dysfunctional cells through endocytic phagosomes ( Beavers et al. 2023 ). The Symbiodiniaceae DNA identified in diseased samples in this study may be a byproduct of this necrosis and degradation of the symbiont. This was especially notable in M. meandrites, which was the coral in this study most susceptible to acute tissue loss and mortality from SCTLD ( Precht et al. 2016 ); this accelerated tissue loss may lead to higher levels of dead and dysfunctional symbionts being produced during visual lesion progression in M. meandrites than in other coral species. Although, this pipeline was initially developed under the one-pathogen-one-disease assumption for humans, our results show that we can identify a consortium of putative pathogens with this method (i.e., Rhodobacteraceae , Flavobacteriaceae , and Vibrio ). However, this pipeline is limited by the amount of prior information about pathogens in the species investigated; for example, human pathogens are well represented in metagenomic classification databases, whereas non-model organism pathogens may not be, which can lead to a higher rate of unclassified or misclassified reads. In addition, disease states are more clearly defined in humans than in non-model organisms, making the assumptions about sample health more reliable in human analyses."
} | 5,031 |
33105722 | PMC7690433 | pmc | 1,484 | {
"abstract": "We demonstrate a nonvolatile memristor based on Co–Al-layered double hydroxide (Co–Al LDH). We also introduce a memristor that has a hexazinone-adsorbing Co–Al LDH composite active layer. Memristor characteristics could be modulated by adsorbing hexazinone with Co–Al LDHs in the active layer. While different, Co–Al LDH-based memory devices show gradual current changes, and the memory device with small molecules of adsorbed hexazinone undergo abrupt changes. Both devices demonstrate programmable memory peculiarities. In particular, both memristors show rewritable resistive switching with electrical bistability (>10 5 s). This research manifests the promising potential of 2D nanocomposite materials for adsorbing electroactive small molecules and rectifying resistive switching properties for memristors, paving a way for design of promising 2D nanocomposite memristors for advanced device applications.",
"conclusion": "4. Conclusions In conclusion, on the basis of drop coating and impregnation methods to fabricate Co–Al LDHs and Co–Al LDH-adsorbed hexazinone-based memristors, the memory behavior of memristors rectified by adsorbed hexazinone was observed. Impressively, Co–Al LDH-based memory devices exhibit gradual changes in current, while the memory device based on adsorbed hexazinone small molecules shows abrupt changes in current. Both memristors show rewritable resistive switching with superior electrical bistability (>10 5 s). The strategies implemented here, including the 2D nanocomposite material that adsorbed electroactive small molecules for adjusting the resistive switching property of hybrid nanomaterials, could principally be extended to other systems of adsorbing electroactive small molecules and offer unique solutions for adjusting the resistive switching behavior of 2D hybrid nanomaterials.",
"introduction": "1. Introduction For the past few years, more and more attention has been paid to memristors in the field of electronic devices [ 1 , 2 , 3 , 4 , 5 , 6 , 7 , 8 , 9 , 10 , 11 , 12 ]. They have been investigated extensively for commercialization on account of their quick response and data processing, and low energy consumption. Typical features of memristors are their two-terminal sandwich structure and resistive switching nonvolatile memories. According to the way the current increases, memristors can be roughly classified as digital or analog. The characteristic feature of a digital memristor is that resistive switching is abrupt, and it has been widely studied for its simple structure, easy three-dimensional stacking, and fast resistive switching speed [ 13 , 14 , 15 , 16 , 17 , 18 ]. For the last few years, analog memristors are typically characterized by gradual resistive switching, which has attracted more and more attention due to the application value of brain-like morphology calculations [ 19 , 20 ], programmable analog circuits [ 21 ], and the like. In particular, analog memristors present features with great promise in developing artificial synapses to implement brain-inspired neuromorphic computing. Broadly speaking, it is generally accepted by scholars that the mechanism of resistive switching is the formation and rupture of conductive filaments. A large number of filamentary memristors exhibit a sudden set or reset process, on behalf of their digital type. The opposite of that type is analog-type memristors [ 22 ]. At present, the mechanism of analog-type memristors is still unclear and controversial. Since the memristor has great application potential in the electronics industry, much research is being devoted to commercialization of usable electronic components, which are largely concentrated in investigating different active layer materials [ 23 , 24 , 25 ]. According to previous reports, a variety of active layer materials have been studied, for example, binary oxides, organic polymers, and biomaterials [ 26 , 27 , 28 ]. In addition, in order to continuously improve the performance of memristor devices, researchers have used a variety of engineering methods to improve device performance, such as illumination effect [ 29 , 30 ], ultraviolet light exposure [ 31 ], multilevel data storage, sub-quantum filamentation, retention modeling by conductive bridge random access memory [ 31 , 32 , 33 , 34 ], defect control [ 35 ], atomic-level control with an optimized high temperature forming scheme [ 36 ], and so on. Therefore, development of new materials for study of memristors and control or adjustment of device performance is of great significance for development of nonvolatile memory. Recently, Co–Al-layered double hydroxide (Co–Al LDH) has attracted more and more attention for 2D composite materials on account of its big specific surface area and short carrier transport diffusion length [ 37 , 38 ]. The higher specific surface area makes it have a better adsorption capacity. Simultaneously, the hexazinone molecule contains an electroactive C=C–C=N group, which would bring about formation of an ingredient bearing two reducible centers; as a consequence, it has good reduction properties [ 39 ]. Given that triazines are electrochemically active, the hexazinone is a famous n-type semiconductor. Furthermore, it could be utilized as the adsorbed component in Co–Al LDH to regulate carrier migration. However, resistive switching biodevices using Co–Al LDH and their adsorption of small hexazinone molecules have not yet been reported. Hence, in this work, the Co–Al LDH-based memristor and Co–Al LDH-adsorbed hexazinone-based memristor were fabricated to study resistive switching behavior and the effect of LDH adsorption of small molecules on their properties.",
"discussion": "3. Results and Discussion Memristors based on Co–Al LDHs or Co–Al LDH-adsorbed hexazinone as resistive switching active layers were fabricated on indium tin oxide (ITO)-coated glass substrates. Al/Co–Al LDH thin film/Al and Al/Co–Al LDH-adsorbed hexazinone thin film/Al structures were used to illustrate the memristors. Figure 1 a depicts the chemical structure diagram of Co–Al LDHs, and Figure 1 b depicts the chemical structure of hexazinone. Co–Al LDHs or Co–Al LDH-adsorbed hexazinone were used as the resistive switching layer in the devices. The schematic device structure of the Co–Al LDH-based memristor is shown in Figure 1 c. The cross section of the resistive switching layer before deposition of top electrode Al was characterized by scanning electron microscopy. Figure 1 d shows the cross section of the resistive switching layer for the Co–Al LDH memristor, and Figure 1 e shows the cross section of the resistive switching layer for the Co–Al LDH-adsorbed hexazinone memristor. FTIR spectra of hexazinone and Co–Al LDHs before and after hexazinone adsorption are shown in Figure 2 . In the FTIR spectrum of hexazinone, as shown in Figure 2 a, the absorption band at 2975 cm −1 corresponds to the stretching vibration of methyl C–H. Absorption bands centered at 2849 cm −1 were caused by the stretching vibration of methylene C–H. Two adjacent bands centered at 1721 and 1635 cm −1 were caused by the stretching vibration of 2-bit C=O and 1-bit C=O, respectively. On the other hand, absorption bands centered at 1554 cm −1 are attributed to the C=N stretching vibration. Absorption bands centered at 1352 cm −1 were caused by C–H deformation vibration of CH 3 O [ 39 ]. The broad strong absorption band centered at 3396 cm −1 originated from the stretching vibrations of surface and interlayer hydroxyl groups [ 42 ], as shown in Figure 2 b. The weaker band at 1648 cm −1 came from the bending mode of water molecules. Bands centered at 589 cm −1 were derived from the stretching and bending vibration of metallic oxygen (M-O) or metallic hydroxyl (M-OH) [ 43 ]. The FTIR spectrum of Co–Al LDHs after adsorption of hexazinone is shown in Figure 2 c. In hexazinone adsorbed on Co–Al LDHs, two bands at 2975 and 2849 cm −1 disappeared, and a new band of hexazinone appeared at 1349 cm −1 , which indicates symmetric vibrations of hexazinone. Resistive switching properties of Co–Al LDHs and the Co–Al LDH-adsorbed hexazinone memory device were accounted for in ambient conditions; for both devices, the tested I–V characteristics exhibited a typical nonvolatile memory property. Electrical characteristics of the Co–Al LDHs memristor were tested under the voltage sweep sequence of 0 V → +6 V, 0 V → +6 V, 0 V → −6 V, and then 0 V → −6 V. When sweeping positive external voltage from 0 to +6 V, the resistance converted gradually from high resistance state (HRS) to low resistance state (LRS). LRS could be maintained during the second positive voltage sweep until an opposite external voltage was applied LRS was gradually converted to HRS, and HRS was maintained during subsequent negative voltage scanning, as shown in Figure 3 a. This indicates that the device has typical bipolar resistive switching behavior. For this memristor, the device maintains in LRS or HRS until an external voltage with opposite polarity is applied, which converts the resistance state to HRS or LRS. Furthermore, subsequent 10 consecutive voltage sweeps are shown in Supplementary Materials Figure S1 . Repeatability of devices is an important prerequisite for their engineering applications. In the experiment, the I–V curves of 49 Co–Al LDHs memristor samples were tested, of which 30 samples had similar resistive properties and the remaining 19 did not. Sample-to-sample I–V characteristics for 30 Co–Al LDH memristors were shown in Figure S2 . Additionally, the I – V plots of the remaining 19 are shown in Figure S3 , as can be seen, and they only have a single resistive state and no resistive switching function. The calculation formula of power is p = I × V , where I represents the current and V represents the voltage. In this device, the current is 0.01 A at most, and the corresponding voltage is 6 V. Therefore, the maximum power consumption of these device is 0.06 watts. Data retention characteristics of the Co–Al LDH memristor were carried out to evaluate the stability of the Co–Al LDHs memristor. Since resistive switching is gradual, we chose different voltages to motivate the device for HRS and LRS with excitation voltages of 0.5, 1, 2, and 4 V, respectively. This means that the current of the device is excited with different maximum voltages, and a reading voltage is used to read the resistance of the device. According to the resistive switching characteristics, the device can be converted to different resistance states when applying either a positive or negative applied voltage, so the excitation voltage can be either positive or negative. A reading constant voltage of 0.1 V was used to read the resistance of device, as shown in Figure 3 b. It shows data retention characteristics of the Co–Al LDHs memristor. It can be seen that each resistance state of the Co–Al LDHs memristor is almost constant and can maintain a clear boundary with the adjacent resistance state. As a representative sample, data retention characteristics of the tenth Co–Al LDHs memristor are shown in Figure S4 . For the I–V curve of 30 Co–Al LDHs memristors in Figure S2 , the cumulative probability for resistance in LRS and HRS under −2 V was collected, as shown in Figure 3 c. The average resistance values required reflect its resistance level in LRS or HRS. Endurance characteristics of the Co–Al LDHs memristor were also carried out, as shown in Figure 3 d. The pulse used had a pulse period and width of 2 and 1 ms, respectively. In order to reflect different resistance states of the Co–Al LDH memristor after excitation pulses of ±0.5, ±1, ±2, and ±4 V, respectively, were used, the 0.1 V reading voltage was used to read the resistance of the device. The Co–Al LDHs memristor was quite stable in every resistance state after 200 cycles. Endurance characteristics of the tenth Co–Al LDHs memristor are shown in Figure S5 . Interestingly, in contrast to the Co–Al LDHs memristor, for the Co–Al LDHs after the hexazinone adsorption-based device, both the set and reset processes were abrupt ( Figure 4 a). In the Co–Al LDH-adsorbed hexazinone memristor, during the first positive voltage sweep from 0 V to the set voltage ( V set ) of 1.65 V the resistance state of the device was converted from its pristine HRS to LRS; this is the “set process” transition. The LRS was maintained during the subsequent positive voltage sweep. When a negative voltage was applied, which converted LRS to HRS at the V reset of −4.05 V (“reset” process), the nonvolatile HRS and LRS were very stable, even when the power was turned off for 10 min or longer. The device could be converted over and over again between HRS and LRS by altering the applied voltage, and the subsequent 10 consecutive voltage sweeps are shown in Figure S6 . In the experiment, the I–V curves of 56 Co–Al LDHs after the hexazinone adsorption-based device were tested, of which 39 samples had similar resistive properties and the remaining 17 did not. Sample-to-sample I–V characteristics for 39 Co–Al LDHs after hexazinone adsorption memristors are shown in Figure S7 . The maximum power consumption of these devices is also 0.6 watts. In practice, the operating voltage can be reduced to reduce power consumption on the basis of ensuring the resistive switching characteristic. Co–Al LDHs after the hexazinone adsorption-based device also exhibited stable data retention characteristics under the reading voltage of −0.1 V, as shown in Figure 4 b, but with a high ON/OFF resistance ratio of 2.5 × 10 4 ; after 10 5 s, the device lost its data retention capability. Data retention characteristics of the tenth Co–Al LDHs after hexazinone adsorption memristors are shown in Figure S8 . Figure 4 c shows the endurance characteristics of the Co–Al LDH-adsorbed hexazinone memristor. For the endurance characteristics test, the set voltage pulse was +5 V, the reset voltage was set as −5 V, each pulse duration was 0.1 ms, and the set and rest pulse were followed by the reading pulse. The reading pulse had a pulse amplitude of 0.1 V. The resistance did not fluctuate significantly for the Co–Al LDH-adsorbed hexazinone memristor in both LRS and HRS during 280 read cycles, and after 280 cycles the resistive switching property was no longer repeated. Additionally, endurance characteristics of the tenth Co–Al LDHs after hexazinone adsorption memristor are shown in Figure S9 . The 10 5 s of data retention and 280 durable cycles of the device are far from meeting the requirements for practical application. Further improvement of data retention characteristics and endurance characteristics of the device is the next step that we need to focus on. For the I–V curve of 39 Co–Al LDHs after hexazinone adsorption memristors in Figure S7 , the cumulative probability for resistance in LRS and HRS were also collected for Co–Al LDHs after the hexazinone adsorption memristor, as shown in Figure 4 d. The cumulative probability for threshold voltage of V set and V reset were also analyzed for Co–Al LDHs after the hexazinone adsorption memristor, as shown in Figure 4 e. The average voltage required to accomplish set and reset operations was stable. The influence of temperature on device performance has always been an important index in engineering applications. Considering that cycloqinone cannot withstand high temperatures, Co–Al LDHs memristor performance was tested at 158 °C, as shown in Figure 5 . Figure 5 a exhibits the I–V curves of the Co–Al LDHs memristor at 158 °C. It can be seen that the Co–Al LDHs memristor still exhibits gradual resistive switching at a high temperature. In contrast to room temperature conditions, the current in HRS and LRS significantly reduced, which may have been caused by the increase of resistance due to the intensified thermal motion of carrier molecules between the graded laminate structure of Co–Al LDHs at a high temperature. Figure 5 b exhibits the data retention capability of the Co–Al LDHs memristor in HRS and LRS at −2 V constant voltage under 158 °C. With a minimum resistance ratio of about 2, the device almost loses its resistive switching capability. Such a low resistance ratio will greatly increase the misreading rate, which is far from meeting practical application requirements. Since the device does not have thermal stability at 158 °C, the device’s resistive switching and data retention characteristics are tested at 85 °C, as shown in Figure 5 c,d. As can be seen, at this temperature the device has good temperature stability. Data retention characteristics and endurance characteristics are important indicators to measure nonvolatile memory. In order to ensure security and reliability of data, the retention time of memory is generally required to be more than 10 years. As it is the next generation of nonvolatile memory, its endurance characteristics should exceed 10 6 . At present, only a few reported examples of resistive switching memory can meet this standard, and there is a big gap between our devices and this standard. In current researches on resistive switching memory, a variety of materials construct different resistive switching memory models, but the mechanism of resistive switching hidden behind different material systems is not clear enough. Resistive switching behaviors of different material systems also have their own characteristics, which indicate that resistive switching mechanisms of different systems are different. At present, there are many models of a resistive switching mechanism proposed by the scientific community. The mechanism can be generally divided into two categories: interfacial effect and local effect. The resistive switching mechanism of interfacial effects includes an interfacial barrier regulator mode and a charge trap charge–discharge mode, while the resistive switching mechanism of local effects mainly includes metal conductive filaments, local valence state transition, electrothermochemical transition, and so forth. Therefore, the practical process of resistance random access memory needs a lot of further research. To uncover the complex resistive switching mechanism of the device, the charge carrier transport of Co–Al LDHs and Co–Al LDH-adsorbed hexazinone were analyzed by model fitting of the I−V curves in the double logarithm coordinate as shown in Figure 6 a–d. For the Co–Al LDH memristor, after data fitting we found that both HRS and LRS obey space charge-limited conduction [ 44 ], as shown in Figure 6 a,b. For the adsorbed hexazinone memristor, the current and voltage relationship for the negative and positive voltage region was fitted under the log–log coordinate system, as shown in Figure 6 c,d. We can see it in terms of the slopes of the HRS (the I–V relationship can be approximately described as I–V with slopes of 1.14 and 1.13 followed by I–V 2 with slopes of 2.35 and 2.04) and LRS (the slopes were 1.02 and 1.13). The fitted current and voltage relationship can be ascribed to Ohm’s law for LRS and space charge-limited conduction for HRS. Additionally, the ln( I )µ V 1/2 relationship fitting was performed for the Co–Al LDH memristor and Co–Al LDH-adsorbed hexazinone memristor, as shown in Figure 6 e,f, indicating that the conduction behavior in the HRS for the Co–Al LDH memristor and Co–Al LDH-adsorbed hexazinone memristor can be ascribed to Schottky emission [ 45 ]. According to previous reports, the conductive filament theory is very popular [ 46 , 47 ]. According to this theory, resistance in LRS is independent of the size of the device. Therefore, device size dependence of the resistance in LRS and HRS for Co–Al LDHs and Co–Al LDH-adsorbed hexazinone-based memristors was performed, as shown in Figure 7 . For each device size, resistance of 10 LRS and HRS tested from five devices were collected. The resistance was read at 0.2 V. According to the above experimental results, the conductive filament theory can be ruled out. Based on the above I – V characteristic fitting, the resistive switching mechanism is mostly attributed to Schottky emission. In consideration of low current density and gradual change in current, in combination with the above analysis for the mechanism of analog resistive switching, as previously reported, formation and migration of oxygen vacancies in LDHs are energetically expedient [ 48 ]. In addition, low electron affinity of oxygen (compared to its first ionization energy) may be conducive to formation of O 2− ions. Hence, we put forward here a simple model based on the drift-diffusion principle to explain the resistive switching process. Gradual change in the resistance state for the Co–Al LDHs memristor is due to drift and diffusion of oxygen ions and vacancies under applied voltage. A schematic diagram of resistive switching of the Co–Al LDHs memristor is shown in Figure 8 . Figure 8 a shows oxygen ion and oxygen vacancy when no voltage is applied to the device. For positive voltage applied to the Al top electrode, O 2− ions drift towards the Al top electrode, forming O vacancies among the Co–Al oxide octahedron, as shown in Figure 8 b. These O 2− ions have a tendency to accumulate on the surface of Co–Al LDHs as a “sheet charge”. While O 2− ions may diffuse back into the dielectric because of the high concentration gradient, as shown in Figure 8 c, the flux by reason of drift keeps most ions intact in the sheet charge. Hence, devices are gradually converted from HRS to LRS. In the negative voltage sweep, the drift effect has a tendency to gradually reduce, and the diffusion flux begins to dominate. This will lead to O 2− ions to diffuse back to passivate the O vacancies, as shown in Figure 8 d, resulting in the drop of device conductivity, which converts devices from LRS to HRS. The switching mechanism in Co–Al LDHs and its small-molecule adsorption material memristor is complicated, and more than one mechanism might run at the same time. Resistive switching is controlled by inherent characteristics of the LDH layer. After Co–Al LDH-adsorbed hexazinone, the sudden resistive switching behavior of the Co–Al LDH-adsorbed hexazinone memristor could be attributed to the strong reduction performance of hexazinone, cooperating with the drift/diffusion of oxygen ions and vacancies, along with the change of Schottky barrier under the applied voltage. Apart from drift/diffusion of oxygen ions and vacancies, when a positive voltage is applied to the top electrode (Al), the Schottky barrier between Al and Co–Al LDH-adsorbed n-type semiconductors hexazinone will be reversed to form a depletion layer, which is formed at the interface between the metal Al and the active layer. The schematic diagram of the change of Schottky barrier under the applied voltage to form a depletion layer is shown in Figure 9 . In addition, a large number of majority carriers (electrons) produced by the ionization of n-type semiconductors of the small molecule hexazinone will migrate toward the top electrode under a positive electrical field. The electrons are attracted by a positive voltage through the depletion layer. As a consequence, the depletion layer was narrowed, leading to reduction in the height of the Schottky barrier. By contrast, when the negative voltage is applied to the top electrode, electrons migrate back to the Co–Al LDH bulk near the interface, which recombines with oxygen vacancies, leading to increase in the Schottky barrier height. After Co–Al LDH-adsorbed hexazinone, more charge carriers migrate toward the electrode accompanied by broadening of the depletion layer. Therefore, conductivity of the device is altered more significantly, which makes the resistive switching process abrupt. Azomethine moieties in hexazinone as donors will deplete further [ 49 ]. At present, the academic community has not reached a consensus on the mechanism of resistive switching. Researchers have proposed a variety of mechanisms to explain the resistive switching behavior, and some microscopic confirmatory experiments are lacking in this work to clarify the resistive switching mechanism. Performance comparison of resistive switching memory between the Co–Al LDH-adsorbed hexazinone-based devices (this work) and other 2D material-based memory devices reported by several research groups was performed, as shown in Table 1 . In comparison with the materials of BiOI, PCBM–MoS 2 nanocomposites, 2D/3D heterostructure-based CH 3 NH 3 PbI 3-x Cl x , and hexagonal boron nitride, our Co–Al LDH-adsorbed hexazinone-based device has a higher ON/OFF resistance ratio of 2.5 × 10 4 . Furthermore, our device has the highest data retention characteristics (10 5 s) compared to other materials in Table 1 ."
} | 6,193 |
30450723 | PMC7379505 | pmc | 1,485 | {
"abstract": "ABSTRACT Large benthic Foraminifera (LBF) are major carbonate producers on coral reefs, and are hosts to a diverse symbiotic microbial community. During warm episodes in the geological past, these reef‐building organisms expanded their geographical ranges as subtropical and tropical belts moved into higher latitudes. During these range‐expansion periods, LBF were the most prolific carbonate producers on reefs, dominating shallow carbonate platforms over reef‐building corals. Even though the fossil and modern distributions of groups of species that harbour different types of symbionts are known, the nature, mechanisms, and factors that influence their occurrence remain elusive. Furthermore, the presence of a diverse and persistent bacterial community has only recently gained attention. We examined recent advances in molecular identification of prokaryotic (i.e. bacteria) and eukaryotic (i.e. microalgae) associates, and palaeoecology, and place the partnership with bacteria and algae in the context of climate change. In critically reviewing the available fossil and modern data on symbiosis, we reveal a crucial role of microalgae in the response of LBF to ocean warming, and their capacity to colonise a variety of habitats, across both latitudes and broad depth ranges. Symbiont identity is a key factor enabling LBF to expand their geographic ranges when the sea‐surface temperature increases. Our analyses showed that over the past 66 million years (My), diatom‐bearing species were dominant in reef environments. The modern record shows that these species display a stable, persistent eukaryotic assemblage across their geographic distribution range, and are less dependent on symbiotic photosynthesis for survival. By contrast, dinoflagellate and chlorophytic species, which show a provincial distribution, tend to have a more flexible eukaryotic community throughout their range. This group is more dependent on their symbionts, and flexibility in their symbiosis is likely to be the driving force behind their evolutionary history, as they form a monophyletic group originating from a rhodophyte‐bearing ancestor. The study of bacterial assemblages, while still in its infancy, is a promising field of study. Bacterial communities are likely to be shaped by the local environment, although a core bacterial microbiome is found in species with global distributions. Cryptic speciation is also an important factor that must be taken into consideration. As global warming intensifies, genetic divergence in hosts in addition to the range of flexibility/specificity within host–symbiont associations will be important elements in the continued evolutionary success of LBF species in a wide range of environments. Based on fossil and modern data, we conclude that the microbiome, which includes both algal and bacterial partners, is a key factor influencing the evolution of LBF. As a result, the microbiome assists LBF in colonising a wide range of habitats, and allowed them to become the most important calcifiers on shallow platforms worldwide during periods of ocean warming in the geologic past. Since LBF are crucial ecosystem engineers and prolific carbonate producers, the microbiome is a critical component that will play a central role in the responses of LBF to a changing ocean, and ultimately in shaping the future of coral reefs.",
"conclusion": "VIII. CONCLUSIONS (1) Geological and modern records of LBF distributions show that diatom‐bearing taxa are the most common, abundant, and dominant taxa across wide environmental gradients, whereas dinoflagellate and chlorophytic species tend to be more restricted in their distribution, and less tolerant to nutrients and terrestrial influences. (2) Modern, cosmopolitan diatom‐bearing species depend less on their eukaryotic microbes for meeting their energetic requirements and show stable diatom symbiont communities in the core of their distribution. By contrast, dinoflagellate‐bearing taxa, are more reliant on their symbiont for survival, and tend to have a flexible association responsive to environmental conditions across their range of occurrence. (3) The occurrence of cryptic speciation in many LBF species can hide host–symbiont specialisation and adaptation capacity of the host to different environments. The capacity of species to adapt to their new environment is a critical component for understanding the role of evolutionary processes in the assembly and dynamics of natural communities. (4) Abiotic factors (e.g. temperature, water clarity, and nutrient availability), and eukaryotic symbionts had an important synergistic contribution to the expansion and contraction of LBF distribution during warming–cooling cycles during the Cenozoic. (5) Interactions between the host, the eukaryotic symbionts, and the prokaryotic endobionts is key to understanding the plasticity, adaptive potential, and resilience of LBF to environmental change. Additionally, the physiological state of both the host and the associates is likely to influence the identity and diversity of the eukaryotic and prokaryotic community. (6) In recent years, we have seen major advances in describing and understanding the role of microbial assemblages in reef fauna, including reef‐building corals, sponges, and benthic Foraminifera, and the role that bacteria and other microorganisms play in maintaining the health of the reefs. LBF are essential ecosystem engineers and prolific carbonate producers, and the study of their microbiome should provide important information on their ability to respond to climate change. (7) Identifying host–prokaryote–eukaryote associations and genetic structure within LBF host populations is crucial to a better understanding of the capacities of LBF species to adapt to their new environment or to shift their distribution range. These are critical components for understanding the role of evolutionary processes in shaping the assembly and dynamics of natural communities.",
"introduction": "I. INTRODUCTION Across the globe, coral reefs are experiencing rapid declines due to deteriorating environmental conditions mainly driven by ocean warming (Pandolfi et al., \n 2011 ; Hughes et al., \n 2017 ). In these environments, symbiotic associations between organisms can provide the partners involved with the capacity to respond to environmental stresses as well as providing robustness under the challenges caused by climate change (Ainsworth & Gates, 2016 ). Associations with prokaryotic and eukaryotic microorganisms can facilitate the success of species across a variety of habitats, playing a fundamental role in the evolution and adaptive capacity of host organisms (Saffo, 1992 ; Cavanaugh, 1994 ), and have been associated with vulnerability when obligatory symbionts are expelled from their host (i.e. bleaching) (Hallock, 2000 ; Hughes et al., \n 2017 ). Many heterotrophic organisms living in oligotrophic waters have evolved obligatory symbioses with photosynthetic microalgae, thus establishing biotrophic mixotrophy (Not et al., \n 2016 ; Selosse, Charpin, & Not, 2017 ). This process is called photo‐symbiosis as it makes photosynthesis indirectly available to the host (Selosse et al., \n 2017 ). However, mixotrophy comes at a cost, as it requires five times more energy and nutrient allocation to maintain the photosynthetic apparatus compared to maintaining a strictly heterotrophic feeding mode (Raven, 1997 ). Nonetheless, the development of mixotrophy allows organisms to occupy previously inaccessible niches, such as nutrient‐poor environments. Photo‐symbiosis is critical to maintaining functioning coral reefs, not only in corals, the best‐known reef‐associated organisms, but also in the (often) overlooked unicellular eukaryotic large benthic Foraminifera (LBF). Symbiosis with eukaryotic taxa (i.e. microalgae) is essential for the health of reef ecosystems, and LBF are responsible for a significant proportion of the carbonate sediment across reef environments worldwide (Langer, 2008 ). From a global carbon perspective, LBF play a fundamental role in carbon sequestration and carbon cycling (Langer, Silk, & Lipps, 1997 ; Langer, 2008 ), in addition to sediment production and reef maintenance (Baccaert, 1986 ; Fujita & Fujimura, 2008 ; Dawson & Smithers, 2014 ; Dawson, Smithers, & Hua, 2014 ; Doo et al., \n 2017 ). LBF species, especially those that produce high‐magnesium tests, serve an important role in maintaining the chemical equilibrium on coral reefs, serving to buffer against daily pH flux from reef metabolism through skeletal dissolution post mortem (Yamamoto et al., \n 2012 ). It is becoming increasingly apparent that other microorganisms such as bacteria and Archaea (henceforth referred to as prokaryotic associates), fungi, and viruses, play a significant and complex role in maintaining the host's health (Peixoto et al., \n 2017 ). Prokaryotic microbial associations can benefit the host by enhancing nutrient cycling (S, C, and N), providing photosynthesis‐dependent nitrogen fixation, enhancing calcification, and in production of antimicrobials and pathogen removal (Knowlton & Rohwer, 2003 ; Lesser et al., \n 2004 ). By contrast, the role of fungi and viruses remains elusive (Lecampionalsumard, Golubic, & Priess, 1995 ; Sweet & Bythell, 2017 ). Identifying specific microbes that provide critical functional contributions to a host organism requires an understanding not only of the endobiotic microbial population, but also of the persistence and stability in time and space of both the microbial functional niches and the microbes that utilise them (Ainsworth et al., \n 2015 ; Hernandez‐Agreda et al., \n 2016 ). These associations with microbial partners likely underpin the capacity of reef organisms to respond to climate change. Ocean warming will influence the biogeographic range of reef species, which could result in poleward expansion as subtropical and temperate marine ecosystems become ‘tropicalised’ (Verges et al., \n 2014 ). The flexibility in these associations will determine the host's capacity to accommodate to local environmental change, as well as allowing adaptations to new environmental conditions following distribution range expansions. The composition of both the prokaryotic microbiome and the eukaryotic symbionts in relation to environmental change has been explored largely in reef‐building corals (LaJeunesse, 2002 ; Ainsworth, Thurber, & Gates, 2010 ; Bourne, Morrow, & Webster, 2016 ). However, many other organisms, such as LBF, also benefit from the intricate interplay between prokaryotic endobionts and eukaryotic endosymbionts, mainly microalgae. Although only ca .10 families of benthic Foraminifera are currently known to have associations with algal symbionts, these families consist of many described species, which are abundant in shallow carbonate platforms worldwide. LBF are a polyphyletic group in which endosymbiosis with microalgae evolved independently in multiple benthic foraminiferal families (Hallock & Glenn, 1986 ; Hallock, 1999 ; Lee, 2011 ). The shell of LBF facilitates the housing of photosynthetic symbionts by morphological adaptations, including canaliculation, flosculisation, and the development of endo‐ and exoskeleton structures or secondary or lateral chamberlets (Renema, 2007 ). Evolutionary radiations indicate that the acquisition of, and change in, algal types were crucial steps in the evolution of large miliolid Foraminifera (Holzmann et al., \n 2001 ). Symbiosis with algae has been suggested to be the driving force in the evolution of these groups of benthic Foraminifera (Lee, 2006 , 2011 ; Lee et al., \n 2010 ). LBF include members of two orders of foraminifera: Rotaliida and Miliolida (Hallock, 1999 ). The order Rotaliida, characterised by an optically radial, bilamellar perforate test (Pawlowski, Holzmann, & Tyszka, 2013 ), includes three modern families: Amphisteginidae, Calcarinidae, and Nummulitidae. The order Miliolida, with an imperforate wall, high‐magnesium calcite test and with randomly oriented crystals refracting light in all directions and resulting in a porcelaneous appearance of the test (Pawlowski et al., \n 2013 ), includes the Alveolinidae, Peneroplidae, Soritidae, and Archaiasidae (Loeblich & Tappan, 1984 ). In general, Rotaliida species are known predominantly to host diatoms, whereas Milioliida also play host to other algal groups, such as chlorophytes, rhodophytes, and dinoflagellates (Lee, 2006 ). Additionally, modern species of both groups have associations with a diverse prokaryotic community, including heterotrophic bacteria, photosynthetic cyanobacteria, and algal plastids, suggesting that Foraminifera are particularly favourable partners for the establishment of symbioses (Lee, 2006 ; Bourne et al., \n 2013 ). At least 47 modern species across 15 genera have been identified as possessing algal symbionts (Lee, 2006 ; Förderer, Rödder, & Langer, 2018 ). Whereas eukaryotic symbiosis has received considerable attention, prokaryotic symbiosis and the role of bacteria in LBF remains largely unexplored (e.g. Webster et al., \n 2016 ; Prazeres et al., \n 2017a ; Prazeres, 2018 ). Not only is the diversity of bacterial communities poorly known, but so is the relationship and role that these bacteria may have in LBF ecology, adaptive potential, and evolution. In this review, we aim to determine how the eukaryotic and prokaryotic microbiome influences the capacity of LBF to occupy new habitats, expand their distribution range, and adapt or acclimatise to shifts in environmental conditions. For the purposes of this review, eukaryotic symbionts and prokaryotic partners are considered algal and bacterial species, respectively. Firstly, we explore the known algal partners and how they influence modern LBF species' biology and ecology. We will also discuss the geographical distribution of fossil LBF species within their environmental context, and link it to their microbiome, particularly to algal symbionts. Finally, we argue that the microbiome (i.e. algal and bacterial species) is likely to be crucial in the resilience, acclimation, and adaptation of LBF in the face of climate change. We discuss how the microbiome could benefit and drive LBF evolution across species distribution ranges: ( i ) by persistent eukaryotic and prokaryotic microbial associations across the distribution of LBF species, which have been reported to be highly species‐specific and to determine ecological niches in LBF; ( ii ) the presence of a variable microbiome responsive to environmental gradients; ( iii ) the presence of a stable, persistent algal symbiont community coupled with flexible bacterial associations; and ( iv ) adaptable algal symbiosis with a persistent bacterial community, which could assist species in crossing biogeographical barriers and adapting to changing environmental conditions. The composition of the microbiome benefits the host in different ways, and different species are likely to utilise different strategies to maintain the holobiont system. Therefore, it is crucial to understand and disentangle how the host and symbiont compartments are likely to interact with, and respond to environmental change."
} | 3,809 |
28515898 | PMC5433980 | pmc | 1,487 | {
"abstract": "Abstract The vast majority of plants obtain an important proportion of vital resources from soil through mycorrhizal fungi. Generally, this happens in exchange of photosynthetically fixed carbon, but occasionally the interaction is mycoheterotrophic, and plants obtain carbon from mycorrhizal fungi. This process results in an antagonistic interaction between mycoheterotrophic plants and their fungal hosts. Importantly, the fungal‐host diversity available for plants is restricted as mycoheterotrophic interactions often involve narrow lineages of fungal hosts. Unfortunately, little is known whether fungal‐host diversity may be additionally modulated by plant–plant interactions through shared hosts. Yet, this may have important implications for plant competition and coexistence. Here, we use DNA sequencing data to investigate the interaction patterns between mycoheterotrophic plants and arbuscular mycorrhizal fungi. We find no phylogenetic signal on the number of fungal hosts nor on the fungal hosts shared among mycoheterotrophic plants. However, we observe a potential trend toward increased phylogenetic diversity of fungal hosts among mycoheterotrophic plants with increasing overlap in their fungal hosts. While these patterns remain for groups of plants regardless of location, we do find higher levels of overlap and diversity among plants from the same location. These findings suggest that species coexistence cannot be fully understood without attention to the two sides of ecological interactions.",
"introduction": "1 Introduction Mycorrhizal fungi play a crucial role for plant survival (Smith & Read, 2008 ). In mycorrhizal interactions, mycorrhizal fungi facilitate the uptake of essential resources for plant metabolism, such as water and soil minerals (Raven, Evert, & Eichhorn, 1999 ). Generally, in exchange, plants transfer photosynthetically fixed carbon to their mycorrhizal partners (Smith & Read, 2008 ). Occasionally, however, plants do not give back carbon, but instead obtain it from the mycorrhizal fungi as replacement for photosynthesis (Leake, 1994 ; Merckx, Bidartondo, & Hynson, 2009 ). This results in an antagonistic interaction between plants and their fungal hosts. Specifically, these interactions are called mycoheterotrophic (MH) interactions and can occur in a single developmental stage (e.g., in orchids, and some ferns and lycopods) or during the entire life cycle of a plant (fully mycoheterotrophic plants) (Merckx & Freudenstein, 2010 ; Winther & Friedman, 2008 ). MH interactions represent a nonmutualistic mode of life that occurs in nearly all major lineages of land plants, involving more than 20,000 plant species (Merckx, 2013 ). In general, the fungal‐host diversity available for these plants is restricted as MH interactions often involve more narrow lineages of mycorrhizal fungi than non‐MH interactions (Bidartondo et al., 2002 ). Unfortunately, little is known whether fungal‐host diversity may be additionally modulated by plant–plant interactions through shared hosts. Yet, this may have important implications for plant competition and coexistence (Bever et al., 2010 ). Recent studies have shown that the diversity of mycorrhizal fungi is strongly associated with plant community composition (Davison, Öpik, Daniell, Moora, & Zobel, 2011 ; Martínez‐García, Richardson, Tylianakis, Peltzer, & Dickie, 2015 ; Peay, Baraloto, & Fine, 2013 ) and habitat conditions (Hazard et al., 2013 ). For instance, in the case of MH interactions, a given group of plant species can be exploiting either closely or distantly related fungal hosts (see Figure 1 ). Additionally, this same group of plants can have either a weak or a strong fungal‐host overlap (see Figure 1 ). The combination of these two factors depends on plant niche and have been shown to be determinant for plant coexistence (Levine & HilleRisLambers, 2009 ; Levins, 1968 ; Rohr et al., 2016 ). According to niche theory (Loreau, 2010 ; MacArthur & Levins, 1967 ), species coexistence is a function of their their niche width and niche overlap (Chesson, 2000 ). Competitive exclusion among species is high when their potential niche overlap is large and their combined niche width is small. Similarly, the chances of co‐occurrence among species in the same niche space is low when their potential niche overlap is small and their combined niche width is large. Species coexistence (co‐occurrence and no exclusion) then is expected to happen when niche overlap and niche width are symmetric (Chesson, 2000 ; Tilman, 2011 ) (see Figure 1 —diagonal). Niche delimitation is never straightforward due to our often lack of a priori knowledge about the resources and functional traits defining the niche dimensions of a species (Kraft, Godoy, & Levine, 2015 ). Defining the niche of fungal hosts of mycoheterotrophic plants is as challenging as for other groups of organisms, but one potential hypothesis is that the higher the fungal‐host diversity of mycoheterotrophic plants, the broader their niche. Thus, species coexistence may be favored under symmetric patterns of fugal‐host overlap and diversity. Figure 1 Illustration of possible fungal‐host patterns among mycoheterotrophic plants. On the vertical and horizontal axes, the figure illustrates, respectively, an increase in fungal‐host diversity and fungal‐host overlap among MH plants. The bottom right panel represents a scenario for plants with high chances of competitive exclusion given by their large fungal‐host overlap and their small fungal‐host diversity (using similar functional traits). The top left panel represents a scenario for plants with low chances of co‐occurring in the same space given by their small fungal‐host overlap and their large fungal‐host diversity (using different functional traits), which could be difficult to find in a common place. The diagonal panels then represent the scenarios for plants with a higher chance of coexistence given by their symmetry between fungal‐host overlap and fungal‐host diversity, which could lead to maximize co‐occurrence (exploit available resources) and to minimize competitive exclusion To work on the above hypothesis, we use a system where the mycorrhizal interaction involves mycoheterotrophic plants. In addition, these plants are associated with arbuscular mycorrhizal fungi (phylum Glomeromycota), which are associated with more than 80% of land plants. Therefore, this association represents one of the most ancient and abundant mycorrhizal interaction among plants on a global scale (Smith & Read, 2008 ; Strullu‐Derrien et al., 2001 ). Here, we investigate MH interactions by analyzing the observed patterns of associations between MH plants and their fungal hosts in a niche framework. In particular, we study how the phylogenetic diversity of arbuscular mycorrhizal hosts varies among individual MH plants, and how this diversity is modulated and shared among groups of MH plants.",
"discussion": "4 Discussion Previous studies have investigated fungal‐host diversity of MH plants in relation to the fungal diversity associated with the surrounding green plants (Bidartondo, Bruns, Michael, Sérgio, & Read, 2003 ; Bidartondo et al., 2002 ; Bougoure, Ludwig, Brundrett, & Grierson, 2009 ; Cullings, Szaro, & Bruns, 1996 ; Roy, Whatthana, Richard, Vessabutr, & Selosse, 2009 ; Yamato et al., 2011 ). However, several MH species present vast geographic distributions despite being locally rare. Therefore, these surrounding plants may not be the exclusive factors determining fungal‐host diversity in MH plants. Indeed, many studies have reported the occurrence of different species of arbuscular mycoheterotrophs in the field without a clear explanation for this phenomenon (e.g. Cheek & Williams, 1999 ; Jonker, 1938 ; Maas & Rübsamen, 1986 ; van de Meerendonk, 1984 ; Merckx, 2013 ; van der Pijl, 1934 ; van Royen, 1972 ). In our study, we have considered potential neighboring effects of MH plants with each other as possible drivers of fungal‐host diversity. Because many unmeasured factors can influence MH interactions, we opted to compare the observed patterns against all the possible fungal‐host combinations (what we called artificially generated groups of plants). We have found that individual MH plants have a tendency to exploit more distantly related fungi than expected by chance. This tendency of targeting distantly related fungi has been described in autotrophic plants (Giovannetti, Sbrana, Avio, & Strani, 2004 ). Nevertheless, it has been suggested that MH plants have more restricted interactions, as they often show higher specificity toward their fungal hosts (e.g. Bidartondo et al. 2002 ; Gomes, Aguirre‐Gutiérrez, Bidartondo, and Merckx 2017 ). For example, in Afrothismia , five closely related MH plants were found to specialize in five closely related lineages of Glomeromycota fungi (Merckx & Bidartondo, 2008 ). In contrast, in Monotropoideae, the five MH species in this clade associate with five different distantly related Basidiomycota fungi, but each within the same fungal lineage (Bidartondo & Bruns, 2005 ). Either way, and despite the processes leading to this extreme level of fungal specificity, it has been suggested that MH plants adapt to the suitable fungal partners that participate in this mycoheterotrophic interaction, and therefore, host‐jumps to distantly related fungal lineages are unexpected (Bidartondo & Bruns, 2002 ). Building on niche theory, our results may reflect a MH plant strategy to increase its fungal‐host diversity or niche width, as species with a wider niche may be more likely to obtain different resources and to establish successfully in new habitats (Levine & HilleRisLambers, 2009 ; Levins, 1968 ; Tilman, Wedin, & Knops, 1996 ). Mycoheterotrophic plants require established mycorrhizal networks to persist (van der Heijden, Martin, Selosse, & Sanders, 2015 ; Sachs & Simms, 2006 ). Although each species tend to increase the phylogenetic diversity of their fungal hosts, it is still a limited fraction of the total diversity of arbuscular mycorrhizal fungi that can be part of this interaction (Douglas, 2008 ; Gomes et al., 2017 ; Merckx et al., 2009 ), suggesting that these fungi appear to be under selection pressure to be resistant to these cheaters (Douglas, 2008 ). Therefore, the ability to increase its fungal‐host diversity may confer an advantage to increase the opportunities to cheat mycorrhizal networks. We have found that in communities of co‐occurring MH plant species in the field the fungal‐host diversity among MH plants appear to increase proportionally to their fungal‐host overlap. This same tendency was confirmed among the artificially generated groups of MH plants showing that the patterns observed are not an artifact of the reduced number of MH communities observed in the field. Moreover, we have found that both fungal‐host diversity and overlap are significantly higher among plants that belong to the same geographic location, which could provide an explanation for the lack of phylogenetic signal on the fungal hosts among MH plants. These results indicate that fungus‐plant interactions can be better explained by understanding plant–plant interactions generated by sharing resources or fungal hosts. Future studies could explain whether this symmetry between fungal‐host diversity and overlap may respond to an ecological mechanism driven by maximizing co‐occurrence and avoiding competitive exclusion among MH plants. A potential bias in our study is the use of ITS2 sequences and future work should consider expanding these sequences (see Supporting Information for more details). Another aspect that deserves particular attention is the influence of abiotic factors that can affect the diversity of fungal hosts for the MH plants. In fact, many other factors can influence diversity, including the surrounding autotrophic plants. Taking everything into account is virtually impossible. However, our findings suggest that species coexistence cannot be fully understood without attention to the two sides of ecological interactions."
} | 3,032 |
27258948 | PMC5042321 | pmc | 1,489 | {
"abstract": "Syntrophies are metabolic cooperations, whereby two organisms co-metabolize a substrate in an interdependent manner. Many of the observed natural syntrophic interactions are mandatory in the absence of strong electron acceptors, such that one species in the syntrophy has to assume the role of electron sink for the other. While this presents an ecological setting for syntrophy to be beneficial, the potential genetic drivers of syntrophy remain unknown to date. Here, we show that the syntrophic sulfate-reducing species Desulfovibrio vulgaris displays a stable genetic polymorphism, where only a specific genotype is able to engage in syntrophy with the hydrogenotrophic methanogen Methanococcus maripaludis . This 'syntrophic' genotype is characterized by two genetic alterations, one of which is an in-frame deletion in the gene encoding for the ion-translocating subunit cooK of the membrane-bound COO hydrogenase. We show that this genotype presents a specific physiology, in which reshaping of energy conservation in the lactate oxidation pathway enables it to produce sufficient intermediate hydrogen for sustained M. maripaludis growth and thus, syntrophy. To our knowledge, these findings provide for the first time a genetic basis for syntrophy in nature and bring us closer to the rational engineering of syntrophy in synthetic microbial communities.",
"conclusion": "Conclusions It is well known that SRBs can act as both hydrogen consumers and producers, taking the latter role in the absence of terminal electron acceptors such as sulfate ( Muyzer and Stams, 2008 ). This is believed to underpin the ability of SRBs to occupy different niches and survive different environmental conditions ( Muyzer and Stams, 2008 ). Our findings suggest that the dual physiology is encoded in DvH as two separate genotypes, which are maintained at different frequencies. As indicated by theoretical studies of phenotypic variability under fluctuating environments ( Kussell and Leibler, 2005 ), it is possible that these DvH genotype frequencies are in alignment with the frequency of the fluctuations in the environment in terms of availability of strong electron acceptors such as sulfate. At the same time, it is also clear that the frequency at which the syntrophic DvH genotype is maintained will be directly linked to the presence of syntrophic partners such as Mm in the environment. It remains to be seen if a stable polymorphism as seen in DvH is a strategy common to other SRBs in the environment. From a more applied point, this study points to engineering of thermodynamic constraints in metabolism as a possible route to the implementation of syntrophic interactions in synthetic microbial communities.",
"introduction": "Introduction Syntrophic interactions represent cases of metabolic cooperation between two phenotypically distinct organisms ( Schink, 1997 ; McInerney et al. , 2008 ; Sieber et al. , 2012 ; Morris et al. , 2013 ). These interactions are common among microbes living in environments that can be readily depleted of strong electron acceptors. In such environments, including anaerobic reactors, animal guts, ocean and lake sediments and soil, the depletion of suitable electron acceptors is expected to increase the abundance of microbes with fermentative metabolism. The low thermodynamic energy associated with fermentative metabolic pathways results in 'thermodynamic inhibition' of microbial growth due to product accumulation ( Schink, 1997 ; Kleerebezem and Stams, 2000 ; Großkopf and Soyer, 2014 , 2016 ). This inhibition of growth in the primary degrading microbes can be lifted by others consuming the inhibitory waste product (mostly hydrogen) ( Seitz et al. , 1988 ; Schink, 1997 ; Sieber et al. , 2012 ). Thus, environmental depletion of electron acceptors provides a setting in which syntrophic interactions can more readily emerge. Besides their central role in natural microbial communities ( McInerney et al. , 2009 ; Sieber et al. , 2012 ), syntrophic interactions also constitute a desired motif in engineered synthetic microbial communities, where they can provide a metabolic basis for stable community formation ( Gilbert et al. , 2003 ; Großkopf and Soyer, 2014 ; Santala et al. , 2014 ). Despite their importance in ecological and engineered settings, our understanding of the molecular basis of syntrophic interactions remains limited. Transcriptomic analyses in a syntrophic model system ( Walker et al. , 2012 , 2009 ) have highlighted several genes that are differentially regulated under conditions of syntrophy. The gene products are involved in primary energy metabolism as well as in secretion of secondary metabolites which are suggested to enhance the growth of the syntrophic partner ( Walker et al. , 2009 , 2012 ). Other studies indicated genes involved in formation of spatial structures to be important for syntophic interactions ( Schink, 1997 ; Kato and Watanabe, 2010 ; Summers et al. , 2010 ). Indeed, biofilm or granule formation ( Summers et al. , 2010 ; Wintermute and Silver, 2010a ; Worm et al. , 2014 ), as well as direct attachment via flagella ( Marx, 2009 ; Shimoyama et al. , 2009 ; Kato and Watanabe, 2010 ), is observed in several co-cultures involving syntrophy or cross-feeding. However, formation of spatial structures like biofilms is also frequently observed in monocultures ( Zhang et al. , 2007 ) and under non-syntrophic conditions ( Nadell et al. , 2009 ), and as such, it is not clear whether biofilm formation is a key driver in the formation and maintenance of syntrophies. At a cellular level, it is well documented that syntrophic associations result in transcriptomic and metabolomic changes in the involved organisms ( Walker et al. , 2009 , 2012 ; Qi et al. , 2014 ). The study of these changes has led to the identification of specific genes, whose deletion completely abolishes the ability to form syntrophic interactions ( Walker et al. , 2009 ). Despite these important insights, it is still not clear whether there are specific genetic alterations that can enhance formation of syntrophies. If such alterations exist, then identifying these could allow for a better understanding of the ecological and evolutionary basis of natural syntrophies and enable engineering of synthetic ones. The latter objective is motivated by observations that synthetic engineering of syntrophic and mutualistic interactions among microbes can increase bioproductivity ( Shou et al. , 2007 ; Wintermute and Silver, 2010b ; Kerner et al. , 2012 ; Mee et al. , 2014 ) and scope of biotechnological applications ( Gilbert et al. , 2003 ; Bernstein et al. , 2012 ; Santala et al. , 2014 ). Here, we aim to identify genetic drivers of syntrophy by focusing on physiological and genetic analysis of the well-established model syntrophic system between a sulfate-reducing bacteria (SRB) Desulfovibrio vulgaris strain Hildenborough (DvH) and a hydrogenotrophic methanogen Methanococcus maripaludis S2 (Mm) ( Stolyar et al. , 2007 ; Hillesland and Stahl, 2010 ). A typical SRB, DvH is equipped with a dissimilatory sulfate reductase and various hydrogenases ( Pieulle et al. , 2005 ), which allows it to utilize hydrogen, while reducing sulfate. The ability to utilize hydrogen makes SRB a competitor for methanogens like Mm in the presence of sulfate. In the absence of sulfate, however, its fermentative metabolism results in the production of hydrogen and opens up the potential of a syntrophic interaction with Mm. We show that this ability to engage in syntrophy with Mm is underpinned by a stable genetic polymorphism in DvH; under a sulfate-free, minimal lactate media, only a specific genotype of DvH can grow syntrophically with Mm. We find that this genotype naturally exists and is characterized by a set of two distinguishing genetic mutations. Both mutations are linked to the lactate oxidation pathway, the energetics of which is a key driver of the syntrophic interaction. In particular, the resulting altered thermodynamics of this pathway increases the tolerated hydrogen pressure of lactate oxidation in DvH, therefore providing substrate for growth of Mm and a stable syntrophy can be formed. These findings provide the first evidence of a specific syntrophy-inducing genetic polymorphism in SRB and point toward energetic determinants of metabolism as a key driver for syntrophy formation in general.",
"discussion": "Results and discussion DvH-Mm syntrophy requires a specific DvH genotype To study the potential genetic basis of syntrophic interactions, we used here the model system of DvH and Mm. In the absence of sulfate, and under minimal, lactate-containing media, the DvH-Mm co-culture has been shown to syntrophically oxidize lactate to produce methane, acetate and CO 2 while relying mainly on hydrogen exchange between the two species ( Stolyar et al. , 2007 ). When we attempted to initialize this system, the initial frequency of co-culture success, assessed with methane formation and OD measurements, was only 1/40 ( Supplementary Figure S1a ). This observation can be explained by the presence of a low frequency phenotypic or genetic variant in the wild-type population of the DvH that is characterized by its ability to form a successful syntrophy with Mm. In the case of phenotypic variability, where a single genotypic background gives rise to multiple phenotypes ( Balaban et al. , 2004 ), it would be expected that DvH clones isolated from the working syntrophic co-culture and re-cultured with Mm should again lead to low frequency in co-culture success. In contrast, in the case of genotypic variability, we would expect DvH clones isolated from the working syntrophic co-culture to give much higher co-culture success when re-cultured with the Mm. To test these possibilities, we isolated 20 clones of DvH both from the working syntrophic co-culture and from the original inoculum culture of DvH. Re-inoculating these 40 DvH clones with Mm (see Materials and methods, Supplementary Figure S1b ), we found that all co-cultures involving clones isolated from the working syntrophic co-culture led to increased OD and methane production after 1 week of growth ( Figure 1 ). For the 20 co-cultures started with clones from the working syntrophic co-culture, the average methane production after 1 week was 1.13 ml (±0.64 ml, n =20), while it was 0.18 ml (±0.03 ml, n =20) for the co-cultures started with clones from the original DvH population. The pattern was consistent after subculturing (1/20 inoculum), with the second generation of cultures producing 1.03 ml (±0.46 ml, n =20) and 0.07 ml (±0.05 ml, n =20) methane for DvH clones isolated from the working syntophic co-culture and the original culture, respectively. For the successful co-cultures, we also observed a positive correlation between the methane production and OD ( r 2 =0.54, V(ml) CH4 =5.23 × OD−0.44); however, the latter is not a good predictor of growth due to a high degree of granule formation in the co-culture and associated patchiness (see Figure 1 , inset). The consistent co-culturing success of DvH clones isolated from the working syntrophic co-culture leads to the conclusion that the observed success in formation of syntrophy is due to a genetic factor and not due to phenotypic variability. Hereafter, we refer to the clone leading to syntrophic co-cultures with Mm as 'DvH-s' and to the other clone as 'DvH-ns', for ‘non-syntrophic'. Genetic analysis of DvH-s vs DvH-ns reveals determinants of syntrophy To better understand the genetic difference between these two genotypes, we isolated five new clones of DvH-s from a successful co-culture and five clones from the original DvH population, which were expected to be DvH-ns. We re-sequenced the genomes of all of these 10 clones and annotated each genome against the reference DvH genome, which contains the main chromosome and a megaplasmid ( Heidelberg et al. , 2004 ) (see Materials and methods ). We found a total of 57 polymorphisms across the 10 clones. There were a total of 26 SNPs on the chromosome ( Figure 2 ) and 2 on the megaplasmid, as well as 27 INDELs on the chromosome ( Figure 2 ) and 2 more on the megaplasmid. A detailed view of the genotype of each strain for all 57 positions is provided in Supplementary File 1 . To verify again the ability of each of these sequenced clones to form a successful syntrophy or not, we co-cultured them with Mm. These experiments revealed that as expected all five DvH-s clones (labeled as DvH-s 1–5) were able to generate a methane producing, syntrophic co-culture with Mm ( Figure 3 ). Among the DvH-ns clones, three resulted in no co-culture growth as expected, while two (initially labeled as DvH-ns 1 and DvH-ns2) did result in relatively successful syntrophic co-culture ( Figure 3 ) (these findings are further discussed below). Taking these physiological results in accord, we analyzed the sequence data for genetic changes that are common to all seven clones that led to successful co-cultures, and that are lacking from the other three clones. Out of all 57 mutations, only 2 were present in all 7 ‘co-culturable' clones, but absent in clones DvH-ns 3–5 ( Figure 2 ), suggesting that these two mutations are essential for syntrophy. The first of these two mutations is a non-synonymous point mutation from G to A in the gene DVU3023 (DvH chromosome location: 3142197), resulting in an amino-acid exchange from Aspartate to Asparagine. The DVU3023 gene is annotated to encode for a sigma-54-dependent response regulator that controls the downstream operon (DVU3025–DVU3033), which includes the genes lactate permease and pyruvate decarboxylase. Notably, both enzymes are involved in the lactate oxidation pathway, with the first enzyme catalyzing the lactate uptake and the second catalyzing conversion of the pyruvate resulting from lactate oxidation. The second mutation is an in-frame deletion that led to the removal of a GAG sequence in the gene DVU2287 (DvH chromosome location: 2381876), resulting in the replacement of a TGA codon with a TCG codon. DVU2287 is annotated to encode for the CooK subunit of Coo, which is annotated as the H + /Na + ion-translocating subunit of the carbon monooxide-dependent, membrane-bound dehydrogenase, which is hypothesized to be involved in the lactate oxidation pathway ( Walker et al. , 2009 ). The TGA codon is known to encode for a stop codon, which can either lead to a truncated gene product or introduction of a Selenocysteine residue during translation ( Valente et al. , 2005 ). Thus, in all seven genotypes able to form syntrophy, this replacement of the TGA codon with the TCG codon either resulted in a gain-of-function at the DVU2287 gene or to an exchange of a Selenocysteine residue with a Serine residue. Interestingly, sequencing data revealed that clones DvH-ns 1 and DvH-ns2, which were picked from the original DvH population but could form syntrophy with Mm, show a mixed signature of alleles consisting of the mutated and non-mutated sequences at this locus ( Figure 2 ). This could be explained by the initial clone picking resulting in a heterogeneous sample of genotypes, or by the presence of multiple genome copies within one cell—a case that is known to be common in SRB ( Postgate et al. , 1984 ). A thermodynamic model highlights potential implications of the cooK mutation Given the well-established role of thermodynamic limitations as the basis of syntrophic interactions and the role of hydrogen formation in this ( Cord-Ruwisch et al. , 1988 ; Schink, 1997 ; Noguera et al. , 1998 ), it is possible that the identified mutation in the Coo hydrogenase relates to the thermodynamics of the lactate oxidation pathway. To explore this possibility, we performed a thermodynamic analysis of this pathway as it is seen in DvH (see Materials and methods and Supplementary Table S1 ). As expected, this thermodynamic view shows the feasibility of lactate oxidation in the presence of sulfate, where electrons can flow from the lactate dehydrogenase (LDH) catalyzed reaction to sulfate reduction reactions, in accordance with the standard reduction potentials of these reactions ( Figure 4 ). This would allow DvH to harvest the available energy for growth. In the absence of sulfate as a terminal electron acceptor, however, lactate oxidation would only be possible if the LDH-catalyzed reaction could be coupled with an appropriate electron accepting reaction. Suitable reactions for this role, and available in the central metabolism of DvH, would include the reduction of acetaldehyde to ethanol and reduction of pyruvate to alanine. Combining the lactate oxidation with either of these reactions would lead to a thermodynamically feasible pathway; however, they would result in the by-passing of the ATP generating step of the lactate oxidation pathway; the conversion of pyruvate to acetate mediated by the enzymes pyruvate oxidoreductase, phosphate acetyltransferase and acetate kinase (POR-PTA-ACK in Figure 4 ). A third option as an electron accepting reaction would be the hydrogenase-catalyzed proton reduction, leading to hydrogen formation ( Walker et al. , 2009 ) ( Figure 4 ). Coupling this reaction with lactate oxidation would allow production of pyruvate, with subsequent pyruvate oxidation via POR-PTA-ACK pathway leading to ATP formation. The free energy of the combined lactate oxidation–proton reduction reaction under standard conditions is, however, strongly positive (43.2 kJ mol −1 ), creating an uphill energy barrier for growth (see Materials and methods , Supplementary Table S1 ). Changing this energy budget in order to make this fermentative reaction thermodynamically feasible would require reducing the hydrogen and pyruvate concentrations. In particular, assuming biologically realistic concentrations ( Bennett et al. , 2009 ) of 1 mmol and 0.1 mmol for lactate and pyruvate respectively, then the hydrogen partial pressure (that is, concentration) required to achieve a negative reaction free energy would be around 0.03Pa. This pressure is well below the thermodynamically allowed minimum hydrogen pressure of 0.2Pa required for hydrogenotrophic methanogenesis ( Schink, 1997 ). Thus, at the possible hydrogen pressures that can lead to syntrophy, the fermentative lactate oxidation metabolism might stall due to thermodynamic limitations. One way to overcome this thermodynamic limitation would be to invest additional energy into the lactate oxidation reaction, with the side effect that a higher hydrogen pressure would result under physiological conditions. As the Coo is annotated as an H + /Na + ion-translocating hydrogenase, we hypothesize that the identified polymorphism in this gene increases the number of ions it can translocate over the membrane per number of hydrogen produced, and thereby use the membrane gradient as a form of cellular energy to invest in lactate oxidation. In particular, each ion translocated per hydrogen produced could result up to a hundred-fold increase in equilibrium hydrogen pressure tolerated by DvH (that is, from 0.03Pa to 3Pa for single ion, assuming 2 pH units membrane gradient). In a view based on electron flow, this scenario could be seen as shifting the reduction potential of the Coo-catalyzed proton reduction reaction to more positive values, and thereby making it thermodynamically feasible for this reaction to accept electrons from LDH-mediated lactate oxidation reaction ( Figure 4 ). We also note that the function of hydrogenases, and in particular their ability to shuttle hydrogen and protons to the reaction center, commonly involves selenocysteine and cysteine residues ( Valente et al. , 2005 ; Arnér, 2010 ). As such, it is tempting to speculate that the mutation observed here in the cooK gene involves the replacement of a selenocysteine residue with serine in the cooK gene product, with direct effect on enzyme kinetics or reduction potential of the catalyzed proton reduction, as seen in some hydrogenases ( Arnér, 2010 ; Gutiérrez-Sánchez et al. , 2010 ). A sequence analysis shows that many SRBs contain a homolog of the CooK subunit and carry a conserved cysteine residue in the identified loci ( Supplementary File 2 ). Physiology of DvH-s is in line with the expectations of the thermodynamic model A key testable prediction stemming from the thermodynamic model is that DvH-s (as well as DvH-ns 1 and 2) clones should produce more hydrogen under sulfate-limited conditions. We verified this prediction by inoculating all 10 sequenced clones on a lactate minimal medium without sulfate, and with 50% and 100% of the sulfate needed to degrade all available lactate via the sulfate pathway (see Materials and methods ). We found that all DvH-s clones, as well as DvH-ns 1 and 2 clones, produce significant hydrogen above background levels in the absence of sulfate or at 50% sulfate ( P ⩽0.01, two-tailed t -test), while DvH-ns 3, 4 and 5 clones only produce hydrogen significantly above background levels at 50% sulfate ( P ⩽0.05, two-tailed t -test) ( Figure 5 ). The cumulative hydrogen pressure achieved by the former clones over the course of 7 days, in the absence of sulfate and at stationary state is ~300 Pa, well above the pressure needed to sustain Mm growth. This explains the ability of the DvH-s clones to engage in syntrophy with Mm. A second prediction stemming from the thermodynamic model discussed above is that hydrogen production of syntrophic DvH-s clones should be affected by modifications in the ion content of the media. The observed cumulative hydrogen pressure of ~300Pa from DvH-s clones shows that indeed there is an energy investment into the lactate oxidation within this genotype. Dependent on physiological equilibrium concentrations of lactate, pyruvate and the magnitude of the ion gradient across the membrane, this investment could be achieved by Coo translocating as low as 1 H + /Na + per hydrogen produced (in Figure 4 we show that this investment needs to be ~−23 kJ mol −1 , which under standard conditions equates to translating 2 single charged ions over a 100-fold ion gradient per hydrogen produced). At any rate, changing the Na + concentration in the media should affect the membrane ion gradient, and thereby the level of energy investment and the steady-state hydrogen pressure. To test this, we transferred triplicates of DvH-s and DvH-ns cultures into Na-buffer at different Na + concentrations, and after 6 days we measured the equilibrium hydrogen pressure. Although these cultures did not show any detectable growth, the DvH-s clones produced consistently more hydrogen than the DvH-ns clones, and such production increased with increasing Na + concentration up to a threshold high concentration ( Figure 6 ). The low hydrogen production at 500 m m Na + indicates that at that level of Na + concentration the salinity tolerance of DvH is exceeded ( Zhou et al. , 2013 )."
} | 5,748 |
37689064 | null | s2 | 1,490 | {
"abstract": "The symbioses that animals form with bacteria play important roles in health and disease, but the molecular details underlying how bacterial symbionts initially assemble within a host remain unclear."
} | 49 |
25626906 | PMC4324313 | pmc | 1,491 | {
"abstract": "ABSTRACT Biofilms are surface-attached multicellular communities. Using single-cell tracking microscopy, we showed that a pilY1 mutant of Pseudomonas aeruginosa is defective in early biofilm formation. We leveraged the observation that PilY1 protein levels increase on a surface to perform a genetic screen to identify mutants altered in surface-grown expression of this protein. Based on our genetic studies, we found that soon after initiating surface growth, cyclic AMP (cAMP) levels increase, dependent on PilJ, a chemoreceptor-like protein of the Pil-Chp complex, and the type IV pilus (TFP). cAMP and its receptor protein Vfr, together with the FimS-AlgR two-component system (TCS), upregulate the expression of PilY1 upon surface growth. FimS and PilJ interact, suggesting a mechanism by which Pil-Chp can regulate FimS function. The subsequent secretion of PilY1 is dependent on the TFP assembly system; thus, PilY1 is not deployed until the pilus is assembled, allowing an ordered signaling cascade. Cell surface-associated PilY1 in turn signals through the TFP alignment complex PilMNOP and the diguanylate cyclase SadC to activate downstream cyclic di-GMP (c-di-GMP) production, thereby repressing swarming motility. Overall, our data support a model whereby P. aeruginosa senses the surface through the Pil-Chp chemotaxis-like complex, TFP, and PilY1 to regulate cAMP and c-di-GMP production, thereby employing a hierarchical regulatory cascade of second messengers to coordinate its program of surface behaviors.",
"introduction": "INTRODUCTION Bacteria are able to live as members of surface-attached microbial communities called biofilms. The formation of a biofilm in pseudomonads begins with cells “reversibly” attaching to a surface and progresses to more stable “irreversible” attachment. The transition from reversible to irreversible attachment is the first committed step in forming a biofilm ( 1 ). Alternatively, once cells contact a surface, they also have the ability to move across the surface using either type IV pilus (TFP)-mediated twitching motility or flagellator-mediated swarming motility ( 2 – 4 ). A comprehensive understanding of these early events remains elusive. The intracellular second messenger cyclic di-GMP (c-di-GMP) is now appreciated as a master regulator for the transition between motile and sessile lifestyles in many bacterial species ( 5 ). In the case of Pseudomonas aeruginosa , studies to date have linked c-di-GMP and its effector proteins primarily to the control of flagellar motility and EPS production; that is, elevated levels of c-di-GMP stimulate biofilm formation by increasing EPS production and inhibiting flagellator-mediated swarming motility ( 4 , 6 – 9 ). The levels of c-di-GMP are modulated by two classes of enzymes: diguanylate cyclases (DGCs) and phosphodiesterases (PDEs). DGCs synthesize c-di-GMP from two molecules of GTP, while PDEs degrade c-di-GMP to pGpG or GMP ( 10 – 12 ). In studying the mechanisms of the early events in biofilm formation, we reported that c-di-GMP levels were elevated 3- to 5-fold in P. aeruginosa PA14 when cells were grown on an agar surface compared to this microbe grown in a liquid broth ( 13 ). Concomitant with this increase in c-di-GMP is the upregulation of the cell surface-associated protein PilY1, which shares sequence similarity with the pilus-associated adhesin PilC of Neisseria . PilY1 also contains a von Willibrand A mechanosensory domain ( 14 , 15 ). Genetic analyses showed that PilY1, as well as the DGC SadC, is required for the increased c-di-GMP in surface-grown cells and, furthermore, that PilY1 functions upstream of SadC in boosting c-di-GMP levels ( 16 ). These data point toward a central role for PilY1 in controlling surface behaviors by this microbe ( 17 ). In this study, we dissected the earliest event in biofilm initiation and explored the regulatory network of surface-induced PilY1 expression. Together, our data revealed a complex regulatory system, whereby P. aeruginosa senses the surface through the Pil-Chp chemotaxis-like complex, TFP, and the FimS-AlgR two-component system (TCS) and employs a hierarchical regulatory cascade of two different second messengers (cyclic AMP [cAMP] and c-di-GMP) to initiate its program of surface behaviors.",
"discussion": "DISCUSSION How cells sense and respond to a surface is a fundamental question in biofilm research. Our data show that the Pil-Chp complex, a chemotaxis-like complex located at the IM, is critical in surface sensing. Mutating the pilJ gene, which codes for an MCP-like protein, leads to a reduction in PilY1 expression, a loss in biofilm formation, and a hyperswarming phenotype. Mutation of ChpB, a methylesterase, presumably results in hypermethylation of PilJ and increased basal signaling through this MCP. In contrast to the pilJ mutant, loss of ChpB results in increased biofilm formation and a nonswarming phenotype, as well as increased PilY1 expression. Given the close regulation between the Pil-Chp complex and TFP synthesis ( 25 – 27 ), and the observation that PilY1 is not deployed until the TFP is assembled, it is tempting to speculate that the TFP or TFP-associated factor(s) may serve as a mechanosensor for a surface, a finding consistent with our observation that loss of pili results in reduced cAMP signaling. The observation that the hyperpiliated pilU mutant, defective in retraction, shows increased cAMP signaling suggests that active extension/retraction of the TFP is not required for cAMP signaling. Our data show that signaling through the Pil-Chp system to stimulate PilY1 production and downstream c-di-GMP-mediated regulation requires both stimulation of cAMP levels via CyaA/B and signaling through the FimS-AlgR two-component system, likely through direct physical interaction of PilJ and FimS. Our findings are in good agreement with previous reports that cAMP-Vfr upregulates transcription at the fimS-algR operon ( 24 ) and that FimS-AlgR is required for pilY1 operon expression, as the phosphorylated AlgR directly binds to the pilY1 operon promoter and activates its transcription ( 22 , 30 ). Our data also show that the TFP assembly complex is intimately involved in PilY1 production, secretion, and signal transduction. The TFP assembly complex is comprised of four interacting subcomplexes, including (i) the IM motor subcomplex (PilC, PilB, PilT, and PilU), (ii) the outer membrane (OM) secretin pore subcomplex (PilQ and PilF), (iii) the alignment subcomplex (PilM, PilN, PilO, PilP, and FimV) that bridges the IM and OM subcomplexes but is not required for pili assembly ( 31 , 39 ), and finally (iv) the pilus itself (PilA) plus the minor pilins FimU, PilV, PilW, PilX, and PilE ( 40 ). We showed that PilY1 is secreted through the TFP apparatus. Furthermore, our published studies show that cell surface-associated PilY1 can signal through SadC to control swarming motility ( 16 ), and this report shows that the PilMNOP alignment complex is also required for PilY1 to stimulate c-di-GMP production via SadC. Taken together, our studies show that PilY1 plays a central role in coordinating the surface behaviors in P. aeruginosa , and this function is exerted via two discrete regulatory pathways: (i) PilY1 in the IM represses its own transcription via modulation of cAMP signaling through Pil-Chp, cAMP-Vfr, and FimS-AlgR, and (ii) PilY1 on the cell surface is required to regulate c-di-GMP levels via the TFP alignment complex and the DGC SadC ( Fig. 1 ). In a broader context, we believe we are now at the beginning stage of answering the question of how cells sense a surface. Here, we propose that P. aeruginosa utilizes a stepwise regulatory circuit to detect and respond to surface growth; this circuit employs a hierarchical regulatory cascade of two second messengers (cAMP and c-di-GMP) to initiate this microbe’s program of surface behaviors. It appears that TFP assembly and the pilus are required for robust cAMP signaling, indicating that the low levels of pili found on planktonic cells participate in initial surface engagement. We propose that the FimS-AlgR system, perhaps with input from a direct physical interaction with PilJ, also is required for sensing surface engagement. Subsequent c-di-GMP signaling does not require the pilus per se but instead requires PilY1, whose expression requires Pil-Chp, cAMP-Vfr, and the FimS-AlgR TCS; and furthermore, PilY1 is secreted in a TFP-dependent manner. Thus, we can propose a hierarchical model of signaling because downstream c-di-GMP-mediated events via PilY1 cannot occur until PilY1 is synthesized, secreted, and localized to the cell surface. Furthermore, previous studies by Harwood and colleagues implicated surface-dependent clustering of WspR and phosphorylation of this protein, likely by a chemotaxis-like system, are required for surface-dependent stimulation of WspR activity ( 41 , 42 ). Given the timing of WspR clustering upon surface contact (~6 to 8 h) and the observation that cAMP signaling is stimulated as early as 2 h, we believe the system we describe here occurs upstream of WspR signaling. A key implication of our findings is that there may not be a single “surface signal” to trigger the surface program in P. aeruginosa . Instead, we suggest that there may be multiple reinforcing signals mediated via numerous sensing proteins (i.e., PilJ, TFP, FimS, PilY1, Wsp), and likely additional downstream factors, to continually reassure the bacterium that it has engaged and remains in contact with a solid surface. Blocking any of these inputs may result in a microbe that is not fully committed to life on a surface and thus offers a number of potential targets for antibiofilm therapeutics."
} | 2,435 |
38179456 | PMC10764424 | pmc | 1,492 | {
"abstract": "Microbial syntrophy, a cooperative metabolic interaction among prokaryotes, serves a critical role in shaping communities, due to the auxotrophic nature of many microorganisms. Syntrophy played a key role in the evolution of life, including the hypothesized origin of eukaryotes. In a recent exploration of the microbial mats within the exceptional and uniquely extreme Cuatro Cienegas Basin (CCB), a halophilic isolate, designated as AD140, emerged as a standout due to its distinct growth pattern. Subsequent genome sequencing revealed AD140 to be a co-culture of a halophilic archaeon from the Halorubrum genus and a marine halophilic bacterium, Marinococcus luteus , both occupying the same ecological niche. This intriguing coexistence hints at an early-stage symbiotic relationship that thrives on adaptability. By delving into their metabolic interdependence through genomic analysis, this study aims to uncover shared characteristics that enhance their symbiotic association, offering insights into the evolution of halophilic microorganisms and their remarkable adaptations to high-salinity environments.",
"conclusion": "Conclusion The comparative genomic analysis revealed intriguing similarities and differences between the halophilic Halorubrum sp. and M. luteus . Both organisms exhibit convergence in terms of the genetic components involved in their adaptation to high-salt environments. The shared presence of gene clusters related to osmo-adaptation and ion transport suggests convergent evolution in response to salinity. However, the presence of distinct genomic elements highlights the influence of their respective evolutionary histories. Along with the metabolic prediction, the data suggest that the two species partially complement each other’s metabolism, as well as greater complexity and environmental dependency than a simple, direct cross-feeding arrangement. Noteworthy, other organisms thriving in the same niche, such as bacteria, archaea, fungi, and protozoa, may also improve the survival, fitness, and metabolism of these strains. Moreover, models derived from metabolic network analysis improve understanding of the ecological complexity and dynamics within microbial communities, offering new insights into metabolism and novel culture-media alternatives. Future work to delve into the dynamic behavior of the co-culture is needed. To increase experimental tractability, time-lapse microscopy can enable us to visualize the physical interactions between archaea and bacteria, providing insights into spatial arrangements and cellular associations. Moreover, conducting stable isotope labeling techniques will help trace the flow of specific nutrients and the transfer of resources.",
"introduction": "Introduction Syntrophy, or cross-feeding, is widely distributed and occurs when a microbial strain consumes substrates produced by other microbial strains, promoting the growth of the cross-fed strain ( Escalante et al., 2015 ). It is central in structuring microbial communities because most microorganisms are auxotrophic ( Zengler and Zaramela, 2018 ), and cross-fed substrates provide essential resources that would be otherwise unavailable. Without syntrophic interactions, most microbial communities would collapse to relatively few prototrophic species. Syntrophic interactions were central in the evolutionary history of life, as eukaryotes are hypothesized to have evolved from a mutually beneficial syntrophy between an archaeon and a bacterium. Archaea and Bacteria are highly divergent, having long phylogenetically distinct evolutionary histories, and archaea are thought to be especially dependent upon syntrophic interactions ( López-García and Moreira, 2019 ; López-García and Moreira, 2020 ). Even so, species from both domains share many features, size, morphology, movement, and defense mechanisms and were long thought to be one taxonomic group ( Medina-Chávez and Travisano, 2022 ). Disentangling syntrophies between archaea and bacteria will shed light on current microbial communities and potentially on the evolutionary origins of eukaryotes. Recent research has unveiled the remarkable diversity within microbial mats from the Cuatro Cienegas Basin (CCB) ( Espinosa-Asuar et al., 2022 ), where similar to other extreme environments, the presence of archaea and extremophilic bacteria is prominent. CCB is an oasis in the Chihuahuan Desert in Coahuila, Mexico, characterized by its oligotrophic states, which involve phosphorus depletion, high salinity, high pH, and high radiation levels ( Souza et al., 2008 ; Medina-Chávez et al., 2020 ). Within CCB, a plethora of microbial diversity has been described in previous reports ( Bonilla-Rosso et al., 2012 ; Peimbert et al., 2012 ; Arocha-Garza et al., 2017 ). Due to its geographic isolation, species have evolved without interruption, resulting in a long and stable period of metabolic compatibility and adaptation for community assembly ( Souza et al., 2018 ). Notably, one halophilic isolate designated as AD140 exhibits a distinctive growth pattern on agar plates. First, yellow-colored colonies grow over a two-week span, followed by red colonies 4 weeks after the initial striking. This is unique since the red colonies do not grow on top of the agar media but rather directly on top the yellow biomass. This is in contrast to other isolates in co-culture where they do not grow in physical contact with each other. Through genome sequencing, strain AD140 was identified as a co-culture between a halophilic archaeon from the Halorubrum genus (red colonies) and a marine halophilic bacterium, Marinococcus luteus (yellow colonies). Despite numerous attempts over the years to separate both strains in axenic cultures, with the use of antibiotics, serial-dilution to extinction, spent media, different media composition, varying ph or temperature, neither microorganism thrive in isolation. These two lineages occupy the same abiotic ecological niche and have developed comparable mechanisms of adaptation and stress response. In addition to thriving in salt-saturated conditions, they demonstrate resilience during periods of desiccation and fluctuations in other environmental factors ( Medina-Chávez et al., 2023 ). The co-occurrence of these two strains as a sufficient symbiont interaction in laboratory settings led us to propose it as an early-stage symbiosis, where their symbiotic continuum depends on their adaptability in an eco-evolutionary framework. The objective of this study was to explore the genomic features of a halophilic archaea and marine bacteria consortium through comparative genomic analysis, with the aim of understanding the metabolic co-dependence that shapes the early stages of symbiosis. Symbiosis through metabolic cross-feeding is enhanced by genome features such as genomic plasticity, horizontal gene transfer events, symbiosis-specific genes, and regulatory mechanisms, leading to changes in the organization of functional elements in the genome ( Hentschel et al., 2000 ; Ochman and Moran, 2001 ; Moran, 2007 ; Juhas et al., 2009 ). Through comprehensive genomic analysis and metabolic network prediction, we describe the characteristics of these two microorganisms and their close interaction based on a set of shared traits. These findings have the potential to illuminate the convergent evolution of halophilic microorganisms and provide insights into the genetic basis of their adaptations to high-salinity environments.",
"discussion": "Discussion Syntrophic interactions are central to microbial communities and were foundational in the history of complex life. Despite this centrality in microbial ecology and evolution, the initial evolution and genomic basis of cross-feeding are challenging to determine. The vast majority of microorganisms cannot be cultured, limiting experimental investigation on most syntrophic interactions. And even for mixed species microbial cultures involving syntrophic interactions, determining shared substrates is daunting as microbes alter their environments in numerous diverse ways. Using a genomic approach through Next Generation Sequencing (NGS) technologies, we assessed the community context for growth of a microbial species pair. Obtained as a single isolate, neither species can be cultivated alone. When grown together in nutrient rich media, the culture reaches a high density in liquid and shows abundant growth on plates. A metabolic network analysis suggests that most auxotrophies in one species could be complemented by crossfeeding from the other species, but not all. This mix of auxotrophies and syntrophies suggests that the source microbial community contains a web of diffuse beneficial syntrophic interactions. The complexity of source microbial community is also reflected in other aspects of their genomes, including genes and genomic islands associated with secondary metabolite production, cell attachment, and horizontal gene transfer. These results suggest that the structure of microbial communities involves numerous competitive and mutualistic interactions, providing opportunities for beneficial symbioses to evolve. Comparing the genomes of symbiotic organisms by identifying conserved genes or specific adaptations unique to symbiotic partners, along with metabolic network predictions, can aid in addressing the remaining questions about symbiotic interactions like: “How do organisms in symbiotic relationships recognize and select suitable partners?” “What are the molecular mechanisms and signaling pathways involved in the recognition and establishment of symbiotic associations?” What are the evolutionary origins of symbiotic relationships? How did these complex interactions among organisms arise, and how have they evolved? Implementing these opens a new perspective into long-standing evolutionary questions about spatio-temporal microbial symbiosis and how cross-feeding affects diversity and speciation. Syntrophic evidence for an early-stage symbiosis Cultures of halophilic archaea and bacteria, like Halorubrum and Marinococcus , can be challenging to cultivate in laboratory settings. In this work, we proposed this syntrophic relation between Halorubrum sp. and Marinococcus luteus as an early-stage symbiosis due to their consistent co-occurrence in laboratory conditions, evolving to thrive as consortia, contributing to the community’s resilience and adaptability to harsh salinity conditions. One of the central questions in this work was whether these species engage in an obligate mutualistic relationship. To address their close relationship, we performed metabolic network modeling. It was found that each species had unique nutrient requirements. Importantly, most of these nutrient requirements could be produced by other species, though that was not true of all Halorubrum requirements. In the model, Marinococcus could not produce 4-hydroxyphenylacetate or thiamine. However, functional annotations coupled with media composition use to gap-fill the network, shows that Marinococcus , has the complete pathway to synthetize thiamine and 4-hydroxyphenylacetate. Furthermore, while both species had the capacity to produce the amino acid that the other species was auxotrophic for, this production itself depended upon the environmental availability of an amino acid that the producer itself was auxotrophic for. For instance, Marinococcus could only produce glutamine when provided with asparagine, and conversely, Halorubrum relied on asparagine to produce glutamine. This cooperative metabolic interaction can lead to increased growth rates, improved resource utilization, expanding ecological niches for the symbiotic microorganisms. Moreover, the prediction of the essential nutrients alone, suggests either the archaea or the bacteria requirements are partially fulfilled by the other, this representation might be attributed to poor annotation in archaeal references available. Archaea are distinguished by the existence of distinctive, modified versions of conventional pathways like the Entner-Doudoroff (ED), Embden-Meyerhof-Parnas (EMP), and novel pentose-degrading and CO 2 -fixing pathways catalyzed by new enzymes ( Sato and Atomi, 2011 ; Bräsen et al., 2014 ). Interestingly, the number of alternative media predicted to allow optimal growth is higher in M. luteus than the archaeon; given the genome plasticity, and the higher number of genes within the metabolic subsystems ( Figure 4 ), M. luteus is probably more efficient utilizing nutrients present in the media. Next, we asked whether each model could produce the compounds which the other is not capable of producing, by optimizing a sink reaction for that compound (see methods). In the model, we found that Marinococcus could produce L-Glutamate (gln__L_c), Undecaprenyl-diphospho-N-acetylmuramoyl-N-acetylglucosamine-L-ala-D-glu-meso-2,6-diaminopimeloyl-D-ala-D-ala (uaagmda_c) and N-Ribosylnicotinamide (rnam_c), but not 4 hydroxyphenylacetate (4hphac_c), or Thiamin (thm_c), however, the genome annotation show the presence of the elements needed for the pathway. Moreover, it could only produce as much N-Ribosylnicotinamide (rnam_c) as it was supplied with (nmn_e), potentially suggesting that this precursor is limiting to both species. On the other hand, Halorubrum can produce both L-Asparagine (asn__L_c) and N1-methylnicotinamide (nmn_e) during growth in a minimal medium. However, it could only produce as much N1-methylnicotinamide (nmn_e) as it was supplied with N-Ribosylnicotinamide (rnam_e). Other mechanisms and processes facilitating interdependence The shift from a free-living lifestyle to symbiotic interaction is occasionally facilitated by horizontal gene transfer (HGT) events, which frequently carry traits that enable the rapid exploitation of the environment ( Madsen et al., 2012 ; Drew et al., 2021 ). Archaea and Bacteria are usually found in mixed population biofilms, offering nutrient supply, protection, and boosting genetic variation through HGT events ( Fröls, 2013 ). HGT processes can be involved in sympatric speciation and in the high rate of genetic recombination on halophilic communities ( Ram Mohan et al., 2014 ). We found genomic evidence of biofilm production in both strains, since there are some gene homologs that participate in chemotaxis, EPS production, pili, and flagella. The initial recognition between the two strains can involve surface interactions through adhesins or pili-like appendages that facilitate attachment to each other or a common surface. These genes were identified in Halorubrum , with the N-terminal domain of the type IV pili from six adhesion pili PilA1-6 . Other studies have demonstrated that the expression of PilA1 , PilA2 , or PilA6 allows the microcolony formation, while Pil3 or Pil4 enhance surface adhesion ( Esquivel et al., 2016 ). Additionally, archaellum genes were identified, along with FlaF and FlaG gene-complex, which is required for archaellum assembly and motility, plus four accessory motor components, FlaC, FlaH, FlaI , and FlaJ ( Shahapure et al., 2014 ; Tsai et al., 2020 ). As for Marinococcus , two motility-related gene clusters fliEFGHIJKLMNOPQR and flgBCDEFGHIJKL were identified, along with their own regulatory systems. Moreover, similar Che gene sets were present, CheW and CheA led to the activation of CheY gene, which acts upon MotAB a flagellar motor complex, in charge of the rotation of bacterial flagella through conformational alterations ( Liu and Ochman, 2007 ; Athmika et al., 2021 ). Interestingly, several biofilm genes found in both strains have more than one copy, this duplication of the genome content offers an abundance of genetic material for evolution and is suggested to be crucial in adapting to novel environments and promoting diversification ( Xu et al., 2020 ). Meanwhile, some genes belonging to operon-complexes were not found, probably due to adaptive evolution trough gene loss, as a rapid response to environmental fluctuations. Genomic islands are also part of those exchanged elements, and another example of their interaction capacity is observed in the genome content of Halorubrum sp. and M. luteus , with accessory functions of symbiosis, secondary metabolites production, and antibiotic resistance (see Supplementary Tables 3 , 4 ). Particularly, Halorubrum comprises seven genomic islands related to symbiosis, while Marinococcus possesses four. According to the metabolic prediction analysis, thiamine constitutes one of the essential nutrients for the Halorubrum strain. Noteworthy, it was identified the presence of a set of islands with thiamine and sugar catabolism capacity, allowing the archaea to acquire and metabolize energy from compounds found in the environment ( Lacour et al., 2006 ). One of the singularities, is the presence of arsenic and cadmium resistant genes in both, Halorubrum sp. (island 19) and Marinococcus sp. (island 3). Previous studies have observed the abundance of arsenic and cadmium in the CCB ( Valdespino-Castillo et al., 2018 ; García-Ulloa et al., 2022 ). Arsenic resistance genomic islands have been characterized particularly in microbes that inhabit oligotrophic environments. Phosphate and arsenate are chemical analogs, substituting each other in chemical reactions ( Strawn, 2018 ). The arsenic metabolism genes can help the microorganisms to use low concentrations of phosphorus under high arsenic concentrations ( Li et al., 2013 ). The scarcity of phosphorus can reinforce arsenic uptake, while high concentrations impede arsenic uptake ( Strawn, 2018 ). Oligotrophic conditions are a main characteristic of the Cuatro Cienegas Basin environment ( Souza et al., 2008 ), strongly marked by a phosphorus limitation. Therefore, the evidence of genomic islands related to arsenic metabolism, might indicate an alternative way of Halorubrum sp. to survive in the CCB environment by up taking the arsenic available. Interestingly, both cadmium resistance and arsenic resistance often co-occur ( Parsons et al., 2020 ). Convergent evolution to an extreme environment The presence of the highly genetically divergent species Halorubrum sp., and M. luteus in CCB suggests that both microorganisms have acquired similar traits, using unique metabolic pathways and osmoregulation mechanisms, allowing them to thrive in high salt concentrations and fluctuating environments through convergent evolution events by evolving specialized transporters, ion channels, or regulatory systems. The Cuatro Cienegas Basin (CCB) has often shown the surprising presence of microorganisms limited to certain distinct environments, and the extant marine environment, along with its community, has been preserved ( Souza et al., 2006 , 2018 ). However, given the ancestral marine state of CCB, we might be able to explain the presence of some unexpected taxa, such as M. luteus , commonly found in salt lakes ( Wang et al., 2009 ; Fox-Powell and Cockell, 2018 ). The functional annotation of both strains revealed the presence of an osmotic regulation system composed of elements coding for compatible solutes such as trehalose, glycerol, glutamate synthesis. These solutes help maintain cell turgor to counteract osmotic stress. Independently, Halorubrum sp. has a complete Trk gene system in charge of potassium uptake, meanwhile, M. luteus has ectoine and betaine biosynthesis, as well as choline uptake regulation pathways, regulated by gbuA , gbuB , gbuC and nsmX genes. While identifying adaptative events can be challenging, the independent evolution of similar traits in multiple lineages is one of the strongest indications that adaptation has occurred. Despite the challenges and limitations of identifying adaptive events, this criterion can help us to understand better the mechanisms driving biological evolution and the relationship between microorganisms and their environment. Part of this suggested evidence is the presence of terpene biosynthetic clusters in both, Halorubrum and Marinococcus strains. This cluster is distinctive for each one; although both clusters have a phytoene synthase gene, the gene content and gene arrangement is different. Phytoene synthase gene is conserved in all terpene and carotenoid-producing archaea and in most carotenoid-producing bacteria. However, carotenoid and terpene biosynthesis pathways are more diverse in bacteria. The study of secondary metabolites in archaea is still rare, and most archaea only have one or two BGCs, usually a terpene or a bacteriocin BGC ( Wang et al., 2019 ). Most of the genes in bacteria associated with terpene biosynthesis belong to the crt gene family ( Sandmann, 2021 ). This correlates with the observed genes in the terpene BGC in M. luteus , as most genes in this cluster belong to the crt family. The terpene biosynthesis is directly involved in the carotenoid biosynthesis pathways, and carotenoid production is common in halophilic archaea and bacteria, as one of its functions stabilizes the cell membrane ( Sandmann, 2021 ). Perhaps surprisingly, we also observed little, or no, evidence of CRISPR-Cas elements. No CRISPR elements were observed in the Marinococcus genome. And in the Halorubrum genome, we have only a low evidence level, which can be sequencing artifacts (see Supplementary Table 5 ). Although the incidence of CRISPR-Cas elements in archaea is high, it has been reported some Halorubrum archaea do not have them. This might be due to the deterioration of CRISPR arrangements in the genome, which can occur during DNA repair processes or because the need for the system is not constant ( Fullmer et al., 2014 ). Likewise, the low concentrations of phosphorus present in the CCB can limit the plasticity of the genome, in addition to the fact that the transformation is reduced under these conditions ( Souza et al., 2008 )."
} | 5,478 |
40128227 | PMC11933406 | pmc | 1,493 | {
"abstract": "Current understanding of how woody plants respond to abiotic stress and how mycorrhizal interactions mitigate this stress is limited, as research has mostly focused on single stress factors. The diverse range of woody plants and mycorrhizal fungi, and the varying intensity and composition of multiple stress factors in different regions worldwide, have made it difficult to study these highly functional symbiotic interactions from a global perspective. Here, we used a top-down approach that involved partitioning known interactions into functional types, and mapping stress tolerances and interactions into overlapping heatmaps. We used a comprehensive dataset of 621 woody species’ tolerance of shade, drought, waterlogging, and cold stress, as well as their mycorrhizal interaction data, to test how stress polytolerance correlates with different functional types of mycorrhiza. We show that single mycorrhizal type associates with shade tolerance, while dual type with cold and waterlogging tolerance. Both arbuscular mycorrhiza and obligate interactions are more abundant in drought stress tolerance conditions, while ectomycorrhiza and facultative interactions are found in more cold and waterlogged stressful conditions. Thus, functionally distinct mycorrhizal interactions form significantly contrasting stress mitigation patterns with woody species, providing insights into both evolutionary and biogeographic patterns related to the development of plant-mycorrhiza interactions.",
"introduction": "Introduction Mycorrhizal symbiosis is an ancient association between plants and fungi, central to plant evolution and expansion. It is crucially important for both herbaceous and woody plants (trees, bushes, shrubs, lianas), the latter of which constitute nearly half of the total vascular plant diversity 1 . Mycorrhizal fungi form a complex and multifaceted symbiosis with plants, providing beneficial stress-mitigating services to plants. This happens mainly by directly supplying plants with additional resources like water and nutrients, but also indirectly mitigating non-resource stress, e.g. by suppressing pathogens in the soil, and retaining soil water more efficiently in dry conditions 2 – 4 . Services provided by mycorrhizal fungi can vary depending on their functional affiliations and biogeography 2 , 5 . Recent literature has begun to address these distinctions, with the emergence of comprehensive studies on the niche of mycorrhizal fungi 6 . In this study we assess various properties of plant-mycorrhizal interactions in woody plants, treating them as traits or properties of plants, who can have either single or dual type interaction; interaction with arbuscular mycorrhiza (AM) and/or ectomycorrhiza (ECM); or whose mycorrhizal interactions are either obligate or facultative. Plants have traditionally been considered to form symbiosis with a single mycorrhizal type. Each mycorrhiza type hosts a particular combination of biotic partners along with distinguished morphological and functional characteristics 2 , 7 . However, numerous plants (89 genera from 32 families) are now known to form interactions with both AM and ECM, which are the two dominant types of mycorrhizal associations 8 . AM and ECM interactions provide different benefits for woody plants. ECM interactions, especially in temperate and boreal forests, likely play a more significant role in young soils with N-limiting conditions exhibiting a relatively restricted organic-nutrient economy. In contrast, AM interactions are more prominent in older forests with weathered soils under P-limiting conditions hosting a more open inorganic-nutrient economy 9 . Thus, we could expect that dual mycorrhizal interactions are more common in intermediate conditions, where both interaction types would be beneficial. These dual mycorrhizal type plants are predominately woody species (84%), distributed across the world where both mycorrhizal types overlap 8 . In addition to the types of mycorrhizal interaction, this symbiosis can also have different statuses, either as obligate or facultative plant symbionts 2 , 10 , 11 . Plants with facultative interactions are expected to better adapt to different environmental stresses and contexts 11 , 12 . They are also known for having higher invasion success 13 and longer dispersal rates 14 , which can be explained by larger niche size of facultative than obligate plant symbionts 15 . Considering the variability of stress mitigating services provided, and the biogeography and the functional properties of both the mycorrhiza fungi and the plants, we can expect the patterns of interactions within this symbiosis to be significantly affected by the properties of both organisms and also their distributions. Abiotic stress polytolerance is a key shaper of species distribution, particularly for sessile organisms like plants 16 – 18 . Lifelong aboveground plant structures, that define woody species, are often concurrently up against multiple abiotic stress factors like temperature extremes, water and light availability, which are considered to be the dominant abiotic stress factors 18 , 19 . Long-term exposure to multiple environmental limitations, inherent of any habitat type, shape woody plants tolerance strategies that are defined by differential contribution of different tolerances to given stress factors to the overall tolerance strategy 20 . However, recent studies have demonstrated that these trade-offs are much less exclusive and strict than postulated so far, leaving significant wiggle room for gaining polytolerance through adaptations 17 , 18 , 21 . This was clearly shown in a comprehensive analysis by Puglielli and others 18 who analyzed abiotic polytolerance patterns of 799 Northern hemisphere woody species (constituting ~ 40% of woody diversity of Northern hemisphere). Using principal component analysis (PCA) to determine the dimensions of woody species’ abiotic stress tolerance, they found that there is a triangular-shaped abiotic stress tolerance space (STS). In this STS the first dimension reflects a trade-off between drought- and cold/waterlogging tolerance strategies, while the second dimension reflects a shade-tolerance strategy spectrum, from low to high shade tolerance, which is independent of the first dimension. This STS can be used as a coordinate system to link woody plants abiotic tolerance strategies with any ecological dimensions 20 , 22 . The traditional view on how mycorrhizal symbiosis benefits woody species suggests that while woody plants require mycorrhiza to survive, herbaceous plants need them to thrive 23 . Most research on mycorrhiza has been focused on how this interaction benefits plant performance (e.g. faster growth), however plant survival aspects and non-trophic mycorrhizal benefits are relatively understudied in comparison 24 . Thus, we can expect woody species to have more nuanced adaptive interactions with mycorrhizal fungi. This assumption is reflected in the higher phylogenetic diversity of ectomycorrhizal fungi that are mainly interacting with woody plants 5 , while arbuscular mycorrhizal fungi, that is relatively omnipresent in both woody and herbaceous plants, have much lower phylogenetic diversity and endemism rates 25 . However, the interactions between plants and mycorrhizal fungi, and the benefits of these interactions, are much less studied in woody species than in herbaceous plants 4 . This study is the first large-scale assessment of how woody plants´ abiotic stress strategies relate to symbiosis types of mycorrhiza. What do we know about mycorrhizal mediation of abiotic stress in woody plants? Our current understandings of how plants respond to abiotic stress and how mycorrhiza mitigates this stress are almost exclusively limited to single stressors 4 , 26 . The main stress factor studied in this context is drought stress – interactions with mycorrhizal fungi are known to improve soil water retention capacity 27 . Hyphae are more efficient than plant´s fine roots in absorbing both water and nutrients from smaller soil pores 2 and directly provide water to plants 28 , but see 29 . On one hand, ECM has shown to have larger and more extensive extraradical mycelium, which can reach farther soil water pockets 4 , 30 . In contrast, AM interactions provide woody species more functions in dealing with drought – enhancing plants´ physiological and biochemical functioning 31 , 32 . Thus, AM interactions are considered more effective at mitigating drought stress than ECM interactions 4 . Actually, gymnosperm-dominated ecosystems—typically characterized by fewer fine roots than those dominated by angiosperms 33 —may have maintained a reliance on AM interactions rather than ECM associations to cope with drought stress, especially in regions that are predominantly or seasonally dry. This is particularly evident for gymnosperms outside the Pinaceae family. Very similar mechanisms are assumed to work in the case of tolerating high and low temperatures, as these stress factors manifest also through obstructing water availability in ecosystems 34 . Accumulation of sugars in plant cell walls helps to maintain osmotic balance, but also to avoid carbon starvation in low photosynthetic periods 35 . As colder climatic conditions are known to favor ECM interactions over AM interactions, on both latitudinal 9 and altitudinal gradients 36 , most probably due to the saprophytic capabilities of ECM fungi, it can be expected that ECM interactions dominate the peripheral zones of STS, in the cold/waterlogging tolerance end, while AM dominates in peripheral zones of STS with more drought stress and moderate cold levels. In turn, facultatively mycorrhizal plants increase following lower temperatures at higher latitudes 11 and elevations 36 . Therefore, it is expected to be associated with adaptations to cold tolerance, given that mycorrhizal associations may be limited in more extreme cold conditions due to a lower photosynthetically active radiation, which could limit the surplus of C for mycorrhizal associations along with possible limitations on the availability of nutrients for plant growth 37 . The relatively high carbon cost for maintaining mycorrhiza typically leads to reduced root colonization of symbiotic fungi 38 . Although this response is known to vary between plant species and their developmental stage 39 , it might dominate in plant species with obligate mycorrhizal interactions that occur in sparser vegetation types; or in particular stages of life, as for example the seedlings of woody species are known to tolerate shade much better than adult individuals of the same species 40 . Mycorrhizal symbiosis seems to regulate or buffer limiting resource uptake in changing light conditions, but there is a lack of mechanistic and quantitative understanding behind these processes 41 ; thus, the universality of this assumption is unclear. Functional diversity of both woody plants and mycorrhizal fungi interacting with them in different regions of the world, makes it difficult to predict the patterns of these interactions from both the adaptational and mitigational point of view as the strength and composition of abiotic stress varies along the environmental gradients 4 . In addition, the possibility of dual mycorrhizal interactions is still debated, and this discussion is heavily related to what constitutes a mycorrhizal interaction – when does a fungal colonization become an interaction 8 , 15 . We followed suit, as most mycorrhizal studies, including data papers, consider colonization as a confirmation of interaction 8 , 11 . In this study, we used a top-down approach, where known interactions are partitioned into functional and biogeographical groups, and then interactions are mapped into the STS, thus generating heatmaps drawing large-scale correlative patterns of these associations. In order to shed light on the patterns of these interactions on a global scale we built a comprehensive dataset of 621 woody species stress tolerance (based on stress tolerance space of shade, drought, waterlogging and cold stress 18 ) and their known species-specific mycorrhizal interactions. We grouped the plant species (based on 17 , 19 ) according to their life form (angiosperms vs. gymnosperms), growth form (deciduous angiosperms; evergreen angiosperms; evergreen gymnosperms), and biogeographic affinity (Europe, North America, East Asia). Mycorrhizal traits were assigned based on the identity of symbionts and their mycorrhizal structures observed in plant roots ( sensu 11 ): single vs. dual type; arbuscular vs. ectomycorrhiza; and obligate vs. facultative mycorrhiza. Ericoid and non-mycorrhizal types were omitted due to infrequent occurrence in the dataset. We expected the following patterns of mycorrhizal interactions in relation to abiotic stress in woody species: (1) dual mycorrhizal type is predominantly associated with woody species tolerating cold and waterlogging stress, as these stress factors indicate overall more severe climatic conditions and shorter vegetation season – therefore symbiotic interactions in these dimensions are more abundant and diverse; (2) similarly, ectomycorrhizal interactions may dominate in the cold and waterlogged periphery of stress tolerance space, while arbuscular mycorrhiza may be more adapted to drought conditions; (3) obligate interactions between mycorrhiza and woody species dominate the STS with the exception of cold conditions, given the multiple possible limitations of cold stress to consistently maintain mycorrhizal associations.",
"discussion": "Results and Discussion Our results confirmed significant contrasting patterns between single vs. dual type (Fig. 1 ), arbuscular vs. ectomycorrhiza (Fig. 2 ), obligate vs. facultative mycorrhizal interaction in relation with the abiotic stress tolerance strategies (Fig. 3 ). In addition to different mycorrhizal interaction patterns among angiosperms and gymnosperms shown in the abovementioned figures, the patterns also differed depending on the biogeographical origin of the woody plant species (Fig S2 – S4 , Table S1 ). \n Fig. 1 Single versus dual type interaction plotted in the abiotic stress tolerance space ( sensu 18 ) of woody species. Panels ( a , b ) show the results based on the complete dataset, while panels ( c – f ) show the results separately for angiosperms and gymnosperms, for single and dual type interactions respectively. Color scale indicates the probabilistic distribution of trait combinations in the functional trait space created by PCA (yellow = high probability; violet = low probability). Contour lines are the quantiles of response variable predictions. Abiotic stress space dimensions are indicated with lines: ST shade tolerance; WT waterlogging tolerance, CT cold tolerance, DT drought tolerance. Kernel density heatmaps for the complete dataset analysis are in (Figure S1 ). \n Single type interactions were associated with the shade tolerance tips of the STS (Fig. 1 a), while dual type interactions with cold and waterlogging (Fig. 1 b). Considering that cold and waterlogging stress indicate shorter vegetation season, which means more frequent abiotic stress conditions 17 , and cold stress also limits water availability, we expected dual type interactions to predominate in these parts of STS. However, this pattern was mainly driven by angiosperm species (Fig. 1 e), while gymnosperms did not extend to that portion of the STS. Further, dual gymnosperms were associated with drought tolerance (Fig. 1 f). One explanation for the differences between single and dual mycorrhizal associations is the varying nature of aboveground and belowground stresses. Belowground, plants encounter a range of stressors—excess water, limited nutrients, or reduced water availability due to lower temperatures—while light is a more specific stress 42 . A single mycorrhizal association may specialize in managing one stress (such as excess light), whereas a dual association can simultaneously address multiple stresses by balancing the benefits of different fungal partners 8 . In turn the higher shade tolerance observed in many single mycorrhizal gymnosperms of the northern hemisphere is the result of a combination of evolutionary history and physiological strategies. Their long-lived, needle-like leaves, efficient low-light photosynthesis, and conservative resource allocation allow them to maintain a positive carbon balance under shaded conditions—a strategy that contrasts with the typically faster but less shade-adapted growth strategy of many angiosperms 43 . In addition, our results reveal divergent trends in interaction duality. Specifically, under conditions of cold and waterlogging stress, angiosperms engage in interactions with both arbuscular mycorrhizal (AM) and ectomycorrhizal (ECM) fungi. In contrast, gymnosperms predominantly employ dual interactions to alleviate drought stress, and potentially nutrient deficiency as well. \n Fig. 2 AM versus ECM type interaction plotted in the abiotic stress tolerance space of woody species. Figure configuration follows (Fig. 1 ). \n Our second hypothesis found support as well: AM was clustered in regions with higher drought stress and moderate cold stress (Fig. 2 a), while ECM is more common in cold and waterlogged peripheral areas of STS (Fig. 2 d). However, it depended significantly on life form—angiosperms were mildly driving the central clustering in AM (Fig. 2 b), while gymnosperms did the same in ECM (Fig. 2 f), even though the latter pattern was statistically not significant. Drought dimension, that is the most studied abiotic stress parameter in plants 4 was noticeable only in case of AMs for gymnosperms (Fig. 2 c), however this pattern was not statistically significant, probably due to relatively small sample size for such clustered analysis. The different patterns reflect, in a way, the traditional view on how mycorrhiza would benefit woody species—“Most woody plants require mycorrhiza to survive, and most herbaceous plants need them to thrive” 23 . While widespread AM generalist species 25 provide woody plants with opportunities to spread, much more endemic and specialist ECM species 5 mitigate the stress for better survival rates. Because of this, we also expect this hypothesis to apply only in the northern hemisphere, and not in the southern, where the dominating abiotic stress patterns are different. \n Fig. 3 Obligate versus facultative type interaction plotted in the abiotic stress tolerance space of woody species. Figure configuration follows (Fig. 1 ). \n As expected, obligate interactions are more abundant in the periphery of STS, especially in the shade tolerance end (Fig. 3 a), as these interactions can render trees more shade tolerant by optimizing nutrient uptake and conserving carbon—a combination that is especially beneficial in the low-light, nutrient-poor conditions typical of many forest understories 2 . In contrast, facultative interactions that were more clustered in the center of STS and in the cold stress end (Fig. 3 d), excel in cold environments likely by flexibly optimizing nutrient uptake and conserving carbon when fungal activity is limited. Despite the lack of mechanistic and quantitative understanding in how mycorrhiza regulates limiting resources in changing light conditions 41 , obligate interactions in both angiosperms (Fig. 3 b) and gymnosperms (Fig. 3 c) were indeed mainly centered around the shade tolerance dimension, though the pattern in case of gymnosperms was not statistically significant. This trend appears to be consistent among woody plant species, albeit specifically within obligately mycorrhizal interactions. The concentration of obligate interactions in the peripheral areas, is relatively similar to ECM (Fig. 2 d), and the woody species with ECM and obligate interactions are indeed correlated (Pearson’s r : 0.36, p = < 0.001). Although obligately mycorrhizal associations avoid cold and waterlogged conditions and ECM species can be adapted to it. Although plants with facultative status are known to be better at long-distance dispersal 14 and their niche tend to be wider 15 , facultative species have not established themselves in all stressful conditions, for example in the shade tolerant end. There seems to be a trade-off, similar to generalist vs. specialist plant species, between tolerating extreme stress with specific adaptations and capability of dispersing and invading new communities and habitats. Overall, the correlative patterns of mycorrhizal interactions in relation to woody species abiotic stress tolerance depend significantly on the life form and the mycorrhizal trait preferences of the plant species. Different combinations of plant species and the functionality of the symbiotic association provide different types of services that are contingent on the tolerance strategy towards a specific abiotic stress factor. Deeper reliance on mycorrhizal interactions, whether expressed in dual type (Fig. 1 ), more species-specific types (Fig. 2 ), or status (Fig. 3 ), is significantly related to tolerating more extreme abiotic conditions, although the specific stress dimension is subject to the specific functional combination of both interaction participants. The ability to form a dual mycorrhizal interaction with AM and ECM fungi provides the necessary resources for angiosperms to survive in cold and waterlogging habitats. These environments are marked by shorter vegetation periods and the simultaneous presence of various stress factors. The benefits of more intimate interaction between ECM (which are more expensive than AM) and angiosperms and gymnosperms are apparent as this interaction mitigates cold and waterlogging stress. At the same time, obligate status may help to mitigate shade tolerance in both angiosperms and gymnosperms. The duality of mycorrhizal interactions might therefore reflect the variability of benefits from these interactions, ranging from nutritional benefits to non-trophic benefits (from soil water holding capacity to pathogen defense) that tend to be more associated with surviving stressful conditions. Similarly contrasting adaptational differences between gymnosperms and angiosperms in the abiotic stress tolerance space have previously appeared in biomass allocation patterns 21 and in the trait dimensions of global spectrum of plant form and function 22 . In summary, woody species occupying the extreme ends of stress dimensions need more abundant and intimate interactions with both AM and ECM in order to facilitate survival, while woody species in areas of STS have adapted to more moderate stress conditions and might use the symbiosis for enhancing the competitive edge. The survival-oriented combinations of mycorrhiza and woody plants seem to have strong roots in evolutionary and biogeographic history, and the thriving-oriented combinations in moderate or symbiotic limiting (cold) stress conditions are more sporadic and voluntary. This means that in the ongoing fast climatic changes, the species located in currently moderate stress conditions might also experience difficulties in finding the optimal mycorrhizal association to interact with, especially in case of ECM interactions, while the woody species in the extreme ends of stress dimensions could turn out to be more resilient to changes, as they already have multiple and intimate symbiotic relationships with mycorrhiza."
} | 5,876 |
36838875 | PMC9960984 | pmc | 1,494 | {
"abstract": "3-Hydroxypropionic acid (3-HP) is a platform chemical with a wide range of existing and potential applications, including the production of poly(3-hydroxypropionate) (P-3HP), a biodegradable plastic. The microbial synthesis of 3-HP has attracted significant attention in recent years due to its green and sustainable properties. In this paper, we provide an overview of the microbial synthesis of 3-HP from four major aspects, including the main 3-HP biosynthesis pathways and chassis strains used for the construction of microbial cell factories, the major carbon sources used for 3-HP production, and fermentation processes. Recent advances in the biosynthesis of 3-HP and related metabolic engineering strategies are also summarized. Finally, this article provides insights into the future direction of 3-HP biosynthesis.",
"conclusion": "6. Conclusions and Future Prospects In recent years, a large number of studies have achieved the efficient conversion of renewable substrates into 3-HP using engineered microorganisms. As outlined in this paper, the design of microbial cell factories that can efficiently produce 3-HP is determined by selecting an appropriate biosynthetic pathway, production chassis, and high-quality carbon source. With continuing research on the microbial synthesis of 3-HP, the range of substrates that can be used by microbial cell factories has been gradually increased. Initially, most studies used glycerol and glucose as substrates, after which lignocellulose, acetate, ethanol, fatty acids, and other alternative substrates were also used for 3-HP synthesis. Researchers are increasingly using renewable feedstocks that are cheaper, easier to obtain, and more environmentally friendly. Combining all reports on the microbial production of 3-HP, a conclusion was drawn that the fewer catalytic reactions required to move from substrate to product, the easier it seems to be to obtain high titers of 3-HP. For each 3-HP production pathway that has been developed, there should be one or more suitable carbon sources (inexpensive, converts to 3-HP by fewer reactions, and can be rapidly assimilated) corresponding to it. There is no doubt that glycerol is the most suitable carbon source for the glycerol pathway, and high 3-HP titers have been achieved in K. pneumoniae and P. denitrificans that are able to naturally and rapidly metabolize glycerol. In particular, the 3-HP productivity of the P. denitrifiers cell factory was excellent, reaching 2.5 g/(L·h) (which is the minimum requirement for the microbial 3-HP production to be economically competitive) [ 50 ]. Reports on the 3-HP production by recombinant P. denitrifiers are scarce, and the production process needs to be further explored. For the malonyl-CoA pathway, it is more appropriate to use substrates that can be efficiently converted to acetyl-CoA, such as acetate and fatty acids. The synthesis of 3-HP from the above substrates is mainly performed in E. coli due to its broad substrate spectrum and clear genetic background. The assimilation of acetate or fatty acids by E. coli still needs to be further strengthened. The β-alanine pathway seems to be less competitive compared to the two pathways mentioned above because of the longer reaction route. A simplified pathway, the oxaloacetate pathway, was recently explored in S. cerevisiae , which showed excellent 3-HP production ability. It is worth testing the effects of the oxaloacetate pathway in other chassis. Finding an inexpensive carbon source that can be efficiently converted to oxaloacetate in a few steps is also essential in the future. In addition, engineered strains need to be matched with appropriate fermentation processes to avoid the accumulation of by-products and to balance production and growth. Dynamic regulation devices have also been developed to address these issues, and good 3-HP titers were achieved in each study [ 9 , 28 ]. Designing or modifying dynamic conditioning devices is an effective strategy to improve the efficiency of engineered strains in the future. Third-generation microbial refineries that synthesize 3-HP directly from CO 2 have been successfully constructed, but the titers are currently at the mg/L level. Further modification of the third-generation microbial refineries for the synthesis of 3-HP to achieve high titers is necessary to unlock the future prospects of environmentally friendly and sustainable production. The efficiency of photosynthesis plays a decisive role in the productivity of cell factories based on cyanobacteria. In the future, it will become possible to optimize the components and pathways of the cyanobacterial photosynthetic system to enhance carbon uptake, sequester carbon, and reduce carbon loss [ 103 ].",
"introduction": "1. Introduction 3-Hydroxypropionic acid (3-HP) is an important bulk chemical with a wide range of applications in chemical synthesis. Since it contains two reactive groups—hydroxyl and carboxyl—one of the important applications of 3-HP is its polymerization to yield poly(3-hydroxypropionate) (P-3HP). Due to its biodegradability and biocompatibility, P-3HP is one of the widely studied substitutes for petrochemical plastics, with great application value and development prospects [ 1 ]. In addition, 3-HP can undergo redox reactions to produce 1,3-propanediol (1,3-PDO), acrylic acid, malonic acid, acrylamide, acrylonitrile, etc. [ 2 ], which are widely used in the production of adhesives, plastic packaging, fibers, and cleaning agents. In particular, acrylic acid occupies the highest market, with a value of USD 12 billion in 2020, which is projected to be around USD 19.2 billion by 2030 [ 3 , 4 ]. Due to the significant commercial application potential of 3-HP, the market value and size are estimated to be more than USD 10 billion/year and 3.6 million tons/year, respectively [ 3 ].Currently, 3-HP production is mainly based on chemical synthesis from fossil feedstocks, but chemical production is costly and environmentally unsafe. 3-HP was listed as one of the most promising bulk chemicals that can be obtained from renewable materials by the U.S. Department of Energy [ 5 , 6 ], which can be achieved using engineered microorganisms. In the past few years, several industrial parties, such as OPX Biotechnologies, Cargill, Novozymes, and BASF, have developed pilot plants for the bio-based production of 3-HP and its chemical conversion to AA [ 4 , 7 , 8 , 9 ]. A series of genetically engineered microbial cell factories to convert renewable substances into 3-HP have been constructed via different biosynthetic pathways [ 3 , 10 , 11 ]. However, microbial 3-HP synthesis processes still has a certain distance to travel meet the market demand. The most important strategy for developing microbial 3-HP synthesis processes is the selection of suitable microbial chassis strains, effective metabolic pathways and economical substrate, but the effects of fermentation processes on overall process economics should also not be ignored. In this paper, we give a brief overview of the main 3-HP biosynthesis pathways that have been used for the development of microbial cell factories. Notably, we summarize the excellent cases of microbial 3-HP production from both chassis and substrate aspects. In this way, readers will have a clearer understanding of the current status of 3-HP production, i.e., what are the 3-HP production pathways, chassis and substrates, and researchers who want to establish 3-HP microbial cell factories will be able to better select and combine them. In terms of chassis, we summarize the excellent chassis used for 3-HP production and introduce their characteristics in the production process. In terms of substrates, we not only summarize the excellent cases of common carbon sources (glycerol and glucose) in 3-HP production, but also outline the cases of other emerging carbon sources, such as xylose, acetate, fatty acids, ethanol, and CO 2 ."
} | 1,983 |
34718804 | PMC8825309 | pmc | 1,495 | {
"abstract": "Abstract Lignin, a polyphenolic polymer, is a major chemical constituent of the cell walls of terrestrial plants. The biosynthesis of lignin is a highly plastic process, as highlighted by an increasing number of noncanonical monomers that have been successfully identified in an array of plants. Here, we engineered hybrid poplar ( Populus alba x grandidentata ) to express chalcone synthase 3 ( MdCHS3 ) derived from apple (Malus domestica ) in lignifying xylem. Transgenic trees displayed an accumulation of the flavonoid naringenin in xylem methanolic extracts not inherently observed in wild-type trees. Nuclear magnetic resonance analysis revealed the presence of naringenin in the extract-free, cellulase-treated xylem lignin of MdCHS3-poplar, indicating the incorporation of this flavonoid-derived compound into poplar secondary cell wall lignins. The transgenic trees also displayed lower total cell wall lignin content and increased cell wall carbohydrate content and performed significantly better in limited saccharification assays than their wild-type counterparts.",
"conclusion": "Conclusion By expressing MdCHS3 in lignifying xylem tissue of hybrid poplar, we have produced transgenic trees with reduced total lignin content and increased cell wall carbohydrate content. These trees display substantially improved saccharification rates after both no pretreatment and dilute acid pretreatment. In addition, MdCHS3- poplar exhibits no differences in growth or biomass yield compared to WT and produces naringenin, a valuable flavonoid compound in xylem tissue. Moreover, we have identified incorporation of naringenin into the poplar wood lignins, demonstrating that if lignin-compatible flavonoid compounds can be produced in lignifying tissue of poplar, these compounds could be incorporated into lignin thereby potentially making them available on a high scale. Moving forward, MdCHS3 -poplars represent a useful genetic background into which many additional flavonoid biosynthetic enzymes may be introduced in order to produce other valuable lignin-compatible flavonoid compounds.",
"introduction": "Introduction Lignin, a major chemical constituent of lignocellulosic biomass, poses a significant barrier to the efficient industrial processing for the production of pulp and paper, specialty chemicals and fibers, and liquid biofuels. However, this complex polyphenolic polymer may serve as a chemical precursor in the development of new bio-based materials, high-value polymers, and chemicals. Although fast-growing woody feedstocks, such as poplar, willow, and eucalyptus, represent abundant and renewable sources of lignocellulosic biomass, narrow profit margins continue to limit the economic feasibility of employing them as dedicated energy crops at an industrial scale ( Mahon and Mansfield, 2019 ). Lignin is typically derived from three canonical monolignols: p -coumaryl, coniferyl, and sinapyl alcohols, which undergo oxidative coupling in the developing cell wall to form polymeric lignin. Efforts to genetically engineer the core monolignol biosynthetic pathway have led to significant changes in both content and composition of lignin, highlighting the remarkable metabolic plasticity of this biosynthetic pathway ( Ralph et al., 2004 ; Leple et al., 2007 ; Coleman et al., 2008a , 2008 b; Sykes et al., 2015 ; Chanoca et al., 2019 ). Moreover, a wide array of noncanonical monolignols has recently been found to naturally incorporate into lignins of different plant species ( Vanholme et al., 2019 ). For example, the stilbenoid compounds, resveratrol, piceatannol, and isorhapontigenin, have all been identified as monomers in lignins of palm fruit endocarps ( del Río et al., 2017 ), and their respective stilbene glycosides have also been identified in the lignins of Norway bark ( Rencoret et al., 2019 ). Hydroxycinnamamides, specifically ferulamides, have been shown to incorporate into plant lignins, behaving as lignin monomers ( Negral et al., 1996 ; del Río et al., 2020 ). Also, the high-value flavonoid tricin, reported to have a wide variety of potential pharmaceutical applications ( Li et al., 2016 ), was found incorporated at the ends of lignins in many monocots ( del Río et al., 2012 ; Lan et al., 2015 ). Recently, disruption of flavone synthase II ( fnsII ) in rice resulted in the accumulation of naringenin, a flavanone precursor to tricin, and the subsequent occurrence of naringenin in the lignin-enriched cell wall fraction, indicating that other flavonoids could be engineered into grass lignins as well ( Lam et al., 2017 ). Chalcone synthase (CHS) catalyzes the first committed reaction in the production of flavonoid compounds by combining p -coumaroyl-coenzyme A (CoA), a precursor in the monolignol biosynthetic pathway, with three malonyl-CoA units to produce naringenin chalcone, which is then cyclized to naringenin ( Figure 1 ). Previous genetic manipulations in plants have shown that the flavonoid and monolignol biosynthetic pathways are tightly linked. For example, RNAi-mediated silencing of an important monolignol biosynthetic gene, hydroxycinnamoyl-CoA:shikimate hydroxycinnamoyl transferase , in Arabidopsis has led to the accumulation of flavonoids ( Hoffmann et al., 2004 ). Similarly, downregulation of a monolignol biosynthetic gene, caffeoyl-CoA O-methyl transferase , in alfalfa ( Medicago sativa L.) resulted in the accumulation of isoflavonoids, the predominant class of flavonoids in legumes ( Gill et al., 2018 ). Conversely, silencing of CHS in maize ( Zea mays ) resulted in drastically reduced levels of the flavonoids apigenin and tricin, yet caused a significant increase in total lignin content of leaves ( Eloy et al., 2017 ). Taken together, these results indicate that CHS plays an important role in directing carbon flux between monolignol and flavonoid pathways. In poplar, transgenic downregulation of 4-coumarate:CoA ligase (4CL), another key monolignol biosynthetic gene, resulted in a reduction of lignin and led to the accumulation of the naringenin and kaempferol in stem extractives of transgenic trees ( Voelker et al., 2010 ). Notably, naringenin and kaempferol were not observed in extractives of wild-type (WT) stem tissue, indicating that CHS expression may be low in stem tissue ( Voelker et al., 2010 ). To confirm this, we evaluated the expression of all six previously identified putative poplar CHS genes ( Zavala and Opazo, 2015 ) in WT poplar trees used for genetic transformation ( Populus alba x grandidentata ). We determined that expression levels were considerably lower in xylem tissue compared to leaf tissue, where flavonoids are known to accumulate ( Supplemental Figure S1 ; Tian et al., 2021 ). Thus, overexpression of CHS in lignifying tissues of poplar, which does not appear to contain high levels of flavonoids, may serve to reduce lignin by redirecting carbon flux away from monolignol biosynthesis while simultaneously producing high-value flavonoids that could be incorporated into the lignins of important potential bioenergy crops such as poplar, adding further value to woody feedstocks. Figure 1 Biosynthesis of naringenin in xylem tissue. Phenylalanine is produced in the plastid via the shikimate pathway and transported into the cytosol where it is deaminated by phenylalanine ammonia-lyase to produce cinnamate, which is then hydroxylated by cinnamate 4-hydroxylase (C4H) producing p- coumarate. p- Coumarate is converted to p- coumaroyl-CoA by 4CL. CHS then combines three molecules of malonyl-CoA with p- coumaroyl-CoA producing naringenin chalcone, which is isomerized to (2 S )-naringenin by CHI. p- Coumaroyl-CoA and p- coumarate are both important precursors in the biosynthesis of monolignols: p- coumaryl alcohol, coniferyl alcohol (CA), and sinapyl alcohol. To this end, we have genetically engineered hybrid poplar ( Populus alba x grandidentata ) to express a previously characterized CHS gene ( MdCHS3 ) derived from apple ( Malus x domestica ) using a xylem-specific promoter ( Yahyaa et al., 2017 ). MdCHS3 (accession number NM_001328985) was selected for expression in lignifying tissues as it shows high substrate affinity for p- coumaroyl-CoA relative to reported K m values for competing poplar enzymes in the lignin biosynthetic pathway ( Wang et al., 2018 ). MdCHS3 was also reported to display greater substrate specificity for p- coumaroyl-CoA over cinnamoyl-CoA making it a good candidate for expression in poplar xylem ( Yahyaa et al., 2017 ). Poplar expressing MdCHS3 in xylem tissue (hereafter referred to as MdCHS3 -poplar) clearly displayed an accumulation of naringenin in xylem methanolic extracts, not inherently observable in WT, and nuclear magnetic resonance (NMR) analysis revealed the incorporation of this flavonoid compound (a flavanone) into polymeric lignins. In addition, the highest-expressing MdCHS3 -poplar lines displayed reduced total lignin, increased cell wall carbohydrate content, yet displayed no changes in growth or biomass compared to their WT counterparts and significantly improved saccharification efficiency after dilute acid pretreatment.",
"discussion": "Discussion Lignin is an important component of plant secondary cell walls, serving to facilitate water transport throughout the plant, support vertical growth, and protect against pests and pathogens ( Weng and Chapple, 2010 ). Disruption of monolignol biosynthesis has led to significant reductions in lignin content and greatly improved biomass processability, yet these modifications often result in growth penalties ( Coleman et al., 2008a, 2008b ; Chanoca et al., 2019 ). This has motivated interest in genetic modifications of woody feedstock that specifically alter the composition of lignin, such as the incorporation of valuable monomers, as a method of improving lignocellulosic biomass ( Mottiar et al., 2016 ; Mahon and Mansfield, 2019 ). Expression of MdCHS3 in hybrid poplar xylem resulted in an appreciable accumulation of naringenin in soluble extracts, in the form of glycosides, as well as accumulation of naringenin in the cell wall ostensibly incorporated into lignin. Poplar has been reported to produce naringenin endogenously in apical tissues, consisting of leaves, and three youngest internodia ( Morreel et al., 2006 ; Morreel et al., 2006 ) as well as bud exudates ( Greenaway et al., 1991 ). Low amounts of naringenin have even been reported in wood of mature poplar species both in its aglycone and glycosylated form ( Pietarinen et al., 2006 ). However, in this study, naringenin was not observed in WT xylem extracts or EL of WT trees, nor does it appear to accumulate in stem tissue of younger trees ( Voelker et al., 2010 ). In addition, we determined that expression levels of endogenous poplar CHS genes in WT xylem were substantially lower compared to MdCHS3 in the lowest expressing transgenic line (line 6) which produced only trace amounts of naringenin in xylem ( Supplemental Figure S1 ). Our data demonstrate that the expression of an exogenous CHS gene is sufficient to substantially increase production of naringenin in poplar xylem without introduction of an exogenous chalcone isomerase ( CHI ), the enzyme responsible for stereospecific ring closure of naringenin ( Figure 1 ; Austin and Noel, 2003 ). The reduction in lignin observed in the highest-expressing lines could indicate that MdCHS3 is drawing significant carbon away from the biosynthesis of monolignols toward the production of naringenin chalcone. However, naringenin itself may play a role in suppression of lignin biosynthesis and contribute to the reduction in lignin observed in MdCHS3 -poplar. Work in Eucalyptus urograndis demonstrated that root supplementation with naringenin altered lignin composition and resulted in downregulation of several lignin-related genes ( Lepikson-Neto et al., 2013 ), and naringenin has been reported to directly inhibit the activity of 4CL, an important monolignol biosynthetic gene, in vitro ( Voo et al., 1995 ). Flavonoids have been found incorporated into the lignins of many grasses and other monocot species ( del Río et al., 2012 ; Lan et al., 2016 ). 2D HSQC NMR analysis herein revealed the presence of naringenin in the lignin fraction of MdCHS3- poplar. To the best of our knowledge, this is the first report of a flavonoid compound being incorporated into poplar wood lignins as the result of genetic engineering. Naringenin was found to incorporate instead of tricin into the lignins of rice with disrupted FNSII expression via reactions occurring at the B ring resulting in 4′-O-β type coupling, which in turn results in β-aryl ether units, and 3′-β type coupling to produce phenylcoumaran units ( Lam et al., 2017 ). Our data are consistent with previous NMR analyses of synthetic lignin polymers generated from radical coupling of naringenin with CA, which shows that the phloroglucinol ring remains intact, suggesting that coupling occurs mainly at the p- hydroxyphenyl B-ring over the phloroglucinol A-ring ( Lam et al., 2017 ). This bears significance as the introduction of lignin monomers capable of single coupling reactions, such as naringenin, into lignifying tissue has been proposed as a strategy to reduce the length of lignin polymers and improve lignin solubilization during pretreatment processing ( Eudes et al., 2012 ; Mottiar et al., 2016 ; Mahon and Mansfield, 2019 ). Due to the likely presence of 4′-O-β, 3′-β, and 3′-5 type linkages from the coupling of monolignols or lignin oligomers with naringenin, degradative methods such as thioacidolysis, which are targeted at cleaving β-ether linkages to release quantifiable monomers, would release only a fraction of the naringenin, that linked only by 4′-O-β linkages. Even for related tricin units that are incorporated into lignins via solely 4′-O-β linkages, the release was found to be 84% at best ( Lan et al., 2016 ); from the original procedure ( Rolando et al., 1992 ), thioacidolysis of poplar lignin releases typical lignin monomers accounting for just ∼24% of the total lignin. Considering that only trace amounts or naringenin were observed by NMR in the lignin fraction of MdCHS3-poplar xylem, thioacidolysis or any other current degradative technique would not be effective for quantifying the incorporation of naringenin into lignin polymers. The occurrence of naringenin in the cell wall space raises questions concerning its export across the cell membrane. A comprehensive model describing the mechanisms of monolignol export from the site of synthesis to the cell wall space has yet to emerge ( Perkins et al., 2019 ), although molecular simulations estimating membrane permeability of tricin indicated that passive diffusion alone is sufficient to facilitate transmembrane efflux ( Vermaas et al., 2019 ). It is, therefore, possible that naringenin is similarly capable of passive diffusion across the membrane in its aglycone form; however, active transport by an unknown endogenous poplar transporter cannot be excluded. We also observed high levels of putative O‒ linked naringenin glycosides in methanolic extracts and, considering that flavonoids are often stored in the vacuole in their glycosylated forms, naringenin may also be entering the cell wall space after release from the vacuole during programmed cell death ( Zhao and Dixon, 2010 ; Perkins et al., 2019 ). Some of the MdCHS3 -poplar lines exhibited a significant increase in alpha cellulose cell wall content ( Table 1 ). Analysis of cell wall carbohydrates released after hydrolysis indicated significant increases in glucose, galactose, and rhamnose in MdCHS3 line 2 compared to WT trees ( Supplemental Table S3 ). The significant improvement in glucose released during saccharification of MdCHS3 -poplar lines is likely due to the combined relative increases in cell wall carbohydrates, including glucose, and the reduction in total lignin ( Table 1 ; Supplemental Table S2 ). Lignin is thought to contribute to recalcitrance of lignocellulosic biomass by competitively binding cellulolytic enzymes and limiting access to cellulose ( Mooney et al., 1998 ; Mansfield et al., 1999 ; Berlin et al., 2005 ), such that genetically modified trees with reduced lignin content often display drastically improved rates of saccharification ( Leple et al., 2007 ; Mansfield et al., 2012a , 2012b ; Sykes et al., 2015 ). Saccharification rates may also be influenced by the reduction in lignin polymer length, as naringenin is only capable of single coupling and therefore prevents any further polymerization once incorporated ( Lam et al., 2017 ). Reductions in polymer length have been previously associated with improvements in saccharification, ostensibly by reducing cross-linking between lignins and cell wall polysaccharides, thereby improving accessibility of hydrolytic enzymes to both the cellulose and hemicellulose moieties ( Eudes et al., 2012 ). In addition, reductions in cell wall acetyl content, as observed in MdCHS3- poplar lines, have also been shown to improve saccharification in hybrid aspen as acetylation is thought to restrict the accessibility of glycanases to cell wall polysaccharides ( Pawar et al., 2017 )."
} | 4,311 |
36282302 | PMC9666320 | pmc | 1,496 | {
"abstract": "Abstract The carboxylic acid propionate is a valuable platform chemical with applications in various fields. The biological production of this acid has become of great interest as it can be considered a sustainable alternative to petrochemical synthesis. In this work, Clostridium saccharoperbutylacetonicum was metabolically engineered to produce propionate via the acrylate pathway. In total, the established synthetic pathway comprised eight genes encoding the enzymes catalyzing the conversion of pyruvate to propionate. These included the propionate CoA-transferase, the lactoyl-CoA dehydratase, and the acryloyl-CoA reductase from Anaerotignum neopropionicum as well as a D -lactate dehydrogenase from Leuconostoc mesenteroides subsp. mesenteroides . Due to difficulties in assembling all genes on one plasmid under the control of standard promoters, the P tcdB - tcdR promoter system from Clostridium difficile was integrated into a two-plasmid system carrying the acrylate pathway genes. Several promoters were analyzed for their activity in C. saccharoperbutylacetonicum using the fluorescence-activating and absorption-shifting tag (FAST) as a fluorescent reporter to identify suitable candidates to drive tcdR expression. After selecting the lactose-inducible P bgaL promoter, engineered C. saccharoperbutylacetonicum strains produced 0.7 mM propionate upon induction of gene expression. The low productivity was suspected to be a consequence of a metabolic imbalance leading to acryloyl-CoA accumulation in the cells. To even out the proposed imbalance, the propionate-synthesis operons were rearranged, thereby increasing the propionate concentration by almost four-fold. This study is the first one to report recombinant propionate production using a clostridial host strain that has opened a new path towards bio-based propionate to be improved further in subsequent work. Key points \n • Determination of promoter activities in C. saccharoperbutylacetonicum using FAST. \n \n • Implementation of propionate production in C. saccharoperbutylacetonicum. \n \n • Elevation of propionate production by 375% to a concentration of 3 mM. \n Supplementary Information The online version contains supplementary material available at 10.1007/s00253-022-12210-8.",
"introduction": "Introduction Propionate is a valuable platform chemical with a wide range of applications. Due to its antimicrobial activity, it is mostly used as a food and feed preservative, an ingredient in cleaning agents, or as an herbicide. It also gains increasing importance in the production of pharmaceuticals, plastics, and cosmetics (Gonzalez-Garcia et al. 2017 ; Samel et al. 2018 ). Moreover, it is considered an important precursor chemical as it is often esterified with short-chain alcohols, olefins, or acetylenes to yield corresponding alcohol or vinyl esters, which themselves have versatile applications (Samel et al. 2018 ). Propionate synthesis is currently achieved via chemical processes, i.e., the carbonylation of ethylene or the oxidation of propionaldehyde (Samel et al. 2018 ). However, bio-based approaches using bacteria as cell factories for propionate production from cheap or waste-derived substrates become increasingly attractive as sustainable alternatives to petrochemical production (Stowers et al. 2014 ). Although not commercially profitable yet, the desire to achieve an environmentally friendly propionate production has led to multiple studies exploring the capabilities of bacteria in that regard. There are different bacterial species that can naturally produce propionate from a range of substrates and via different pathways, e. g. Propionibacterium sp. via the Wood-Werkman cycle, or Clostridium and Megasphaera sp. via the acrylate pathway (Gonzalez-Garcia et al. 2017 ). Especially Propionibacterium sp. such as Propionibacterium acidipropionici and Propionibacterium freudenreichii have been studied extensively to develop fermentation strategies that allow the turnover of cheap substrates such as glycerol or molasses to propionate (Dishisha et al. 2012 ; Feng et al. 2011 ). Furthermore, these bacteria have also been engineered to improve propionate yields and overcome typical fermentation obstacles such as low acid tolerance (Jiang et al. 2015 ; Wang et al. 2015 ). Aside from propionibacteria, other non-native propionate-producing microorganisms have been engineered for propionate production, including Escherichia coli (Akawi et al. 2015 ; Gonzalez-Garcia et al. 2020 ; Kandasamy et al. 2013 ), Lactobacillus plantarum (Balasubramanian and Subramanian 2019 ), and Pseudomonas putida (Ma et al. 2020 ; Mu et al. 2021 ). Surprisingly, clostridia have never been considered hosts for recombinant propionate production although they are organisms with a versatile metabolism enabling them to use diverse carbon sources, including lignocellulosic hydrolysates and waste-derived substrates, and convert them into various products (Cho et al. 2015 ; Tracy et al. 2012 ). Furthermore, multiple tools are available to genetically modify clostridia for optimized production of native or recombinant compounds thus making them promising host strains for the production of commodity chemicals such as ethanol, isopropanol, 2,3-butanediol, or fatty acid esters (Cho et al. 2015 ; Feng et al. 2021 ). Clostridium saccharoperbutylacetonicum is a well-characterized solventogenic bacterium, which is genetically accessible and has high growth rates in favorable medium. Since it is a known hyper-butanol producer, it has mostly been employed for butanol production (Jiménez-Bonilla et al. 2021 ). However, C. saccharoperbutylacetonicum has also successfully been used for the production of hydrogen (Singh et al. 2019 ), isopropanol (Wang et al. 2019 ), 1,3-butanediol (Grosse-Honebrink et al. 2021 ), as well as caproate and hexanol (Wirth and Dürre 2021 ), thus highlighting its potential as a host for production of recombinant compounds. Here, we report the approach to convert C. saccharoperbutylacetonicum into a propionate producer by the implementation of the acrylate pathway from An. neopropionicum and a D -lactate dehydrogenase from L. mesenteroides subsp. mesenteroides (Fig. 1 ). For that purpose, a two-plasmid system harboring two propionate-synthesis operons (PSOs) was constructed, and gene expression was controlled by the sigma factor-inducible P tcdB promoter from C. difficile . In order to identify promoters that are suitable for mediation of gene expression in C. saccharoperbutylacetonicum , a promoter study using FAST was conducted. Fig. 1 Schematic overview of glycolytic and acidogenic pathways in C. saccharoperbutylacetonicum based on Jones and Woods ( 1986 ) coupled with acrylate pathway for propionate production (grey box; based on Hetzel et al. 2003 ) from An. neopropionicum and D -lactate dehydrogenase from L. mesenteroides subsp. mesenteroides (stoichiometrically incorrect). LdhD, D -lactate dehydrogenase (LEUM_1756); Pct, propionate CoA-transferase (CLNEO_17700); Lcd, lactoyl-CoA dehydratase (CLNEO_17730-17710); Acr, acryloyl-CoA reductase (CLNEO_21740-21760); ABE, acetone-butanol-ethanol",
"discussion": "Discussion The data presented clearly show that heterologous expression of the acrylate pathway leads to propionate production in C. saccharoperbutylacetonicum . In the first growth experiment, only traces of propionate (0.7 mM) were obtained, indicating that the carbon flux through the established pathway was not high. Although induction of gene expression led to elevated lactate concentrations in comparison to control strains, possibly a result of the heterologously expressed D -lactate dehydrogenase, lactate turnover to propionate seemed to be limited. The acrylate pathway consists of multiple steps starting with the activation of D -lactate to its CoA-derivative D -lactoyl-CoA, which is subsequently dehydrated to the toxic compound acryloyl-CoA. These reactions are catalyzed by the propionate CoA-transferase and lactoyl-CoA dehydratase, respectively. The resulting acryloyl-CoA is reduced to propionyl-CoA by an acryloyl-CoA reductase. Finally, propionate is released as the product by a CoA transfer from propionyl-CoA to D -lactate by the Pct (Fig. 1 ; Hetzel et al. 2003 ). Considering the fact that the PSOs encoding all mentioned enzymes were controlled by two different promoters, which FAST studies showed to be highly different in strength (Figs. 2 , 3 , and S4 ), the concern of a metabolic imbalance is valid. Lcd-encoding genes were under the control of the stronger P bgaL promoter, thus possibly leading to a higher expression level compared to the acr genes, which were controlled by the weaker P tcdB promoter. Based on these assumptions it would follow that the Lcd turnover is higher than the Acr turnover, thereby leading to an acryloyl-CoA accumulation. Furthermore, the Acr apparently has a low catalytic efficiency, which, according to reports is so low that native producers must compensate for this by producing high amounts of the enzyme (Hetzel et al. 2003 ; Kandasamy et al. 2013 ). Therefore, it is likely that acryloyl-CoA was accumulated rather than converted by Lcd and Acr reactions. Due to the electrophilic properties of acryloyl-CoA (Herrmann et al. 2005 ), it seems logical that an accumulation of this toxic compound would cause a high stress level for the production host, which could manifest itself in an impaired growth, a delayed or complete lack of acid reassimilation and solventogenesis, and a low productivity, all of which was observed in the growth experiments conducted in this study. Also, since the acrylate pathway is an electron sink, a deficiency in reducing equivalents could be a consequence (Kandasamy et al. 2013 ). With the metabolism out of balance due to the high burden exerted by the accumulation of pathway intermediates such as acryloyl-CoA and depletion of the NADH pool, high strain performances cannot be expected. Furthermore, FAST studies conducted via flow cytometry revealed that C. saccharoperbutylacetonicum cultures were heterogeneous even in the case of feg expression mediated by the strong and native P bld promoter (Figs. 3 and S4 ). Especially the use of the P tcdB - tcdR system with tcdR expression driven by P bgaL , which reflects the situation created in the propionate-producing strains, seems to limit the amount of productive cells as only 29% showed fluorescence in presence of TF Lime (Fig. 3b ). When transferring this observation to the strains carrying PSOs under control of P tcdB , it is reasonable to assume that only a small portion of the culture expressed the acrylate pathway genes and thus was able to form propionate. Such heterogeneity of bacterial populations in connection with an induced, plasmid-based gene expression has been shown several times and hypothesized to be caused by the uptake mechanism and concentration of the inducer, plasmid instability, or plasmid loss (Binder et al. 2016 ; Flaiz et al. 2021 ; Siegele and Hu 1997 ). Further factors to contribute to culture heterogeneity are cell morphogenesis or sporulation, both of which are especially apparent in clostridial cultures (Jones et al. 2008 ; Tracy et al. 2010 ). Most recently, Flaiz et al. ( 2022 ) also demonstrated heterogeneity of C. saccharoperbutylacetonicum cultures expressing feg in a P tcdB -dependent manner with tcdR controlled by P bgaL . Again, culture heterogeneity was observed throughout all growth phases, although in this case, the number of fluorescent cells was higher compared to the present study (Flaiz et al. 2022 ). Nevertheless, the impact of culture heterogeneity on the production behavior of C. saccharoperbutylacetonicum cannot be neglected and has to be considered a major contributor to the low observed productivity aside from the postulated metabolic imbalance. The rearrangement of acrylate pathway genes to overcome the postulated metabolic imbalance and circumvent the bottleneck created by the Acr led to an increase in propionate production by almost four-fold. Interestingly, introduction of the optimized PSOs led to a shift of metabolic products from solvents to acids. While the non-induced C. saccharoperbutylacetonicum strain accumulated 83.8 mM acetate and 69.4 mM butyrate, the induced strain first produced 57.9 mM lactate, which was partially reassimilated and used for butyrate and propionate formation. The reason for this shift is not completely clear, however, can possibly be explained by the increased demand of the strain for ATP to maintain and express the enlarged PSOs. Both acetate and butyrate formation are important energy sources as they involve the formation of one ATP via substrate-level phosphorylation (Boynton et al. 1996 ; Hartmanis 1987 ), thus increasing the amount of available ATP. With an increased turnover of glucose to acids, the demand for reducing equivalents, especially oxidized ferredoxin but also NADH, increases in parallel as these are needed by the pyruvate:ferredoxin oxidoreductase, 3-hydroxybutyryl-CoA dehydrogenase, and butyryl-CoA dehydrogenase reactions (Jones and Woods 1986 ). In addition, the acrylate pathway consumes NADH during several steps. This raises the question of how the strain can adapt its metabolism to meet this increased need for reducing power. Under normal conditions, carbon flow from glucose to each of the acids and solvents happens in a particular ratio so that carbon and redox balances are closed (Jones and Woods 1986 ). However, solventogenesis is a process involving a high turnover of NADH to form ethanol and butanol (Jones and Woods 1986 ). Thus, a reduction of solventogenesis could save a substantial amount of NADH to be invested in acidogenesis including propionate formation via the acrylate pathway. Since C. saccharoperbutylacetonicum strains harboring optimized PSOs only produced little solvents (Fig. 5 ), it seems as if solventogenesis was indeed spared for the benefit of the energetically more favorable acidogenic pathways. To meet the increased need for oxidized ferredoxin, the strain can use its Rnf complex to recover oxidized ferredoxin and form NADH, which is accompanied by the generation of an ion gradient (Poehlein et al. 2017 ). This in turn can be used by the ATPase for further ATP formation and to fill up the ATP pool. Although propionate production was successfully increased, the overall titer of 3 mM is still low. Considering the partial propionate reassimilation observed in both growth experiments, propionate titers could have been higher had it not been reduced to propanol. The formation of propanol is most probably due to the uptake of propionate by the acetoacetyl-CoA:acetate/butyrate CoA-transferase, which has a broad substrate spectrum including propionate (Hartmanis et al. 1984 ). The resulting propionyl-CoA can then be converted to propanol by aldehyde and alcohol dehydrogenases. Aside from propanol, other by-products such as solvents and butyrate limit the level of produced propionate as their formation requires both carbon and reducing equivalents. Other studies targeting heterologous propionate production using different host strains also reported challenges leading to mixed results. While E. coli engineered with the Sleeping beauty mutase operon and carrying multiple gene deletions or recombinant Ps. putida cultivated in fed-batch mode achieved a maximum of approx. 160 and 823 mM propionate from 326 mM glycerol and 850 mM L -threonine, respectively (Akawi et al. 2015 ; Mu et al. 2021 ), bacterial strains modified with the acrylate pathway also only produced rather low amounts of propionate. These ranged between 0.01 mM for L. plantarum (Balasubramanian and Subramanian 2019 ) and 3.7 mM for E. coli (Kandasamy et al. 2013 ), the latter of which is comparable to the propionate concentration produced by C. saccharoperbutylacetonicum [pMTL83151_P tcdB _L_P tcdB _AA][pMTL82251_P bgaL _LPTT]. The low propionate titers could be due to metabolic imbalances leading to the accumulation of pathway intermediates, redox deficiencies, and low activities of recombinant enzymes as previously hypothesized by other groups (Balasubramanian and Subramanian 2019 ; Kandasamy et al. 2013 ). Despite these hurdles, there are options that might lead to an improved propionate production using C. saccharoperbutylacetonicum through which it could possibly outperform recombinant E. coli and Ps. putida strains or at least reach the same production level. One such option is the introduction of PSOs into strains harboring tailored mutations to improve carbon and redox balances. Since carbon and redox equivalents are predominantly invested in C 4 -producing pathways in C. saccharoperbutylacetonicum , manipulations in these metabolic branches might be promising. A deletion of 3-hydroxybutyryl-CoA dehydrogenase, crotonase, or aldehyde and alcohol dehydrogenases in either C. saccharoperbutylacetonicum or its close relative C. acetobutylicum led to a reduction or complete abolishment of butyrate or solvent formation (Baur 2022 ; Cooksley et al. 2012 ; Lehmann and Lütke-Eversloh 2011 ). Simultaneously, saved reducing equivalents were used for lactate or ethanol formation (Baur 2022 ; Lehmann and Lütke-Eversloh 2011 ). If such a strain carried the reductive acrylate pathway, this might lead to increased propionate concentrations. Other options for modifications to manipulate the carbon and electron flow in favor of the acrylate pathway would be the deletion of the hydrogenase gene cluster or the global regulator Spo0A. The ferredoxin hydrogenase of C. saccharoperbutylacetonicum produces hydrogen while simultaneously oxidizing reduced ferredoxin (Dada et al. 2013 ). Thus, a deletion of the hydrogenase could save reduced ferredoxin to be turned over by the Rnf complex, which as mentioned before, can use this for NADH generation. Deletion of spo0A was shown to reduce both solventogenesis and sporulation in many clostridial strains including C. saccharoperbutylacetonicum (Atmadjaja et al. 2019 ; Harris et al. 2002 ; Schwarz et al. 2017 ). Since sporulation is also a major contributor to culture heterogeneity (Tracy et al. 2010 ), a combination of spo0A deletion with the PSOs could result in increased propionate titers. Another option to overcome culture heterogeneity would be chromosomal integration of the PSOs as this would lead to a plasmid-independent expression and thereby eliminate plasmid loss or instability as possible limiting factors. Whether such an optimized C. saccharoperbutylacetonicum strain could then outperform native propionate producers such as propionibacteria remains questionable as even higher propionate titers were achieved by the cultivation of these bacteria in bioreactors using more sophisticated approaches that thus far have never been employed using C. saccharoperbutylacetonicum . The highest ever reported propionate concentrations are 1 M using P. acidipropionici (Liu et al. 2016 ) and 1.2 to 1.8 M using P. freudenreichii (Chen et al. 2013 ; Feng et al. 2011 ), when strains were cultivated in fed-batch mode with high cell density or immobilized cells and glucose or hydrolyzed sugar cane molasses as substrates. These concentrations were achieved with productivities of 4.3 to 7.7 mM h −1 , which is 54- to 96-fold higher than the productivity of C. saccharoperbutylacetonicum [pMTL83151_P tcdB _L_P tcdB _AA][pMTL82251_P bgaL _LPTT] (0.08 mM h −1 ). Nevertheless, it is imaginable that fed-batch or continuous cultivation of further optimized C. saccharoperbutylacetonicum could lead to another increase in propionate concentration and thus make it a strain that can very well compete but probably not outperform native producers with respect to the propionate concentration and productivity. Overall, C. saccharoperbutylacetonicum was successfully engineered as a propionate producer. Although propionate titer was rather low with 3 mM, different options for further strain engineering and cultivation are conceivable to increase the performance of recombinant C. saccharoperbutylacetonicum strains. Successfully engineered and improved strains could then possibly be considered for commercial propionate production using sustainable resources such as lignocellulosic hydrolysates."
} | 5,114 |
36540398 | PMC9743417 | pmc | 1,498 | {
"abstract": "Resistive switching has provided a significant avenue for electronic neural networks and neuromorphic systems. Inspired by the active regulation of neurotransmitter secretion, realizing electronic elements with self-adaptive characteristics is vital for matching Joule heating or sophisticated thermal environments in energy-efficient integrated circuits. Here we present energy-adaptive resistive switching via a controllable insulator–metal transition. Memory-related switching is designed and implemented by manipulating conductance transitions in vanadium dioxide. The switching power decreases dynamically by about 58% during the heating process. Furthermore, the thresholds can be controlled by adjusting the insulator–metal transition processes in such nanowire-based resistive switching, and then preformed in a wide range of operating temperatures. We believe that such power-adaptive switching is of benefit for intelligent memory devices and neuromorphic electronics with low energy consumption.",
"conclusion": "Conclusions In summary, we have demonstrated energy-adaptive RS with adjustable thresholds by the IMT features in VO 2 nanowires. The switching thresholds are electric-related and thermal-dependent while the IMT-based RS can also adapt to different thermal environments or temperature variations. The self-adjusted switching powers demonstrate a well performed energy-adaptive resistive switching. Ongoing investigations are carrying out for in-depth manipulating and understanding in VO 2 based energy-efficient memory. We think that such thermal-adaptive switching helps reduce energy consumption, and shows promise in the construction of advanced Mott memory and correlated in-memory devices.",
"introduction": "Introduction Advanced information devices are fundamental and required in intensive information data and modern electronic applications. 1–5 Resistive switching (RS) has exhibited excellent prospects in low operation power and flexible architecture towards neuromorphic techniques and artificial intelligence. 6–12 These neuromorphic RS have shown promise in deep learning, 13,14 autopilot systems, 15,16 and man–machine interaction systems. 17,18 Notably, the approximately 14 billion neurons in the human brain only have a power consumption of ∼15 W. 19,20 It is still critical for realizing low energy consumption, high computing ability and ultrahigh storage density in such advanced bionic electronic devices. However, Joule heating is unavoidable and may result in redundant energy consumption. 21,22 In contrast to the low potential changes of sodium and potassium ions in nerve conduction, 23,24 the thermal effect induced by electronic transport may unquestionably alter switching behaviors in RS units. Recently, energy-adaptive devices have received huge attention and sparked intense interest in advanced RS and energy-efficient components. 25–29 Active response and self-regulation will contribute to the realization of feedback, monitoring and passive control networks in natural neurons. 30–32 Significantly, emergence of energy-adaptive is precisely to meet the requirements of complex environments, which highly require dynamic and controllable features. Unique insulator–metal transitions (IMT) has emerged in Mott insulators ( e.g. VO 2 , NbO 2 ), 33–36 and been recently adopted in the constructions of advanced RS devices. The IMT-based RS differs from those originated from redox in oxides or conductive filaments. 37,38 Extra fields ( e.g. electric and thermal) 39–42 can induce an ultrafast resistance changing with multiple orders of magnitude. 43 Here, we propose a switching power-adaptive RS with controllable thresholds towards intelligent memories. Diverse conductivity states get regulated though the electro-excited IMT process in VO 2 nanowires (VO 2 -NWs). Variable thresholds show adaptability to the fluctuating thermal field, and correspondingly decreased switching power may effectively depress energy consumption and Joule heating. Moreover, energy-adaptive can be manipulated at higher operating temperature by adjusting field-dependent IMT processes in VO 2 -NWs. We believe that energy-adaptive switching will provide more opportunities for designing and implementing advanced in-memory and neuromorphic components with low power consumption.",
"discussion": "Results and discussion \n Fig. 1a and b illustrated the principles of operation of the power-adaptive RS. To match the varied temperature induced by Joule heating, the self-adjusting RS got introduced via IMT characteristic. Here we simulated the low and high Joule heating by adjusting the working temperature. In an adaptive process of power-scaling, the corresponding working power decreases from P T 1 to P T 2 when the working temperature related with Joule heating increases from T 1 to T 2 . Fig. 1 IMT-based power-adaptive RS in (a) low and (b) high Joule heating environment simulated by low and high temperatures respectively. The insert schematic images represented the effect of Joule heating induced by working power on the temperature. (c) Electrically excited RS (red line) and the differential conductance dI/dV (blue line). The insert image represented the optical microscopy photograph of the two-terminal device based on VO 2 -NW. (d) Thermal excited RS in VO 2 -NW (red line) and related derivative of the resistance (blue line). (e) Cycles of RS under a pulsed voltage of 2 V with a period of 12 s at 300 K, 320 K and 330 K. (f) Retention property of HRS and LRS under different working temperatures. Joule heating related temperature ( T RS ) could be quantitatively determined and then generally analyzed by the heat equation: 44 1 where C was the heat capacitance of the VO 2 nanowire, P was the working power of the RS, T 0 was assigned to the (measured) substrate temperature and κ was the thermal coupling constant between the substrate and the nanowire. In the equilibrium state the eqn (1) could be transformed into the following forms: 2 From the eqn (2) , the adaptive reduced switching power might produce a low T RS through an effectively feedback regulation. And such negative feedback was benefit for reducing the energy dissipation. The IMT-based RS behaviors got confirmed under both electric and thermal fields as shown in Fig. 1c and d . The insert image of Fig. 1c illustrated the two-terminal RS device constructed with VO 2 -NW and gold electrodes. We also confirmed the fine lattice structure and components in VO 2 -NW used for the construction of IMT-based RS by scanning electron microscope, Raman spectroscopy, X-ray diffraction spectroscopy and X-ray photoelectron spectroscopy. Here, the switching devices were fabricated on the silica/silicon substrates in our cases. In a typical RS device, we used the crystalline nanowires with a length of about 50 μm and a width of around 200 nm (Fig. S1 † ). The monoclinic VO 2 is verified by Raman shifts at 195, 225 and 617 cm −1 (Fig. S2a † ), and help for producing an obvious resistance changing. 45–47 In addition, a good crystallization and a preferred orientation of (011) plane were found in VO 2 -NW (Fig. S2b and c † ). 48–52 All these were helpful for the realization of a single-domain IMT process and the regulation of electron transport in the VO 2 -NWs. 53 Fig. 1c and d illustrated that there were three different regions with changed conductance states, i.e. , insulator region, transition region and metal region in a typical IMT process. Symmetrical electrodes with chrome and gold were adopted to inhibit heating caused by undesired rectification effect, and then ohmic contact was confirmed by the current voltage curves in Fig. S3. † The electric-related behaviors of IMT-based RS were further observed in the insulator/transition regions ( Fig. 1c ). In these cases, a compliance current of 10 −3 A was set to protect the device from breakdown under high potential. Moreover, the time-dependence of the fabricated RS was carried out under voltage pulses with amplitude of 0.2 V and 2 V. In the wake of the increasing electric field strength, the high resistance state (HRS) abruptly switched to the low resistance state (LRS). As suggested in the inset in Fig. 1e , the electro-induced IMT-based RS was fast and repeatable when we simulated Joule heating and used a working temperature (300 K–330 K). The retention characteristics ( Fig. 1f ) of the resistance states were further investigated and verified at an operating temperature of 300 K, 320 K or 330 K. These three groups of temperatures were determined based on the insulation region and transition region as shown in Fig. 1d . Continuous stabilizations of the HRS and LRS kept a stable ON/OFF ratio about 10 3 . Subsequently, we further demonstrated the dynamic and thermal-adaptive thresholds of RS during IMT processes. Well-performed bidirectional resistive switching processes were illustrated in Fig. 2a . The sweeping voltage was from 0 V to −3 V, and then to 3 V, and finally to 0 V. From 280 K to 340 K, the threshold varied from 2.4 V to 0.3 V until the system completely turned into metallic state. Significantly, the IMT-based RS did not require a forming process due to the carrier triggered Mott transition. The switching window about 1 V at 300 K was shown in Fig. 2b . And the enhanced electronic correlation in the transition region would make the switching window narrow. In order to understand such thermal-adaptive and controllable thresholds, we mapped the dynamic differential conductance as a function of the temperature and bias voltage. The slanted colored stripes indicated the minimum values of negative differential conductance, that was, the critical points of RS. The blue stripe represented the voltage range from the threshold to the corresponding voltage arriving at 10 −3 A. The d I /d V maps with high contrast sharply displayed the linear trend of threshold value from 298 K to 318 K ( Fig. 2c ). A similar trend was also found in a transition region in a temperature range from 320 K to 332 K ( Fig. 2d ). Fig. 2 (a) Current–voltage characteristics of the two-terminal device at different working temperatures with compliance current of 10 −3 A. (b) The differential conductance d I /d V at 300 K. (c and d) Maps of differential conductance (d I /d V ) as a function of the temperature and bias voltage. Boundary between bright and dark red region indicated critical point of the variable conductance. The sweeping voltage direction was from 0 V to −3 V (c) insulator region and (d) transition region under a compliance current of 10 −3 A. Critical threshold voltages were extracted for quantitative analysis about the thermal-adaptive thresholds. The voltage variations and a linear fitting were displayed in Fig. 3a when the operating temperature increased with a step of 2 K. For the insulator region from 298 K to 318 K, the varied thresholds matched an equation of V T = −14.81 + 0.04446 T . As a contrast, the equation of V T = −8.0 + 0.02286 T was for the transition region from 320 K to 332 K. In insulator region, the small relative error R 2 of ∼0.9877 revealed the high reliability and matching degree. Correlated thresholds varied from −1.7 V to −0.8 V. The step of 2 K corresponded to a threshold attenuation coefficient of about 0.090, while the related fitting threshold attenuation coefficient was about 0.089. As to the transition region, a R 2 of 0.8929 indicated a susceptible threshold in strong electronic correlation region. Such difference was about 0.005 with a dynamic range of 0.3 V. Fig. 3 (a) The experimental (black discrete square) and the fitting curve (red solid lines) of threshold value as a function of the working temperature. (b) Sweeping current–voltage characteristics at 302 K and 322 K. (c) Resistance–power curves acquired at different working temperatures. (d) The functional relationship between working temperature and switching power. As to the cases at 302 K and 322 K, the measured thresholds were well agreement with the calculated one ( Fig. 3b ). The calculated V T =302K was −1.38 V while the measured is −1.40 V (the black dot in Fig. 3b ). When the switching processes were performed in the transition region, the calculated V T =322K was −0.64 V while the measured with −0.60 V (the red dot in Fig. 3b ). All these results indicated the universality of the fitting equations, and that the dynamic thresholds were thermal-adaptive. Furthermore, the electrically induced IMT switching shown in Fig. 3c indicated trend of the varied power. From 1 μW to 10 μW, the stable resistance at different working temperatures revealed that stable thermal performance of the IMT-based switching. As shown by the red arrow, high operating temperature brought a skewed to the left of the resistance–power curves, that was, Joule heating for motivating switching got reduced. Corresponding to the enhancive Joule heating, increasing working temperature effectively reduced the switching power from 151.3 μW to 63.6 μW around 58.0% shown in Fig. 3d . In the real and practical computing processes, a larger working temperature range of the thermal-adaptive RS was important and required because a higher frequency operation might induce a higher Joule heating and higher thermal fluctuation. Therefore, we attempted to construct a stable IMT-based RS worked at a higher operating temperature up to 390 K. As shown in Raman spectra in Fig. 4a , the weakening and blue shifts around 612 cm −1 indicated the increase of defects when we adjusted the defects in VO 2 nanowires by altering oxygen partial pressure. Such defects might lead to a higher transition temperature as displayed in Fig. 4b . In the range from 365 K to 390 K, the bidirectional switching processes were confirmed ( Fig. 4c ). Thermal-adaptive threshold voltages varied from −1.1 V to 0.6 V and finally to −0.4 V. Fig. 4 (a) Raman spectra of the VO 2 nanowires deposited under different pressures. (b) Derivative resistance under variable temperatures of the two-terminal device based on VO 2 nanowires with a transition point of 390 K. (c) I – V curves of the two-terminal device based on VO 2 with a transition temperature of 390 K. (d) Map of differential conductance (d I /d V ) as a function of the temperature and bias voltage. (e) Plots of threshold voltage and (f) switching power versus working temperature and the fitting curve (red solid lines). In Fig. 4d , the Map of differential conductance also suggested a linear function between the working temperature and bias voltage, which were similar with these illustrated in Fig. 3 . The similar trend of the varied threshold indicated a stable reproducibility of the adaptive threshold voltages. Further analysis about the threshold voltage and switching power were displayed in Fig. 4e and f . A depressed switching power was from 4.8 μW to 4.2 μW. All these suggested the adaptive voltage and power scaling RS could be flexibly controlled and even programmed when we exploited the IMT processes in a widening range of operating temperature. Furthermore, we analyzed the mechanism behind IMT-based energy-adaptive RS in briefly. It had been confirmed that the competition between coulomb repulsion energy and kinetic energy of the carrier determined the IMT process in strongly correlated system. 54 When the magnitude of external perturbation and coulomb repulsion energy were basically consistent, the delocalized carrier corresponded to the high kinetic energy of the increased density of free electrons would turn the system to be metal state. 55 In our case, the density of free electrons was directly related to field assisted carrier injection or increased working temperature. As to insulator region, the initial insulation state got quickly restored when bias voltage was removed. In the equilibrium state, the decreased thresholds corresponded to a stable resistance value of the same magnitude, that was, the energy used to excite the IMT got reduced. But for the transition region, coexisting insulating state and metallic state 56 promoted the delocalization energy to the same order of magnitude with coulomb repulsion energy. Higher ground states of kinetic energy got maintained accompanied by nonlinear attenuated resistances. With the decreased bias voltages, switching power also shown a nonlinear trend, and the system recovered to the intermediate state faster with a shrink hysteresis window."
} | 4,117 |
37296173 | PMC10256809 | pmc | 1,499 | {
"abstract": "Glucose is the most abundant monosaccharide, serving as an essential energy source for cells in all domains of life and as an important feedstock for the biorefinery industry. The plant-biomass-sugar route dominates the current glucose supply, while the direct conversion of carbon dioxide into glucose through photosynthesis is not well studied. Here, we show that the potential of Synechococcus elongatus PCC 7942 for photosynthetic glucose production can be unlocked by preventing native glucokinase activity. Knocking out two glucokinase genes causes intracellular accumulation of glucose and promotes the formation of a spontaneous mutation in the genome, which eventually leads to glucose secretion. Without heterologous catalysis or transportation genes, glucokinase deficiency and spontaneous genomic mutation lead to a glucose secretion of 1.5 g/L, which is further increased to 5 g/L through metabolic and cultivation engineering. These findings underline the cyanobacterial metabolism plasticities and demonstrate their applications for supporting the direct photosynthetic production of glucose.",
"introduction": "Introduction Glucose is the most abundant monosaccharide molecule in nature. The breakdown of glucose provides energy and carbon materials in cells throughout all domains of life, powering the cellular machinery by diverse glycolytic pathways, including the Embden–Meyerhof–Parnas (EMP) pathway, the oxidative pentose phosphate (OPP) pathway, and the Entner–Doudoroff (ED) pathway 1 , 2 . As monomers, glucose, and its derivatives are also involved in the synthesis of various macromolecules and cellular components 3 , 4 . Moreover, glucose also serves as an important feedstock in the biorefinery industry, supporting the cultivation of multiple microbial cell factories for green biomanufacturing of fuels, chemicals, and pharmaceuticals 5 – 7 . In nature, glucose is mainly synthesized through the photosynthesis of plants and algae and exists as monomers of polysaccharides in plant/algae biomass, e.g., cellulose and starch. The plant-biomass-sugar route dominates the current massive glucose supply, whose economic feasibility is influenced by multiple parameters, such as plant-cultivation cycles, biomass-collection radius, and pre-treatment costs 8 – 11 . Against the backdrop of the global climate crisis and worsening food shortages, developing more efficient, continual, and industrial glucose production routes would be valuable 12 , 13 . In recent years, direct conversion of carbon dioxide into glucose, glucose precursors, and glucose polymers has been achieved by chemical-biochemical, electrochemical-biological, and in vitro cascade enzymatic routes 14 – 16 . In contrast, continuous glucose production has not been successfully linked directly to photosynthesis. In photoautotrophs, e.g., higher plants and algae, glucose is synthesized as storage for carbon and energy and plays important regulatory roles. Glucose metabolism possesses complex interactions with photosystems, disturbs the synthesis and metabolism of pigments, and might even inhibit the photosynthetic activities 17 – 19 ; thus, free glucose is rarely synthesized or accumulated in excess in photosynthetic cellular metabolism. In cyanobacteria, a group of oxygenic prokaryotic microalgae, some progress has been made to facilitate the direct synthesis and secretion of natural or non-natural sugars through genetic manipulations 20 – 22 . However, the photosynthetic production of glucose has not yet been well-studied. The recombinant strains can only produce limited amounts of glucose accompanied by the production of other sugars, suggesting that more detailed mechanisms of glucose metabolism in photoautotrophs remain to be disclosed 23 , 24 . In this work, we aim to engineer a direct and stable conversion of carbon dioxide to glucose through cyanobacteria photosynthesis. In a model cyanobacterium Synechococcus elongatus PCC 7942 (hereafter PCC 7942 for short), we identify the native glucokinase activity as the bottleneck restricting the metabolism potential for glucose synthesis. Targeted knockout of two glucokinase genes disturbs the carbohydrate metabolism and activates a metabolic flux towards glucose through the sucrose metabolism network, which is generally considered as a specialized response to osmotic stress. The enhanced glucose synthesis promotes the enrichment of a specific spontaneous genomic mutation on the chromosome of PCC 7942, which facilitates efficient glucose secretion. By implementing multiple omics approaches combined with systematic genetic manipulations, we clarify the pathways and mutations leading to glucose synthesis and secretion and optimize the glucose synthesis performances of the recombinant strains. Through subsequent metabolic engineering and cultivation optimization, the glucose secreted by the engineered strain surpasses 5 g/L during long-term cultivation, accounting for up to 70% of the fixed carbon source.",
"discussion": "Discussion In this work, we unlocked the potential of cyanobacterial photosynthesis to convert carbon dioxide directly into glucose. Photosynthesis is the most extensive biochemical process on Earth, and photoautotrophs convert solar energy and carbon dioxide into primary organic carbon to support the maintenance of the biosphere 36 . Currently, the refining of sugar-rich plant/algal biomass serves as the dominating route to produce sugar in large quantities, meeting the needs of the food and medicine areas as well as the biorefinery industry 8 , 9 . The “one-pot, one-step” mode of cyanobacterial photosynthetic production could serve as a more continual, stable, and efficient sugar alternative than the “plants-biomass-sugar” route 37 . Most cyanobacteria species can naturally synthesize and accumulate sugar-type compatible solutes to resist abiotic environmental stress 38 . The accumulation and secretion of these compatible sugar compounds (such as sucrose and trehalose) by cyanobacteria can be optimized through genetic manipulation 20 , 21 , and biotechnology application scenarios such as the development of artificial consortia are explored and expanded 39 , 40 . However, glucose, the most representative and important monosaccharide molecule, cannot be efficiently synthesized and accumulated in cyanobacteria. Recently, glucose secretion (accompanied by sucrose secretion) has been observed in a PCC 7942 mutant with deleted xpk gene (which encodes phosphoketolase). However, this glucose synthesis can only be triggered in the dark with salts stress conditions, and the titer is rather limited, reaching approximately 3 mM by high-density cells (the initial OD 730 is 20, and the final average productivity of 0.027 g/L/OD 730 ) 24 . In another, earlier reported work, the titer of salt-induced glucose synthesis (accompanied by fructose secretion) reached only 30 μM (productivity of 0.027 g/L/OD 750 ) in PCC 7942 by overexpressing sucrose hydrolase and hexose transporters 23 . In contrast, continuous synthesis of glucose during the rapid photoautotrophic growth phase was achieved in this work, with productivity higher than 0.27 g/L/OD 730 , by simply removing the glucokinase activities in Synechococcus cells, meaning tenfold higher photosynthetic productivities in the absence of any heterologous catalytic enzymes or transporters and independent of any environmental stress inductions. Although a majority of the isolated cyanobacteria species perform autotrophic metabolism and are deficient to absorb and utilize exogenous glucose as the carbon source, some cyanobacteria strains are capable of using glucose 41 , 42 . As for the cyanobacteria strains able to perform heterotrophic or mixotrophic metabolism using glucose, glucokinase plays an essential role in phosphorylating glucose and initiating the diverse glycolytic processes 43 , 44 . However, glucokinase genes are ubiquitous in the vast majority of the sequenced cyanobacterial genomes (Supplementary Data 5 ). For example, Synechococcus elongatus PCC 7942, which has been shown to lack the capacity to absorb and utilize exogenous glucose, has two glucokinase genes on its genome, implying that glucokinase has more physiological functions than glucose utilization. Through systematic genetic modifications, we identified an active and stable glucose-phosphoglucose cycle in PCC 7942, which was maintained in the sucrose metabolism pathway. Previously, sucrose synthesis and accumulation were generally recognized as a salt-stress responsive mechanism to resist hyperosmotic stress 31 , 45 . In contrast, we demonstrated in this work that the synthesis-degradation cycle of sucrose in PCC 7942 remains active independent of salts stress induction, and the glucokinase activity performs as a “floodgate” to avoid the overaccumulation of glucose, which is an intermediate in the sucrose metabolism network and has a potential impact on the photosynthetic metabolism. Besides sucrose metabolism, some other non-specific reactions may also contribute to the synthesis and accumulation of glucose in PCC 7942. Thus, it could be assumed that intracellular glucose synthesis may be continuously maintained in PCC 7942, whereas glucokinase was crucial to recycle glucose into the central metabolism network by a phosphorylation process, and this mechanism can potentially rescue the potential energy loss and metabolism disturbance caused by glucose secretion and accumulation. In addition, although cell growths and photosynthesis were not inhibited (Supplementary Fig. 16 ), glucokinase--deficient strains showed improved sensitivity to 150 mM NaCl (Supplementary Fig. 21 ), indicating that the stable glucose-phosphoglucose cycle contributes to the rapid response of PCC 7942 toward salt stress. Previously it has been proven that the ionic effect is important for inducing sucrose accumulation in PCC 7942, whereby an elevated ion concentration directly activates the sucrose-synthesizing enzyme Sps and simultaneously inhibits the sucrose-degrading enzyme Inv 31 . The maintenance of the sucrose metabolism cycle permitted the direct activity-regulation of ion concentrations on the enzymes, and the glucokinase activities could guarantee the rapid re-utilization of sugars (fructose and glucose) from hydrolyzed sucrose (activity of Inv would be reduced rather than completely removed by elevated ion concentrations). An important issue in microbial metabolic engineering is the adaptation of chassis cells to artificial pathways and genomic modifications. The in-plugged and rewired metabolic activities can disturb carbon distribution, co-factor supply, and redox balances of the native cellular metabolism, and are prone to stimulate the responses of the host cells on levels of physiology, metabolism, and even genetics 46 , which should be relieved through metabolic designs and engineering. Transporter engineering has become a universal strategy to increase the adaptability of microbial chassis cells to heterologous pathways, which could promote product secretion, relieve the metabolic burden of intracellular over-accumulation and reduce the potential inhibitory effect of the products 47 . An alternative solution was suggested in this study. To achieve effective glucose transport, we introduced a heterologous transporter GalP on the chromosome of PCC 7942, which reduced the potential intracellular metabolic or physiological stress and facilitated the convenient and smooth acquisition of the genetic transformants (SZ17). In comparison, without introducing heterologous transporters, a spontaneous mutation ( Synpcc7942_1161 -G274A) was generated and enriched during the long-term cultivation process and endowed the mutant strain with even better capacities for glucose synthesis and secretion. Combining artificial genetic modifications with metabolic stress-induced spontaneous genomic mutations might rewire the metabolic flow more effectively. In summary, we identified the key factors restricting the natural potential of cyanobacteria for the photosynthetic production of glucose. Without heterologous catalysis or transportation genes, a large portion of photosynthetically fixed carbon dioxide in the engineered strain with deficient glucokinase activities and a spontaneous genome mutation was rewired into the secretory production of glucose. The discoveries shed light on developing and industrializing more directional and continuous glucose production systems with solar energy and carbon dioxide."
} | 3,126 |
30527541 | null | s2 | 1,500 | {
"abstract": "Cyanobacteria are photosynthetic prokaryotes that are influential in global geochemistry and are promising candidates for industrial applications. Because the livelihood of cyanobacteria is directly dependent upon light, a comprehensive understanding of metabolism in these organisms requires taking into account the effects of day-night transitions and circadian regulation. These events synchronize intracellular processes with the solar day. Accordingly, metabolism is controlled and structured differently in cyanobacteria than in heterotrophic bacteria. Thus, the approaches applied to engineering heterotrophic bacteria will need to be revised for the cyanobacterial chassis. Here, we summarize important findings related to diurnal metabolism in cyanobacteria and present open questions in the field."
} | 201 |
39727760 | PMC11672995 | pmc | 1,501 | {
"abstract": "Developing a durable multifunctional superhydrophobic coating on polymeric films that can be industrially scalable is a challenge in the field of surface engineering. This article presents a novel method for a scalable technology using a simple single-step fabrication of a superhydrophobic coating on polymeric films that exhibits excellent water-repelling and UV-blocking properties, along with impressive wear resistance and chemical robustness. A mixture of titanium precursors, tetraethylorthosilicate (TEOS), hydrophobic silanes and silica nano/micro-particles is polymerized directly on a corona-treated polymeric film which reacts with the surface via siloxane chemistry. The mixture is then spread on polymeric films using a Mayer rod, which eliminates the need for expensive equipment or multistep processes. The incorporation of silica nanoparticles along with titanium precursor and TEOS results in the formation of a silica–titania network around the silica nanoparticles. This chemically binds them to the activated surface, forming a unique dual-scale surface morphology depending on the size of the silica nanoparticles used in the coating mixture. The coated films were shown to be superhydrophobic with a high water contact angle of over 180° and a rolling angle of 0°. This is due to the combination of dual-scale micro/nano roughness with fluorinated hydrocarbons that lowered the surface free energy. The coatings exhibited excellent chemical and mechanical durability, as well as UV-blocking capabilities. The results show that the coatings remain superhydrophobic even after a sandpaper abrasion test under a pressure of 2.5 kPa for a distance of 30 m.",
"conclusion": "4. Summary and Conclusions The present study presents a novel approach for fabricating superhydrophobic coatings on polymeric films that exhibit UV-blocking and wear-resistant properties. The challenge of superhydrophobic coatings that are multifunctional as well as durable is addressed, using a scalable and cost-effective single-step method. Different surface morphologies were achieved in relation to the size of the SiO 2 particles incorporated. Coatings containing only TiO 2 or small-sized SiO 2 nanoparticles seem to organize as tunnels with nanogaps between them, thus increasing the surface area and roughness of the coating. The surface morphology of coatings containing larger-sized SiO 2 nano/micro-particles appears to contain tunnels composed of silica–titania network structures alongside raspberry-like particles consisting of a SiO 2 particle core and a silica–titania rough shell. Both forms of coating morphologies and structures show significantly high water contact angles, as the droplets did not stay on the surface but immediately rolled away, exhibiting a rolling angle of 0° and a contact angle of 180°. Several coatings were made with different compositions to examine how SiO 2 particles of various sizes affect the surface morphology, roughness and functionality of the resulting coating. It was shown that the incorporation of nano/micro-particles of various sizes affects surface roughness and increases the coating durability from both environmental and chemical wear. These coatings retained their superhydrophobic properties even after being rubbed against sandpaper for a significant distance of 30 m under a pressure of 2.5 kPa. The addition of a titanium precursor to the coating’s solutions imparts UV blocking functionality, making the coating suitable for outdoor use. The combination of mechanical durability, UV resistance, and ease of application makes this coating system highly suitable for a wide range of industrial and commercial applications.",
"introduction": "1. Introduction Artificial systems inspired by nature, or bionics, have been used and adapted for convenience in our everyday lives. Examples of these types of technologies include adhesive materials that can be removed without leaving residue, imitating the feet of the gecko, hook and loop fastening mechanisms, imitating burrs, and imaging applications such as sonar, radar and ultrasound, imitating bat echolocation. Another example of bionics is artificial superhydrophobic surfaces, which are surfaces characterized by their ability to repel water. This adaptation has been inspired by the leaves of the lotus flower and has long been of academic and industrial interest. Superhydrophobic surfaces, like that of the lotus leaf, are defined by their ability to repel water and are characterized by having a rough surface topography and hydrophobic surface chemical groups. The hydrophobic surface chemistry contains non-polar bonds that lower the surface energy and thus repel water molecules rich in polar hydrogen bonds. Superhydrophobic surfaces have a broad range of applications such as self-cleaning, anti-corrosion and medical devices, making it a focal point of both academic and industrial research [ 1 ]. A surface is characterized as superhydrophobic when the water contact angle (WCA), the angle between a tangent of a water droplet and a solid surface, is greater than 150° and the contact angle hysteresis is lower than 10°. Superhydrophobicity results from a delicate interplay between surface roughness and surface free energy. The rough surface topography enhances water repellence by entrapping air pockets within the surface cavities, reducing the interaction between water and the surface [ 2 , 3 , 4 ]. Many theoretical models describe surface wetting behavior. The Cassie–Baxter model [ 5 ] explains air pockets trapped within the rough surface cavities, creating a composite interface of solids and air beneath the water droplet, repelling the water and lowering the WCA. This phenomenon, referred to as the “lotus effect”, allows water droplets to roll off easily while carrying dirt along [ 6 ]. By decreasing water interactions and entrapping air, surface roughness is integral to achieving extreme water repellence [ 7 ]. Surface free energy is another critical factor affecting surface–water interactions. Surfaces treated with low-surface energy materials, such as fluorinated compounds or long chains of hydrocarbons, reduce the interaction between the water droplets and the surface due to their inherent hydrophobic nature [ 1 ]. A surface structure composed of nano and micro scale roughness with a layer of a low-surface energy material is crucial in achieving the desired superhydrophobic characteristics [ 8 , 9 ]. Various materials, techniques and methods have been employed for implementing superhydrophobic properties in a surface. These methods include techniques such as electrochemical deposition [ 10 , 11 , 12 ], wet chemical reactions [ 13 , 14 , 15 , 16 ], electrospinning [ 17 , 18 , 19 , 20 ], the dip coating technique [ 21 , 22 , 23 , 24 , 25 ], spray coating [ 26 , 27 , 28 , 29 ], layer-by-layer deposition [ 30 , 31 , 32 , 33 , 34 ], lithography [ 35 , 36 , 37 , 38 ], chemical vapor deposition (CVD) [ 39 , 40 , 41 , 42 ] and nanoparticle self-assembly [ 43 , 44 ]. These approaches create a coating layer with hydrophobic surface chemicals and a roughened topography. However, developing durable, cost-effective superhydrophobic coatings remains a significant challenge due to issues like mechanical wear and environmental exposure. Recently, silica (SiO 2 ) and titanium dioxide (TiO 2 ) have emerged as promising materials for superhydrophobic coatings due to their ability to form robust structures. SiO 2 facilitates nanoscale roughness and bonds with low-energy materials, enhancing water repellence [ 45 , 46 , 47 , 48 , 49 , 50 ]. Silica is also used as a basis for intrinsically hydrophobic materials that, when cut or abraded, remain hydrophobic without the need for any surface chemical treatment [ 51 , 52 ]. The Stöber process is a well-known method for synthesizing SiO 2 which typically uses tetraethylorthosilicate (TEOS) as the monomer for the SiO 2 matrix in acidic or basic aqueous conditions [ 53 ]. Under these conditions, TEOS undergoes hydrolysis by water, which forms silanol groups (Si-OH) followed by condensation reactions to form the SiO 2 network [ 54 ]. Titanium dioxide (TiO 2 ) is valued for its strong covalent bonding properties, providing resistance to mechanical stress. Titanium precursors such as titanium (IV) butoxide undergo hydrolysis and condensation reactions, like that of SiO 2 , to form a TiO 2 network [ 55 ]. Combining silica and titanium precursors forms Si–O–Ti bonds, which results in a stable composite surface structure where SiO 2 maintains surface roughness and TiO 2 strengthens the coating adhesion to the surface, enhancing coating adhesion and durability [ 56 , 57 ]. Despite progress in the fabrication of superhydrophobic surfaces, maintaining coating durability against mechanical stress and environmental exposure remains a challenge [ 58 , 59 , 60 ]. Wear and abrasion can reduce surface roughness, and low-energy materials may degrade over time. Industrially, high costs, complexity, and scalability constraints limit the widespread adoption of superhydrophobic coatings. Additionally, weak physical bonds, such as Van der Waals and hydrogen bonds, result in coatings that are easily removed and require frequent reapplication. To address these challenges, this research uses a modified in situ Stöber process containing tetraethylorthosilicate (TEOS), titanium (IV) butoxide (TBT) and per-fluorodecyltriethoxysilane (FTES). The monomer solution is thinly spread on a corona-treated polypropylene (PP) surface using a Mayer rod. These monomers react with the surface of an oxygenated PP film in acidic aqueous conditions, thus forming a strong and covalently bound matrix between the monomers and the oxygenated surface. This matrix of strong covalent bonds imparts high durability to the coating as opposed to the physical bonds of the coatings described above. Similar research has been performed on this type of modified in situ process to impart other qualities to surfaces, for example, superhydrophobic, antifog and flexible electronics applications [ 61 , 62 , 63 , 64 ]. This method offers several advantages: (1) the formation of strong covalent bonds be-tween the surface and coating, which increases the durability of the coating against mechanical stress, (2) no need for expensive equipment prior to or after the coating process, (3) the need for a minimal amount of monomers to be used in the coating, (4) the use of corona treatment to activate the film surface, which negates the need for chemical binding agents, such as primers or adhesives, prior to the coating process, and lastly, (5) the short amount of time needed for the reaction to be completed, which reduces industrial costs. The superhydrophobic surface layer, which has a micro/nano hierarchical structure, functions not only as water repellent, but also displays self-cleaning ability and UV blocking properties, which makes it suitable for various applications. Additionally, this process is versatile, applicable to various polymeric substrates such as polyethylene terephthalate, polyethylene, and polycarbonate. While this study focuses on PP, the method can be extended to other polymeric films.",
"discussion": "3. Results and Discussion In the present study, we developed a simple and straightforward method for applying superhydrophobic coatings to polymeric films using a one-step coating process using a Mayer rod. This method is also applicable to most thermoplastic polymers as well as various other substrates such as paper and cardboard, without the need for prior corona treatment. This method allows for extremely rapid sample preparation, as the polymerization occurs simultaneously with the application of the coating on the polymeric surface. As a result, wide films can be coated within seconds. Additionally, this approach is highly cost-effective and economical, as it minimizes material waste, reduces processing time, and eliminates the need for complex equipment. As previously discussed, experimental results in the literature demonstrate that a surface exhibits superhydrophobic properties due to a combination between its structural characteristics and its chemical composition. Surface roughness plays a critical role in enhancing hydrophobicity, often leading to superhydrophobicity when combined with low-surface-energy materials. The presence of micro- and nanoscale structures amplifies the water repellence of the surface by reducing the water contact angle due to air trapped between the formed cavities. This effect also contributes to the self-cleaning properties of the superhydrophobic surface. Consequently, surface roughness is a crucial factor in the development of such coatings. We achieved a high surface roughness due to a combination between SiO 2 nano/micro-particles and silica/titania, which forms smaller-scale structures on top of the SiO 2 nano/micro-particles. The addition of SiO 2 particles of various sizes to the coating results in a coating morphology consisting of hierarchical structures ( Figure 1 ), increasing roughness (Table 3), and an increase in the coating durability. 3.1. Surface Morphology and Chemical Composition of TiO 2 and TiO 2 -SiO 2 Composite Coatings on Polypropylene Films As previously mentioned, water contact angle is directly affected by the surface roughness and is higher when the surface consists of hierarchical structures. The surface morphology of the various coatings and their microstructures was analyzed by SEM. Figure 1 shows SEM images of samples 1–4, as marked in Table 1 . The coating in sample 1 ( Figure 1 , sample 1, images a–c) was synthesized using titanium isopropoxide and hydrochloric acid, resulting in roughness of the surface. In this sample, the TiO 2 coating is uniform upon the surface in both coverage and roughness. It seems like the structure of sample 1’s coating is composed of many TiO 2 clusters with a diameter of about 100 nm stacked one on another, forming a very porous structure that contributed to the water-repellent properties. The highly textured surface plays a pivotal role in affecting the superhydrophobic effect by promoting the Cassie–Baxter model, describing air pockets trapped in the cavities between the surface and the water droplets, resulting in an increase in the water contact angle. Samples 2–4 contain, in addition to the TiO 2 , also SiO 2 formation, due to the presence of TEOS. The SiO 2 nano/microparticles have an average size of 70 nm, 250 nm and 500 nm, respectively. These particles contain a hydrophilic surface, allowing the formed TiO 2 -SiO 2 network to bond with -OH groups on the SiO 2 particles’ surface, resulting in covalent bonding between the SiO 2 particles and the formed TiO 2 -SiO 2 nucleation centers. The resulting morphology of the TiO 2 -SiO 2 network on the SiO 2 particles is called raspberry-like due to the core–shell structure of a large center covered with smaller particles. Due to this hierarchical morphology, the coatings exhibit extremely high WCA and 0° rolling angles (Table 3). The raspberry-like particles are randomly scattered along the surface of the substrate along with the TiO 2 -SiO 2 amorphous structures between them, covering the entire area. To illustrate the raspberry-like morphology and the development of the hierarchical structure, a higher-magnification image of sample 3 is presented in Figure 2 . The SEM micrographs in Figure 1 and Figure 2 distinctly reveal the hierarchical coating topography on the films’ surface. These images indicate that the raspberry-like particles create a topography with micro- and nanoscale features capable of entrapping air, thereby producing superhydrophobic surfaces characterized by exceptionally high water contact angles (WCAs) and low water roll-off angles (WRAs). The presence of titania and Ti-Si functional groups on the surface can be confirmed by FTIR/ATR analysis ( Figure 3 ) and XPS analysis ( Figure 4 , Table 2 ). Figure 3 a shows the FTIR spectra of a corona-treated uncoated PP film, the sample 1 coating of TiO 2 on PP and the sample 4 coating of both silicon and titanium on PP. In the baseline spectrum of an uncoated PP, which serves as a reference, there is minimal absorption below 1000 cm −1 and the characteristic C-H bending and rocking vibrations of the polymer backbone are expressed as weak peaks in the region between 700 and 900 cm −1 . The FTIR spectra of samples 1 and 4 contain absorption bands in the lower region of the spectra around 450–750 cm −1 , which is attributed to the presence of titanium in the coatings. Specifically, the peaks around this area at 480, 670 and 735 cm −1 are attributed to Ti-O bond vibrations, symmetric O-Ti-O stretch, and non-symmetric vibrations of Ti-O-Ti during tetrahedral coordination, respectively [ 69 ], as the broad peak at 1042 cm −1 is usually referred to as the Si-O-Ti bridge vibrations [ 70 ]. The spectra of sample 4 contain very significant absorption bands between 1000 and 1300 cm −1 , where both Si-O-Si and Si-OH stretching vibrations occur. The absorbance peak observed at 1080 cm −1 , specifically in the spectrum of sample 4, may correspond to the Si-O-Si stretching vibrations, indicating the presence of silane crosslinking, which barely exist in both other graphs due to the lack of silica particles and TEOS. Those peaks can be attributed to Si-O-Si bonding forming a vast SiO 2 network during polymerization directly on the PP surface. When comparing the spectra of samples 1 and 4, it seems that the incorporation of SiO 2 in the coating leads to a more structured and interconnected coating, based on the increased intensity of the peaks in this region and their complexity. The FDTS presence is also indicated by the C-F stretching vibrations’ significant peaks at 1810 and 1210 cm −1 , corresponding to the C-F stretching vibration and bending band in the fluorinated alkyl chains, respectively. The peak at 1230 cm −1 , appearing as a shoulder to one of the C-F peaks, can be ascribed to the stretching vibrations of the Si-OH bonds, which also appears on sample 1’s spectra due to the presence of silicon atoms originating from FDTS [ 71 ]. The coated films show characteristic peaks related to the functional groups of the titanium, fluorosilane, and SiO 2 compounds, confirming the successful deposition of the superhydrophobic coatings. The presence of C-F and Si-O-Si bonds, alongside the Ti-O bonds, highlights the formation of a complex, functionalized surface on the polymeric substrate. Assessment of the coatings’ effectiveness in blocking UV radiation was carried out by transmittance measurements. The UV transmission spectrum of samples 1, 4 and plain PP is described in Figure 3 b. Since all coatings are translucent and have no color, the transmittance in the visible wavelength area between 400 and 700 nm is around 100%. The transmittance spectrum of the uncoated PP film indicates a higher transmittance in relation to the coated samples, especially around 250 nm, reaching almost 50% transmittance. It is well established that PP has a limited ability to block UV radiation, especially in the UVA region (320–400 nm) [ 72 ]. In contrast to the PP film, both coated samples demonstrate significant UV absorption, not higher than 10% in the region below 280 nm. The contribution of the coating to the UV radiation absorbance is visible right below 350 nm, where the PP absorption remains around 80%, while the coatings show a steep increase in absorbance. The enhanced UV-blocking capability of this coating can be attributed to the presence of titanium, which is known for its strong absorption of UV light, particularly in the UVB and UVC regions. Si also plays a role by contributing to the film’s overall stability and durability without significantly affecting the transmittance in the visible range. The presence of silica and titania in one coating combines the UV blocking properties of the titania as well as the increased durability-resistant properties of silica. Sample 1 exhibits better UV absorption, especially in the UVC region (below 280 nm) where the transmittance is about 6%. For all samples, the transmittance is lower only in the region of the UV section. However, the transmittance increases when reaching the visible radiation area. This demonstrates that, in addition to the excellent ability of the coatings to block UV radiation, they do not harm the transparency and clarity of the PP film in the visible spectrum. XPS is a widely used technique for quantitative surface analysis. Here, XPS was carried out to determine the surface composition of each coated PP film and to assess the influence of the amount of titanium in relation to the silica particles on the coating structure, as seen in Figure 4 a. Across all spectra, peaks representing silicon, titanium, oxygen and silica were prominent and consistent with the coating materials used ( Table 2 ). Carbon signals were observed in all samples and in similar percentages. The typical X-ray penetration depth is up to 5 nm, meaning that the carbon presence in the samples originates mostly from the PP substrate and trace amounts from the fluorinated silanes present in the coating. The titanium signal corresponding to the Ti2p binding energies confirms the presence of titanium dioxide (Ti 4+ ); the Ti2p 3/2 peak at 459.8 eV of Ti(IV) in TiO 2 can be seen in all samples in Figure 4 c. Comparing Figure 4 a,b allows us to consider that the percentage of Ti2p is higher as the silica content is lower, meaning that the signal corresponds to a greater surface coverage of titania with lower presence of the SiO 2 nanoparticles. This also corresponds with the SEM images showing a partial surface coverage of the titania with increased size of silica particles. The signal for silicon is strongest in sample 4, which contains the largest silica nanoparticles (500 nm), as opposed to the smaller 70 nm SiO 2 nanoparticles used in sample 2. Sample 1, containing TiO 2 without SiO 2 particles, exhibits traces of Si2p, which corresponds to the Si atoms of FDTS. These findings align with the expected chemical compositions and underscore the flexibility of the coating methods in tailoring surface properties for diverse applications. 3.2. Wettability Properties The wettability properties of the coatings were studied by sessile water contact angle measurements of the films. Surface roughness plays a crucial role in affecting the water contact angle and superhydrophobicity. This feature can be controlled by changing the size of silica nanoparticles used in each coating. Incorporating silica nanoparticles allows varying degrees of surface roughness depending on the size of the particles. The size of the SiO 2 nano/micro-particles combined with TiO 2 or with the TiO 2 -SiO 2 network formed on the surface was designed to form a dual-scale roughness to achieve superhydrophobicity. Table 3 shows the AFM measurements of the topography and average surface roughness values of each coated film alongside the water contact angle and rolling angle. Incorporation of larger-sized SiO 2 particles increases the surface roughness. Despite some differences in surface roughness, all samples showed excellent superhydrophobicity, as the water droplets immediately rolled off the horizontally placed surface ( Figure 5 ). Sample 1, containing only TiO 2 , had the lowest surface roughness of 5.4 nm as opposed to the larger surface roughness that was observed for sample 4. This is due to the addition of SiO 2 nano/micro-particles to the coating in which smaller nanoparticles (70 nm) were added to sample 1 and larger particles (500 nm) were added to sample 4. According to the SEM images presented in Figure 1 , the morphologies of the coatings are dependent on the size of the nano/micro-particles added to the coating solution. The addition of larger SiO 2 particles to the coating solution presents as raspberry-like structures, while the addition of smaller particles presents as tunnel-like structures on the surface. These structures increase the surface roughness of the coating and amount of air that can be trapped within the cavities, which results in extremely high water contact angles and very low rolling angles. 3.3. Self-Cleaning Ability The self-cleaning performance of superhydrophobic samples 1–4 was evaluated using sand as a model contaminant. Figure 6 illustrates the self-cleaning process, where sand powder was unevenly distributed over the coated samples and plain PP film, followed by the application of water droplets onto the contaminated surfaces. On the coated samples, the rolling water droplets effectively captured and removed the sand particles, leaving the surfaces clean. In contrast, on the plain, uncoated PP film, the sand particles were displaced but remained mixed with residual water on the surface, resulting in a dirtier appearance. 3.4. Mechanical Properties There is no single standardized test for durability properties of superhydrophobic coatings. The absence of a universally accepted method to test such coatings makes the comparison of results across various methods very difficult. Despite significant advances in the development of superhydrophobic coatings, one major gap is their lack of wear resistance and durability. Maintaining water repelling properties under physical stress remains a challenge that is mostly overcome by multi-step processes that involve long, complex, and expensive methods such as multi-step electrodeposition [ 73 ], LbL (layer-by-layer) assembly of nanoparticles and polymers [ 74 ] and many more [ 75 , 76 , 77 ]. These methods offer improvement in superhydrophobic performance but are difficult to perform and control in large-scale applications. Here, we report on a single-step method that requires no special or expensive equipment and yields excellent superhydrophobic properties. The issue of the coating’s durability was addressed by first performing surface preparation on the PP films prior to the coating and then by reacting TIIP, TEOS and SiO 2 nano/micro-particles to the surface, allowing for covalent bonding between the PP surface and the coating. As mentioned in the experimental section, the durability of each coating on PP films was assessed using the sandpaper abrasion test. A weight was placed on the back of a coated film facing down on a 240-grit silicone-carbide sandpaper. This weight applied a pressure of 2.5 kPa between the coating and the sandpaper. Durability resistance results of each coating can be seen in Figure 7 . The graphs in Figure 7 presents the WCA measurement after each rubbing cycle applied to each coating. Sample 1, containing only TiO 2 , shows the lowest durability resistance in relation to the other samples containing both TEOS and SiO 2 nano/micro-particles. Although this sample exhibits a relatively constant CA, after 60 cm, there is a gradual degradation and the sample loses superhydrophobicity after 300 cm. Sample 3, containing 250 nm SiO 2 particles, experienced a dip at the beginning of the test, but showed an increase in in WCA back to 180°. Samples 2 and 4 contain SiO 2 nano/micro-particles with larger sizes of 70 and 500 nm, respectively. Both show very impressive durability resistance and maintain their superhydrophobic properties. The test was performed with samples 2–4 up to a distance of 30 m without any WCA reduction. The combination of corona–discharge pretreatment on the PP films prior to the coating application along with incorporation of TEOS and nano/micro-particles seems to improve the adhesion of the coating through silane chemistry. The reactivity of the PP increases due to the introduction of polar groups on the surface [ 50 ]. The reactive silanol groups that form on the SiO 2 particle surface and the hydrolyzed form of TEOS can react with the functional groups on the corona-treated surface. This forms strong covalent bonds and results in enhancing the overall adhesion of the coating [ 67 ]. The presence of the SiO 2 particles increases the coating’s durability to mechanical scratching effectively. TEOS also facilitates the formation of a crosslinked SiO 2 and a silica–titania network within the coating, which provides improved resistance to chemical and physical degradation [ 78 , 79 ]. TEOS acts as a binder that integrates all the components (the SiO 2 particles and the titanium precursor) into a cohesive network, improving both the superhydrophobicity and the durability of the coatings [ 80 , 81 ]. The decrease in the hydrophobicity of the coatings at the first stages can be related to the partial removal of the coating’s top layers in a nonuniform way, as the sandpaper scratches the surface, forming certain patterns of stripes and grids. Removal of the top coating layer decreases the content of fluorocarbons on the surface, thus resulting in a less hydrophobic surface, which is expressed as an initial lowering of the contact angle. However, Figure 7 shows an increase in the contact angle after applying more sandpaper tests to the coated substrate. This can be attributed to the formation of a new surface topography of the grid-formed structure and the nano-microscale of the coating. At some point, the sandpaper reveals SiO 2 particles that can slow down the mechanical wear; thus, the new morphology remains stable even after many applications of the sandpaper test. This phenomenon highlights the importance of hierarchical surface designs in durable superhydrophobic coatings, where both surface chemistry and multi-scale roughness contribute to long-lasting water repellency even under mechanical stress. The roughness supplied by the SiO 2 particles combined with the silica–titania network provides mechanical stability and resilience while maintaining the hierarchical roughness and morphology of the coating, allowing the superhydrophobic properties to be stable. The formation of covalent bonds between the TEOS-derived silica network and the functional groups introduced by corona discharge, combined with the mechanical reinforcement provided by SiO 2 particles, significantly improves the coating’s adhesion, structural integrity and resistance to wear. In addition to the mechanical durability testing, the coating’s chemical stability and durability was tested by soaking each type of coated film in solutions with different pH levels, each with and without detergent, for 12 h. The water contact angles of the coatings were tested before and after the soaking, as can be seen in Figure 8 . All samples showed no discernible change after 12 h of soaking and no change in coating uniformity in all pH values with or without detergent. This multifunctional approach offers a promising pathway for creating durable superhydrophobic coatings with potential applications in various industries."
} | 7,721 |
36262310 | PMC9574501 | pmc | 1,502 | {
"abstract": "Summary Memristor-based Pavlov associative memory circuit presented today only realizes the simple condition reflex process. The secondary condition reflex endows the simple condition reflex process with more bionic, but it is only demonstrated in design and involves the large number of redundant circuits. A FeO x -based memristor exhibits an evolution process from battery-like capacitance (BLC) state to resistive switching (RS) memory as the I-V sweeping increase. The BLC is triggered by the active metal ion and hydroxide ion originated from water molecule splitting at different interfaces, while the RS memory behavior is dominated by the diffusion and migration of ion in the FeO x switching function layer. The evolution processes share the nearly same biophysical mechanism with the second-order conditioning. It enables a hardware-implemented second-order associative memory circuit to be feasible and simple. This work provides a novel path to realize the associative memory circuit with the second-order conditioning at hardware level.",
"conclusion": "Conclusion Memristor evolution involving the BLC stage and RS memory behavior is systemically examined. The electron/ion accumulated at interface and the concomitant redox reaction is responsible for the BLC stage. Electron/ion accumulation, migration, and diffusion are enhanced by the increasing I-V sweepings. The RS memory behavior and coexistence of NDR are ascribed to the growth of the different types of conduction paths. The corresponding dynamics in simulation are also explored by constructing a flexible compact model of the developed memristor. Furthermore, a memristor-based second-order associative memory circuit with second-order conditioning is designed and implemented, which opens a novel path for the deep integration of physical memristors into neuromorphic computing systems.",
"introduction": "Introduction Memristor, which is characterized by the nonvolatile resistance switching, has made great breakthrough in ultra-high data storage ( Wang et al., 2020a , 2020b ; Sun et al., 2021a , 2021b ), circuit system ( Kim et al., 2021 ; Wan et al., 2014 ), and neuromorphic computing because of its non-linearity, low-power consumption, and synaptic bionic ( Li et al., 2020 ; Wang et al., 2019 ; Zhou et al., 2021 ; Martin et al., 2022 ). Different applications require different types of memristor ( Zhou et al., 2019a , 2019b , 2019c , 2019d ; Ma et al., 2020 ). The digital-type memristor that is featured by high resistance ratio, fast switching speed, and long retention time is suitable for the ultra-high data storage and digital logic circuit, while the analog-type memristor, which is characterized by the high non-linearity, reliable nonvolatility, and multi-conductance states, is desirable for neuromorphic computing and associative memory circuit ( Ma et al., 2020 ; Zhou et al., 2022 ; Zhang et al., 2020 ). The ion dynamic process during the non-linear resistance change endows the memristor with unique merit to mimic the brain function. For instance, the in-sensor computing was realized using the coupling effect between photon and ion ( Zhang et al., 2020 ; Wang et al., 2020a , 2020b ; Mennel et al., 2020 ); the third-order nanocircuit element of the neuromorphic computing was developed by controlling the crystallization kinetic process ( Zhou et al., 2018 ; Hu et al., 2021 ; Kumar et al., 2021 ); and the neuro-transistor could faithfully mimic the neurotransmitter release by the ion diffusion process ( Wang et al., 2018 ). Utilizing the ion/electron accumulation, diffusion, and migration in the switching function layer, the RS evolution processes including the non-standard faradic capacitance (NFC), battery-like capacitance (BLC), and RS memory behavior were discovered in our previous work ( Sun et al., 2019 ; Zhou et al., 2020 ; Sun et al., 2020a , 2020b ; Sun et al., 2021a , 2021b ). The ion/electron dynamic processes offer the unique route to the memristor to faithfully mimic the function of human brain, such as forgetting and learning, which are the core processes of associative memory ( Ziegler et al., 2012 ; Bannur et al., 2022 ; Bannur and Kulkarni, 2020 ). The learning and forgetting functions of the classical Pavlov associative memory were demonstrated in both digital and analog memristor ( Bannur et al., 2019 ; Liu et al., 2016 ; Wang et al., 2018 ; Sun et al., 2020a , 2020b ). Memristor-based Pavlov associative memory circuit was designed to realize associative cognitive functions ( Zhang and Zeng, 2021 ). In addition, the electron vision of Pavlov associative memory was realized in the AgInSbTe and Ag/HfO x /ITO memristor ( Li et al., 2015 ; Pei et al., 2020 ; Ji et al., 2021 ). After combining the merits of analog memristor, the full-function Pavlov associative circuits were developed ( Wu et al., 2016 ; Yang et al., 2018 ; Wang et al., 2018 ). Importantly, the group of Hong stressed that the Pavlov associative memory with second-order conditioning, which included two types of learning and forgetting processes, exhibited high performance on image classification ( Du et al., 2021 ). Both simulation and real memristor for the associative circuit, thus, have made milestone progress during past decades. However, the associative circuit presented today still suffers from the limitation in the complexity of circuit and physical realization, because the circuit design does not take the elaborate RS dynamics into account. In this work, an elaborate RS dynamic involving the evolution process from the BLC to RS state is demonstrated in the FeO x -based memristor. Utilizing the different interaction between ion and electron for the BLC and RS state, an associate memory circuit verified by Pavlov associative memory with second-order conditioning is realized at hardware level.",
"discussion": "Results and discussion The memristor with the structure of Ag/FeO x /Fe-based alloy is prepared to examine the evolution of RS memory behavior, as shown in Figure 1 A. The cross-section FE-SEM image illustrates the thickness of the FeO x switching layer is ∼300 nm (top side of the Figure 1 A). The FE-SEM image exhibits that the FeO x switching layer is composed by nanosheets. Each nanosheet has an average width of ∼100 nm and an average thickness of ∼5 nm ( Figure 1 B). The 1.7 Å of inter-planar spacing observed in the HR-TEM image is contributed by the lattice plane of [422] ( Figure 1 C). The lattice planes of [211], [220], [311], [222], [400], [422], [511], and [440] obtained from the XRD measurement are consistent with the feature of γ-Fe 2 O 3 (JCPDS: 39-1346). Thus, the main component of the switching layer is Fe 2 O 3 ( Figure 1 D). The non-defined 2θ diffraction peaks located at 51.77° and 65.59° are also detected, implying that the Fe 2 O 3 nanosheet possibly contains another iron-based component. The binding energy of the core level of Fe 2p is 710.7 and 724.5 eV, resulting in a splitting energy of 13.8 eV between 2p 3/2 and 2p 1/2 ( Figure 1 E). The splitting energy of 13.8 eV is mainly contributed by the Fe-O bond in the Fe 2 O 3 . The binding energy of 529.8 eV for the O 1s core level is well agreed with the Fe-O bond in the Fe 2 O 3 ( Figure 1 F). Based on the characterization and analysis, the function layer is composed of Fe 2 O 3 . Figure 1 Memristor preparation and characterization (A) Schematic diagram the memristor with the structure of Ag/Fe 2 O 3 /Fe-alloy. The FeO x switching layer is about 300 nm, the scale bar is 100 nm. (B) FE-SEM image of self-assembled FeO x nanosheets. (C) HR-TEM image of the FeO x nanosheet characterized by the crystalline inter-planar spacing of 1.7 Å contributed by the lattice plane of [422] of the Fe 2 O 3 . (D) The XRD pattern of the FeO x film. (E and F) XPS spectra of the core level of Fe 2p and O 1s. (G) Schematical diagram the evolution process of memristor. Evolution process of the memristive system, which contains the pure capacitance state, non-standard faraday capacitance, BLC, and RS state ( Zhou et al., 2020 ; Sun et al., 2021a , 2021b ), is schematically demonstrated in Figure 1 G. To further examine the evolution process of the Ag/Fe 2 O 3 /Fe-based alloy memristor, the current-voltage ( I-V ) sweepings were operated. One can see that the memristor presents a stable BLC state ranging from 1 st to 200 th voltage sweeping ( Figure 2 A). The BLC state is featured by an obvious oxidation peak located at ∼0.2 V and a reduction peak located at ∼1.2 V. The redox peaks are weaking as the increase of the I-V sweeping from 300 th to 500 th and the current also presents an obvious increase from ∼6 to ∼30 μA ( Figure 2 B). It notes that the redox peaks nearly disappear when the I - V sweepings increase from 600 th to 782 th and the corresponding current reaches to the compliance current level ( Figure 2 C). Therefore, the redox process is possibly submerged by the increasing current. Entering 900 th ∼ 903 th sweepings, the redox process is thoroughly suppressed by the high current, but a pinched I-V hysteresis is detected ( Figure 2 D). The observation of the pinched I-V hysteresis implies that the Ag/Fe 2 O 3 /Fe-based alloy presents the RS memory behavior ( Sun et al., 2020a , 2020b ; Zhou et al., 2019a , 2019b , 2019c , 2019d ). The RS memory behavior is maintained during 905 th ∼ 917 th \n I-V sweepings ( Figure 2 E). The RS memory becomes stable during the 990 th ∼ 1000 th \n I-V sweepings ( Figure 2 F). It is worth noting that the negative differential resistance (NDR) effect is also detected during this range. Figure 2 The converter from battery-like capacitance state to resistive switching (RS) behavior triggered by the cycling endurance (A) Redox-based capacitance behavior during 200 I-V sweepings. (B) The degraded redox-based capacitance behavior during the 300 th ∼500 th I-V sweepings. (C) The degraded redox-based capacitance behavior becomes very weaken during the 600 th ∼782 th I-V sweepings. (D and E) Evolution from capacitance state to RS memory behavior. (F) Coexistence of the negative differential resistance and RS memory behavior. The Ag/Fe 2 O 3 /Fe-based alloy memristor has experienced three stages: i) BLC stage at very beginning state, ii) RS memory behavior, and iii) coexistence of the RS memory behavior and NDR effect. It notes that the BLC and NDR effect involve the ion/electron-based redox reaction process. To calculate the number of charge (Q) of redox process, the memristor is set to short-circuit once it reaches to the oxidation peak current density (J p-oxi ) or reduction peak current density (J p-red ). After integrating the area of discharge current versus time, the Q was obtained. The redox process was described by the Randles-Sevcik equation ( Valov et al., 2011 ; Valov et al., 2013 ): (Equation 1) J p − r e d o x = 2.99 · 10 5 ⋅ Z 3 2 · C r e d o x · α ϑ D r e d o x where the J p , Z, C redox , α , ϑ , and D redox denote the redox peak current value (A/cm 2 ), the number of electrons transferred during redox process, ions concentration (mol/cm 3 ), charge transfer coefficient, bias voltage scan rate, and ions diffusion coefficient (cm 2 /s), respectively. The J p-oxi and J p-red can be given by integration area of discharge curve. At very initial stage, the active Ag electrode is oxidized to Ag + , thus the number of transferred electrons in oxidation process is one (Z = 1). The charge transfer coefficient of α is ∼0.5 and the bias voltage is 500 mV/s. In the Equation 1 , the D redox is given by the Nernst-Einstein relation ( Valov et al., 2013 ). (Equation 2) μ r e d o x = D r e d o x Z e K B T The μ redox , K B , and T are the ion mobility (cm 2 /Vs), boltzman constant, and temperature, respectively. Here, the laboratory temperature is 27°C (T = 300 K) and K B is 1.38 × 10 −23 J/K. According to the Q redox obtained from the integration of discharge versus time evolution, the C redox can be calculated from the Equations 1 and 2 . Therefore, the ion concentration of C redox versus both D redox and μ redox can be given for the evolution process from BLC stage to RS memory behavior. To the BLC stage, the peak current density of the J p-oxi (1.36 × 10 −4 ) and J p-red (1.1 × 10 −4 A/cm 2 ) is calculated by the corresponding peak current (inset of the Figure 3 A) and the area of the top electrode (πr 2 ). After integrating the area of the discharge current versus time curve ( Figures 3 B and 3C), the Q oxi (1.28 × 10 −5 C) and Q red (3.9 × 10 −6 C) are obtained. The ion concentration of C oxi (0.13 mol/cm 3 ) and C red (0.012 mol/cm 3 ) is further calculated according to the transfer charge of the redox process. According to the Equation 1 , the ions diffusion coefficient for the oxidation process is 4.89 × 10 −17 cm 2 /s (D oxi ), while the reduction process is 3.76 × 10 −15 cm 2 /s (D red ). From the Equation 2 , the corresponding ion mobility is 1.73 × 10 −15 cm 2 /Vs and 1.44 × 10 −13 cm 2 /Vs for the oxidation (μ oxi ) and reduction (μ red ) process, respectively. Figure 3 Ion concentration during the evolution process (A) Evolution from the redox-based capacitance phase to the RS memory phase. (B)–(E) Current versus time under short circuit for the memristor at different current peaks. (F) Ion concentration versus migration rate and ions diffusion coefficients. Entering the RS memory stage, the corresponding charge is obtained from the integration ( Figures 3 D and 3E). Comparing with the BLC stage, the C red and C oxi for the RS memory stage increase to 0.42 and 4.05 mol/cm 3 , respectively. It will lead to the decrease of D redox and μ redox ( Figure 3 F). In particular, the μ oxi is 3.98 × 10 −16 cm 2 /Vs and the D red is 2.64 × 10 −16 cm 2 /s, which decreases one order of magnitude compared with BLC stage. The calculation demonstrates that i) the Ag/Fe 2 O 3 /Fe-based alloy with a low ion concentration at the BLC stage, but the ion holds a high mobility and diffusion; ii) entering the RS memory stage, the ion concentration sharply increases, but the mobility and diffusion are weakened. It notes that the C oxi is higher than C red for both BLC stage and RS memory behavior. Considering the oxidation reaction, the ionized Ag is described as following ( Wan et al., 2019 ; Wan et al., 2020 ; Zhou et al., 2019a , 2019b , 2019c , 2019d ; Yan et al., 2022 ): (Equation 3) A g ⇌ A g + + e According to Equation 3 , the Ag electrode is ionized to be Ag + , and then electric-field driven diffuses into the Fe 2 O 3 layer. In addition, the oxygen vacancy ( V 0 ) of the Fe 2 O 3 is sensitive to water molecule. According to the half-cell theory, the reduction reaction occurs at the counter electrode. Therefore, the redox reaction between water and Fe 2 O 3 film is described as following ( Yan et al., 2022 ; Liao et al., 2022 ; Messerschmitt et al., 2015 ): (Equation 4) H 2 O + O 0 × + V 0 ⇌ 2 O H − where the O 0 × and V 0 denote the oxygen in lattice and oxygen vacancies, respectively. According to Equations 3 and 4 , the current increases because the electron/ion diffusion and migration is enhanced, which is described by ( Valov et al., 2013 ; Zhou et al., 2018 ): (Equation 5) σ = q η h . c h . + q η e ′ c e ′ + 2 q η V O .. c V O .. + q η O H O . c O H O . where the σ, q, c, μ, h, e, V o .. , and O H o . denote the electric conductivity, elemental charge, ions concentration, ions mobility, holes, electrons, oxygen vacancy, and proton bound to oxygen in the metal oxide, respectively. Therefore, the C oxi and C red are respectively contributed by the Ag + and OH − for both BLC stage and RS memory behavior. According to our previous studies, the evolution process was dominated by ion/electron accumulation at interface, reaction, and migration ( Zhou et al., 2018 , 2019a , 2019b , 2019c , 2019d ). In other words, if the accumulation, reaction, and migration orderly happen under external stimulation, the evolution processes will happen regardless of the type of stimulations. To further verify this assumption, the I-V sweepings are measured under different moisture level because the water splitting on the surface can accelerate the evolution. Being similar with our previous studies ( Zhou et al., 2019a , 2019b , 2019c , 2019d , 2020 ; Sun et al., 2020a , 2020b ), the evolution processes are expectedly observed ( Figure 4 A). It notes that the similar evolution processes are observed under the voltage sweepings and high relative humidity. Therefore, the evolution processes for the memristive system are verified. To investigate the stability of the evolution process, resistance states for the BLC and RS are investigated. One can see that the transition state between the BLC and RS memory stage exhibits slightly volatile ( Figure 4 B). It stresses that the FeO x -based memristor is nonvolatility and is modulated by compliance current after entering the RS memory and the coexistence stage ( Figure 4 C). Figure 4 Physical mechanism for the evolutions (A) The memristor shows the evolution process from the BLC, RS to NDR as the relative humidity increasing from 0% to 85%. (B) The retention time measurement of the transition state between BLC and RS stage. (C) The RS memory behavior is similar with most of memristive devices modulated by the compliance current once our memristor complicated the evolution process. (D) Physicochemical dynamic process for the evolution RS memory behaviors. Under ultra-low bias voltage, the device exhibits the capacitance state due to the interface ions. A typical RS behavior is triggered by increasing voltage sweepings. Based on above results and analysis, the evolution process can be comprehended. At very beginning stage, the Ag + and OH − ion respectively distributes at the interface of Ag/FeO x and FeO x /Fe-alloy, resulting in the memristor exhibiting capacitance effect. The increase of ion concentration provides enough ions to diffuse and migrate in the switching layer, and finally forms a conduction filament, which directly leads to the memristor entering the RS memory behavior. The oxygen vacancy at surface (V o- sur ) and subsurface (V o- sub ) not only facilitates the water split to generate OH − but also penetrates the switching layer to form nanofilament. In this case, the three types of mechanisms including the Ag filament, OH − , and V o conduction path lead to a coexistence of the RS memory behavior and NDR effect ( Figure 4 D). According to the electron/ion dynamic, a second-order associative circuit that verified by the Pavlov associative memory is proposed. A flexible compact model is proposed to explore the dynamics in simulation. The simulation includes two current sources ( G m and G x ), and the 1F capacitor ( C x ). The model is connected as the circuit schematic, in which the terminals plus and minus denote the top and bottom electrode of memristor. Notably, at the device level, 1F capacitance may be unrealistic; while at the model level, the 1F capacitor makes the circuit structure and mathematical expression easier and simpler. According to the equivalent circuit, the relationship between current and voltage is described as following: (Equation 6) i G m ( t ) = { a 1 x ( t ) n 1 v ( t ) b 1 + a 2 ( 1 − x ( t ) n 2 ) ( 1 − e − m 1 v ( t ) ) , v ( t ) ≥ 0 a 3 x ( t ) n 3 v ( t ) b 2 + a 4 ( 1 − x ( t ) n 4 ) ( 1 − e − m 2 v ( t ) ) , v ( t ) < 0 where v ( t ) denotes the applied voltage, i Gm ( t ) denotes the current passing through the memristor, a 1 , a 2 , a 3 , a 4 , b 1 , b 2 , n 1 , n 2 , n 3 , n 4 , m 1 , and m 2 are the fitting parameters of the model. The I-V response curve of a mathematical model can approximate the actual physical model. The a 1 x ( t ) n 1 and a 3 x ( t ) n 3 terms are used to stimulate the dynamics of the kγ/ ( γ+ 1) term in the space-charge limited current mechanism. The a 2 (1- x ( t ) n 2 )(1 -e -m 1 v ( t ) ) and a 4 (1- x ( t ) n 4 )(1 -e -m2v ( t ) ) terms are used to represent the Schottky tunneling current. The x ( t ) denotes the state variable for characterizing the conductivity of the device, where x ∈ [0,1]. The value of x ( t ) is derived by integrating the current i Gx ( t ) over time through capacitor C x in the equivalent circuit. The change in the state variable is mathematically expressed as: (Equation 7) i G x ( t ) = d x d t = { α 1 ( e β 1 v ( t ) − e − β 2 v ( t ) ) f o f f ( x ) , v ( t ) > 0 α 2 ( e β 3 v ( t ) − e − β 4 v ( t ) ) f o n ( x ) , v ( t ) ≤ 0 where α 1 and α 2 are the fitting parameters. The β 1 , β 2 , β 3 , and β 4 denote the voltage control parameter. The f off ( x ) and f on ( x ) as the window functions are employed to ensure the state variable x ( t ) at the range of [0, 1]. They are described as following: (Equation 8) { f off ( x ) = exp [ − exp ( x − a off w c ) ] , v ( t ) > 0 f on ( x ) = exp [ − exp ( a on − x w c ) ] , v ( t ) ≤ 0 where w c , a off , and a on are the fitting parameters. Based on above theoretical description and mathematical derivation, the proposed memristor-based Spice model is obtained, in which the detail sub-circuit description is shown in Table S1 . The gradient descent and minimized the relative error function value were used to measure the fit between the real memristor and circuit model. Here, the error function is selected as the relative root mean squared error (RRMSE): (Equation 9) E r r m s = 1 N ⋅ ( ∑ k = 1 N ( V k − V r e f , k ) 2 V r e f 2 + ∑ k = 1 N ( I k − I r e f , k ) 2 I r e f 2 ) where N is the total number of samples, and V k and V ref,k denote the k th voltage applied to the terminals of the memristor and the circuit model, respectively. The I k and I ref,k represent the k th current through the memristor and circuit model, respectively. The V ref and I ref are the Euclidean norms of voltage and current of the circuit model, respectively. The corresponding fitting result and parameter setting are shown in Figure 5 A, where the solid spheres and solid lines represent the experimental data and model data, respectively. The whale optimization algorithm is used to obtain all the parameters in Equations 6 , 7 , and 8 . The initial value of state variable x ( t ) is 0. Figure 5 B illustrates the simulation results of the memristor after 300 th switching cycles. When the state variable x ( t ) approaches to zero, the memristor exhibits a BLC behavior. After obtaining the RRMSE from the theoretical fitting, a 0.106% is obtained for the experimental data. When the switching cycles increase to 600 th and 782 th , the state variable x ( t ) gradually increases as the maximum current increases to 12 μA. In this case, a typical RS behavior is observed in positive voltage region ( Figure 5 C). In this region, the fitted curve is well matched with the target data and the corresponding RRMSE is 0.227%. The coexistence of NDR and RS memory behavior is observed when the x (t) approaches to one ( Figure 5 D). By comparison, the peak current value of the memristor increases from 12 to 100 μA, and the RRMSE is 0.514%. Figure 5 Equivalent circuit and simulations for the evolution from capacitance state to RS behaviors The related programming code shows in Table S1 . (A) The equivalent circuit model for the Ag/Fe 2 O 3 /Fe-alloy memristor. (B–D) Simulation (line) versus experimental results (dot): (B) the capacitance state, (C) the typical RS behavior, and (D) the coexistence of the negative differential resistance and RS memory behavior. The second-order Pavlov conditioning behavior emerges along with the establishment of the classic Pavlov conditioned reflex ( Figure 6 A). The fabricated memristor exhibits an impressive evolution process from the BLC state to RS state, which shares the nearly same biophysical mechanism with the second-order conditioning. It enables a fully hardware-implemented second-order associative memory circuit to be feasible and simple ( Figure 6 B). The U1A and U2A are two XOR logic gates. The C amp means the voltage amplification circuit with the magnification A of 7. The M 1 and M 2 are the two memristor cells in BLC state and RS state, respectively. Figure 6 A fully hardware-implemented second-order associative memory circuit for neuromorphic computing (A) The collected information of classic/secondary Pavlov conditioned reflex experiments. (B) Input signals of the proposed second-order associative memory circuit. (C) The Ag/Fe 2 O 3 /Fe-alloy memristor-based second-order associative memory circuit. (D) Outputs of the proposed second-order associative memory circuit. The initial memristance of 10 6 ( R H1 ) and 10 4 ( R L1 ) is selected in BLC state, respectively. The R 1 , R 2 , R 3 , and R 4 are four fixed resistors, satisfying R 1 = R 3 = R L2 ≪ R 2 = R 4 ≪ R H1 . Meat signal S M , ring signal S R , and light signal S L are three input voltages with two different states, which can be respectively represented by the high-level state V H (2 V) and the low-level state V L (0 V). The transistor plays the role of switch with a control voltage g . The output voltage V out depends on the sum of V out1 and V out2 . The output voltage V out = 2 V ( V dd ) indicates the dog secretes saliva. Contrarily, the dog does not secrete saliva when the output voltage V out = 0 V(GND). To further demonstrate the merit of the proposed second-order associative memory circuit, simulation and analysis are conducted. One can see that the proposed circuit not only realizes the classical Pavlov conditioned reflex but also has the second-order conditioning behavior, in which the first conditioned reflex is realized by the BLC state, and the second conditioned reflex is realized by the RS state ( Figure 6 C). Classic Pavlov conditioned reflex is divided into three phases: the premise phase, first learning phase, and first forgetting phase. To simulate these three phases, the control voltage g (green line) and light signal S L (black line) remain in low-level state V L . During the premise phase (0–200 ms), the meat signal S M (red line) and ring signal S R (blue line) are respectively assigned to V H and V L . In this case, the output voltage V out = V dd = V H implies the dog secretes saliva. Contrarily, the S M and S R are respectively assigned to V L and V H ; the output voltage obeys the V out = GND = V L , indicating the dog does not secrete saliva. M 1 (orange line) and M 2 (cyan line) are maintained in the BLC state and RS state, respectively. During the first learning phase (200–700 ms), the S M and S R are assigned to V H , implying that the meat and ring are both provided to the dog. M 1 quickly switches from BLC state to RS state and the output V out = V dd = V H , which means that the dog secretes saliva. Once the S M is removed, the first forgetting phase will start along with the entire second-order Pavlov conditioning. Similarly, the second-order Pavlov conditioning is also divided into three phases: the premise phase, second learning phase, and second forgetting phase ( Figure 6 D). To simulate these three phases, the control signal g and S M are respectively operated to the V H and V L , implying that the meat is removed. During the premise phase (0–200 μs), the S R and light signal ( S L ) are operated to V H and V L , respectively. In this case, the output voltage satisfies the V out = V dd = V H ; thus, the dog secretes saliva. Contrarily, the S R and S L are respectively assigned to V L and V H . In this case, the output voltage V out = GND = V L implies the dog does not secrete saliva. During the second learning phase (200–650 μs), the S R and S L are assigned to V H , indicating that the ring and light are both provided to the dog. Here, the M 1 and M 2 are 0.5 and 7.6 kΩ, respectively. The dog secretes saliva ( V out = V dd = V H ) and the associative memory between the ring and light is established. During the second forgetting phase (650–1,200 μs), the S R and S L are respectively operated to V L and V H . The M 1 remains unchanged at 1.5 kΩ, while M 2 increases to 2 kΩ. The V out decreases from V H to V L , indicating the dog does not secrete saliva. In other words, the second-order conditioning established before is disappeared. Conclusion Memristor evolution involving the BLC stage and RS memory behavior is systemically examined. The electron/ion accumulated at interface and the concomitant redox reaction is responsible for the BLC stage. Electron/ion accumulation, migration, and diffusion are enhanced by the increasing I-V sweepings. The RS memory behavior and coexistence of NDR are ascribed to the growth of the different types of conduction paths. The corresponding dynamics in simulation are also explored by constructing a flexible compact model of the developed memristor. Furthermore, a memristor-based second-order associative memory circuit with second-order conditioning is designed and implemented, which opens a novel path for the deep integration of physical memristors into neuromorphic computing systems. Limitations of the study This work is mainly focused on the physical mechanism of the memristor and related the second-order associative memory circuit with second-order conditioning with the evolution process (From BLC to the coexistence of NDR and RS memory behavior). The circuit model and simulation only represent the specific I-V curve; thus, it is limited in terms of the cycle-to-cycle stability. In addition, the evolution can be triggered by the electric field, or testing environment, these factors have not considered in the physical models. The charge releasing is roughly estimated by the releasing charge by the short circuit. Large number of peripheral circuits and ideal capacitors are used during the simulations. Large-scale memristor array integration and parameter control are required to further realize the complex function and application in the future."
} | 7,547 |
22163213 | PMC3233664 | pmc | 1,504 | {
"abstract": "Nowadays, many software solutions are currently available for simulating neuron models. Less conventional than software-based systems, hardware-based solutions generally combine digital and analog forms of computation. In previous work, we designed several neuromimetic chips, including the Galway chip that we used for this paper. These silicon neurons are based on the Hodgkin–Huxley formalism and they are optimized for reproducing a large variety of neuron behaviors thanks to tunable parameters. Due to process variation and device mismatch in analog chips, we use a full-custom fitting method in voltage-clamp mode to tune our neuromimetic integrated circuits. By comparing them with experimental electrophysiological data of these cells, we show that the circuits can reproduce the main firing features of cortical cell types. In this paper, we present the experimental measurements of our system which mimic the four most prominent biological cells: fast spiking, regular spiking, intrinsically bursting, and low-threshold spiking neurons into analog neuromimetic integrated circuit dedicated to cortical neuron simulations. This hardware and software platform will allow to improve the hybrid technique, also called “dynamic-clamp,” that consists of connecting artificial and biological neurons to study the function of neuronal circuits.",
"introduction": "Introduction In recent years, a new discipline called neuromorphic engineering has emerged which challenges classical approaches to engineering and computer research. There are two main aspects to neuromorphic engineering: neuromorphic modeling, which reproduces neuro-physiological phenomena in order to increase the understanding of the nervous systems, and neuromorphic computation, which uses the neuronal properties to build neurally inspired computing hardware. Neuromorphic engineering proposes to fill the gap between computational neurosciences, and low-power consumption engineering (Mead, 1989 ). As alternatives to software-based solutions (Brette et al., 2007 ; Davison et al., 2009 ) and parallel graphics processors (GPUs) to alleviate the significant computational cost (Wang et al., 2011 ), neuromorphic systems are often based on custom integrated circuits (IC; Indiveri et al., 2011 ) and systems (Misra and Saha, 2010 ). A neuromorphic system could be digital, analog, or mixed. Brüderle et al. ( 2011 ) describes a methodological framework for a fully automated translation between PyNN domain and appropriate hardware configurations. In our case, we chose an analog implementation of neuron models while the communication between neurons is digital. The main advantage of the analog implementation, compared to its numerical simulation, arises from the locally analog and parallel nature of the computations. Neuroscientists provide biological measurements to computational neuroscientists who then propose a model for simulation, and for studies of the single cell or neural network dynamics. The chip designer uses this model for the design of analog neuromimetic ICs (see Figure 1 ). Figure 1 Design methodology (left-hand gray arrows) and validation of an integrated neuromimetic circuit (Right-hand gray arrow) . We wish here to reproduce the behavior of biological neurons to extend the hybrid technique, also called “dynamic-clamp,” (Le Masson et al., 1995 ) to Micro-Electrode Arrays (Bontorin et al., 2007 ). This technique consists of connecting artificial and biological neurons to create a real-time loop (Sorensen et al., 2004 ). A review of the motivation for using the hybrid technique to study biological cells can be found in Destexhe and Bal ( 2009 ). This technique has also been used with intracellular recordings to study the transmission of information from the retina (implemented by an analog retinal model of neuron) to the cortex through a hybrid thalamic network composed of a biological thalamic relay cell recorded in vitro and a model reticular cell providing feedback inhibition onto the biological neuron (Le Masson et al., 2002 ). The real-time feature of the hybrid technique extended to the Micro-Electrode Arrays would allow us to study the properties of larger biological networks. This extension of the hybrid technique has never been done. In previous work, we designed several neuromimetic chips (Levi et al., 2008 ) included Galway chip that we used for this paper. Galway includes analog operators to compute Hodgkin–Huxley (HH) formalism and multi-synapses to build neural network. Our analog IC was optimized for reproducing a large variety of neuron behaviors thanks to tunable parameters. However the IC does not guarantee that the use of parameters extracted from a biological cell will reproduce the exact behavior of that cell. We choose to compensate for the process variation and the device mismatch through the tuning process. To demonstrate that our neuromimetic IC can emulate the most important properties of biological neurons, we use electrophysiological recordings as a reference. To extend the hybrid technique, the aim is to build a biophysically realistic network model of the cerebral cortex which incorporates the diversity of intrinsic cell properties in the cortex by making use of reconfigurable integrated circuits (Saïghi et al., 2011 ). Such ICs are designed with two goals in mind: firstly to enable the construction of bio-realistic networks, and secondly to offer the possibility of dynamically tuning the model parameters. ICs are organized to form a simulation toolbox, so that a large variety of models can be implemented in real-time. Although our choice implies a costly design (Galway contains 105 pads, around 50000 components, its area is 10.5 mm 2 , and is power consumption is 550 mW), it is an interesting alternative to digital computation in simulation platforms for computational neurosciences in terms of simulation time–cost. For the simulation of one neuron, hardware system is not really relevant but seeing further, for a large scale network, the hardware system will be in real-time contrarily to the NEURON software. In addition, these ICs lead to neuromorphic network models that are typically highly scalable and able to emulate neural networks in real-time or much faster, independently of the underlying network size. The choice of neural model is important for designing an analog tunable neuromimetic circuit. All models have advantages and drawbacks (Izhikevich, 2004 ). In our case, where the hybrid technique should allow one to better understand the biological phenomena, the chosen model has to be the most biologically plausible. We have no flops (floating point operations per second) limitation for our design thanks to the analog computation. Within the family of biologically plausible point neuron models, there is a group of conductance-based models, in which ionic and synaptic currents charge and discharge a capacitor representing the neuron membrane. All of these models find their origins in the HH model (Hodgkin and Huxley, 1952 ) which will be described in the next section. Moreover, conductance-based models and real-time processing at the sample level will be helpful for the hybrid technique. The neuroscientists can dynamically play with the model parameters that have a biophysical meaning and observe the effects on the biological cells. Our chip, based on our library of analog operators, has a large range of validity domains for the parameter to reproduce different kind of neurons. The tuning of conductance-based analog neuromimetic chips has already been investigated by a few researchers (Shin and Koch, 1999 ; Simoni et al., 2004 ; Rasche and Douglas, 2007 ; Yu and Cauwenberghs, 2010 ). However, none of them compare their results with biological data. Simoni et al. ( 2004 ) and Yu and Cauwenberghs ( 2010 ) validate the tuning notably thanks to internal variables of the model which are not usually recorded in biological cells. Rasche and Douglas ( 2007 ) and Shin and Koch ( 1999 ) focus on the control of the firing rate versus stimulation. This dependency between frequency and input currents is used for the study of the network dynamic. However these neuromimetic designs were never compared to biological data. Moreover, the variety of implemented cell types is limited to the fast spiking (FS) neuron (Yu and Cauwenberghs, 2010 ) and also the regular spiking (RS) neuron (Shin and Koch, 1999 ; Rasche and Douglas, 2007 ). Only Simoni et al. ( 2004 ) presents more complex behaviors. Additionally, the chip tuning of neuromimetic circuits is also a current issue for other kinds of applications or models. Russel et al. ( 2010 ), Orchard et al. ( 2008 ) use a genetic algorithm to reproduce the activity of the Central Pattern Generator thanks to an adaptive integrate and fire model, while Brüderle ( 2010 ) has developed a technique for reproducing the statistical characteristics of the adaptive exponential integrate and fire model. To fulfill our requirement, we selected the four most prominent electrophysiological classes: “FS,” “RS,” “intrinsically bursting” (IB), and “low-threshold spiking” (LTS) neurons, inspired from the classification of Connors and Gutnick ( 1990 ). This subdivision corresponds to classifying cells according to three qualitative criteria: (1) presence or absence of spike-frequency adaptation; (2) presence or absence of burst discharges from depolarizing stimuli; (3) presence or absence of burst (or any other type of) discharge following hyperpolarizing inputs (rebound response). In this paper, we present their implementation in our analog/digital neuromimetic chip dedicated to the simulation of cortical networks using the data and the electrophysiological recording presented in Pospischil et al. ( 2008 ). In Section “Materials and Methods,” we present the methods that include the HH formalism, the implemented neural model, and the chip tuning method. The material part of this section presents the neuromimetic chip and the whole system for the parameter optimization. This is followed by the Section “Results” where we compare the biological and artificial neural behavior. Finally, in the last section, we discuss the outlooks of this research for improving the hybrid technique.",
"discussion": "Discussion The understanding of neuronal circuits is a great challenge which involves a large number of researchers from different disciplinary fields. All strategies, including neuromorphic engineering, contribute to that understanding. Even though Mead ( 1989 ) defined neuromorphic engineering as the use of the characteristics of analog components for neuronal computation, modern neuromorphic designers implement analog, digital, or mixed systems. Among all neuromorphic designs, we notice that a few groups use conductance-based-model, which belongs to the most biologically plausible single point model, with parameters tuned to understand the biology (Shin and Koch, 1999 ; Simoni et al., 2004 ; Rasche and Douglas, 2007 ; Yu and Cauwenberghs, 2010 ). However these neuromimetic designs were never compared to biological data. Additionally, we notice that the chip tuning of neuromimetic circuits is also used with integrate and fire model (Orchard et al., 2008 ; Russel et al., 2010 ) and adaptive exponential integrate and fire model (Brüderle, 2010 ) to reproduce biological network activities. In our case, we wish to insert silicon neurons among biological neurons. We decided to qualitatively compare the dynamics of our silicon circuit to biological cells with a level of details that was never done with silicon neuron. To reach our goal, we propose a simplified version of the HH formalism and the appropriate parameter sets of the FS, RS, IB, and LTS neurons that can be implemented in our analog neuromimetic chip. The models considered here are the simplest types of biophysical models where the intrinsic properties arise from voltage-dependent conductances which are described by differential equations (HH type models). The main motivation for this model type is the strong correspondence of their parameters with those in biology. We used the fixed time constant for the gating variables to simplify the model and validated this simplified model thanks to the comparison with software simulations for the four neuron types. We then tuned our chip following those models and a full-custom and dedicated technique. This optimization method, based on the DE algorithm proposed in Buhry et al. ( 2011 ), is an alternative to the estimation methods associated with voltage-clamp measurements. In any event, we observed a large discrepancy for all parameters which confirms the necessity of the tuning step. As with any circular problem, we chose an arbitrary starting point to solve it. Even though the optimization phase of our tunable chip is time-consuming, we will save time in the emulation phase thanks to three features. First, our chip requires only one tuning of its parameters. The parameter sets are then stored in a database and the required parameters are loaded into the chip as needed to emulate any given cell type. Second, the model parameters can be modified on-line at any time. Third, the neuron type can also be modified on-line by connecting/disconnecting an ionic channel and/or modifying a few parameters. For example, the burst in IB cell can be modulated by the value of the conductance or time constants of the calcium current. All these on-line changes take only a few microseconds. This will enable the user to alter parameters in order to study their effect on the dynamic of the biological network. Finally, we tested the parameters obtained by comparing the behavior of the membrane voltage in the recordings of biological cells with membrane voltages simulated with our chip. This comparison is doable thanks to the translations rules between biological and hardware neurons based on the chip characteristics. We directly compare the behavior of our chip with biological recordings. Our results show that our system is able to reproduce the main features of four common classes of cortical cells. In the near future, we plan to tune hundreds of neurons and create a large artificial network composed of 10% of IB and LTS, 20% of FS and the remainder of RS neurons. These neuromimetic chips will be embedded in a full-custom system that will allow us to scale up the neuromimetic network to about 100 real-time and biophysically realistic neurons. This system will help be useful for understanding neural network dynamics using the hybrid technique. The hybrid technique has already provided some valuable results at the single cell and small network levels. Any extension of the hybrid technique to larger networks should contribute to the understanding of neuronal circuit function. The following characteristics are necessary: (a) operate in biological real-time, (b) use biophysically realistic models, and (c) permit tuning. Thanks to the parallel nature of the computations, the analog design can easily have real-time characteristics even for the most accurate neural models. However, for the system to be truly useful, the chip designers must propose an associated tuning technique for it. All of these specifications are included into our system. Then compared with the original hybrid technique where one artificial neuron was connected to one biological cell (Le Masson et al., 2002 ), a network this large will be enough to employ the hybrid technique with Micro-Electrode Arrays composed of 64 electrodes. Due to the parallel nature of the circuit, the real-time properties of our system will easily be preserved independently the number and the complexity of the emulated neurons. This extension of hybrid technique has never been done."
} | 3,943 |
39920822 | PMC11806602 | pmc | 1,505 | {
"abstract": "Polyhydroxybutyrate (PHB) is a class of biodegradable polymers generally used by prokaryotes as carbon sources and for energy storage. This study explored the feasibility of repurposing used soybean oil (USO) as a cost-effective carbon substrate for the production of PHB by the strain Gordonia terrae S-LD, marking the first report on PHB biosynthesis by this rare actinomycete species. This strain can grow under a broad range of temperatures (25–40 ℃), initial pH values (4–10), and salt concentrations (0–7%). The findings indicate that this strain can synthesize PHB at a level of 2.63 ± 0.6 g/L in a waste-containing medium containing 3% NaCl within a 3 L triangular flask, accounting for 66.97% of the cell dry weight. Furthermore, 1 H NMR, 13 C NMR, and GC–MS results confirmed that the polymer was PHB. The thermal properties of PHB, including its melting (T m ) and crystallization (T c ) temperatures of 176.34 °C and 56.12 °C respectively, were determined via differential scanning calorimetry analysis. The produced PHB was characterized by a weight-average molecular weight (M w ) of 5.43 × 10 5 g/mol, a number-average molecular weight (M n ) of 4.00 × 10 5 g/mol, and a polydispersity index (PDI) of 1.36. In addition, the whole genome was sequenced, and the PHB biosynthetic pathway and quantitative expression of key genes were delineated in the novel isolated strain. In conclusion, this research introduces the first instance of polyhydroxyalkanoate (PHA) production by Gordonia terrae using used soybean oil as the exclusive carbon source, which will enrich strain resources for future PHB biosynthesis. Supplementary Information The online version contains supplementary material available at 10.1186/s13068-025-02613-w.",
"conclusion": "Conclusions A newly isolated salt-tolerant Gordonia terrae S-LD can be utilized to biosynthesize PHB from used soybean oil with intriguing physicochemical properties. This finding is highly important, as it may favor process economics. Under waste-containing medium conditions, a maximum PHB concentration of 2.63 g/L was achieved after 96 h of cultivation at a 3% salt concentration. In addition, the whole genome and PHB synthesis pathway of this strain were sequenced and elucidated. These observations suggest that Gordonia terrae S-LD may be a promising candidate for PHB production using different oil-based substrates. Further investigation into Gordonia terrae is anticipated to determine whether PHB production can be more economically viable for subsequent large-scale industrial applications.",
"introduction": "Introduction Polyhydroxybutyrate (PHB) is a bioplastic synthesized by microorganisms, can be fully degraded in the natural environment, and possesses advantages comparable to those of traditional nondegradable plastics. These advantages make PHB a promising material for a variety of applications, such as medical devices, packaging materials, and disposable plastic products [ 1 , 2 ]. Despite its many advantages, PHB faces significant challenges that hinder its widespread adoption. The primary issue is its high production cost, which stems from the complex fermentation processes required to produce PHB and the relatively low yields compared with those of traditional petroleum-based nondegradable plastics. The reasons for the high cost of PHB production are mainly due to the high cost of the carbon source required for production, and the limited number of strain species currently available for demonstration purposes. Most PHA producing bacteria usually take up simple sugars (e.g., glucose and fructose) as their main carbon source [ 3 ]. However, the cost of industrial grade glucose is approximately 35–50 USD/kg. Ongoing research aims to improve production efficiency and reduce the cost of the carbon source, with some studies focusing on the use of waste materials, such as oil-based substrates, agroindustrial residues, and domestic sludge, as feedstocks for PHB production [ 4 ]. Soybean oil consumption accounts for the highest consumption of edible vegetable oils in China, accounting for approximately 29.6%. Thus, waste soybean oil may be a cost-effective oil-based substrate for synthesizing PHB and has great potential for large scale applications. Studies have demonstrated that Cupriavidus necator can synthesize PHB using mixed waste cooking oil as the sole carbon source, yielding 7.6 g/L PHB [ 5 ]. This research not only offers a solution for mixed waste oil management but also contributes to sustainable plastic production. Researchers have successfully utilized active sludge from waste water treatment, which is rich in microorganisms capable of synthesizing PHB, to produce PHBV (a copolymer of PHB and hydroxyvalerate) [ 6 , 7 ]. These studies highlight the advantages of using microbial processes to synthesize PHB using oil-based substrates as a carbon source, which can assist in solving environmental problems while reducing the cost of PHB production. However, many waste cooking oils or wastewaters have relatively high salinities or other extreme physicochemical properties, which strongly affect their bioavailability and bioconversion. In recent years, halophilic Halomonas spp., which constitute the next generation of industrial biotechnology [ 8 ], have also been studied and reported to be useful for the large-scale production of PHA [ 9 ]. Extremely halophilic bacteria can grow under harsh conditions (such as extreme salinity or osmotic pressure), where most microorganisms are unable to proliferate, and consequently, they are more resistant to contaminants during fermentation. These methods allow for the open continuous production of PHA, reducing the sterilization process and energy consumption [ 10 ]. However, such extremely halophilic bacteria generally need to utilize sugars as a carbon source to synthesize PHB. In addition, the corrosion of fermentation equipment caused by high salt concentrations and the difficulty of treating large quantities of high-salt wastewater may also limit the development of PHB synthesis by extremophiles on a large scale [ 11 – 13 ]. In this study, Gordonia terrae is reported to synthesize PHB during cultivation on used soybean oil containing 3% NaCl. It can grow under a broad range of temperatures, initial pH values, and salt concentrations. The physio-chemical properties of PHB synthesized from used soybean oil were examined. Based on the whole genome, the PHB biosynthetic pathway and quantitative expression of key genes were delineated in the novel isolated strain. This provides strain resources and a process basis for future large-scale production applications of PHB synthesis using fried soybean oil as an economically feasible carbon substrate.",
"discussion": "Discussion To the best of our knowledge, few studies on the production of PHB by the rare Actinomycete genus Gordonia have been conducted. Most studies on Gordonia sp. have focused on the degradation of alkanes, n-hexadecane, and benzenes, and the production of emulsifiers or carotenoids [ 23 ]. In this study, a salt-tolerant strain of Gordonia terrae that can utilize used soybean oil to synthesize PHB under 3% salt was successfully screened. Moreover, this strain has a wide range of growth conditions. This study represents the first report on PHB production by Gordonia terrae , particularly utilizing used cooking soybean oil as the sole carbon source. Numerous studies have confirmed that microorganisms can synthesize PHB using various carbon sources, such as sugars, corn, potatoes and more [ 4 ]. The use of inexpensive carbon and energy sources can significantly reduce the cost of synthesizing PHB while reducing environmental pollution problems [ 24 ], and used soybean oil is one of the best representatives of these inexpensive carbon and energy sources. Cupriavidus necator H16, also known as Ralstonia eutropha , is currently the most extensively researched and typical example of a strain that can be used to synthesize PHA from waste oil [ 25 ]. The highest cell dry weight (CDW) of 9.44 g/L and the maximum PHA accumulation of 7.22 g/L were obtained with C. necator when low-cost vegetable oil was used [ 26 ]. Ruiz et al. used the Pseudomonas chlororaphis and used waste cooking oil (WCO) to synthesize PHA, reaching yields of 0.09–0.14 g PHA/g oil substrate [ 27 ]. Most of the conventional wild strains found thus far are essentially intolerant to salinity. However, most of the various waste oils or oily wastewaters produced in daily life have high salinities (≥ 3%). Therefore, it is necessary to explore some salt-tolerant strains for potential application strains in the synthesis of PHA using waste oil-based substrates. The newly screened Gordonia terrae can not only synthesize PHB in the presence of 3% salt, but also tolerate a salt concentration gradient of up to 7%, which greatly broadens the scope of use of this strain in subsequent large-scale applications. The molecular weight of PHB synthesized by C. necator using waste rapeseed oil as a carbon source was 5.77 × 10 5 g/mol and the PDI was 2.66 [ 28 ]. C. necator was grown on extracted spent coffee ground oil to synthesize PHB with a lower M w of 4.27 × 10 5 g/mol and a PDI of 2.51 [ 29 ]. According to the literature, the molecular weight of PHB synthesized from different types of waste oil is generally between 0.4 and 3.0 × 10 5 g/mol, and the PDIs range from 1.2 to 3.0 [ 30 ]. The PHB produced by Gordonia terrae bioconverted from used soybean oil was characterized by a weight-average molecular weight (M w ) of 5.43 × 10 5 g/mol, a number-average molecular weight (M n ) of 4.00 × 10 5 g/mol, and a polydispersity index (PDI) of 1.36. The PDI of the PHB polymers produced in our study is within the range reported in the literature, whereas M w is considered relatively high. A higher PHB molecular weight will increase the strength and hardness of the material and improve its physical and chemical stability, making it suitable for application scenarios that require high stability. Moreover, the PDI value of PHB synthesized by this strain is relatively smaller, indicating that the molecular weight distribution of the PHB synthesized by this strain is relatively uniform, which makes subsequent applications easier. In addition, the microbial synthesis of PHB is a multifaceted process that can be influenced by various factors, including the type of carbon source and the metabolic pathways employed. Understanding and optimizing these pathways are crucial for improving the efficiency and cost-effectiveness of PHB production, especially for new strains and new synthetic pathways. Some bacteria, such as Pseudomonas aeruginosa and Aeromonas caviae, utilize the β-oxidation pathway to synthesize PHB. In this pathway, the intermediate enoyl-CoA is converted to (R)−3-hydroxyacyl-CoA by enoyl-CoA hydratase, which is then polymerized by PHA synthase (PhaC) into PHB. Bacteria such as Pseudomonas putida can synthesize PHB through the de novo fatty acid synthesis pathway. Acetyl-CoA enters the fatty acid synthesis pathway, and the intermediate (R)−3-hydroxyacyl-ACP is converted to (R)−3-hydroxyacyl-CoA by 3-hydroxyacyl-ACP: CoA transacylase, which is then polymerized into PHB by PHA synthase(PhaC) [ 31 ]. In this study, analysis of the changes in medium-to-long-chain fatty acids during PHB synthesis, revealed that soybean oil used for 24 h is decomposed into medium-to-long-chain fatty acids under the action of lipase. On the basis of our analysis of the genome of the strain and the elucidation of the synthesis pathway, it is speculated that these medium–long-chain fatty acids are decomposed into short-chain fatty acids through β-oxidation, and these short-chain fatty acids accumulate and subsequently synthesized into PHB. Notably, some strains, such as Comamonas testosteroni and Pseudomonas putida , can utilize other oils to synthesize medium chain length poly (3-hydroxyalkanoates) (mcl-PHAs). They can utilize the intermediate metabolites in the β-oxidation of fatty acids, such as 3-ketoacyl-COA and (R)-enoyl-CoA, to synthesize medium-chain PHAs. However, this strain was not able to directly convert medium-chain fatty acids into medium-chain PHAs. We speculate that this may be due to the substrate specificity of PHA synthase. However, this study provides a good research point and possibility for our subsequent metabolic engineering efforts. In subsequent work, gene knockout and heterologous expression or modification of enzymes should be used to further verify the polymer synthesis pathway. Advances in metabolic engineering have enabled increased PHB accumulation in microorganisms. The choice of carbon source significantly affects PHB production. Previous studies have compared the efficiency of glucose, acetic acid, and ethanol as carbon sources for PHB production, highlighting the importance of optimizing carbon source utilization in the production process. Moreover, recent studies have shown that recombinant Escherichia coli can use acetic acid and lactic acid as carbon sources for PHB production [ 32 ]. In this study, the used soybean oil employed as the substrate matrix, which is actually a relatively complex substance, mainly consisted of 18-carbon and 16-carbon fatty acids. This composition prompted us to pay attention to the issues of substrate selection and metabolic pathway diversification when we carry out future metabolic engineering of this strain. This study also demonstrates the flexibility of the metabolic pathways in the utilization of different carbon sources. The transformation of waste into new raw materials is a fundamental aspect of the circular economy framework [ 33 ]. The higher cost of PHB synthesis impacts its market competitiveness, especially compared to less expensive, non-biodegradable alternatives. Addressing these challenges is essential for ensuring the commercial viability and environmental impact of PHB as a sustainable alternative to traditional plastics. Moreover, improper disposal of used soybean oil and other waste oils causes serious environmental pollution or endangers human health. For large-scale PHB production, the use of inexpensive carbon sources or energy is a critical prerequisite. This strain has good salt tolerance and can grow well over a wide temperature and pH range. Therefore, there is great potential for this strain to be scaled up for the bioconversion of waste oils. The ability of Gordonia terrae to degrade hydrocarbons and BTEX compounds has been reported [ 34 ]. Moreover, higher molecular weight PHB has better oil–water separation effect while improving its stability and pollution resistance. Since the molecular weight of PHB synthesized by this strain is high, it is easier to separate and purify it during subsequent large-scale application. These characteristics determine the great advantages of this strain in future large-scale applications. Of course, in subsequent research we must also consider the planned scale-up process and purification issues and even conduct a life-cycle assessment [ 35 ]."
} | 3,777 |
35200427 | PMC8869736 | pmc | 1,508 | {
"abstract": "The ever-increasing use of plastics, their fossil origin, and especially their persistence in nature have started a wave of new innovations in materials that are renewable, offer the functionalities of plastics, and are biodegradable. One such class of biopolymers, polyhydroxyalkanoates (PHAs), are biosynthesized by numerous microorganisms through the conversion of carbon-rich renewable resources. PHA homo- and heteropolyesters are intracellular products of secondary microbial metabolism. When isolated from microbial biomass, PHA biopolymers mimic the functionalities of many of the top-selling plastics of petrochemical origin, but biodegrade in soil, freshwater, and marine environments, and are both industrial- and home-compostable. Only a handful of PHA biopolymers have been studied in-depth, and five of these reliably match the desired material properties of established fossil plastics. Realizing the positive attributes of PHA biopolymers, several established chemical companies and numerous start-ups, brand owners, and converters have begun to produce and use PHA in a variety of industrial and consumer applications, in what can be described as the emergence of the “PHA industry”. While this positive industrial and commercial relevance of PHA can hardly be described as the first wave in its commercial development, it is nonetheless a very serious one with over 25 companies and start-ups and 30+ brand owners announcing partnerships in PHA production and use. The combined product portfolio of the producing companies is restricted to five types of PHA, namely poly(3-hydroxybutyrate), poly(4-hydroxybutyrate), poly(3-hydroxybutyrate- co -3-hydroxyvalerate), poly(3-hydroxybutyrate- co -4-hydroxybutyrate), and poly(3-hydroxybutyrate- co -3-hydroxyhexanoate), even though PHAs as a class of polymers offer the potential to generate almost limitless combinations of polymers beneficial to humankind. To date, by varying the co-monomer type and content in these PHA biopolymers, their properties emulate those of the seven top-selling fossil plastics, representing 230 million t of annual plastics production. Capacity expansions of 1.5 million t over the next 5 years have been announced. Policymakers worldwide have taken notice and are encouraging industry to adopt biodegradable and compostable material solutions. This wave of commercialization of PHAs in single-use and in durable applications holds the potential to make the decisive quantum leap in reducing plastic pollution, the depletion of fossil resources, and the emission of greenhouse gases and thus fighting climate change. This review presents setbacks and success stories of the past 40 years and the current commercialization wave of PHA biopolymers, their properties, and their fields of application.",
"conclusion": "7. Conclusions As mentioned in the introduction, we are currently witnessing a significant wave of activities in PHA development and commercialization. While most of the commercial activities are focused on scl -PHA as bulk polymers such as P(3HB), P(3HB- co -3HV), P(3HB- co -4HB), and P(3HB- co -3HHx) and to a minor extent on P(4HB) and highly amorphous mcl -PHA copolyesters, most of these biopolymers are processed exclusively into single-use products. The next several years will draw a clearer picture about which concepts in the marketing of PHAs are indeed future-fit, be it in terms of market requirements, customer acceptance, sustainability, or economic feasibility. However, it is quite clear that the long-term success of commercial PHA needs to be grounded on some defined solid fundaments: robust and powerful microbial production strains, optimized and simple cultivation facilities, amply available renewable feedstocks, and sustainable and inexpensive downstream processing technologies, pointing to its ease of production and acceptable price points that would allow its proliferation to serve various markets In any case, considering the currently already well-established status of biotechnological PHA manufacturing and the far-developed understanding of the metabolic background of PHA biosynthesis, we can expect that the current wave of commercialization of PHAs has indeed started. This is in contrast to the first commercialization efforts for PHAs made several decades ago: That time, we witnessed some short-term efforts and successes in biopolymer production during periods of crude oil shortage and exploding prices of fossil resources, but rapidly declining interest in renewable-sourced polymers, chemicals, and solvents as soon as the price for fossil resources dropped again. Now, the time is ripe for a continued market presence of these biopolymers. During the next few years, it will be of utmost interest to follow which types of PHA biopolymers, industrialization concepts, production strains, raw material sources, cultivation modes, and applications (single-use or durable) will prevail in the long term.",
"introduction": "1. Introduction Plastics based on fossil resources have proven their valuable role in increasing our quality of life in various sectors, as shown by their wide-spread application as packaging materials for food and other perishable goods; in the medical and pharmaceutical field; and in the transportation sector, e.g., in automobiles or aircraft, where plastics have enabled novel technological and safety-related improvements. Thus, it is undisputed that plastics, which ubiquitously accompany us in our daily lives, have made our society more convenient. However, the persistence of fossil plastics at their end of life, the insufficiency of collection and recycling systems, and their leakage into terrestrial and aquatic environments, ultimately leading to microplastic pollution of the eco- and biosphere, are omnipresent threats to all life on earth [ 1 ]. A UN study in 2005 concluded that plastic waste in oceans would result in the formation of microplastics and that these microplastics constituted the next environmental threat so great that it had the potential to be the next epic threat [ 2 ]. These and the topic of greenhouse gas emissions connected to the production and incineration of fossil plastics have raised significant awareness among consumers and finally among policymakers. While regulations are being put in place to reduce the use of plastics, not all of these measures are necessarily beneficial to the environment and practical from a convenience standpoint. For a real cure to the plastic pollution predicament, real, sustainable solutions are needed [ 1 ]. It is important that such solutions forgo the negative environmental and logistical impacts of plastics while retaining their benefits. Alternatives that fulfill all of these criteria have been provided by nature, and such materials already exist. Polyhydroxyalkanoate (PHA) biopolyesters, produced by and playing multifaceted metabolic roles in numerous bacteria and archaea, are expedient examples of materials that bridge the desired benefits of plastics without endangering the environment. For illustration, a recent study by Dilkes-Hoffman et al. reports the complete degradation of PHA bottles in marine environments within 1.5 to 3.5 years, in contrast to the decades or even centuries that disintegration of petro-plastic bottles would take [ 3 ]. Similar to other biopolymers such as carbohydrates, nucleic acids, and proteins, PHA biopolymers have been established as macromolecules embedded into the closed cycles of producing and degrading materials in nature: PHAs are produced by living organisms (“biosynthesized” materials) and they biodegrade. Moreover, PHAs are produced from renewable raw materials, thus originating from natural substrates instead of fossil resources. Importantly, PHAs are biocompatible to humans and other life forms and are readily metabolized to non-toxic compounds when ingested by living organisms [ 4 ]. The ecological concerns of fossil plastics, together with ongoing limitations of fossil resources, have now opened the door for PHA biopolymers to play a front-running role for industry and society while maintaining nature’s cycle of circularity and sustainability. The skyrocketing crude oil prices in the 1970s prompted the first commercialization efforts in PHAs. However, much of that effort slowed down after the recovery of crude oil prices, although the scientific research continued. Price and availability were identified as the primary obstacles to the continued development and commercialization of PHAs. They were identified as not being price competitive to well-established fossil plastics, and certain challenges in their processability also had to be overcome. The material properties of PHAs do not exactly match those of the fossil-based competitors; in other words, they were not drop-ins for fossil plastics, although they cover a substantial spectrum of their property profile [ 5 ]. Now, many decades later, after having accumulated significant knowledge on production, processability, and end-of-life outcomes of PHA, we finally are on the threshold of serious and sustained commercialization efforts of these biopolymers. It is now better understood how the production price can be lowered by resorting to inexpensive or even near zero-cost carbon sources [ 6 , 7 ], which natural microbes are best suited to produce the various types of PHA from a given substrate [ 8 ], how microorganisms can be tailored using systems biology and metabolic engineering approaches [ 9 , 10 ], how bioprocesses can be rendered to consume less energy by running PHA production under low-sterility or nonsterile conditions with extremophilic microorganisms [ 11 ], and how to optimize downstream processing for recovery of intracellular PHA [ 12 ]. In addition, a large body of knowledge has been generated on fine-tuning the (co)polymer composition and thus tailoring the product properties during the bioprocess [ 13 ] and on facilitating PHA processing by blending appropriate chemical additives and other polymers [ 14 ]. Indeed, a growing number of companies spread over different global regions have started commercial-scale PHA biopolymer production for processing towards vendible items by melt extrusion, injection molding, 3D printing, electrospinning, etc. [ 15 , 16 ]. This new wave of PHA commercialization is increasingly becoming an integral part of current concepts of the bioeconomy and circular economy, which, as postulated by the European “Green Deal” [ 17 ], are characterized by the replacement of “end-of-pipe products” such as fossil plastics, especially for single-use applications, by biodegradable alternatives based on renewable carbon to drastically reduce plastic pollution and greenhouse gas emissions and to curb global warming [ 1 ]. In fact, calculations based on a plethora of life cycle studies estimate that replacement of 1 kg fossil plastic by PHA could salvage on average CO 2 emissions by 2 kg and around 30 MJ of fossil resources on an energy basis [ 18 ]. Material properties of PHA biopolymers are dependent on the type and distribution of various monomeric building blocks; PHAs are a versatile group of biomaterials, with characteristics that range from elastomeric to semicrystalline thermoplastic-like polymers [ 19 ]. Despite the discovery of more than 150 different hydroxyalkanoate (HA) building blocks that constitute the PHA biopolyester family, only a limited number of PHA copolymer types have reached industrial maturity. As shown in the subsequent sections, we currently witness considerable activities in different regions globally towards commercialization of a few PHA biopolymers, namely the homopolyester poly(3-hydroxybutyrate) (P(3HB)); the copolyesters poly(3-hydroxybutyrate- co -3-hydroxyvalerate) (P(3HB- co -3HV)), poly(3-hydroxybutyrate- co -4-hydroxybutyrate) (P(3HB- co -4HB)), and poly(3-hydroxybutyrate- co -3-hydroxyhexanoate) (P(3HB- co -3HHx)); and, to a minor extent, the homopolyester poly(4-hydroxybutyrate) (P(4HB)) and some medium-chain-length PHA ( mcl -PHA) copolyesters. The chemical structures of these biopolyesters are illustrated in Figure 1 . While it is challenging to estimate the current volume of PHA produced industrially, it has, to the best of our knowledge, not yet exceeded 10,000 t annually at the time of submission of the present article (January 2022); however, capacity expansions of over 1.5 million t have already been announced for the next 5–10 years, and an additional 1 million t are in the planning stages. Compared with the estimated global plastic production of roughly 400 Mt per year, the share of PHA is negligible [ 20 ]. Therefore, this review is dedicated to drawing a clear current and developing picture of a very old biopolymer platform that is transforming into the new “PHA industry”. The intent here is to take the reader on a journey through the history of PHA commercialization attempts, highlighting the obstacles and stumbling blocks that often made this path difficult and the targeted applications where PHA is intended to be commercialized."
} | 3,265 |
30460107 | PMC6138339 | pmc | 1,509 | {
"abstract": "ABSTRACT A typical colony of Neotropical army ants (subfamily Ecitoninae) regularly raids a large area around their bivouac by forming a narrow directional column that can reach up to one hundred meters in length. The raid is finished and then relaunched 12–17 times, each time toward different orientation. After completing all bouts the colony relocates to a new area. A hypothetical alternative to this foraging mode is raiding radially and symmetrically by expanding the search front in every direction like a circular bubble. Using an existing agent-based modeling software that simulates army ants’ behavior, we compared the two possible modes of foraging in different food distributions. Regardless of the food patch abundance, the radial raiding was superior to the directional raiding when food patches had low quality, and the directional raiding was favorable when the patches were rich. In terms of energy efficiency, the radial raiding was the better strategy in a wide range of conditions. In contrast, the directional raiding tended to yield more food per coverage area. Based on our model, we suggest that the directional raiding by army ants is an adaptation to the habitats with abundance of high-quality food patches. This conclusion fits well with the ecology of army ants.",
"introduction": "Introduction The army ants are specialized collective predators that always forage in large groups (Kronauer 2009 ). Their colony can form a swarm of many thousands of hunters, advancing in a column over a hundred meters long (Couzin and Franks 2003 ). In the ‘nomadic phase,’ they move their camp every day, but when a colony enters ‘statary phase,’ it launches multiple successive ‘raids (Willson et al. 2011 ).’ The raids occur about once a day, and they avoid the recently exploited direction (Willson et al. 2011 ). After depleting a region with about 14–17 raids, the colony relocates to a new area (Willson et al. 2011 ). Their nomadism depends on their skills to form a ‘bivouac,’ a huge ball of ants that temporarily shelter the young and the queen (Anderson et al. 2002 ). This set of behaviors has been evolutionarily conserved in more than 200 species of this clade over two continents (Brady 2003 ). However, from the exploratory viewpoint, their columnar raid formation is an unusual choice. The search front is narrow (as short as one tenth of the column length; Couzin and Franks 2003 ), and only the minority of the foragers are exposed to the novel environment. The remaining majority run over the same path as their predecessors did, contributing almost nothing to the search. The successive raids will increase the final coverage area, but still it seems inferior to non-directed search patterns. For example, an Eciton burchelli raid can employ 200,000 individuals (Couzin and Franks 2003 ), and if they were to radiate uniformly from the colony in every direction, they could form an unbroken ring as large as 63,000 body-lengths in diameter. Even with minor workers, this would be nearly 200 meters wide, and it would not miss any single food item within the expanding ‘bubble’. This would provide much better coverage than the column raiding does. On the other hand, while being inferior in terms of exploration, the directional column raiding enables instant mass transportation after discovery. If the target food source is far away, the colony may save a considerable time by skipping the return trip of the discoverers and the dispatch trip of the recruited transporters. However, to justify having hundreds of thousands of potential transporters following the search front, the colony needs to ensure that the frontiers will find a very rich food patch. Otherwise, it could end up in a waste of time and energy for a very little gain. Therefore, the distribution of food should be a major parameter affecting the advantages of the directional column raiding. To test the raiding performance under different food distributions, we chose a simulation software (Brown 2008 ) aimed at modeling Eciton species, the popularly studied new world army ants. We modified the program to enable comparison between the naturally occurring ‘directional raiding’ and the hypothetical ‘radial raiding’ strategies. We expected that the radial raiding would provide better coverage, but the directional raiding would yield more food in a certain range of food distributions.",
"discussion": "Discussion Although the simulation showed that the directional raiding is generally coverage-efficient, this mode of foraging is not very energy-efficient (the second and third columns of Figure 2 ). These trends are likely to arise when a large crowd of ants is concentrated in a small number of food patches. In this situation, most of the individuals are active in the already visited area rather than a new unexplored territory, leading to a more thorough search and the higher coverage efficiency. However, the movement efficiency may be negatively impacted by collisions between individuals due to the high density. Why do the army ants raid directionally? We believe that the coverage efficiency is unlikely to be the ultimate reason, because it is difficult to find a selective pressure that may adaptively constrain the raid coverage. Neotropical army ants are the top predator of the ecosystem (O’Donnell et al. 2007 ) except when they rarely encounter the anteaters (Willson et al. 2011 ), and unlike the Afrotropical Dorylus (Wilson 1971 ), inter-colonial conflicts are easily resolved without much mortality (Willson et al. 2011 ). They also have a set of behaviors specifically tuned to access difficult terrains, such as the ‘living bridge (Reid et al. 2015 ; Graham et al. 2017 )’ or the ‘pothole plug (Powell and Franks 2007 ),’ implying that they gain benefit by expanding their activity range. Finally, they are nomadic species without permanent shelter, and they frequently relocate to a newer area (Kronauer 2009 ; Willson et al. 2011 ; Garnier and Kronauer 2017 ) suggesting again that they do not pursue smaller coverage. Can selection in a foraging context explain why the army ants raid directionally? Our model demonstrated that the directional column raiding was not a good foraging strategy to search for scattered small food sources. The model parameters were determined from the observation, so the natural selection could have optimized them for the directional raiding. In contrast, the radial raiding behavior in our simulation did not involve any further optimization to the new foraging regime. Only with the diversification of the initial departing directions, just one simple alteration of the model parameter, the colony gained substantial energetic reward in a wide range of test conditions. Compared to the radial raiding, the 15 directional raids were often inadequate to provide coverage over the full circle of range available to the colony, and left many food patches unexploited. However, if the food patches were of very high quality, the directional raiding had advantages in various aspects of efficiency. Unlike the radial raiding, the directional raiding could maintain the density of the search front even after a considerably long expedition. This would allow fast and instant concentration of the workforce into a resourceful patch, draining it within a short time. After that, the subsequent raids are unlikely to re-visit the depleted patch. On the other hand, in a radial raiding, it was difficult to recruit the remotely scattered foragers to the discovered patch. The discoverers could lay a pheromone trail back to the bivouac, but the information could not reach the majority of the outside foragers until they come back home. This bottlenecked the transition from exploring to transporting jobs. The previous research on army ants support the adaptive value of directional raids in habitats with high food patch quality. Army ants, both the neotropical and the Afrotropical groups, are believed to have evolved from a common Gondwanan ancestral clade that preyed on social insect colonies (Berghoff 2003 ; Brady 2003 ; Brady et al. 2014 ), and numerous species still maintain the diet (Berghoff 2003 ; Ramirez and Cameron 2003 ; Powell and Clark 2004 ; Le Breton et al. 2007 ; Souza and Moura 2008 ; Kronauer 2009 ; Powell 2011 ; Dejean et al. 2014 ). Others have their diet diversified, but they also generally opt for large preys or rich litter patches (O’Donnell et al. 2005 ; Kaspari et al. 2011 ). A study reported that some army ant species generally cherry-pick higher quality patches, only skimming the most convenient 25% of the animal biomass and leave the rest intact (Kaspari et al. 2011 ). Interestingly, in our simulation, the conditions favorable to the directional raiding were identical to the conditions of less exhaustive exploitations ( Figure 2 (d,h,l)). To sum up, the directional raiding is a trait closely related to highly rich resources that are not easily exhaustible, both in the real world and in our simulation. Then, why the majority of other ant species that rely on rich food patches, e.g. the leafcutter and honeydew-harvesting ants, do not utilize the directional column raiding? In these species, a large number of reserve recruits waiting in the nest compensates the downside of the undirected search (Jaffe and Deneubourg 1992 ). When a recruitment signal is given, the reserves follow the pheromone trail to the newly discovered food source, allowing massive and concentrated exploitation (Jaffe and Deneubourg 1992 ; Shaffer et al. 2013 ). Most ant species have highly varied recruitment strategies based on this principle, implying the evolutionary flexibility and universal utility of this behavioral scheme (Hölldobler and Wilson 1990 ). In contrast, army ant foragers could not benefit from this recruitment scheme, because they leave the bivouac at a much faster rate and save less reserve in the colony (Deneubourg et al. 1989 ; Sole et al. 2000 ; Brown 2006 ). This extreme scout-reserve imbalance is probably for overcoming the specialized prey defense mechanisms (Dejean and Corbara 2014 ; Dejean et al. 2014 ; Kessler et al. 2016 ). The model does not include some other possible advantages of directional column raiding. First, the large number may allow the ants to overcome physical obstacles collectively by forming self-assembled bridges (Reid et al. 2015 ; Graham et al. 2017 ). Second, their large number and density might serve as a protection against predation or competitive aggression. Third, large and mobile prey e.g. living vertebrates could be quickly overwhelmed right after the discovery. However, these additional benefits from directional raiding by a large number of individuals may not always be high. First, the self-assembled structures are costly commitments of many potentially active foragers, and their traffic enhancement is actually not very great (Powell and Franks 2007 ; Brunelle 2011 ). During colony migration, such bridges or rafts would be worth constructing because the queen, pupae, larvae and eggs need to be transported. However, in the foraging context it is difficult to think of a situation where a heavy investment in traffic infrastructure is more important than increasing the coverage area. Second, as noted previously, the selective pressure from predation and competitive aggression is quite low for new world army ants (O'Donnell et al. 2007 ; Willson et al. 2011 ). Finally, although some species of army ants do hunt vertebrates (O’Donnell et al. 2005 ), most army ants primarily feed on social insect colonies (Berghoff 2003 ). Social insect colonies are immobile food sources and successfully exploitable with non-army-ant behaviors, e.g. by the termite-eating Matabele ants (Villet 1990 ) and the slave-making social parasites (Alloway 1979 ; Hasegawa and Yamaguchi 1994 ). Therefore, we excluded the aforementioned factors from the simulation and focused on the effect of the food distribution only. In summary, this study illustrates that food distribution alone is sufficient to create ecological situations in which the natural selection may favor column raiding over the radial searching. The future studies should consider variations in different movement parameters as well as in the diet and the colony size to further investigate adaptive value of the raiding behavior."
} | 3,095 |
30793292 | null | s2 | 1,510 | {
"abstract": "How unicellular organisms optimize the production of compounds is a fundamental biological question. While it is typically thought that production is optimized at the individual-cell level, secreted compounds could also allow for optimization at the group level, leading to a division of labor where a subset of cells produces and shares the compound with everyone. Using mathematical modeling, we show that the evolution of such division of labor depends on the cost function of compound production. Specifically, for any trait with saturating benefits, linear costs promote the evolution of uniform production levels across cells. Conversely, production costs that diminish with higher output levels favor the evolution of specialization-especially when compound shareability is high. When experimentally testing these predictions with pyoverdine, a secreted iron-scavenging compound produced by Pseudomonas aeruginosa, we found linear costs and, consistent with our model, detected uniform pyoverdine production levels across cells. We conclude that for shared compounds with saturating benefits, the evolution of division of labor is facilitated by a diminishing cost function. More generally, we note that shifts in the level of selection from individuals to groups do not solely require cooperation, but critically depend on mechanistic factors, including the distribution of compound synthesis costs."
} | 351 |
34442639 | PMC8399598 | pmc | 1,512 | {
"abstract": "To detect the change during coral–dinoflagellate endosymbiosis establishment, we compared transcriptome data derived from free-living and symbiotic Durusdinium , a coral symbiont genus. We detected differentially expressed genes (DEGs) using two statistical methods (edgeR using raw read data and the Student’s t -test using bootstrap resampling read data) and detected 1214 DEGs between the symbiotic and free-living states, which we subjected to gene ontology (GO) analysis. Based on the representative GO terms and 50 DEGs with low false discovery rates, changes in Durusdinium during endosymbiosis were predicted. The expression of genes related to heat-shock proteins and microtubule-related proteins tended to decrease, and those of photosynthesis genes tended to increase. In addition, a phylogenetic analysis of dapdiamide A (antibiotics) synthase, which was upregulated among the 50 DEGs, confirmed that two genera in the Symbiodiniaceae family, Durusdinium and Symbiodinium , retain dapdiamide A synthase. This antibiotic synthase-related gene may contribute to the high stress tolerance documented in Durusdinium species, and its increased expression during endosymbiosis suggests increased antibacterial activity within the symbiotic complex.",
"introduction": "1. Introduction The dinoflagellate of the Symbiodiniaceae family live symbiotically with a variety of marine invertebrates, including clams, sea slugs, sea anemones, foraminifera, and corals [ 1 , 2 , 3 ]. Among these, the symbiotic relationships between symbiodiniacean algae and cnidarians have been studied extensively. Symbiotic algae provide photosynthetic products to corals and receive nitrogen in exchange [ 4 , 5 , 6 ]. Published evidence indicates that the activity of symbiotic Symbiodiniaceae is under the control of the host corals [ 7 , 8 , 9 ]. In coral cells, algae are present in host-derived acidified vesicles that have carbon-concentrating mechanisms and activate photosynthetic capacity [ 7 ]. A recent transcriptome analysis found that dinoflagellate genes involved in molecular chaperoning as well as sugar and ammonia transportation were suppressed during the establishment of endosymbiosis with Aiptasia and coral planula larvae [ 10 , 11 ]. Gene expression analyses of actin, Ca 2+ ATPase, and H + APTase in Symbiodiniaceae also revealed that their expression patterns differed considerably between the non-symbiotic and symbiotic states [ 12 , 13 ]. However, despite these recent advances, many aspects of the changes dinoflagellate undergo during coral endosymbiosis establishment remain unclear. Previous studies have established a model endosymbiosis system consisting of monoclonal alga and juvenile corals, and transcriptome data for these coral–alga complexes have been published, with a focus on coral gene expression [ 11 , 14 , 15 , 16 ]. By contrast, the gene expression levels of dinoflagellate during coral endosymbiosis have not been investigated due to a lack of transcriptome data for the non-symbiotic state. Since the coral–alga model system facilitates the investigation of dinoflagellate gene expression and proliferation processes over time, it is well suited for examining changes in dinoflagellate during endosymbiosis. In this study, we obtained transcriptome data for cultured Durusdinium that were previously used in an infection experiment with juvenile corals [ 14 ]. Therefore, in order to comprehensively investigate the changes in Symbiodiniaceae associated with the transition to the symbiotic state, we attempted to detect differentially expressed genes between the free-living and endosymbiotic states. Following the inoculation of juvenile corals with Durusdinium , its rate of increase was greater than that of Cladocopium , with about 300 Durusdinium cells per polyp detected by the 10th day of endosymbiosis, and about 600 cells per polyp detected by the 20th [ 14 ]. Here, we identified and functionally annotated differentially expressed genes (DEGs) between the non-symbiotic and symbiotic states, and performed phylogenetic analyses for a part of the DEGs to confirm that the DEGs were derived from Durusdinium .",
"discussion": "3. Results and Discussion In this study, we attempted to clarify the gene expression changes taking place in dinoflagellate during coral–alga endosymbiosis establishment. We prepared and sequenced a cDNA library derived from the symbiont culture, which isolated 25,068 contigs containing ORFs. Low reads derived from algae engaged in endosymbiosis with corals and free-living cultured algae were mapped against these contigs, and DEGs between these states were identified. The edgeR analysis identified 8543 DEGs, representing 34% of the candidate alga-derived transcripts. To validate these results, we used bootstrap replicates of RNA-seq data to detect DEGs between the non-symbiotic ( n = 2) and symbiotic states ( n = 2). Differences in the mean (among 100 × 2 replicates) expression of each gene between the two states were detected using the Student’s t -test ( p < 0.025). The p value was set to correspond to the number (8642) of DEGs detected in the edgeR analysis (q = 0.01). A total of 4587 DEGs were common between both groups. Finally, we selected 1214 genes with log 2 (fold change) > 1 between the free-living and symbiotic states in the edgeR analysis ( Figure S1 ). The top 50 genes showing expression changes due to endosymbiosis with the lowest FDR included ribosomal proteins, heat-shock proteins, and chlorophyll-binding proteins ( Figure S1 ). We also searched for GO molecular function terms that were enriched in these 1214 genes ( Figure S2 ). The most enriched GO terms for these upregulated genes included protein–chromophore linkage and photosynthesis. The upregulated DEGs indicated that algae have enhanced photosynthetic activity during endosymbiosis with corals, which is consistent with previous reports of endosymbiosis in Symbiodiniaceae [ 7 ]. Seven processes were related to downregulated DEGs, including microtubule-based processes, mRNA splicing via spliceosome, and protein folding. Genes with decreased expression in microtubule-based processes include genes encoding tubulin, which is a component of flagella. This result may reflect the fact that algae lose their flagella inside corals [ 18 ]. Furthermore, a large number of ribosomal and chaperone proteins were detected among the downregulated DEGs, suggesting that some translational and protein-folding functions were inactivated following endosymbiosis establishment. Decreased expression of the chaperone gene has also been reported in the genera Symbiodinium and Cladocopium [ 11 ], and may represent a typical response to coral endosymbiosis establishment. It should be noted, however, that some of the DEGs detected include genes that were altered due to environmental differences between the two states. The time of year when the symbiotic and non-symbiotic states were cultured, as well as changes in the light environment, salinity, and CO 2 in the coral cells, may have affected the genes whose expressions were altered. In order to investigate more specific changes in a symbiotic organism, we must more closely replicate the exact conditions of the culture strain and the symbiotic state. Among the top 100 DEGs, two genes encoding dapdiamide A synthase (TRINITY_DN38519_c0_g1_i5.p1 ( Figure 1 ) and TRINITY_DN38519_c0_g1_i1.p1) were found to be upregulated. Dapdiamide A synthase adds valine to the carboxylate of fumaramoyl-DAP to form dapdiamide A, an antibiotic, in Pantoea agglomerans [ 19 ]. Few studies have reported on the antibiotic synthase in Symbiodiniaceae; however, a recent large-scale transcriptome analysis identified dapdiamide A synthase in Symbiodinium [ 20 ]. Therefore, we performed a phylogenetic analysis to investigate whether the gene encoding dapdiamide A synthase is derived from Symbiodiniacea or from bacteria ( Figure 2 ). One of the genes (TRINITY_DN38519_c0_g1_i5.p1) encoding dapdiamide A synthase was used for a BLASTp query against the NCBI database, and 117 sequences were selected for phylogenetic inference. The distribution of eukaryotic dapdiamide A synthase was restricted to large phylogenetic groups including stramenopiles, haptophytes, and alveolates (Symbiodiniaceae). In the ML tree ( Figure 2 ), most of the eukaryotic sequences formed two separate clades, A and B. Clade A comprises sequences derived from bacillariophytes (stramenopiles), haptophytes, and Symbiodiniaceae. Nine Symbiodiniaceae sequences were monophyletic, with 86% support, and its sister group was shared by Emiliania huxleyi (haptophyte) sequences. These branching patterns suggest that the eukaryote–eukaryote lateral gene transfer of dapdiamide A synthase occurred between Emiliania and Symbiodiniaceae. In addition, close relationships between two Symbiodinium sequences as well as archaean (100%) and Aureococcus (pelagophyte) (no support) sequences on the lower part of the tree suggest other types of lateral gene transfer involving Symbiodiniaceae. However, these genes were detected only in the genera Symbiodinium and Durusdinium of Symbiodiniaceae in this study. Durusdinium species exhibit high stress resistance, and have been reported to confer this property to their host corals [ 21 ]. Our results show that both dapdiamide A synthase genes from Durusdinium were upregulated during endosymbiosis establishment, which may enhance antibacterial action and confer stress tolerance to the host coral. In this study, Durusdinium genes that exhibited expression changes during coral endosymbiosis establishment were selected using two analysis methods for functional analysis. The weakness of this study is the small number of replicas and the different timings of the fixation of symbiotic and non-symbiotic dinoflagellate. In future gene expression research, it will be necessary to improve these areas. In addition, transcriptome data do not necessarily correlate with protein expression data, thus requiring proteome analysis to elucidate the entire internal symbiotic process. The roles of these genes in dinoflagellate adaptation to the host coral environment need to be further investigated; such data could be useful in clarifying the evolutionary process of symbiont trait acquisition."
} | 2,578 |
40278457 | PMC12029552 | pmc | 1,513 | {
"abstract": "This comprehensive review systematically explores the molecular design and functional applications of nano-smooth hydrophilic ionic polymer surfaces. Beginning with advanced fabrication strategies—including plasma treatment, surface grafting, and layer-by-layer assembly—we critically evaluate their efficacy in eliminating surface irregularities and tailoring wettability. Central to this discussion are the types of ionic groups, molecular configurations, and counterion hydration effects, which collectively govern macroscopic hydrophilicity through electrostatic interactions, hydrogen bonding, and molecular reorganization. By bridging molecular-level insights with application-driven design, we highlight breakthroughs in oil–water separation, anti-fogging, anti-icing, and anti-waxing technologies, where precise control over ionic group density, the hydration layer’s stability, and the degree of perfection enable exceptional performance. Case studies demonstrate how zwitterionic architectures, pH-responsive coatings, and biomimetic interfaces address real-world challenges in industrial and biomedical settings. In conclusion, we synthesize the molecular mechanisms governing hydrophilic ionic surfaces and identify key research directions to address future material challenges. This review bridges critical gaps in the current understanding of molecular-level determinants of wettability while providing actionable design principles for engineered hydrophilic surfaces.",
"conclusion": "4. Conclusions and Outlook 4.1. Conclusions In this review, we have underscored the critical role of molecular-level material design in modulating surface wettability. We have provided a comprehensive analysis of methodologies for engineering nanoscale smooth hydrophilic polymer surfaces, encompassing not only fabrication techniques but also the fundamental molecular determinants governing their behavior. Furthermore, we have systematically evaluated applications such as oil–water separation, anti-fogging, anti-icing, and anti-waxing based on varying hydrophilicity requirements by considering the collective effects of both polar and nonpolar moieties on surfaces. Through illustrative case studies, we have further demonstrated how molecular design strategies can be tailored to meet specialized functional demands. The molecular mechanisms underlying the hydrophilicity of charged polymer surfaces can be elegantly distilled as follows ( Figure 13 ): the hydrophilic nature of these surfaces is predominantly dictated by those ionic polar groups. Specifically, the type, density, and counterion species of ionic terminal groups collectively govern surface hydration efficiency. Moreover, the geometric configuration of adjacent non-polar groups modulates the persistence of surface hydration, ensuring stable hydrophilicity across diverse environmental conditions. Thus, achieving optimal surface hydration necessitates a holistic consideration of the interplay between polar and non-polar moieties, ensuring that the hydration of polar groups is sufficiently robust to mask underlying hydrophobic domains. 4.2. Outlook While significant progress has been made in elucidating the molecular mechanisms of hydrophilicity in ionic polymer surfaces, numerous intricate aspects remain ripe for further exploration. Several promising avenues for future investigation include:\n A. Surface Reconstruction Induced by Adjacent Non-Polar Groups: The extent and likelihood of surface reconstruction driven by adjacent non-polar groups can be inferred from discrepancies between the ideal ( θ l / e c a l ) and experimentally observed ( θ l / e e x p ) contact angles ( δ l / e ), as well as disparities in the polar ( γ p ) and non-polar ( γ d ) components of surface energy. Existing thermodynamic indicators (e.g., deviation in adhesion W Figure 14 a) [ 111 ] have been developed to assess molecular-scale reconstruction in non-ionic polar surfaces. However, for charged polar surfaces, particularly in the context of dynamic reconstruction processes, a self-consistent, quantitative thermodynamic framework related to wettability remains elusive. B. \n Molecular-Scale Roughness of Layer-by-Layer (LBL) Self-Assembled Multilayers: \n Ionic group-functionalized surfaces constructed via LBL assembly—such as those composed of strong polyelectrolytes ((PDDA/PSS) n )—offer a unique platform for investigating counterions and charged moieties at the molecular scale. However, in practical applications, molecular-scale roughness in LBL-assembled surfaces can escalate due to the intrinsic properties of membrane materials or the number of assembly layers [ 168 ]. In such cases, the surface composition resembles that of self-assembled monolayer (SAM) films with multiple polar components. For instance, as shown in Figure 14 b, Gooding et al. demonstrated that in SAMs terminated with binary polar end groups, the intrinsic anti-cell adhesion properties were lost when charged end groups (EG6) underwent inward surface reconstruction due to surface potential shifts [ 169 ]. Yet, the wettability and anti-fouling performance of LBL multilayers with molecular-scale roughness remains an underexplored frontier. C. \n Quantitative Characterization of Counterion Hydration Capacity \n : \n Although the viscosity coefficient B η offers preliminary qualitative insights into whether ions “structure” or “disrupt” water networks, the wettability differences among ions with identical B η values remain quantitatively underexplored. For instance, despite sodium (Na + ) and chloride (Cl − ) ions exhibiting comparable B η values and similar positions in the Hofmeister series, they demonstrate markedly different selectivity in ion channel transport within cells ( Figure 14 c), contributing to the distinct distribution of sodium and potassium ions in nerve and muscle cells and the generation of potential differences [ 170 ]. D. \n Enhancing Surface Hydration and Mitigating Hydrophobic Defect Exposure \n : \n Recent studies have introduced collective molecular parameters f p θ and f p M to classify surface hydrophilicity based on the packing density of polar moieties. When polar groups achieve the maximum density of two-dimensional random packing spheres (approximately 0.67), the surface hydration capacity meets basic hydrophilic requirements. At the theoretical maximum density of two-dimensional close-packed spheres (approximately 0.9), hydrophobic regions are effectively masked, enabling superhydrophilicity for highly demanding applications. However, these studies primarily focus on interactions between polar and non-polar groups. When surfaces incorporate multiple polar and non-polar functionalities or feature intricate configurations, their synergistic effects on wettability warrant further in-depth investigation. E. \n Quantitative Self-Cleaning Applications in Air and Underwater \n : \n Current quantitative assessments of hydrophilic self-cleaning applications primarily focus on the environment of air. However, in underwater anti-fouling applications, while superior wettability is crucial, the dissociation and charge variations of certain contaminants complicate the mechanisms governing hydrophilic ionic surfaces. For instance, medical instruments, ship hulls, and reverse-osmosis membranes frequently encounter proteins bearing both positive and negative charges [ 171 ]. This necessitates the consideration of not only surface hydrophilicity but also electrostatic interactions between charged surfaces ( Figure 14 d). Consequently, molecular designs must meet more stringent requirements. Beyond highly hydrophilic zwitterions ( Figure 14 e), researchers have discovered that surfaces functionalized with natural proteins (e.g., bovine serum albumin, BSA) exhibit exceptional anti-protein adhesion properties ( Figure 14 e) [ 172 , 173 ]. Thus, biomimetic surface designs inspired by natural proteins may represent a promising strategy for next-generation anti-fouling materials. F. \n Challenges in Hydrophilic Material Stability and Environmental Responsiveness \n : \n With the continuous evolution of industrial and biomedical technologies, the demand for advanced hydrophilic materials is surging. However, maintaining stable hydrophilicity under extreme conditions—such as high temperatures, high pressures, and exposure to strong acids or bases—remains a formidable challenge. Additionally, certain applications require materials to be biodegradable and environmentally responsive. As sustainable development gains momentum, the design of eco-friendly materials has become an imperative research focus. Developing hydrophilic polymers that exhibit both long-term stability and controlled degradation presents a novel challenge in ionic polymer material science. Figure 14 ( a ) Schematic showing the competition between macroscopic wettability and adhesion work of the wetting phase ( W f ) and environmental wettability work ( W e ). Ref. [ 111 ] Copyright from Elsevier @2023. ( b ) Schematic of an intelligent surface with charged functional groups on the distal end and molecules terminating with RGD peptides for cell adhesion, which can dynamically switch by applying +300 or −300 mV to allow or prevent cell proximity to the peptide. Ref. [ 169 ] Copyright from Wiley @2012. ( c ) Schematic illustrating the distinct selectivity of sodium (Na + ) and chloride (Cl − ) ions in ion channel transport within cells, despite their comparable B η values and similar positions in the Hofmeister series, highlighting their role in the differential distribution of sodium and potassium ions in nerve and muscle cells and the generation of potential differences (downloaded from: Bioelectricity—Human Physiology (uoguelph.ca)). ( d ) Schematic showing protein adsorption on surfaces with low and high charge densities. Ref. [ 171 ] Copyright from Elsevier @2011. ( e ) I. Schematic of anti-fouling/scaling via hydration layers. Ref. [ 173 ] Copyright from Wiley @2022. II. Antithrombotic results in vitro and ex vivo from a rat model with pCBM/pSB hydrogel coating and bare PVC tubing. Ref. [ 174 ] Copyright from Springer Nature @2022. III. Schematic of anti-organic contaminant FO/RO permeation membrane achieved by modifying superhydrophilic nanoparticles. Ref. [ 175 ] Copyright from ACS @2012. ( f ) Schematic of hydrophilic coating surface preparation by thiol-ene click chemistry to synthesize BSA@PSBMA. Ref. [ 173 ] Copyright from the Wiley @2022. In conclusion, the exploration of molecular mechanisms governing nanoscale smooth hydrophilic ionic polymer surfaces remains a dynamic and rapidly evolving domain, replete with both challenges and opportunities. With the continuous advancement of molecular-level characterization techniques—including X-ray and neutron diffraction, sum-frequency generation (SFG) spectroscopy, nuclear magnetic resonance (NMR) spectroscopy, and molecular dynamics (MD) simulations—we are confident that future research will progressively unravel the intricate mysteries of ionic surface hydrophilicity, ultimately providing invaluable insights for the future molecular design and real-world application of next-generation hydrophilic materials.\n\n4.1. Conclusions In this review, we have underscored the critical role of molecular-level material design in modulating surface wettability. We have provided a comprehensive analysis of methodologies for engineering nanoscale smooth hydrophilic polymer surfaces, encompassing not only fabrication techniques but also the fundamental molecular determinants governing their behavior. Furthermore, we have systematically evaluated applications such as oil–water separation, anti-fogging, anti-icing, and anti-waxing based on varying hydrophilicity requirements by considering the collective effects of both polar and nonpolar moieties on surfaces. Through illustrative case studies, we have further demonstrated how molecular design strategies can be tailored to meet specialized functional demands. The molecular mechanisms underlying the hydrophilicity of charged polymer surfaces can be elegantly distilled as follows ( Figure 13 ): the hydrophilic nature of these surfaces is predominantly dictated by those ionic polar groups. Specifically, the type, density, and counterion species of ionic terminal groups collectively govern surface hydration efficiency. Moreover, the geometric configuration of adjacent non-polar groups modulates the persistence of surface hydration, ensuring stable hydrophilicity across diverse environmental conditions. Thus, achieving optimal surface hydration necessitates a holistic consideration of the interplay between polar and non-polar moieties, ensuring that the hydration of polar groups is sufficiently robust to mask underlying hydrophobic domains.",
"introduction": "1. Introduction The molecular-scale integration of hydrophilic ionic motifs on solid surfaces has emerged as a cornerstone of advanced material design, where the synergistic action of polar functional groups and surface charge determines the interface wettability, hydration stability, and efficiency. These ionic architectures leverage synergistic hydrogen bonding and electrostatic interactions with water molecules to engineer surfaces with programmable hydrophilicity, enabling applications ranging from self-cleaning coatings to bioadhesive interfaces. Nature offers masterclasses in such molecular engineering: phospholipid bilayers employ densely packed ionic headgroups (e.g., phosphatidylcholine and phosphatidylserine) to construct ≈1 nm hydration layers that regulate ion channel permeability ( Figure 1 a) [ 1 , 2 , 3 , 4 , 5 , 6 , 7 , 8 , 9 ], while cartilage surfaces exploit ultra-dense carboxylate arrays to achieve superlubricity (friction coefficient ≈ 0.001) under megapascal pressures via hydration-mediated boundary lubrication ( Figure 1 b) [ 10 , 11 , 12 , 13 , 14 , 15 ]. These biological systems exemplify how precise control over interfacial hydration states—governed by ionic group density, charge distribution, and supramolecular assembly—can impart specialized functionalities through water-structuring phenomena. Inspired by these natural paradigms, synthetic strategies have harnessed ionic surface engineering to transcend conventional material limitations [ 16 , 17 , 18 , 19 ]. Layer-by-layer polyelectrolyte assemblies on liposomes (PAA/PAH, Figure 1 c) enhance drug bioavailability through charge-modulated hydration stability [ 16 ], while carboxylate/sulfonate-enriched hydrogels replicate cartilage-like tribological performance via ionic group density optimization ( Figure 1 d) [ 17 ]. Further innovations, such as pH-responsive hyaluronic acid coatings on nanocrystals (HA@Cur-NCs, Figure 1 e), demonstrate how dynamic hydration layer restructuring enables tumor-targeted drug release with extended circulatory half-lives [ 18 ]. Yet, despite these advances, critical knowledge gaps persist in correlating molecular-scale ionic features—polarity, spatial configuration, and counterion mobility—with macroscopic wettability. Current models often oversimplify hydrophilicity as a static function of chemical group polarity or density, neglecting dynamic factors such as molecular reorientation under environmental influences and hydration efficiency related to the counter-ion coordination with the ionic polymer. For instance, polyelectrolyte multilayers (PDDA/PSS) with identical roughness (<3 nm) exhibit environmental polarity-dependent wettability divergences: Based on the contact angle measurements, the wettability of the PDDA surface shows minimal variation between nonpolar and polar phases, whereas the PSS surface exhibits significant wettability differences ( Figure 1 f) [ 19 ]. It can be observed that for nanoscopically smooth surfaces, wettability differences arise not only from material-specific hydrophilicity but also from other factors affecting the same hydrophilic groups. This anomaly highlights the limitations of conventional models in explaining surface hydrophilicity at the molecular scale. To address these challenges, this review begins by examining conventional theoretical models, emphasizing the importance of deconstructing surface wettability at the molecular level and the suitability of ionic polymer surfaces as model systems. We then introduce current fabrication techniques for ionic polymer surfaces and establish a framework to compare their mechanical advantages and limitations. Furthermore, from a molecular perspective, we elucidate how ionic polymer surfaces modulate macroscopic hydrophilicity through microscopic control, focusing on the types of ionic groups, molecular configurations, and the hydration effects of counterions. Moreover, building upon this foundation, the discussion pivots to application-driven design principles, where industrial and biomedical requirements—amphiphilic membrane systems for oil–water separation, stimuli-responsive anti-icing/anti-fogging interfaces, and bioinspired antifouling coatings—are translated into tailored molecular blueprints. Case studies illustrate how zwitterionic architectures, pH-triggered hydration layers, and biomimetic charge gradients enable breakthrough functionalities. Finally, we consolidate the molecular mechanisms underlying hydrophilic ionic polymer surfaces, and by bridging molecular-scale insights with macroscopic properties, we propose promising research directions to address future challenges in materials science. Figure 1 The critical role of ionic functional groups in controlling surface wettability across diverse biomaterials . ( a ) Phospholipid bilayers leverage ionic groups to form hydration layers, influencing surface interactions and wettability. Ref. [ 20 ] Copyright from Elsevier @2022. ( b ) Cartilage surfaces utilize charged macromolecules (e.g., hyaluronic acid, lubricin, and phospholipids [ 15 ]) to achieve superior hydration and lubrication under high pressure, showcasing the dynamic control of wettability. Ref. [ 21 ]. Copyright from Wiley@2021. ( c ) Polyelectrolyte-modified liposomes employ ionic groups to enhance drug delivery efficiency and biocompatibility, illustrating how surface chemistry regulates wettability for sustained release. I. Basic structure of liposomes. II.Plasma concentration-time curves of PTX suspension, PTX liposomes, PAA-PTX liposomes, and PAH-PAA-PTX layered formulations. Ref. [ 16 ]. Copyright from Elsevier @2012. ( d ) Hydrogels with abundant ionic groups (e.g., carboxylate/sulfonate) exhibit super-lubrication properties, mimicking natural cartilage and highlighting the role of ionic groups in modulating wettability and tribological performance. I. Schematic of anionic hydrogel “CS-Fe3+” rich in carboxylate/sulfonate. II. Schematic of tribological testing of layered “CS-Fe” hydrogel and the relationship between sliding velocity and friction coefficient (COF) under low, medium, and high load conditions. III. Experimental design simulating osteoarthritis (OA) and cell viability of L929 cells and chondrocytes co-cultured with RAW264.7 cells regulated by “CS-Fe” hydrogel. Ref. [ 17 ]. Copyright from ACS @2022. ( e ) Curcumin nanocomposites (Cur-NCs) demonstrate pH-dependent release and plasma concentration profiles, emphasizing how ionic groups on surfaces can be tailored to control wettability and drug delivery kinetics. I. Release behavior of Cur and Cur-NCs at different pH values. II. Appearance and average plasma concentration-time curves of Cur, Cur-NCs, and HA@Cur-NCs. Their respective blood elimination half-lives are also noted. Ref. [ 18 ]. Copyright from RSC @2019. ( f ) Polyelectrolyte multilayers (e.g., PDDA/PSS) exhibit distinct wettability behaviors in air and oil phases, revealing how the spatial arrangement of ionic groups influences surface interactions and wettability. I. Schematic of molecular properties and wetting behavior of PDDA and PSS in air and oil phases. II. Contact angle values of freshly prepared (square), 72-h aged at room temperature (triangle), and 72-h aged at 60 °C (circle) PEMs with varying bilayer numbers in air. III. and IV. Contact angle and water-in-oil contact angle curves for (PDDA/PSS) 3.5 (square) and (PDDA/PSS) 4 (circle) in oil over time. Ref. [ 19 ]. Copyright from Wiley @2015."
} | 5,075 |
26322321 | PMC4543078 | pmc | 1,514 | {
"abstract": "This data article includes size exclusion chromatography data of soluble eADF4(C16), an engineered spider silk variant based on the core domain sequence of the natural dragline silk protein ADF4 of Araneus diadematus, in combination with light scattering; the protein is monomeric before assembly. The assembled mature fibrils were visualized by transmission electron microscopy (TEM) and atomic force microscopy (AFM). Sonicated fibrils were used as seeds to by-pass the nucleation lag phase in eADF4(C16) assembly. We also provide data on the sedimentation kinetics of spider silk in the presence of different NaCl concentrations revealing very slow protein aggregation in comparison to the fast assembly triggered by phosphate ions published previously [1] . Experiments in the Data article represent supporting material for our work published recently [1] , which described the assembly mechanism of recombinant eADF4(C16) fibrils."
} | 234 |
35542133 | PMC9082298 | pmc | 1,516 | {
"abstract": "Herein, we report the fabrication of a superhydrophobic surface with a new and effective silica nanocomposite. A facile synthesis was developed by spraying the as-prepared silica suspension on a glass substrate, where the SiO 2 nanoparticles were composed of methylated aerogel particles wrapped by hydroxyl-terminated polydimethylsiloxane (PDMS). Three types of methylated silica aerogel nanoparticles with different surface roughness and porosities were prepared using specific precursors and methylation agents. The coating of the silica aerogels (sodium silicate and trimethylchlorosilane) wrapped in PDMS was exceptionally superhydrophobic with a superior water contact angle of 169.80 ± 3° and a sliding angle of less than 4°. The semi-transparent coating maintained its excellent water repellency at 350 °C for least 4 h and exhibited durable superhydrophobic properties for 6 months at ambient conditions. Additionally, the coating also showed good mechanical stability and remarkable self-cleaning behaviour.",
"conclusion": "4. Conclusions A facile and scalable method to prepare superhydrophobic coatings was reported in detail. The semi-transparent coatings exhibited excellent durability, self-cleaning, and superhydrophobicity as well as excellent thermal and mechanical stabilities. The advancing water contact angle and sliding angle were 169.80 ± 3° and 4°, respectively, and these were remarkable results. The coating maintained water repellency under 350 °C for at least 4 h and possessed durable superhydrophobicity for 6 months at ambient conditions. Additionally, it remained superhydrophobic after at least four rounds of abrasion with sandpaper under more than 2000 Pa. This novel coating with self-cleaning, antifouling, low-drag, and anti-smudge properties can be used in applications where superhydrophobicity is important.",
"introduction": "1. Introduction Superhydrophobic surfaces, with a water contact angle greater than 150° and a sliding angle (the difference between the advancing and receding contact angles) of less than 10°, 1 have received increasing attention from both academic and practical fields such as self-cleaning, 2,3 anti-corrosion, 4,5 oil-water separation, 6,7 and anti-icing. 8,9 The prevalent fabrication approaches for superhydrophobic surfaces include etching, 10 physical/chemical vapour deposition, 11,12 electro-spinning, 13 nano/microparticle assembly, 14,15 the template method, 16 magnetron sputtering 17 and hot-pressing. 18 However, problems remain for most of these methods such as high-energy consumption and the need for expensive and sophisticated instruments. Additionally, expensive poisonous agents (fluorinated components) were used to achieve superhydrophobic surfaces in many studies. 19–21 Compared to fluorosilanes, PDMS not only has low surface energy but is also less toxic, cheaper, thermally stable and durable than many polymers, suggesting that PDMS may serve as both a structural precursor and a raw modifying material. 22,23 Due to its low intrinsic surface energy and moldability, many kinds of nanomaterials and polymers such as ZnO, 24 carbon nanowalls 25 and epoxy resin 26 were doped with PDMS. Due to suitable preparation process, controllable particle size and good reactivity with organic groups, SiO 2 nanoparticles are the most widely applied materials in superhydrophobic surfaces. 27,28 Because the aggregation of nanoparticles results in multi-scale roughness and PDMS provides good hydrophobicity with non-polar –CH 3 groups, the composites containing SiO 2 nanoparticles and PDMS have attracted much attention in recent decades. Few papers investigating the thermal and mechanical stabilities of durable superhydrophobic coatings on glass have been reported, which constraints wide applications of superhydrophobic coating. Herein, silica aerogel/PDMS composites were synthesized using a typical sol–gel method and subsequent PDMS chemical modification. Interface adhesive (Qsil 216) was used to enhance the adhesion between the superhydrophobic coating and the glass substrate. The preparation was simple and scalable by spraying Qsil 216 and silica aerogel/PDMS composites in order. The semi-transparent superhydrophobic coating was found to be self-cleaning and durable with excellent thermal and mechanical stabilities.",
"discussion": "2. Results and discussion 2.1 Surface morphology and chemical composition \n Fig. 1 shows a schematic illustration of the superhydrophobic coating. MSA possessed great quantities of non-polar groups (–CH 3 ), but it was difficult to form pure MSA coating on the glass base due to the weak cross-linking interaction. It has been recognized that PDMS with high viscosity and low surface energy has good connectivity for its end hydroxyl groups and long polymer chains, which are flexible and rotatable. Thus, P-MSA exhibited stable and controllable morphology and overcame the fragile features of MSA. Fig. 1 Schematic of the chemical composition of MSA and P-MSA and the micromorphology of the superhydrophobic coating. \n Fig. 2 shows the presence of rough micromorphology. Comparing Fig. 2(a), (d) and (g) with Fig. 2(b), (e) and (h) , respectively, it was found that the aerogel particles modified by PDMS stacked more intensely and irregularly, implying that the coatings possessed better film-formation properties than before. As seen in Fig. 3(a), (d), and (g) , whether or not TMCS or HTMS was added, the particles aggregated more seriously and even became indistinguishable due to the presence of a silane coupling agent. As shown in Fig. 2(b), (e) and (h) , the P-MSA-II sample exhibited the most serious aggregation and the smallest porosity. It was speculated that the residual hydroxyl groups on the surface of MSA-II were more than those on MSA-I and MSA-III because of the vigorous hydrolysis of TMCS. Fig. 2 SEM and AFM images of (a) MSA-I, (b) and (c) P-MSA-I, (d) MSA-II, (e) and (f) P-MSA-II, (g) MSA-III, and (h) and (i) P-MSA-III coatings. Fig. 3 FTIR spectra of (a) MSA-I, (b) P-MSA-I, (c) MSA-II, (d) P-MSA-II, (e) MSA-III, and (f) P-MSA-III coatings. As we know, the hierarchical micro/nanoscale structure on the surface is essential for superhydrophobic coatings besides chemical composition with low surface energy. 29 This was achieved by pores, clusters of silica aerogel bulks and silica aerogel particles, which are marked in Fig. 2(b), (d) and (e) , respectively. The globular entities and pores were less than 300 nm in diameter, whereas the sizes of the clusters of silica aerogel bulk were approximately more than 1 μm in the micron level. Notably, globular entities were coarse and composed of small mastoid nanoparticles. The three-dimensional micromorphologies of P-MSA-I, P-MSA-II and P-MSA-III coatings obtained using AFM are shown in Fig. 2(c), (f) and (i) , respectively, and these results confirmed the existence of a hierarchical micro/nanoscale structure, corresponding to previous analysis results based on SEM images. FTIR spectra showed the functional groups of different aerogels before and after PDMS modification ( Fig. 3 ). The peak at 2960 cm −1 was related to the –CH 3 groups, and the broad band centered at 3400 cm −1 was ascribed to O–H stretching in the Si–OH groups. The bands located at 1103, 953, and 801 cm −1 were associated with the Si–O–Si asymmetric bond stretching vibration, the Si–OH stretching vibration, and the network Si–O–Si symmetric bond stretching vibration, respectively. 16,30 It was clearly observed that many hydroxyl groups remained in three nanocomposites due to the modification of PDMS. Additionally, the intensity of the peaks corresponding to the methyl group increased. PDMS, which was loaded with many methyl groups, provided good hydrophobic properties. Most importantly, PDMS provided many abundant hydroxyl groups due to which the coatings possessed good adhesion with substrates such as glass and paper. 2.2 Static and dynamic wetting properties of three coatings As shown in the lower-left corner of Fig. 2(b), (e), and (h) , it is clear that the water droplets adopted a perfect ball shape, which implied excellent superhydrophobicities of the corresponding coatings. Furthermore, the hydrophobicities of the three P-MSA coatings are quantitatively shown in Table 1 including advancing angle, receding angle and contact angle. The water contact angles of the three samples were higher than 150°. The results were consistent with the droplet shapes in Fig. 2(b), (e), and (h) . It was speculated that the superhydrophobic property resulted from the synergistic effect of MSAs and PDMS. On the one hand, nanoparticles tended to agglomerate due to high surface area and surface energy. The inter-particle forces within the agglomerates originated from van der Waals, capillary, and electrostatic forces. 31 MSAs, made up of aggregated nanoparticles, provided a multi-scale hierarchical structure with granular bulges and pores, which could trap sufficient air pockets on the rough interstices between liquid and solid. 16 Therefore, the water drop could remain on the layer of air with minimum solid fraction in contact. On the other hand, the guest phase (MSAs) was uniformly and completely coated by the host matrix (PDMS) ( Fig. 2 ), which gave rise to the enhancement of hydrophobic methyl groups ( Fig. 3 ). The contact angle hysteresis ( θ CAH = θ Adv − θ Rec ) was less than 10° among the three samples and thus, the droplets could easily slide away from the coating. This resulted in the non-continuous solid/liquid/gas three-phase contact line. The spherical droplets on the surface of the superhydrophobic coatings showed Cassie–Baxter's state with almost no sticking towards the surface. 32 Furthermore, these superhydrophobic coatings were intact under normal circumstances for a period of more than 6 months. The naming, precursor and methylation agent of methylated silica aerogels (MSAs) Silica aerogels Precursor Methylation agent MSA-I Sodium silicate TMCS MSA-II TEOS TMCS MAS-III TEOS HMDS 2.3 Self-cleaning properties Alternative surfaces that clean themselves by water flow are known as self-cleaning surfaces. Especially on superhydrophobic surfaces, the accumulated dust particles can be easily washed off by rolling water drops. 33,34 The self-cleaning property of the superhydrophobic P-MSA-I coating was verified, as shown in Fig. 4 . The P-MSA-I coating was contaminated at first by dust particles, as shown in Fig. 4(a) . A hanging droplet of nearly 15 μl was produced from a syringe and slowly dragged on the dust-accumulated area. The droplets effectively collected the dust particles while rolling on the surface rapidly due to the extremely low sliding angle. Fig. 4 Optical images of (a) the heavily dusty surface and (b) the self-cleaned surface of the superhydrophobic P-MSA-I coating. 2.4 Abrasion resistance stability Objective comparison of the abrasion resistance of superhydrophobic surfaces has been hampered by the lack of a single, standardized test method. 35 There are many methods to analyse the mechanical properties of superhydrophobic coatings including the pencil hardness test, the sandpaper test, the scotch tape test, and the water drop/jet impact test. 36–38 The methodology illustrated in Fig. 5 was applied. Sand paper (800 mesh) served as an abrasive surface, and the P-MSA-I coating was tested facing this abrasive material. While a pressure (∼2178 Pa) was applied vertically to the glass, the glass slide was moved in one horizontal direction. The advancing contact angle, receding contact angle and hysteresis changes of the coating were then measured after abrasion (10 cm in abrasion length), as shown in Fig. 6 . Fig. 5 Schematic of the abrasion test employed to evaluate the robustness on the superhydrophobic coating. Fig. 6 The effect of the sand paper abrasion cycle on the advancing contact angle, receding contact angle and contact angle hysteresis in (a) the absence of Qsil 216 and (b) the presence of Qsil 216. The pure P-MSA-I coating showed poor mechanical stability, and the superhydrophobic surface was damaged completely after two abrasion cycles. However, the P-MSA-I/Qisl 216 surface remained superhydrophobic after going through at least four abrasion cycles. Even so, the contact angle hysteresis was less than 10°, which indicated that a small loading of Qsil 216 on the glass has better durability against mechanical force. PDMS, a soft polymer, could be easily scratched away from the surface. Thus, the optimum material should be obtained between the inorganic phase basement and the polymer to achieve improved mechanical stability of the coating, which in our case was the coating prepared by the addition of Qsil 216 on the surface of the basement. It is well-known that glass is a dense silicate material with extremely poor conductivity; the chemical composition and surface texture are different from those of metal and fabric. As a result, many polymers cannot be integrated with glass substrates by electrostatic interaction or molecular diffusion. Due to the presence of many activated functional groups, such as –Si–OH groups, in the curing process, firm adhesion can occur because of strong intermolecular forces and chemical bonds between the adhesive and glass. Additionally, P-MSAs can adhere well on the surface of Qsil 216, indicating that the adhesive has lower surface tension than PDMS, which is more propitious to wetting the surface of glass. 2.5 Optical transparency As shown in Fig. 7 , pure glass transmission was about 85% in the visible range from 350 to 770 nm. When P-MSA-I was coated on the glass, the transmission decreased to 72%. Even when the P-MSA-I/Qsil 216 composite was coated, the transmission decreased to 60%, which was 70.95% of the pure glass transmission. The semi-transparent superhydrophobic glass met most of the requirements for practical applications. Fig. 7 Optical transmission spectra of pure glass, the glass coated by P-MSA-I and the glass coated by P-MSA-I/Qsil 216 composite. 2.6 Thermal stability The thermal stability was tested at the gradient of 50 °C from 150 °C to 400 °C, and the tests were carried out at a rate of 10 °C min −1 under air atmosphere for 4 h under ultimate temperature. From Fig. 8 , it can be observed that the coating maintained superhydrophobic features up to 350 °C. At 400 °C, the local area of the P-MSA-I coating became hydrophilic. It was speculated that the continuous high-temperature conditions resulted in the loss of weight and therefore caused irreversible damage to the microstructure. The effects of temperature on the advancing contact angle, the receding contact angle and the contact angle hysteresis are shown in Fig. 8 . The coating possessed excellent superhydrophobic features ( θ Adv , θ Rec ≥ 155°) and remarkable self-cleaning properties ( θ CAH ≤ 10) ( Fig. 9 ). Fig. 8 Optical images of water droplets on the glass coated by the P-MSA-I/Qsil 216 composite after high temperature calcination for 4 h. Fig. 9 The effects of temperature on the advancing contact angle, the receding contact angle and the contact angle hysteresis."
} | 3,786 |
31310633 | PMC6640817 | pmc | 1,517 | {
"abstract": "Because of disturbance and plant species loss at the local level, many arid ecosystems in the western USA benefit from revegetation. There is a growing interest in improving revegetation success by purposefully inoculating revegetation plants with mutualistic endophytic fungi that increase plant stress tolerance. However, inoculant fungi must compete against fungi that indigenous to the habitat, many of which may not be mutualistic. Our overall goal, therefore, is to learn how to efficiently colonize revegetation plants using endophytic fungal inoculum. The goal will be facilitated by understanding the factors that limit colonization of plants by endophytic fungi, including inoculum dispersal and host compatibility. We analyzed endophytic fungal communities in leaves of Bromus tectorum and Elymus elymoides (Poaceae), Chrysothamnus depressus and Artemisia tridentata (Asteraceae), Alyssum alyssoides (Brassicaceae) and Atriplex canescens (Amaranthaceae), each occurring in each of 18 field plots. We found that dispersal limitation was significant for endophytic fungal communities of Atriplex canescens and Bromus tectorum , accounting for 9 and 17%, respectively, of the variation in endophytic fungal community structure, even though the maximum distance between plots was only 350 m. Plant species identity accounted for 33% of the variation in endophytic fungal community structure. These results indicate that the communities of endophytic fungi assembling in these plant species depend significantly on proximity to inoculum source as well as the identity of the plant species. Therefore, if endophytic fungi are to be used to facilitate revegetation by these plant species, land managers may find it profitable to consider both the proximity of inoculum to revegetation plants and the suitability of the inoculum to targeted host plant species.",
"introduction": "Introduction Revegetation may become a management necessity when severe habitat disturbance leads to a significant loss in plant cover. This is certainly the case in the arid, western region of the USA, where revegetation is becoming increasingly necessary as a consequence of native species losses due to land use change [ 1 ] and increased fire frequency following invasion by non-native grasses [ 2 – 5 ]. However, revegetation may be difficult in the physically stressful habitats characteristic of the arid, western USA [ 6 ], which suffer from water stress and extreme temperatures [ 7 ]. Endophytic fungi have been found in all plant species investigated thus far [ 8 – 10 ]. They frequently form complex communities within plant tissues comprising dozens of species [ 8 , 11 , 12 ]. In some cases, the vigor of the plant is significantly improved as a consequence of colonization by these fungi, especially under stressful conditions. For example, endophytic fungi may increase plant resistance to herbivory [ 13 , 14 ] and disease [ 9 , 15 , 16 ], and tolerance to heat [ 16 – 19 ], cold [ 20 ] and drought [ 16 , 19 – 21 ]. It is not surprising, therefore, that there is significant interest in utilizing mutualistic, endophytic fungi to facilitate revegetation, particularly in stressful habitats [ 22 – 25 ]. While it is appealing to inoculate revegetation plants with beneficial endophytic fungi, such an approach may not be effective if inoculant taxa do not compete well against indigenous endophytic fungi, especially because many indigenous taxa will not be particularly beneficial to their plant hosts. Depending on the combination of fungal taxon, plant taxon and environment, endophytic fungi range from mutualistic [ 8 , 18 , 26 ] to latent pathogenic [ 27 , 28 ] and latent saprotrophic [ 29 , 30 ]. Because desirable inoculant endophytic fungi will have to compete with indigenous endophytic fungi, effective and low-cost inoculation strategies will require understanding and overcoming the major constraints to colonization of plants by inoculant strains. Two of the potentially important factors that influence the colonization of plant tissues by endophytic fungi are dispersal limitation from inoculum source to target plants, and compatibility of inoculant fungi with the target plant species. Significant dispersal limitation [ 31 ] and biogeographical pattern in the distribution of endophytic fungi [ 12 ] suggest that proximity to a source of inoculum influences the probability of a fungal taxon colonizing plant tissue. In addition, colonization of a particular host plant by a particular endophytic fungus is limited by the degree of compatibility between fungus and plant [ 32 ], so a given source of inoculum may produce significantly different communities of endophytic fungi depending on the plant species [ 11 , 33 ] or plant genotype [ 34 – 37 ]. In this study our goal was to characterize the impacts of both plant species identity and fungus dispersal-limitation on endophytic fungal community structure in leaves of six plant species common in the eastern Great Basin of the USA. In some previous studies, the distinction between plant species identity and dispersal-limitation could not be made because plant species identity was confounded by spatial location in the environment [ 12 , 38 ]. In order to distinguish between plant species identity and spatial location, we sampled six plant species in each of 18 small field plots to greatly reduce dispersal limitation among plant species within plots in order to quantify dispersal limitation among plots.",
"discussion": "Discussion Colonization of plants by mutualistic, endophytic fungi may have large, positive effects on plant tolerance to environmental stresses [ 17 , 20 ], which are commonly experienced in many parts of the arid, western region of the USA [ 2 , 3 ]. Therefore, revegetation success in that region may be increased when plants are colonized by these mutualistic fungi [ 56 ]. However, deliberate inoculation of plants with mutualistic fungi may not be as effective as hoped for in a field setting where natural sources of inoculum also contribute to the endophytic fungal communities, given the fact that not all such endophytic fungi are beneficial to plants [ 27 , 28 , 30 , 57 ]. To maximize the effectiveness of inoculant fungi, it is important to understand the constraints to their use, including the constraints imposed by inoculum dispersal limitation and host compatibility. In some previous studies, it was impossible to separate the effects of plant species identity from dispersal limitation because plant species identity was confounded by geographic location [ 12 , 38 ]. Because we sampled all six plant species from each of the 18 small plots (approximately 16 m 2 ), each of the six plant species within a plot was presumably exposed to the same inoculum sources. In our study, therefore, plant species identity and spatial location were not confounded, and our results may constitute some of the best evidence for the roles of plant species identity and dispersal limitation in the determination of endophytic fungal community structure. We found that there was a significant dispersal limitation in endophytic fungal communities of Atriplex canescens , accounting for 9% of community variation, and of Bromus tectorum , accounting for 17% of community variation. In the other four plant species, however, there was no significant dispersal limitation among the plots. The fact that for four plant species there was no significant dispersal limitation is, perhaps, not completely unexpected given the fact that the maximum distance between plots in this study was only 350 meters. We previously studied dispersal limitation of endophytic fungal communities of Quercus gambelii , another inhabitant of the eastern Great Basin [ 31 ]. In that study the maximum distance between sites was 15 km and dispersal limitation accounted for only between 3 and 8% of the variability in community structure. The surprising result is that in the current study the endophytic fungal communities of Atriplex canescens and Bromus tectorum leaves did exhibit a significant dispersal limitation across a maximum distance of only 350 meters. This suggests that, at least for these two species, the proximity of revegetated plants to a source of endophytic fungal inoculum may significantly influence the structure of endophytic fungal communities, even on a relatively small spatial scale. Proximity of revegetation plants to the inoculum source may thus be an important consideration when using this technology for revegetation purposes. We also found that plant species identity accounted for 33% of the variation in endophytic fungal community structure. In other words, the six plant species possessed endophytic fungal communities that were quite different from each other. In some cases, the differences were caused by fungal OTUs that colonized some plant species but not others. Others have also found that the identity of the plant, either at the level of species [ 11 , 33 ] or genotype [ 34 – 37 ] influences the structure of endophytic fungal communities. Obviously if beneficial inoculant fungi had limited host breadth, inocula may have to be separately developed for different plant species. Nevertheless, there were 18 fungal taxa that were capable of colonizing leaves of all six plant species, although sometimes at markedly different frequencies. This indicates that among several of the endophytic fungal taxa in the eastern Great Basin, there is broad host plant compatibility. Therefore, it may be possible to develop a single beneficial inoculant fungus with broad host compatibility, which would simplify using inoculants to improve revegetation. We conclude that if we implement endophytic fungal inoculation schemes to improve revegetation success, we must take into consideration both inoculum dispersal limitation and plant-fungus compatibility in order to achieve high levels of effectiveness and cost-efficiency. While plant species identity was more important than dispersal limitation in this study, the relative impacts of these two factors may depend on spatial scale; over larger spatial distances dispersal limitation is expected to increase in importance. Their relative impacts may also depend somewhat on year to year variation in average windspeed, and on the timing of plant establishment and leaf growth, which may relate to factors such as rainfall and temperature. One limitation of this study is that it was carried out during a single growing season and one might expect variation in the relative importance of dispersal limitation and species identity to vary by year. Our results suggest a few additional hypotheses that warrant future testing. First, among the six plant species of our study, phylogenetic distance was a significant determinant of the structure of endophytic fungal communities, accounting for some 29% of total variability. As there is less phylogenetic distance among plant species within a family than among plant families, it was not surprising to find significant variation in the structure of endophytic fungal communities among plant families, and no significant variation among plant species within a plant family (either the Poaceae or the Asteraceae). A significant effect of plant phylogeny on the structure of endophytic fungal communities was also consistent with the fact that, for example, some fungal taxa were most frequently occurring in the Poaceae, or were most frequently occurring in the Poaceae and Asteraceae and less frequently occurring in the nonmycorrhizal species ( Atriplex canescens and Alyssum alyssoides ). The impact of plant phylogeny has not been significant in every study. For example, Vincent et al. [ 33 ] suggested that tree species relatedness was not a significant factor determining the structure of endophytic fungal communities. Possibly plant phylogeny is important only at higher fungal taxonomic levels [ 58 ]. In any case, our study included only six plant species and thus offered only a limited ability to test the role of plant phylogeny. We feel, therefore, that the plant phylogeny hypothesis warrants proper testing in the future. The implication of this, however, may be a plant may serve as an inoculum source of a range of fungal taxa that are compatible with other members of the same family. Second, our results also suggest the hypothesis that mycorrhizal status of plant species significantly influences the structure of endophytic fungal communities. The Amaranthaceae and Brassicaceae are both generally nonmycorrhizal or weakly mycorrhizal [ 59 – 62 ]. While their mycorrhizal status differs from that of the Asteraceae and Poaceae, which are both generally mycorrhizal, the Brassicaceae and Amaranthaceae are more closely related to the Asteraceae than Poaceae is to the Asteraceae. Yet the endophytic fungal communities of Atriplex canescens and Alyssum alyssoides were more dissimilar to those of the Asteraceae than those of the Poaceae were to those of the Asteraceae. This distinction between the nonmycorrhizal and the mycorrhizal plant species suggests that the ability of plants and mycorrhizal fungi to engage in the necessary molecular dialog to effect root colonization may be important in structuring foliar fungal communities. Our sample size was too small to test this hypothesis, but we feel that our result warrant further exploration of the mycorrhizal status hypothesis. Third, some fungal taxa were most frequently occurring in Atriplex canescens , or most frequently occurring in Alyssum alyssoides . Among the fungal taxa that most frequently occurred in Atriplex canescens and Alyssum alyssoides were unknown species in the Pleosporales, and this order of fungi appeared to be unexpectedly diverse in Atriplex canescens and Alyssum alyssoides leaves. This suggests that some fungal lineages may have adapted to colonize nonmycorrhizal plant taxa and since adaptively radiated. That hypothesis may also warrant further testing. Fourth, the endophytic fungal community of Alyssum alyssoides (Brassicaceae) was significantly different from that of Atriplex canescens (Amaranthaceae). While both families are largely nonmycorrhizal, they apparently utilize different mechanisms to limit mycorrhizal colonization [ 62 ], with the possible involvement of systemic mustard oils in the Brassicaceae but not in the Amaranthaceae [ 63 ]. Further experimentation into the mechanisms by which plants regulate mycorrhizal and endophytic fungi also appears to be warranted. Finally, Bromus tectorum is a problematic invasive plant species in much of the arid western portion of the United States [ 64 ]. Its success as an invasive species may be partly determined by enhanced vigor associated with colonization by particular endophytic fungi [ 4 , 65 ]. Therefore, invasion of habitat by Bromus tectorum may be facilitated by the presence of sources of inoculum for mutualistic endophytic fungi [ 4 ]. As we learn more about the nature of interactions between specific host plants and specific endophytic fungi, it may be possible to manipulate endophytic fungi inoculum sources to disfavor Bromus tectorum while benefiting native plant species."
} | 3,800 |
38802437 | PMC11130337 | pmc | 1,519 | {
"abstract": "Coevolution describes evolutionary change in which two or more interacting species reciprocally drive each other’s evolution, potentially resulting in trait diversification and ecological speciation. Much progress has been made in analysis of its dynamics and consequences, but relatively little is understood about how coevolution works in multispecies interactions, i.e., those with diverse suites of species on one or both sides of an interaction. Interactions among plant hosts and their mutualistic ectomycorrhizal fungi (ECM) may provide an ecologically unique arena to examine the nature of selection in multispecies interactions. Using native genotypes of Monterey pine ( Pinus radiata ), we performed a common garden experiment at a field site that contains native stands to investigate selection from ECM fungi on pine traits. We planted seedlings from all five native populations, as well as inter-population crosses to represent intermediate phenotypes/genotypes, and measured seedling traits and ECM fungal traits to evaluate the potential for evolution in the symbiosis. We then combined field estimates of selection gradients with estimates of heritability and genetic variance–covariance matrices for multiple traits of the mutualism to determine which fungal traits drive plant fitness variation. We found evidence that certain fungal operational taxonomic units, families and species-level morphological traits by which ECM fungi acquire and transport nutrients exert selection on plant traits related to growth and allocation patterns. This work represents the first field-based, community-level study measuring multispecific coevolutionary selection in nutritional symbioses.",
"conclusion": "Conclusion In this study, we provide evidence for natural selection in the mycorrhizal symbiosis between ECM fungi and Monterey Pine during one of the most extreme drought events on record in California. These results contribute to the growing body of evidence quantifying selection in multispecies interactions, especially bolstering our understanding of how coevolutionary selection operates in multispecific mutualisms. In particular, we demonstrate selection on plants for altered compatibility with specific fungal OTUs and families, with the direction and nature of this selection reflective of the apparent specificity of the fungi involved. We further demonstrate selection for particular fungal traits associated with the ability of the fungi to explore and acquire nutrients from the soil and the potential for genetic correlations between plant traits and specific fungal OTUs and exploration types. In total, this research represents an important first step in understanding multispecies coevolution; however, in order to fully understand this phenomenon in mycorrhizal interactions, common gardens that measure selection need to be replicated in other populations, which would allow estimation of geographic mosaics of coevolutionary selection 4 , 13 in these multispecific mutualisms.",
"introduction": "Introduction Coevolution describes evolutionary changes in which two or more interacting species reciprocally alter each other’s evolution, but the importance of coevolutionary processes for shaping evolutionary diversification has been an area of debate within evolutionary biology, especially for interactions involving numerous species 1 – 4 . Studies have shown that interspecific selection—selection by one or more species on the traits of another species—can lead to ecological speciation by driving adaptive differentiation among populations, which can lead to sustained evolutionary change in species at multiple spatial and temporal scales 5 . Thus, studies of interspecific selection can provide insight into the fundamental processes generating and maintaining biodiversity, including genetic and phenotypic diversity within and between species. Such studies also represent building blocks in our understanding of coevolution. Indeed, for most species interactions and especially for diverse multi-species interactions, interspecific selection has rarely been measured in one direction, much less reciprocally 6 . Analyzing interspecific selection and coevolution in diverse multi-species interactions (multispecific coevolution) presents a unique challenge because there exists a wide range of scenarios both for the number of traits involved and for the degree of independence of evolutionary dynamics in any particular species pair that makes it difficult to tease apart directionality and sources of selection. For example, in an interaction between a single plant host and a diverse guild of herbivores, the outcome of interspecific selection may vary along a continuum such that host plant traits all experience selection from the herbivore community in the same direction or those same host plant traits may experience selection from individual herbivores within the same community in different directions. Additionally, different herbivore species may also alter each other’s evolutionary dynamics with the host. This entire multispecific coevolution process has been called diffuse coevolution and has received very little empirical attention despite the ecological importance of multispecies communities 7 , 8 . The empirical attention it has received has come from antagonistic interactions, such as plant–herbivore systems (e.g., 2 , 9 – 12 ). These studies have sometimes found that genetic covariance among host traits constrain evolutionary responses of the host to diverse antagonists such that hosts are never able to fully obtain maximum fitness 11 . Interspecific selection in mutualisms may operate differently than in antagonistic systems, and may also depend on whether the mutualism is symbiotic or free-living. Evidence to date suggests that symbiotic multispecific mutualisms may evolve to favor complementary sets of non-competing symbionts, while free-living (non-symbiotic) multispecific mutualisms may evolve to favor the accumulation of species that share a core set of mutualistic traits, rather than specializing on a partner species or mode of interacting with the host 13 . Individual plants are often concurrently associating with diverse rhizosphere microorganisms 14 , 15 such as mycorrhizal fungi, which are common symbionts of over 80% of terrestrial plants 16 . Because mycorrhizal symbioses are diverse, multispecies interactions in which multiple fungi can associate with the same plant and vice versa 17 , 18 , it is not clear whether we expect them to evolve towards a core set of shared mutualistic traits (as predicted for purely free-living mutualisms) or towards a set of complementary non-competing symbionts that have unique modes of interaction with each other (as predicted for intimate symbiosis 13 ). Experiments with pines and ectomycorrhizal (ECM) fungi suggest that pine populations have evolved preferences for particular fungal species 19 , 20 and that some of these interactions may be controlled in plants by independent loci of large effect 21 . Estimates of natural selection by ECM fungi on plant traits could lend insight into how interspecific selection may operate in such multispecific mutualistic species interactions, yet we lack direct field estimates of natural selection in these diverse interactions. Monterey Pine ( Pinus radiata D.Don) is a locally dominant conifer, the native range of which consists of small, isolated populations spanning a broad latitudinal gradient 22 . Post-Pleistocene native populations of Monterey pine are restricted to a small set of geographically separated sites along the west coast of California (USA) and two islands off Baja California (Mexico) 23 . The geographic isolation of these populations 24 provides an opportunity to study how isolated sets of diverse interactions evolve in different contexts. Moreover, the native populations of Monterey pine not only harbor different communities of ECM fungi 25 , and also exhibit significant genetic differentiation in compatibility with particular species of ECM fungi and in several growth/allocation traits, including growth rate, biomass allocation among shoots and roots, and root coarseness 19 , 20 . However, it is unknown whether and how natural selection, including interspecific selection from ECM fungi, may have driven the diversification of those traits. A field experiment where plants have access to a wider array of ECM fungi than in a greenhouse inoculation experiment could be used to test that hypothesis. In order to directly estimate the effect of phenotypic variation on plant fitness in these interactions, we conducted a common garden experiment in which seedlings from P. radiata phenotypes representing a broad suite of possible plant traits were grown in a single location, minimizing the influence of environmental variation on plant fitness. We first used traditional quantitative genetic approaches to evaluate the extent to which plant and fungal traits are heritable. We then estimated selection gradients of plant and fungal traits on three proxies for plant fitness using traditional selection analysis 26 for relative growth rate and total biomass and using logistic selection analysis 27 for plant survival.",
"discussion": "Discussion In this study, we provide evidence for one side of coevolutionary selection in the diverse mutualism between ECM fungi and Monterey pine from a field study, i.e., selection of ECM fungi on Monterey pine. Despite the potential importance of coevolution for driving trait diversification 4 , 13 , there are relatively few examples that quantify reciprocal natural selection in diverse species interactions, i.e., multispecific coevolution. Instead, most studies quantify selection in pairwise host-parasite/predator/competitor interactions (i.e., 11 , 39 , 40 ). Indeed, we are not aware of any examples of studies estimating reciprocal selection forces in a diverse mutualism 6 . This paucity of evidence for coevolution in multispecies interactions may stem from the assumption that the complex nature of biotic selection in diverse interactions may prevent or override the effects of coevolution, making them difficult to measure 41 . It also may reflect limitations in the way coevolution has been traditionally defined, focusing on pairwise interactions; rather, it may be important to recognize that in diverse mutualisms, guilds (groups of species with similar traits) may have converged on core coevolving traits, and thus whole guilds may exert selection on another species in aggregate 6 . In addition, even if it may be difficult to measure responses to selection in individual members of diverse guilds, we suggest that analysis of guild-level traits can lend insight into how the traits of guilds of species may exert interspecific selection. As such, we considered how guild-level traits of the ECM fungal community, including their abundance, diversity, composition, and exploration morphology, may exert (as fungal traits) or respond to (as plant traits) selection on the plant. By combining field estimates of selection gradients with the genetic variance–covariance matrix for multiple traits of the mutualism, we found evidence that the presence of certain fungal OTUs, families, and exploration types can alter the evolutionary response of the plant to other mycorrhizal fungi. Selection on abundance of fungal OTUs Fungal species within diverse assemblages of mycorrhizal fungi can be important sources of selection as individual fungal species may exert selective pressure on particular plant traits and plants may select for particular species of fungal taxa as well, influencing the resulting composition of the ECM fungal community. For example, plants may exhibit a degree of specificity in recognizing fungal partners by sanctioning or rejecting fungi if they are less beneficial 42 , 43 . While previous research with Monterey pine under controlled conditions has suggested that this species has evolved independently in response to different single species of ECM fungi 19 , 20 , here we show that in a field setting, multiple fungal OTUs and a single fungal family are sources of selection on Monterey pine morphological traits; however, the nature and direction of that selection is driven by the likely specificity of the fungi involved. The OTU R. californiensis had a particularly important role in selection; however, the strength and nature of that selection depended on plant fitness proxy. Selection was negative when the proportion survived was used as a plant fitness proxy but positive when total biomass was used as a plant fitness proxy. This difference in selection patterns based on fitness proxy suggests selection in plants for increased compatibility with R. californiensis would increase seedling biomass but decrease seedling survival. While we failed to identify a signal of significant selection when RGR was used as a fitness proxy, analysis of the G-matrix indicated positive genetic correlations of R. californiensis abundance with RGR, suggesting the potential for selection between the two traits . While not widely studied, R. californiensis was first identified in Monterey pine and California live oak forests, and public collection records (e.g., mycoportal.org and mushroomobserver.org) are largely restricted to coastal California, suggesting at least the potential of host specificity for this species 44 ; more research is needed to understand whether this apparent specificity is genetically based or simply represents range restriction. In contrast, we identified several generalist ECM fungal taxa that drove selection in the same way regardless of fitness proxy. Specifically, the abundances of two OTUs, Tomentella1 and T. sublilacina, and their associated family, Thelephoraceae, experienced positive selection regardless of plant fitness proxy. Despite the consistent nature of fungi from the Thelephoraceae family to demonstrate patterns of natural selection with plant traits, G-matrix analysis only identified fungi from the OTU Thelephoraceae3 as experiencing significant genetic correlations with plant traits (negative with both biomass and RGR) while the other six OTUs from this family in the study failed to demonstrate significant genetic correlations with plant traits. Thus, apparent positive selection of ECM fungi in the Thelephoraceae on plant traits can best be interpreted as interspecific selection by fungal traits on plant traits, rather than simply correlated evolution of multiple plant traits. Positive selection for fungi from the genus Tomentella is perhaps not surprising given that these fungi are widespread, dominant species in mature forest stands, sporulate in the soil organic horizon, and can establish from the spore bank shortly after disturbance 45 – 47 . These characteristics suggest that selection may favor plant compatibility with fungi from this genus because they can provide benefits for the plant under a variety of conditions. However, positive selection on plants for increased compatibility with these fungi during the extreme drought conditions of this experiment may also indicate that Monterey pine may adapt to extreme climatic conditions via evolution of increased association with Tomentella and other Thelephoraceae fungi. It is perhaps unsurprising that the most abundant OTU recovered from our seedlings, Tomentella1, was involved in mycorrhizal mediated selection, as the net selective pressure exerted by mycorrhizal fungi on a particular plant trait may be dominated by the numerically most abundant member of the community 48 . Our results suggest that the specificity of fungi involved in plant-mycorrhizal interactions has the potential to drive natural selection in opposing ways; however, for many mycorrhizal fungi we lack an understanding regarding their fidelity 49 . This research further emphasizes the need to bridge this important knowledge gap. Selection on exploration types Exploration types, which reflect the species-level morphological traits by which ECM fungi acquire and transport nutrients, provide an integrated assessment of fungal function and may provide insight into how guilds of ECM fungi are exerting selection pressures 25 , 31 , 34 , 35 . In this study, there was selection by fungi from the contact exploration type on four different plant traits, suggesting they play an outsized role in the selection process. Fungi from the contact exploration type are hydrophilic but their ranges seem to be restricted by mean annual precipitation 50 , suggesting they may be important in dry conditions for plants to acquire water. The range of Monterey pine is coastal, but the soils where Monterey pine exist are generally dry as the pine acquires a large portion of its water budget from the annual fogbank, particularly in the summer when rainfall is limited 51 , 52 ; these conditions were amplified in our study, which took place during an extreme drought event 53 . Taken together, these pieces of evidence suggest that selection on Monterey pine in these conditions has come to favor associations of the pine with ECM fungi that may alleviate water stress. We also demonstrated instances of genetic correlations of exploration types with plant traits without identifying significant contemporary selection. For example, no plant traits experienced significant selection due to fungi from the short exploration type in any of the natural selection models despite positive genetic correlations of fungi from this exploration type with RGR and negative correlations with diameter. This could be because fungi from the short exploration type have previously exerted correlational selection on plant traits, such that the fitness of certain combinations of traits represented peaks on the adaptive landscape under different historical environmental conditions 54 . Alternatively, these genomic covariances of ECM fungal exploration types with plant morphological and fitness traits may result from correlated selection from other unmeasured environmental variables on both the mycorrhizal traits of the plant and these other plant traits. Selection on overall ECM fungal diversity In support of correlational selection as a driving factor in natural selection of plant-fungal relationships is our finding that fungal symbiont diversity itself was an important source of selection. Diversity indices represent a quantitative measure for how many different ECM fungal species are present on the root, and thus may capture the effect of multiple mycorrhizal species as selective agents on the plant, and/or the outcome of selection (both from the environment and through interactions with other biota) on plants for their compatibility with individual fungal species. Interestingly, no particular combination of diversity and plant traits maximized both plant biomass or plant survival, suggesting the potential for antagonistic selection between plant traits and fungal community diversity, and for growth-survival trade-offs in plants. These patterns may reflect the complex nature of biotic selection, particularly for interactions whose function can vary from mutualistic to parasitic depending on resource availability 55 , 56 . Selection under drought conditions This experiment took place during the hottest and driest period on record in the state of California 53 and thus it is likely that selection favored combinations of plant and fungal traits more suited for desert-like conditions, as found more often on the Mexican islands (Cedros and Guadalupe) compared to the California mainland populations of Monterey pine (Supplementary Fig. S1 20 ). Indeed, we found a significant advantage in survival for seedlings with an island (Guadalupe or Cedros) genetic background compared to seedlings with a mainland background (Cambria, Monterey, Año Nuevo), suggesting maladaptation of mainland genotypes to extreme drought conditions. Specifically, the odds of mortality for seedlings with a pure island background or hybrid between an island and mainland pine were 0.70 and 0.69 times the odds of mortality compared to a pure mainland genetic background. This result suggests that selection, especially when survival was used as a plant fitness proxy, favored traits that promote survival in hotter, drier climates over the wetter, cooler climates historically present at Cambria. Understanding selection on mycorrhizal relationships in Monterey pine and other trees during such an extreme climatic event is particularly important as climate change models predict increases in temperature and decreases in precipitation in the future for this region 57 . Moreover, these results lend insight into the microevolutionary processes that may underlie recently identified macroevolutionary patterns of dependent evolution between drought adaptation and mycorrhizal strategies in land plants 58 . One of the biggest advantages of genotypic selection analysis is that it allows for the correction of the role of the environment on traits and thus decreases the possibility that the covariance between the environment and the trait(s) of interest leads to false conclusions regarding whether selection is acting on that trait 59 , 60 . This is perhaps most important in our study due to the extreme drought environment experienced by the plants and fungi used in this study. However, because we used family means to correct for environmental bias, outcomes of selection identified here are more likely to reflect genetic correlations rather than phenotypic correlations determined by the environment 60 ."
} | 5,397 |
40109102 | PMC11923606 | pmc | 1,520 | {
"abstract": "Division of labour (DoL) is most prominently observed in eusocial insects but also occurs in much smaller cooperative groups where all individuals could potentially perform any task. In such groups, previous experience and learning are the most important mechanisms underlying specialization. Using behavioural simulations, we investigate the dynamics of task specialization in groups of various sizes and with different constraints on the choice of task. We assume that individuals choose tasks by weighing their own competence to perform a task against the group requirement of how much that task needs to be performed. We find that task specialization occurs even if individuals choose tasks based solely on the group’s needs rather than their own competence. As large groups are less affected by demographic stochasticity, they can more accurately distribute labour across tasks, and individuals become more effective due to a reduced need to switch between tasks. This effect is enhanced if groups must perform a larger number of tasks. However, from an evolutionary point of view, individuals in larger groups develop a greater responsiveness to group requirements than those in small groups when labour variation carries a fitness penalty and thus will more readily switch between tasks. Small groups thus seem less able to distribute labour optimally over tasks through increased switching, and therefore evolve to ignore task imbalances up to a higher level before the threshold to switch between tasks is crossed. Further, we find that selection on learning ability is stronger in small than in large groups. We conclude that the reason why DoL may emerge more readily in large groups might not be due to a group-size effect on optimal decision-making, but rather because of a lower degree of variation of the labour distribution as a consequence of demographic stochasticity. This article is part of the theme issue ‘Division of labour as key driver of social evolution’.",
"introduction": "1 . Introduction Division of labour (DoL), the uneven distribution of tasks among the members of a cooperative group, is a widespread natural phenomenon in social animals. DoL occurs most conspicuously in eusocial insects [ 1 ] but is also seen in a range of other group-living organisms such as other arthropods, birds, mammals and even bacteria [ 2 – 9 ]. Understanding DoL is of great interest not only because of its pervasive influence on the lives of our own species [ 10 , 11 ], but also because it constitutes a crucial step in the recurring evolutionary transitions of group formation, functional differentiation of its members and integration towards higher levels of organization [ 12 , 13 ]. The forms that DoL can take vary widely across the tree of life. The most extreme form of DoL are the sterile, morphologically differentiated castes found in some eusocial insect species, in which individuals are specialized for functions such as reproduction, defence or foraging that are pursued for their entire lives [ 14 ]. Although even individuals belonging to specialized morphs can often still perform a number of different tasks, such levels of specialization come at a large opportunity cost [ 15 , 16 ]. Another form of DoL is temporal polyethism, where individuals specialize into different tasks in accordance with their age [ 17 ]. Examples can again be found among the eusocial insects, such as the succession of tasks performed by honeybees ( Apis mellifera ) but also in group of vertebrates that form hierarchies based on size and age. In the social cichlid Neolamprologus pulcher , depending on circumstances, large group members may specialize in digging out the breeding shelter, while small individuals specialize in the protection of eggs [ 18 ]. In contrast to caste-based DoL, temporal polyethism is more plastic, with individuals being able to respond to group needs in switching between tasks [ 17 ]. DoL can also be spontaneous, in that individuals specialize on different tasks regardless of caste or age. Such DoL may take advantage of prior individual variation but more importantly arises through the general tendency of animals to develop habits and learn from experience [ 19 , 20 ]. The ubiquity of this mechanism allows DoL to arise spontaneously even in solitary species when experimentally grouped together [ 21 , 22 ]. Spontaneous specialization and learning can also play an important role in DoL in many eusocial insects [ 20 ], and indeed eusocial insects such as bees stand out for their exceptional learning abilities [ 23 , 24 ]. The plasticity of learned behaviour, however, also means that this form of DoL can be ephemeral and rather inconspicuous. Group size has a crucial influence on the emergence of DoL [ 25 ]. Superficially, caste specialization and temporal polyethism seem to be largely limited to eusocial insects, especially those termites, bees and ants that form huge colonies with tens of thousands of individuals, whereas spontaneous DoL occurs in small groups of vertebrates as well. DoL more readily emerges when group size is experimentally increased [ 26 – 29 ]. Several theoretical studies have predicted that increased group sizes can promote the emergence of DoL [ 29 – 33 ]. Cooper & West [ 13 ] argued that group size can enhance DoL if the efficiency benefits of specialization increase with group size. Ulrich et al . [ 29 ] showed that task specialization increases with group size in a simple fixed-threshold model, due to increased homeostasis and reduced stochasticity in individual task choice. Nakahashi & Feldman [ 33 ] predicted an increase of DoL with increasing group size due to what effectively amounts to resource competition between group members, while Pacala et al . [ 30 ] and Gautrais et al . [ 31 ] predicted that group size will increase DoL due to enhanced uptake of information and interaction between individuals. The most widely employed approach to modelling DoL is the response-threshold model, which assumes that individuals perform tasks if a certain stimulus threshold is surpassed [ 31 , 32 , 34 – 36 ]. Between-individual variation in this threshold due to genetic variation or phenotypic plasticity can then give rise to DoL [ 37 ], as long as within-individual variation is not overwhelming [ 38 ]. Response-threshold models often incorporate a feedback mechanism, whereby the repeated execution of a task lowers the threshold and thereby allows individual specialization to arise even in the initial absence of between-individual variation [ 35 ]. However, this widely modelled mechanism (cf. [ 39 ]) does not incorporate any improvement in task performance due to experience or learning. It seems plausible that individuals prefer to perform tasks that they are good at, while at the same time acting to reduce the group’s collective labour demands across the various tasks. Especially in smaller groups, where individuals can track the overall condition and needs of the group, the importance of the decision-making of the individual may have been under-emphasized. To address this gap, in this study we investigate the relationship between group size and DoL using a novel model that highlights the role of learning and the relative importance of individual efficiency versus group labour requirements in spontaneous task specialization. We build an individual-based simulation model in which individuals cooperate to perform a number of tasks and improve their task competence through learning. Demographic stochasticity and the group requirement to work on multiple tasks may induce task switching. We use this model to assess how task switching and DoL depend on group size, the number of tasks and the rate at which individuals improve their task performance (e.g. through practice and learning).",
"discussion": "4 . Discussion DoL, as measured by the (low) average number of task changes per lifetime, emerged under all parameter combinations of our model. As individuals are initially naive, they choose tasks based on the labour requirements of the group and therefore distribute themselves over the tasks, not unlike independent foragers that distribute themselves over a number of resource patches. The group may thereby arrive at a distribution at which no individual can reduce the variation of the labour distribution by changing its own task, similar to an ideal free distribution where no individual can improve its intake by switching between patches [ 40 ]. While individuals remain in a task, they gather competence through experience and thus become less likely to switch between tasks ( figure 1B ). The faster individuals learn, the quicker this process progresses and therefore we observe that high learning rates reduce the number of task switches for a given responsiveness. As Ravary et al . [ 19 ] showed, individual experience can have a lasting effect on task choice and DoL. In our model, we observed that when individuals choose tasks based on previous experience, they readily distribute over tasks in such a way that task switches only rarely occur. But even when individual experience does not greatly inform task choice, the need to balance the group’s labour across tasks yields a distribution of individuals over tasks that leads to the emergence of task specialization. The necessity to switch between tasks arises when additional individuals join the group, or established individuals die. New individuals are initially without experience, and this may lead to a situation where the established individuals remain in their dedicated tasks, while the newcomer starts switching between tasks to reduce inequalities in the labour distribution ( figure 1B ). A similar pattern has been observed in wood-eating termites that can switch between different castes during early developmental stages but later become specialized and lose this ability [ 41 ]. The developmental constraints that mark this transition in termites, however, are absent from our model; even old individuals can freely adopt any task and occasionally will do so. This particular form of temporal polyethism—where young, inexperienced individuals readily switch between tasks according to group needs, while older ones remain specialized—is something we expect to be particularly common in social vertebrates, where experience may play an important role in task performance. Empirical evidence is currently scarce. In cooperatively breeding cichlids, experiments have revealed that small (young) helpers respond more flexibly to territory intrusions than larger (older) helpers by adjusting their threat behaviour and shelter maintenance to the actual type of challenge [ 42 ], and task performance of helpers is influenced by their age (size) and preferred location in the territory [ 18 ]. In humans, a decrease in job-switching frequency with age is well documented ([ 43 ] and citations therein). What are the effects of group size on the processes described above? First of all, a larger group size under the assumption of a constant mortality rate of its members reduces the amount of demographic stochasticity (electronic supplementary material, figure S1). A single mortality event is more impactful on the distribution of labour in a group of 5 than in a group of 33. Thus, the labour variation is generally higher in small than in large groups ( figure 2B ), and to balance the labour distribution individuals start switching sooner (electronic supplementary material, figure S6) and lose more task competence ( figure 2A ) in small than in large groups. Only at high responsiveness and a larger number of tasks, this group-size effect may change to the opposite, where frequent small-effect changes of the group composition induce more task switching in large groups ( figure 2A , middle and right panel). These findings do not, however, indicate that individuals in large groups have a stronger evolved tendency for task specialization. Indeed, the evolutionary optima for large groups yield a higher responsiveness to group requirements. Due to the lower overall labour variation, however, this results in a higher realized task specialization. In our simulations, the number of tasks is generally lower than the number of individuals. If the number of tasks exceeds the number of individuals, some tasks would not be fulfilled, which implies that task switching would increase. If responsiveness is sufficiently low, however, individuals might still specialize if the benefit gained from their experience outweighs the disadvantage caused by an imbalance in labour distribution. Modelling a scenario where some tasks are left unattended due to group size limitations should consider also for how long the effects of task performance will prevail (see e.g. [ 44 ]). The influence of group size on DoL has been studied empirically in a number of species [ 26 – 29 , 45 – 47 ], but the mechanisms underlying group-size effects are largely unclear. Pharaoh’s ants Monomorium pharaonis show a transition from unorganized to organized foraging behaviour above a colony size of 500 individuals, which was hypothesized to be due to the volatility of trail pheromones and the associated inability of small colonies to form stable trails [ 46 ]. Such effects of scale may be important in the evolution of colony size, but this does not relate to the behavioural mechanisms scrutinized in our model. Wasps show frequent transitions between pulp foraging, water foraging and building in small colonies, and an increasing tendency of task specialization in large colonies [ 48 ]. In our model, the different tasks are not explicitly integrated with each other (task partitioning, [ 49 ]), but in wasps, the need to balance these three tasks for the group to succeed may resemble the situation in our model in which the fitness cost of labour variation is high. The mechanism by which specialization arose in wasp societies is unclear [ 48 ]. Theoretical models have investigated group-size effects from various angles, including threshold reinforcement [ 31 ], fixed response thresholds [ 29 , 32 ], interaction effects [ 30 ] and task partitioning [ 49 ]. All these models have in common that the effect of group size on DoL is positive. The larger number of individuals in response-threshold models [ 29 , 31 , 32 ] reduces the variation in stimulus levels that triggers individual activity. Thus, only a few individuals repeatedly pursue a task, either due to innate variation [ 29 , 32 ] or reinforcement [ 31 ]. Although the assumptions of our model differ in important respects (individuals continuously perform tasks, improve their task performance through experience and are sensitive to their own task competence), we recover similar results in the form of a reduction of labour variation in large groups that leads to a decreased number of task switches and therefore increased DoL. An important difference, however, is that the effect on DoL in the other models is derived from the group size itself, whereas in our model it arises through the increased demographic stochasticity characterizing small groups; if the group composition were stable over the long run, task specialization and DoL would emerge regardless of group size. Anderson & Ratnieks [ 49 ] considered the effect of group size in a task-partitioning model, where group members together perform a number of functionally integrated tasks. The larger number of individuals leads to a smoother transition between sequential tasks and reduces queuing delay (for instance when a wasp that foraged for pulp waits to transfer material to a nest-mate; [ 48 ]). Their model presupposes a regime of DoL that gains efficiency through increased group size. The effect of queuing delays is related to variation in the degree to which certain tasks are performed by the group, and therefore represents a situation where, in terms of our model, variation of the labour distribution is strongly penalized. We found that for a given variation penalty, the responsiveness to switch between tasks increases with group size, but that the overall number of switches decreases. Concerning the effect of learning rates on DoL, we found that the learning rate is under stronger selection in small than in large groups. This is because individuals in small groups are more often forced to switch between tasks and therefore regularly perform tasks for which they have no previous experience. The rate of learning thus has a stronger influence on fitness in small groups where individuals have to learn and relearn tasks more often than in large groups where individuals have a higher level of task specialization. One might therefore suggest that all else being equal, DoL and its coordination are cognitively more demanding in smaller groups, in contrast to the social brain hypothesis that predicts larger group sizes to increase cognitive demands and brain sizes [ 50 ]. Our model, however, only considers the DoL aspect of group living, not the role of conflict and social interactions. A key limitation of our modelling approach is that we disregarded potential conflicts of interest between group members by assuming a complete alignment between individual and group fitness, suggesting that group-level adaptation may prevent intra-group conflict [ 51 ]. This might be a reasonable assumption if all tasks are equally costly to perform, and there is no cost to individual fitness in switching between tasks. Our study cannot provide insights, however, into how group size may affect DoL via the negotiation of conflicts between individual and collective interests, which is a drawback as the conflict potential may be particularly prevalent in smaller social groups below the threshold of eusocial colonies. Future work should therefore address the role of negotiation and conflict resolution on the emergence of DoL in groups of varying sizes."
} | 4,474 |
39277566 | PMC11401808 | pmc | 1,522 | {
"abstract": "At the intersection of computation and cognitive science, graph theory is utilized as a formalized description of complex relationships description of complex relationships and structures, but traditional graph models are static, lack the dynamic and autonomous behaviors of biological neural networks, rely on algorithms with a global view. This study introduces a multi-agent system (MAS) model based on the graph theory, each agent equipped with adaptive learning and decision-making capabilities, thereby facilitating decentralized dynamic information memory, modeling and simulation of the brain’s memory process. This decentralized approach transforms memory storage into the management of MAS paths, with each agent utilizing localized information for the dynamic formation and modification of these paths, different path refers to different memory instance. The model’s unique memory algorithm avoids a global view, instead relying on neighborhood-based interactions to enhance resource utilization. Emulating neuron electrophysiology, each agent’s adaptive learning behavior is represented through a microcircuit centered around a variable resistor. Using principles of Ohm’s and Kirchhoff’s laws, we validated the model’s efficacy in memorizing and retrieving data through computer simulations. This approach offers a plausible neurobiological explanation for memory realization and validates the memory trace theory at a system level.",
"conclusion": "Conclusion We modeled the memory mechanism of the cerebral cortex by leveraging the known principles of synaptic plasticity and the behavior of engram cells, which have been extensively studied in neurobiology. Specifically, the model simulates the strengthening of synaptic connections based on Hebbian plasticity and the dynamics of neuron activation patterns as described in the engram theory of memory storage. Additionally, the electrophysiological principles guiding our model are grounded in the propagation of action potentials and synaptic transmission laws, such as Ohm’s and Kirchhoff’s laws, which have been validated through computer simulations. These simulations align with empirical observations of neuron activation and memory recall dynamics in biological neural networks, such as those demonstrated in Kitamura et al.’s work on memory engram cells [ 14 ]. The MAS model in the algorithm is active, with each agent capable of personalized adaptive learning based on local neighborhood information, modeling the memory system of biological neural networks. This breaks away from the reliance of the traditional graph model on a global view for operation, lacking autonomous parallel distributed processing capabilities, and is closer to actual biological neural networks. Inspired by the capability of biological neurons to transmit electrical signals, this adaptive learning behavior was simulated through microcircuits centered on variable resistors, successfully realizing the simulation of the memory and retrieval processes in the entire MAS network on a computer. In simulations, the model could distribute memory instances across the MAS, transforming them into connected paths, and achieving path memory and retrieval within the network. The model’s generalizability was verified, and it was shown to be capable of achieving memory functions in different topological structures of MASs. Tests determine the capacity of networks of various scales, verifying factors influencing capacity size, as network scales, edge agent activation ratios, and dispersion of memory instances all affect network capacity. Particularly, referencing the structure of the cerebral cortex in higher primates, the memory capabilities of networks with six layers of depth were verified, exploring the relationship between the capacity of six-layer networks and their width. The process of memorizing and retrieving instances simulates the biological processes of memory and recall. Experimental results showed certain similarities between the model and biological neural networks. For example, the number of iterations required for a MAS network to memory different instances and achieve the same retrieval effectiveness varies, as does the time required for humans to remember different content. Furthermore, our model universally possesses memory functions in MASs of different sizes and topological structures, similar to how the microscopic differences in people’s brain networks do not affect the effectiveness of memory functions. Our model performed better and had a larger capacity in networks with smaller depths, with capacity increasing linearly with width, resembling the flat morphology and horizontally expansive cortical structure of the brains of higher primates. In summary, we proposed a self-learning MAS model that does not require a global perspective, based on the simulation of biological neural networks. The MAS achieves memory functionality while decentralized, based on local information and adaptive learning capabilities. The model demonstrates biological plausibility by aligning its simulated memory processes with known neural activity patterns, the activation of engram cells during memory recall. It also serves as an inspiration for further research into memory neural mechanisms, enriches neural computational models, and offers new perspectives and ideas for neural computation research.",
"introduction": "Introduction Deficiencies in research on brain memory mechanisms Memory is a cognitive function produced by the activity of the brain’s cortical neural system, and the ability to form memory is the foundation for accumulating knowledge and making reasoning judgments. As an extremely complex information processing system, humans have not yet developed a complete theoretical system for the brain’s memory mechanisms. Exploring how the brain encodes, stores, and retrieves information remains a significant challenge in memory research. Modern neuroimaging technology is rapidly advancing, and through methods such as fluorescent tagging and two-photon imaging, humans are gradually unveiling the neural system’s connectivity and other details. However, such analyses can only offer limited insights, and we are still unfamiliar with the neural mechanisms and information flow directions underlying memory activities. From a neurobiological perspective, memory refers to neural system activities or physical changes in neuronal connections triggered by external stimuli or brain states [ 1 ]. Neurobiological research on memory mechanisms focuses on synaptic plasticity [ 2 – 5 ], neurotransmitters [ 6 , 7 ], and electrochemical signals [ 8 , 9 ] at the cellular and molecular levels, while cognitive psychology likens memory to an information processing system responsible for encoding, storing, and retrieving information. These theories provide a reasonable starting point to map the brain’s coarse-scale organization using functional imaging technologies including EEG and fMRI [ 10 ]. Neurobiology and cognitive psychology investigate the brain’s memory mechanisms from different levels, yet there remains a significant gap in our understanding of the actual biological processes behind memory, making it challenging to accurately model the brain’s memory mechanisms. Neurobiology provides a rich array of dynamic components, and cognitive psychology outlines the global functional layout of the brain, yet neither fully elucidates the complete process of brain information processing. Many unknown details still exist about how biological neural networks encode, store, and retrieve information, which cannot be fully explained by electrophysiological or anatomical experiments, nor by cognitive experiments alone. The challenge in memory research at the level of biological neural networks is to develop a computational model of brain information processing that adheres to neurobiological constraints and can execute memory tasks. Modern digital computers possess powerful storage capabilities, and comparing them with the brain can further our understanding of what is still missing in memory research. Table 1 compares computer storage and brain memory management, from which the implementation process of computer storage, from the bottom to the top layer, is clear. However, the implementation process of brain memory involves numerous unknowns, which motivates us to simulate and emulate the brain’s memory process based on neurobiological mechanisms.\n Table 1 Comparison of computer storage management and brain memory management Category Computer storage management Brain memory management Minimum Functional Unit Data are stored in binary form, with hard drives using polarity of magnetic particles, charges in capacitors, etc., to represent 1s or 0s In neural cells, membrane possesses resting and action potentials. In resting state, signal is inhibited; if activated, it generates an action potential, conducting signal Storage Device An HDD records data by changing the distribution of the two polarities of magnetic particles using an electromagnet on read/write head; an SDD stores 0s and 1s by altering the number of electrons in the floating gate layer through the application of an electric field Memory refers to neural system activities or physical changes in neuronal connections triggered by external stimuli or brain states [ 1 ]. Many studies have begun to define engrams as basic units of memory, but there are unresolved issues such as how structure of engrams affects quality of memory, how multiple engrams interact with each other, and how engrams change over time [ 11 ] Data Allocation Operating systems allocate disk space for files with blocks as the basic unit, using methods such as contiguous allocation, linked allocation, and indexed allocation We have a very limited understanding of how a single memory item is structurally stored at the level of biological neurons, and how multiple memory items are allocated among neural cells (groups) in cerebral cortex Data Access Control circuits translate logical addresses into physical addresses. Read/write head of an HDD moves to the cylinder, track, and sector where the data are located for data access; for an SDD, controller directly performs read and write operations on storage units at physical address When brain stores or recalls a memory, we know little about how it locates engram of memory content among vast number of neural cells in cerebral cortex, replacement of old memories with new ones, and neuronal network mechanisms during recall The memory trace theory The concept of how memory is stored and retrieved has always been a focal point of exploration for neuroscientists. Semon introduced the term “engram” to describe the neural substrate that stores memories [ 12 ]. He suggested that experienced events activate a group of neurons to produce chemical or physical changes, thereby forming engrams, with the cells generating memory traces known as engram cells. The reactivation of these engrams can induce the recovery of memories. Hebb proposed the theory of cell assemblies, positing that cells activated by the same event and having intrinsic connections would form cell assemblies, and the synaptic connections between these assemblies would be strengthened [ 13 ]. Based on the memory trace theory and the theory of cell assemblies, we have gained further understanding of the neural basis of memory storage and retrieval-that the encoding and storage of memory information depend on the concurrent activation of neurons during memory formation. During the memory storage phase, external stimuli cause a group of neurons in the brain to discharge together, initiating changes in relevant signaling pathways and gene expression. These neurons undergo lasting chemical or physical changes, and the memory information is believed to be stored in the network formed by these neurons. Thanks to a combination of various techniques such as molecular and cellular neurobiology, physiological recording and multiphoton imaging, transgenic and viral vector-mediated gene insertion, and optogenetics and chemo-genetics, neuroscientists have begun to identify and manipulate memory engram cells. Kitamura and colleagues demonstrated that memories are stored in engram cells through engram cell labeling and optogenetic manipulation, showing that engram cells reactivate during memory retrieval, and artificial activation or inhibition of this group of engram cells can directly trigger or suppress memory expression, proving that fear memories exist in engram cells [ 14 ]. Many laboratories’ research results have begun to define engrams as the basic unit of memory. Recent studies on memory engram cell populations suggest that the memory trace of a given memory is not necessarily located in a single anatomical location but is distributed across multiple locations connected by a specific memory pattern, thus forming memory engram cell pathways [ 15 ]. Currently, three types of evidence support the rise of the memory trace theory: Observational studies provide correlational evidence between the physiological and structural characteristics of neurons in specific brain regions and memory behaviors In functional loss studies, animals or humans with physical or chemical damage to specific brain regions exhibit impairments in certain aspects of memory behavior. The use of transgenic, optogenetics, and chemo-genetic techniques to identify specific subgroups of cells related to specific memory behaviors. The ability to identify and manipulate engram cells and the whole-brain engram complex has advanced the study of the memory neural substrate. However, many unknown details remain about how biological neural networks store information, such as how engram structures affect memory quality, how multiple engrams interact, and how engrams change over time [ 11 ]. Directed graph and MAS for brain modeling The brain can be conveniently represented as a network of neurons and their directed interconnections, and memory maybe represented by those connection patterns according to the trace theory, making directed graph a mathematical tool for studying its structure and memory system. Directed graph, as a branch of graph theory, which provides tools for processing and analyzing network structures. Tracing back to Euler’s solution to the Seven Bridges of Konigsberg problem in the 18th century [ 16 ], it has been applied such as to design solutions for the Traveling Salesman Problem(TSP) [ 17 ], construct knowledge graphs [ 18 ], and create databases using graph structures [ 19 ], where graphs usually serve as structured representations of data or knowledge, inherently lacking dynamic behavior, with their functionality reliant on externally applied algorithms. In these applications of graph theory, algorithms typically operate on the graph structure from a global view, meaning that the executor of the algorithm (like a CPU) has access to global information and makes decisions. The modeling of brain by directed graph is centralized, needs a “God’s eye” view, while biological neural networks lack such a global view or central controller and are characterized by decentralization, consisting of many simple units that are only connected to their neighbors. But if we upgrade the static nodes in the directed graph to dynamic agents, creating a network system by multiple agents, it can better align with biological neural networks and more accurately model and simulate biological memory processes. We propose an algorithm that does not rely on a “God’s eye” view, focusing on implementing a memory function in a multi-agent system (MAS). Just as neurons have only local connections, agents in a MAS can only see their connected neighbors. Each agent makes decisions based on its local field of information. Agents are no longer passive data storage nodes, but are active, autonomous units, which can adaptively learn how to respond in different contexts, simulating the behavior of biological neurons. Passive nodes with a single global algorithm and active agents with numerous independent small algorithms represent two completely different paradigms. Contrasting these two modes in Fig. 1 , we see the difference between traditional centralized processing and the proposed decentralized processing. In the latter, each agent stores information and can also process and transmit information, forming complex dynamic patterns across the entire MAS. Fig. 1 Comparison of centralized and decentralized modes We focus on the information processing mechanisms behind brain memory activities, using MAS to comprehensively model and simulate memory, abstracting memory instances into directed paths in the MAS. Through MAS and the pervasive learning algorithms inherent in each agent, it dynamically learns and optimizes these paths, achieving a memory mechanism based on MAS, enabling efficient information memory and retrieval. MAS can provide a detailed implementation algorithm for imprinting hypothesis at the directed graph network level. Perhaps providing an intermediate level between the low-level electrochemical level and the high-level behavioral level, offering a novel perspective on the brain’s memory mechanisms."
} | 4,308 |
19537995 | PMC2990307 | pmc | 1,523 | {
"abstract": "Division of labor in social groups is affected by the relative costs and benefits of conducting different tasks. However, most studies have examined the dynamics of division of labor, rather than the costs and benefits that presumably underlie the evolution of such systems. In social insects, division of labor may be simplistically described as a source-sink system, with external tasks, such as foraging, acting as sinks for the work force. The implications of two distinct sinks – foraging and waste-heap working – for division of labor were examined in the leaf-cutting ant Atta colombica . Intrinsic mortality rates were similar across external task groups. Exposure to waste (a task-related environment) led to a 60% increase in the mortality rate of waste-heap workers compared to workers not exposed to waste. Given the small number of workers present in the waste-heap task group, such increases in mortality are unlikely to affect division of labor and task allocation dramatically, except perhaps under conditions of stress.",
"introduction": "Introduction Division of labor plays a central role in the organisation and success of social groups ( Oster and Wilson 1978 ). By enabling groups to coordinate their response to challenges (e.g., Gordon 1986 ), division of labor is assumed to promote ergonomic efficiency ( Oster and Wilson 1978 ) and provide ecological benefits (e.g., predator evasion; McGowan and Woolfenden 1989 ) from which all group members can benefit. Consequently, an understanding of the behavioural rules and causal factors underlying division of labor will enhance our overall understanding of the evolution of social groups. Analyses of behavioural rules and causal factors have tended to employ a cost/benefit approach (pioneered by Oster and Wilson 1978 ), where the benefits of a particular aspect of division of labor are compared to its costs at the individual level, the group level, or both (e.g., Bednekoff 1997 ). However, there is little hard evidence for the underlying assumption that costs and benefits are crucial to the structuring of division of labor ( Schmid-Hempel 1992 ; but see Clutton-Brock et al. 1998 , 1999 ). Social insects have complex and well-studied systems of division of labor ( Wilson 1975 ; Gordon 1996 ; Beshers and Fewell 2001 ), and studies of insect societies have generated both ultimate and proximate explanations for how division of labor is structured ( Beshers and Fewell 2001 ). Proximate explanations have included genetic variation ( Fuchs and Moritz 1998 ), behavioural thresholds ( Beshers et al. 1999 ), interaction patterns ( Gordon 1999 ), age ( Wilson 1976 ), physiology ( Powell and Tschinkel 1999 ; Blanchard et al. 2000 ), and source-sink models of worker movement ( Tofts 1993 ). Assuming that division of labor and worker allocation are adaptive, their structure should be the product of the associated costs and benefits of particular strategies. That costs are important is suggested by observations such as high levels of worker inactivity ( Cole 1986 ), changes in foraging in response to overabundant food ( Rissing 1989 ) or mortality pressure ( Gentry 1974 ) and general flexibility in worker allocation to different tasks as the environment changes ( Gordon 1986 , 1987 , 1989 , 1991 ; Calabi and Traniello 1989 ; Crosland and Traniello 1997 ). However, the costs and benefits associated with different tasks, or patterns of task allocation and division of labor, have rarely been measured and may not always be tractable [but see for foraging costs De Vita (1979) , Porter and Jorgensen (1981) , Schmid-Hempel and Schmid-Hempel (1984) , Weier and Feener (1995) , Fewell et al. (1996) and for trail maintenance costs Howard (2001) ]. Insect societies normally have a workforce functionally split into an internal work group (or innendienst ), tending the brood and performing other intra-nidal duties, and an external work group ( aussendienst ) gathering forage and defending the nest or territory ( Hölldobler and Wilson 1990 ). The usual pattern of temporal polyethism is that internal workers become external workers towards the end of their lives ( Hölldobler and Wilson 1990 ). However, an asymmetry in mortality rates between external and internal workers (with external workers generally facing higher mortality rates than internal workers; Schmid-Hempel and Schmid-Hempel 1984 ) can result in a pull of workers from the internal source to the external sink. This unidirectional pull may be an important factor in the structure and organisation of division of labor ( Tofts and Franks 1992 ; Tofts 1993 ). While external workers may themselves be divided into different task groups, in most species the final group and thus the main sink are the foragers ( Porter and Jorgensen 1981 ; Gordon 1986 ). Consequently, any change in the mortality rate of external workers may lead to a relatively simple change in the linear flow of workers from internal to external tasks. Systems with more complex flow patterns between sources and sinks are likely to provide us with novel insights into the structuring and organization of division of labor. Leaf-cutting ants with an external waste heap (e.g. Atta colombica ( Weber 1972 ) where waste is composed mainly of discarded fungus garden material) embody just such a complex system, with two distinct and well-characterised external sinks – foragers and waste workers ( Hart and Ratnieks 2002 ). At any one time, foragers represent 88.8% of external workers, with waste workers making up the remainder ( Hart and Ratnieks 2002 ). It is likely that waste heap management increases nest hygiene and reduces disease transmission within the colony, and that a failure to allocate workers to waste management tasks would be highly detrimental to the colony ( Hart and Ratnieks 2001 , 2002 ). Waste workers are further sub-divided into waste transporters, carrying waste from inside the nest to the external waste heap, and waste heap workers (0.2% of the total external work force), which remain on the heap and work with the waste, presumably speeding its decay. Transitions between foraging and waste heap working are very rare and the reverse transition has not been observed ( Hart and Ratnieks 2002 ). Thus, internal workers may follow one of two routes to an external task – either to become a forager or to become, first, a waste transport worker and then a heap worker. Consequently, the dynamics of worker flow from internal to external tasks should depend, to some extent, upon mortality rates in the two spatially-separated and physically-differentiated tasks of foraging and waste-working. While from a static numerical perspective foraging would appear to be the most important sink in this system, a relatively high rate of mortality in waste workers could have a significant impact on the flow of workers between internal and external task groups. Mortality rates in external workers have two components. First, the intrinsic rate of mortality due, for example, to worker age. Second, additional mortality imposed by the task-related environment, for example, the energy cost of doing a task or exposure to predators. In this study, we take the first steps in addressing how mortality rates across task groups may affect division of labor. Taking A. colombica as our model system, we asked a) do intrinsic mortality rates vary across external task groups?, and b) does exposure to a task-related environment result in increased mortality rates among waste-heap workers? The results are discussed in relation to field observations of division of labor in this species, and more generally in the context of the evolution of division of labor in social insects.",
"discussion": "Discussion Working on the waste heap of a leaf-cutting ant colony is clearly a costly business. Worker mortality rates were nearly 60% higher when ants were exposed to, and allowed to work, waste material than when left in control conditions. In this study two sources of mortality were analysed in external workers. First, intrinsic mortality rate did not differ across task groups. If this were the only source of mortality, it would indicate that foraging is by far the largest sink for a colony’s workers. In these experiments, ants had no access to food, but work by Silva et al. (2003) shows that foragers of Atta sexdens , a closely related species to A. colombica , are likely gaining nutrition from either leaves, fungal garden material, or both, which acts to prolong their lifespan. Consequently, for foragers our experimental protocol almost certainly underestimated potential lifespan. While it is unclear whether waste transport workers have any source of nutrition (their activity leads to them being isolated from the rest of a colony’s workers, to prevent contamination of the fungal garden), waste heap workers have no access to either source of nutrition. These considerations suggest that mortality in waste heap workers may be more important as a sink for a colony’s workers than our results suggest (see below for further discussion). In the second experiment, the first step was taken towards understanding task-specific additional mortality, which showed that waste-heap workers have a 60% increased mortality rate when exposed to their task-related environment. This increase in mortality either results from an increase in metabolic activity due to sorting and working waste, or from exposure to pathogenic organisms present in waste (including mites and fungi, Bot et al. 2001 ), or, most likely, a combination of both factors. It would be interesting to conduct further experiments using sterilised waste material to separate these two sources of task-related mortality. In addition, a complete description of this system would require further laboratory and field studies to measure task-related mortality rates in foragers and waste transporters. What are the implications of these results for understanding division of labor? This work, together with Hart and Ratnieks (2002) , provides the first demonstration of a division of labor system with two costly and thus potentially important worker sinks. In Atta colombica , foragers and transport/heap workers do not belong to distinct morphological castes ( Hart and Ratnieks 2002 ) and thus presumably draw on a common pool of reserves within the nest. Consequently, decisions about allocating workers to foraging must be traded off against allocating workers to waste management work. This contrasts with previously studied systems, where only one sink exists and transition of workers occurs from midden work (equivalent to heap work in this study) to foraging work ( Porter and Jorgensen 1981 ; Gordon 1989 ). While we have shown that such a trade-off exists in our study system, how important is it? A simple acceptance of the mortality data from experiment 1 would suggest that mortality due to heap work is unlikely to play a major role in regulating this trade-off, given the small number of workers involved in this task. However, if we have overestimated mortality in foragers, as the work of Silva et al. (2003) would strongly suggest, it is possible that the increased rate of mortality in active heap-workers may indeed play an important role in the flow of workers between tasks. Further work, on both this and the importance of waste transport work vs. foraging, is needed. The mortality rates suffered during heap work also affect the movement of workers between waste transport and waste heap work. The population of waste transporters is 55 times larger than that of heap workers ( Hart and Ratnieks 2002 ). This suggests that either waste transporters have remarkably low task-related mortality rates or, more likely, many workers die before having the opportunity to become heap workers. If the latter is true, then waste transport, not heap work, would be the main second sink for workers and assessment of mortality rates in waste transporters and foragers is essential to really understand the dynamics of this two sink system. However, it is under conditions of colony-stress that waste-related mortality is likely to have the biggest impact. Hart et al. (2002) showed that large numbers of heap workers were only present when the danger posed to a colony by the dangerous parasitic fungus Escovopsis was high (we note that this also supports the idea that the costs of heap work are sufficient for colonies to modulate task allocation with respect to them). In such conditions, where large number of heap workers are needed, there may yet be effects for the whole system, with heap workers pulling more workers through from the transport group, and potentially reducing the number of workers available for foraging. We also found a significant effect of worker size on mortality rates, with larger ants living longer. It seems likely that this effect was due to larger ants having greater energy stores and thus being able to resist mortality for longer. Such lifespan/size patterns have been found under other circumstances ( Porter and Tschinkel 1985 ), leading to the suggestion that larger ants are more valuable to a colony as they can do a task for longer than their smaller sister workers ( Hölldobler and Wilson 1990 ). So far, we have implied that the impact of worker sinks on division of labor and task allocation is a simple relationship, with smaller sinks drawing fewer workers through and vice versa. However, in reality task allocation and division of labor are complex dynamic systems, which require both positive and negative feedback loops in order to function. It seems likely to us that if mortality becomes too high, tasks that previously acted as sinks may then inhibit the movement of workers from one task group to another, leading to the shutting down of colony functions ( Whitford and Bryant 1979 ; MacKay 1982 ; Greene and Gordon 2003 ) and potentially the initiation of new activities, such as nest migration away from the cause of mortality ( Hart 2002 )."
} | 3,507 |
26648921 | PMC4663275 | pmc | 1,526 | {
"abstract": "The microbial-mediated anaerobic digestion (AD) process represents an efficient biological process for the treatment of organic waste along with biogas harvest. Currently, the key factors structuring bacterial communities and the potential core and unique bacterial populations in manure anaerobic digesters are not completely elucidated yet. In this study, we collected sludge samples from 20 full-scale anaerobic digesters treating cattle or swine manure, and investigated the variations of bacterial community compositions using high-throughput 16S rRNA amplicon sequencing. Clustering and correlation analysis suggested that substrate type and free ammonia (FA) play key roles in determining the bacterial community structure. The COD: NH 4 + - N (C:N) ratio of substrate and FA were the most important available operational parameters correlating to the bacterial communities in cattle and swine manure digesters, respectively. The bacterial populations in all of the digesters were dominated by phylum Firmicutes, followed by Bacteroidetes, Proteobacteria and Chloroflexi. Increased FA content selected Firmicutes, suggesting that they probably play more important roles under high FA content. Syntrophic metabolism by Proteobacteria, Chloroflexi, Synergistetes and Planctomycetes are likely inhibited when FA content is high. Despite the different manure substrates, operational conditions and geographical locations of digesters, core bacterial communities were identified. The core communities were best characterized by phylum Firmicutes, wherein Clostridium predominated overwhelmingly. Substrate-unique and abundant communities may reflect the properties of manure substrate and operational conditions. These findings extend our current understanding of the bacterial assembly in full-scale manure anaerobic digesters.",
"introduction": "Introduction Anaerobic digestion (AD) represents an efficient process for the treatment of various kinds of organic waste along with biogas production (Alvarado et al., 2014 ). The biological process involves four sequential steps: substrate hydrolysis, fermentation, acetogenesis and methanogenesis, which requires the cooperation of bacteria and archaea (Ali Shah et al., 2014 ). Archaea, especially methanogens, are key players during methanogenesis, thus attracting much attention. However, bacterial populations are essential in anaerobic digesters treating insoluble organic materials, such as animal manure, since the hydrolysis step is often the bottleneck of AD process due to the nature of complex and recalcitrant substrates (Werner et al., 2011 ; St-Pierre and Wright, 2014 ; Carballa et al., 2015 ). In addition, bacteria also take charge of the critical syntrophic metabolism coupled to methanogenesis (Morris et al., 2013 ), so that stable performance can be achieved during the AD processes. Multiple factors including digester design, substrate and operational conditions influence microbial community structures (Lin et al., 2013 ; Town et al., 2014 ). Substrate is recognized as a key factor affecting fermentation efficiency, as well as the microbial community composition (Zhang et al., 2013 ; Ziganshin et al., 2013 ). Cluster analysis of the bacterial and archaeal communities shows that reactors treating similar substrates group together (Sundberg et al., 2013 ). It is postulated that substrate type determines the observed differences in phylogenetic structure based on a meta-analysis of 16S rRNA gene sequences retrieved from 79 digesters treating various substrates (Zhang et al., 2014 ). Nonetheless, microbial populations in anaerobic manure digesters can display high variations even at the digestion of a common core substrate (St-Pierre and Wright, 2014 ). Operational conditions including temperature and ammonia content could impact bacterial community structure. It is reported that bacterial communities clustered based on factors rather than the input materials in lab-scale thermophilic digesters (Town et al., 2014 ). That is probably because high temperature imposes much stronger influences than other operational conditions on the communities (Ziganshin et al., 2013 ). Animal manure is widely used as substrate in anaerobic digesters, which often contains high free ammonia (FA) due to high protein content (Deublein and Steinhauser, 2008 ; Riviere et al., 2009 ). FA has an inhibiting or even toxic effect on prokaryotic communities because it may passively diffuse into cells, causing proton imbalance and potassium deficiency (Sprott and Patel, 1986 ; Chen et al., 2008 ). FA also inhibits pH sensitive species (Chen et al., 2008 ). Syntrophic acetate oxidization (SAO) performed by SAO bacteria is observed to become important under high ammonia content (Schnurer et al., 1999 ; Schnurer and Nordberg, 2008 ). Therefore, the selectivity of ammonia to different microbial populations could be the mechanism structuring prokaryotic communities in anaerobic digesters treating animal manure. In anaerobic digesters, core communities [represented by operational taxonomic units (OTUs)] are commonly found in different digesters with relative high abundances (Riviere et al., 2009 ; Saunders et al., 2015 ). In addition, core communities of anaerobic digesters were found within microbial populations that are capable of performing substrate hydrolysis, fermentation and syntrophic metabolism (St-Pierre and Wright, 2014 ; Rui et al., 2015 ). They may vary depending on different substrate (Riviere et al., 2009 ; Nelson et al., 2011 ; St-Pierre and Wright, 2014 ). Therefore, the elucidation of core and unique communities among different full-scale anaerobic digesters might be useful to indicate important traits of AD process, and to identify putatively important organisms for microbial management in AD (Saunders et al., 2015 ). Previously, core and unique OTUs were identified in 7 different full-scale anaerobic digesters with the clone library method (Riviere et al., 2009 ). Three anaerobic digesters shared 132 core OTUs (St-Pierre and Wright, 2014 ). However, information is still limited due to limited samples of full-scale biogas reactors. Core and unique OTUs can be better determined by using more independently-operated full-scale anaerobic digesters and high-throughput methods (Saunders et al., 2015 ). In China, there are 3717 large-scale (digester volume >500 m 3 ) and 18,853 medium-scale (digester volume of 50–500 m 3 ) biogas plants that have been established by the end of 2009 (Jiang et al., 2011 ). Of these, swine and cattle manure are two most popular substrates. Few studies have been conducted to identify the potential core and unique bacterial populations, as well as the factors driving the assembly of the bacterial communities, among multiple full-scale anaerobic digesters treating animal manure. In this study, we collected 20 sludge samples from independently-operated full-scale anaerobic digesters at different geographical locations across China. The objectives were to identify: (i) important factors shaping the bacterial communities, and (ii) the potential core and unique bacterial populations in digesters treating cattle and swine manure.",
"discussion": "Discussion Although our studied digesters were operated under different operational conditions, substrate types and geographical locations, potential core and unique OTUs were identified. Further, substrate type and free ammonia (FA) were revealed as the most dominant factors differentiating bacterial communities in these digesters. In this study, the observed clustering of the samples from different full-scale digesters could be attributed to substrate type. This is consistent with findings that substrate shapes microbial community structure in AD systems (Sundberg et al., 2013 ; Wagner et al., 2013 ; Ziganshin et al., 2013 ; Zhang et al., 2014 ). A wide variety of components in manure can be utilized to produce biogas in AD systems, e.g., protein, cellulose and lipid. Though the feedstock in both types of digesters is animal manure, different chemical natures and microbial communities in manure inoculums could contribute to the variations of community structure in the AD systems. Indeed, swine manure is a kind of protein-rich organic substrate (Hansen et al., 1998 ), while cattle manure is often composed of cellulose-rich material since the feedstock is mainly silage with high C:N (ASABE, 2005 ). In this study, cattle manure digesters did show significantly higher C:N compared to swine manure digesters ( p < 0.05, data not shown). Nevertheless, substrate type could not explain the observed segregation of the bacterial communities within both cluster members. A previous study also reported high variations of microbial populations in anaerobic manure digesters at the digestion of a common core substrate (St-Pierre and Wright, 2014 ). Operational parameters may cause such variations. In this study, correlation analysis revealed that pH, FA, and NH 4 + - N were all significantly correlated with Shannon's diversity and the observed number of OTUs. However, further analysis revealed that sludge pH did not affect the clustering in both Cluster 1 and Cluster 2 samples. Rather, NH 4 + - N and FA were highly related to the clustering of the samples (Figure 1B ). FA is very toxic to methanogenic community (Chen et al., 2008 ). We also observed that the responses of different bacterial taxa to FA were not the same. Thus, the selection of different prokaryotic taxa by free ammonia is likely an important mechanism shaping prokaryotic community structure in manure AD systems. Excessive FA is detrimental to AD process because high FA content not only changes pH in the digesters, but also causes proton imbalance and potassium deficiency in microbial cells (Sprott and Patel, 1986 ; Chen et al., 2008 ). In this study, though the sludge pH was neutral in these digesters, FA content highly varied, and more than 50 mg l −1 was detected in several swine manure digesters (s9, s10, s11, s12, s14, and s19). However, the FA content in the swine manure digesters is less likely to cause acute ammonia inhibition (Rajagopal et al., 2013 ). Alternatively, it may select specific bacterial populations that can better tolerate higher FA. For example, members of phylum Firmicutes, especially Clostridium sensu stricto , showed a positive correlation with FA (Figure 4 and Table S6 ). Firmicutes are ubiquitously involved in substrate hydrolysis, fermentation and acetogenesis (Nelson et al., 2011 ; De Vrieze et al., 2015 ). Several known species which are capable of syntrophic acetate oxidation (SAO) at elevated total ammonia concentrations are affiliated to this phylum (Schnurer et al., 1996 ; Westerholm et al., 2010 ; Sieber et al., 2012 ). Therefore, Firmicutes probably play more essential roles under high free ammonia content. In contrast, many populations affiliated to Proteobacteria, Chloroflexi, Synergistetes, and Planctomycetes showed negative correlations with FA ( p < 0.05; Figure 4A , Table S6 ), suggesting that they may be inhibited by high FA content. Many Synergistetes and Chloroflexi members are able to perform syntrophic metabolism in association with hydrogenotrophic methanogens during AD process (Sekiguchi et al., 2003 ; Yamada et al., 2006 ; Sieber et al., 2012 ). Dominant genera Smithella and Syntrophorhabdus in phylum Proteobacteria are able to convert propionate and aromatic compounds into acetate by syntrophic association with hydrogenotrophic methanogens (de Bok et al., 2001 ; Qiu et al., 2008 ). In this study, they were less represented in digesters with high FA content. In addition, other syntrophic microbes, e.g., Pelotomaculum, Syntrophomonas , and Desulfobulbus showed decreased trends, even though such changes were not significant at p = 0.05. The overall results indicated that syntrophic metabolism by these microbes are likely inhibited when FA content is high. High FA content is also known to trigger the metabolic shift toward SAO (Schnurer et al., 1999 ; Schnurer and Nordberg, 2008 ). A limited number of mesophilic syntrophic acetate oxidizers have been isolated, e.g., Clostridium ultunense (Schnurer et al., 1996 ), Syntrophaceticus schinkii (Westerholm et al., 2010 ), and Tepidanaerobacter acetatoxydans (Westerholm et al., 2011 ). However, we did not observe the emergence of these species in most digesters, possibly suggesting that SAO is not a dominant pathway. This is likely caused by the fact that FA content (1.34–149.23 mg l −1 ) in our studied digesters did not reach the ammonia inhibition threshold (Hansen et al., 1998 ), and thus characterized SAO species were not observed. Alternatively, some uncharacterized heterogeneous SAO bacteria are possibly responsible for SAO in reactors with increased ammonia content (Werner et al., 2014 ). Further simultaneously in-depth studies of methanogenic and bacterial communities and their interactions are warranted. The core communities play critical roles in AD processes and the concept might be useful to identify putatively important organisms for microbial management in AD (Saunders et al., 2015 ). The core bacterial communities were defined as those commonly found in anaerobic digesters, and six core OTUs were identified within phyla Chloroflexi, Betaproteobacteria, Bacteroidetes, and Synergistetes (Riviere et al., 2009 ). In line with St-Pierre and Wright ( 2014 ), we identified different core OTUs mainly distributed in phylum Firmicutes. This is probably due to different substrates used in anaerobic digesters, which support differential microbial populations in the engineered AD systems (Zhang et al., 2014 ). Indeed, phylum Chloroflexi may predominate in digesters treating municipal wastewater or sewage sludge (Riviere et al., 2009 ; Nelson et al., 2011 ; Sundberg et al., 2013 ), while Firmicutes are dominant in most manure digesters or co-digesters of mixed substrates (Liu et al., 2009 ; Sundberg et al., 2013 ; St-Pierre and Wright, 2014 ). As a result, Clostridium in phylum Firmicutes, which contains various genes encoding cellulose and hemicellulose-digesting enzymes for the degradation of complex plant fibers (Deublein and Steinhauser, 2008 ; Zhu et al., 2011 ), gained dominance in the core OTUs of these digesters. Other core OTUs mainly included genera Turicibacter, Sedimentibacter, Saccharofermentans , and Syntrophomonas . These bacterial populations were recognized as important players in substrate hydrolysis, fermentation, acetogenesis, and syntrophic metabolism (Bosshard et al., 2002 ; Chen et al., 2010 ; Vanwonterghem et al., 2014 ). Due to the combined high abundances of core communities in all the digesters, they may be targets for manipulation of microbial activities to achieve an efficient performance and stability in manure anaerobic digesters. Substrate-unique and abundant OTUs were identified, while a majority of these OTUs were unclassified at genus level. Substrate-unique and abundant OTUs may reflect the variations of manure quality, and the differences in the digestive tracts of rumen and non-rumen animals. This is supported by the fact that cattle-unique and abundant OTUs were exclusively and positively correlated with C:N and COD ( p < 0.05), while swine-unique and abundant OTUs positively correlated with FA and NH 4 + - N ( p < 0.05). The C:N and FA strongly select unique bacterial populations that can be well adapted in these anaerobic digesters. Our findings are based on mesophilic digesters treating cattle and swine manure. When adding more digester samples with different substrates or operational parameters, such as chicken manure and high temperature, the revealed key factors shaping bacterial community structure may change. If one environmental factor outcompetes other factors, it may decouple the relationships between microbial communities and other factors (Rui et al., 2015 ). Indeed, temperature is recognized as a key factor to shift microbial community structure in AD systems (Sundberg et al., 2013 ; De Vrieze et al., 2015 ). Overall, our study revealed that substrate and free ammonia play key roles in determining the bacterial community structure. The selection of different prokaryotic taxa by substrates and free ammonia is likely an important mechanism shaping prokaryotic community structure in manure AD systems. Core communities may be responsible for the central function in AD systems, while substrate-unique and abundant communities may reflect the selection effects largely exerted by substrate quality and operational conditions. These findings provide further understanding of the bacterial assembly in full-scale manure anaerobic digesters. Conflict of interest statement The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest."
} | 4,246 |
38543395 | PMC10976177 | pmc | 1,527 | {
"abstract": "The increasing number of IoT devices has led to more electronic waste production, which harms the environment and human health. Self-powered sensor systems are a solution, but they often use toxic materials. We propose using biocompatible peanut skin as the active material for a self-powered humidity sensor (PSP-SPHS) through integration with a peanut-skin-based triboelectric nanogenerator (PSP-TENG). The PSP-TENG was characterized electrically and showed promising results, including an open circuit voltage (162 V), short circuit current (0.2 µA), and instantaneous power (2.2 mW) at a loading resistance of 20 MΩ. Peanut skin is a great choice for the sensor due to its porous surface, large surface area, eco-friendliness, and affordability. PSP-TENG was further used as a power source for the PSP-humidity sensor. PSP-SPHS worked as a humidity-dependent resistor, whose resistance decreased with increasing relative humidity (%RH), which further resulted in decreasing voltage across the humidity sensor. This proposed PSP-SPHS exhibited a good sensitivity (0.8 V/RH%), fast response/recovery time (4/10 s), along with excellent stability and repeatability, making it a potential candidate for self-powered humidity sensor technology.",
"conclusion": "4. Conclusions In this study, a bio-waste peanut skin powder (PSP) was utilized as the active layer for two distinct devices: a resistive humidity sensor and a triboelectric nanogenerator (TENG). These devices were effectively integrated to showcase a self-powered humidity sensor (SPHS). PSP was characterized by field emission scanning electron microscopy (FESEM) and Fourier transform infrared spectroscopy (FTIR) to deeply understand its surface and material characteristics. A contact separation mode TENG was fabricated using PSP and PTFE thin films as a triboelectric pair. Aluminum (Al) and copper (Cu) tape served as electrodes, and the PET sheet acted as a supporting layer on both sides of the TENG. The PSP-TENG generated an open circuit voltage of 162 V, a short circuit current of 0.2 µA, and a power of 2.2 mW, respectively, at a loading resistance of 20 MΩ. The output voltage of PSP-TENG was measured at various loading resistances and was found to increase proportionally with increasing load resistance. The rectified output of the PSP-TENG was able to charge different capacitors (10 µF to 8 nF) and powered 24 red LEDs. The fabricated PSP-based humidity sensor was resistive in nature and had a sandwich structure. The PSP thin film was sandwiched between Al and Cu electrodes. This proposed PSP-based humidity sensor was a humidity-dependent resistor whose resistance decreased with increasing %RH. The output voltage of the PSP-TENG changed as the resistance of the PSP resistive humidity sensor varied in response to the changes in humidity, utilizing the PSP-TENG as a power source. We successfully demonstrated a humidity-dependent output voltage for the proposed PSP-SPHS. The SPHS exhibited a good sensitivity of 0.8 V/%RH, along with rapid response and recovery times of 4 s and 10 s, respectively. This device also demonstrated excellent stability and repeatability, making it as a promising candidate for SPHS technology.",
"introduction": "1. Introduction Humidity sensors are widely used in several industries, including agriculture, pharmaceuticals, cosmetics, textiles, food, and beverage. They play a significant role in continuous environmental monitoring in the Internet of Things (IoT) industry and smart homes [ 1 , 2 , 3 , 4 , 5 , 6 ]. Furthermore, humidity sensors integrated into e-skin systems contribute significantly to expanding the capabilities of wearable monitoring devices by enhancing sensory perception, improving monitoring accuracy, and ensuring biocompatibility and transparency for diverse applications in healthcare, military, and environmental sectors [ 7 , 8 ]. Typically, humidity sensors rely on replaceable batteries for power, which can be costly and inconvenient to replace. This leads to the accumulation of electronic waste (E-waste), posing environmental burdens. Improper disposal of E-waste results in ecosystem contamination, landfill space depletion, resource wastage, increased energy consumption, and global environmental impacts [ 9 , 10 , 11 ]. Researchers have begun to develop self-powered humidity sensors to cut down the need for external power supplies [ 12 , 13 ]. Presently, several self-powered humidity sensors (SPHSs) exist, including piezoelectric nanogenerators [ 14 ], triboelectric nanogenerators [ 15 ], humidity-induced voltage sources [ 13 ], solar cell integrated humidity sensors [ 16 ], and potentiometric humidity-transduction [ 12 ]-mechanism-based SPHSs. Among these different types of SPHSs, triboelectric nanogenerator (TENG)-based SPHSs received more attention due to their simple device structure, versatility, high sensitivity, scalability, easy fabrication, high performance, and economic feasibility [ 17 , 18 , 19 , 20 ]. Recently, many researchers reported TENG-SPHS, using different types of materials like PTFE [ 21 ], chitosan [ 22 ], reduced graphene oxide (RGO)-polyvinylpyrrolidone (PVP) [ 18 ], tin disulfide nanoflowers/RGO [ 15 ], chitosan/amido-graphene oxide [ 23 ], titanium oxide nanotube arrays [ 24 ], graphene oxide (GO) paper [ 25 ], GO nanoribbons [ 26 ], fluorinated ethylene propylene (FEP)/nylon-66 (PA66) [ 19 ], chitosan/activated carbon [ 27 ], and perfluorosulfonic acid ionomer (PFSA)/fluorinated ethylene propylene (FEP) [ 25 ]. The main objective of promoting SPHSs is to replace conventional sensors that require a constant external power source like replaceable batteries, which leads to E-waste and harms the environment. Alternatively, SPHSs can produce their own power from ambient energy (humidity, mechanical, thermal, etc.) to fulfill their requirements. While the demonstrated SPHSs function effectively, the materials utilized often lack biocompatibility, which remains a significant unresolved concern for us [ 12 , 16 , 22 , 24 , 25 ] due to which biodegradable materials need to be studied for TENG [ 28 , 29 ] and humidity sensors [ 30 , 31 , 32 , 33 ]. Cellulose and lignin, which are biocompatible materials, demonstrate impressive performance in TENGs [ 34 , 35 , 36 , 37 ] and also exhibit good humidity-sensing properties [ 36 , 37 , 38 ]. With the motivation to find a solution, we explored the properties of peanut skin, a bio-waste rich in cellulose (46%) [ 38 ] and lignin (23%) [ 39 ], for use in a SPHS. This study presents a successful demonstration of a biocompatible SPHS integrated with a TENG device by utilizing peanut skin powder as the primary material. The TENG, comprising a biocompatible PSP (tribopositive layer) and PTFE (tribonegative layer), effectively generated a high voltage (162 V) and power (2.2 mW). The resistive humidity sensor based on PSP exhibited notable characteristics, including good sensitivity (0.8 V/%RH), fast response/recovery times (4/10 s), high stability, and good repeatability. These findings established PSP as a promising material for self-powered humidity sensor technology.",
"discussion": "3. Result and Discussion Figure 2 a–e exhibit the SEM images of the PSP film captured at various resolutions, enabling a detailed examination of its microstructure. An analysis of these images revealed the presence of a uniform PSP film characterized by an extensive surface area, pronounced roughness, a multitude of pores/voids with varying dimensions (which significantly contribute to the film’s permeability and absorption capacity), and irregular pathways for charge carrier conduction. These features of PSP proved it to be a highly suitable material for the targeted applications of TENGs and humidity sensors. In the context of TENGs, the described features of a PSP thin film have significant importance due to their optimized charge transfer, promotion of efficient energy conversion, increase in charge separation, and enhancement in power generation capabilities. These characteristics contribute to the overall effectiveness and performance of the nanogenerator in harnessing mechanical energy and converting it into usable electrical energy. This is supported by research on the use of natural materials, such as sunflower husks [ 41 ] and rice paper [ 42 ], in the development of high-performance, biocompatible TENGs for biomechanical energy harvesting. In the context of resistive humidity sensors, the described features of PSP thin films also have a significant impact on their properties. The large surface area of the PSP sensing thin film played a crucial role in enhancing its sensitivity to humidity. This is due to the increased availability of active sites for water vapor adsorption, allowing for improved detection and response to moisture variations. Additionally, the rough surface of the sensing thin film further amplified its effective surface area, promoted greater contact area with the surrounding environment, and enhanced its ability to adsorb water molecules. As a result, the sensitivity and responsiveness to humidity changes were significantly improved. The presence of pores/voids within the sensing thin film also contributed to its performance. These structures enhanced the moisture absorption capacity of the PSP thin film and enabled it to effectively respond to humidity changes. Furthermore, the sensitivity to different humidity levels was enhanced as the pores/voids provided additional sites for water vapor adsorption. The irregular paths for charge conduction within the sensing thin film had a significant impact on its electrical conductivity and response to humidity. These paths influenced the flow of charge carriers, thus affecting the film’s electrical resistance. By measuring the changes in electrical resistance, the sensing thin film can accurately determine the humidity level. Overall, the combination of a large surface area, roughness, pores/voids, and irregular charge conduction pathways contributed to the film’s enhanced sensitivity, responsiveness, and ability to measure humidity levels. The distinctive microstructure of the PSP thin film, as evident from its FESEM images, established its suitability as both a tribopositive layer of a TENG device and a sensing thin film for resistive humidity sensors. The film’s microstructural features, including a large surface area, roughness, numerous pores/voids, and irregular charge conduction pathways, collectively contributed to its functional characteristics and enabled its effective utilization in these applications. The energy-dispersive X-ray spectroscopy (EDS) analysis of the PSP thin film was performed to analyze its elemental composition. EDS analysis revealed the presence of carbon (C) at 57.44%, oxygen (O) at 40.65%, magnesium (Mg) at 0.74%, sulfur (S) at 0.46%, and potassium (K) at 0.71%, as shown in the obtained EDS graph in Figure 2 g. The elemental composition obtained from the EDS analysis was consistent with the expected composition of cellulose [ 43 ], hemicellulose [ 44 ], and lignin [ 45 ], as they are primarily composed of carbon, oxygen, and other trace elements such as Mg, S, and K. Therefore, the EDS analysis results supported the likelihood of the presence of cellulose and lignin in PSP, which are important for its structural and functional properties. The inclusion of cellulose, hemicellulose, and lignin within the sensing thin film of a humidity sensor significantly enhances its moisture adsorption capacity, stability, and durability, rendering it suitable for deployment in sensitive environments and biocompatible applications. The presence of carbon within the film enables efficient charge conduction, which is a critical factor in the resistive humidity sensing mechanism, while the presence of oxygen influences the film’s electrical properties and sensing capabilities. In the context of a triboelectric nanogenerator (TENG), the presence of cellulose, hemicellulose, and lignin within the tribopositive layer contributes to improved triboelectric performance, mechanical integrity, and compatibility with environmentally friendly and biocompatible applications. Additionally, the presence of C in the tribopositive layer enhances electrical conductivity, facilitating efficient charge transport and enhancing the output performance of the nanogenerator. Figure 2 f presents the energy-dispersive X-ray spectroscopy (EDS) mapping of the peanut skin powder (PSP) film. The EDS analysis of PSP confirms the presence of the C K series, O K series, Mg K series, and K K series at a magnification level of 50 µm, as illustrated in Figure 2 h–k. Figure 3 illustrates the FTIR analysis of PSP thin films. The peaks obtained from the FTIR analysis of peanut skin at different wavemubers of 1000, 1200, 1680, 2800, 2900, and 3300 cm −1 were indicative of the presence of cellulose, hemicellulose, and lignin in the skin. The specific peaks corresponded to the characteristic vibrational modes of these components. The peak at 1000–1200 cm −1 was associated with the stretching vibrations of C-O and C-C in cellulose and hemicellulose. These natural plant-based elements were suitable for moisture adsorption in a humidity-sensing thin film. The peak at 1680 cm −1 was attributed to the stretching vibration of C=O in lignin. The presence of this peak indicated the structural stability and durability of the thin film, essential for long-term humidity sensor performance. The peaks at 2800–2900 cm −1 corresponded to the stretching vibrations of C-H in cellulose, hemicellulose, and lignin. These observed peaks are proof of such organic compounds in PSP thin films that were helpful in enhancing the sensing film’s electrical conductivity for the conduction of charge carriers during resistive humidity sensing and TENG. The peak at 3300 cm −1 was related to the O-H stretching vibration in cellulose and hemicellulose [ 46 , 47 ]. The observed peak signified the enhancement of the film’s sensitivity to changes in the humidity levels. The schematic of the electrical characterization setup is shown in Figure 1 c. The output voltage and current of the PSP-TENG were measured by an oscilloscope Keysight (DSOX3014T) and a precision source measurement unit (KEYSIGHT 2911A), respectively. The fabricated device was a contact separation mode triboelectric nanogenerator (TENG) with an effective contact area of 25 cm 2 . The schematic diagram and conduction mechanism of the TENG are depicted in Figure 4 a and Figure 4 b, respectively, where the conduction mechanism is divided into four steps. The schematic diagrams illustrate the different stages of the PSP-TENG operation. In the initial neutral state (first stage), no charge was present on the surfaces. When compressed, positive and negative charges were generated and distributed on the PTFE and peanut skin powder films based on their triboelectric tendency (second stage). Upon release, the separated charges formed a dipole moment, driving the flow of electrons from the bottom Al electrode to the top Cu electrode (third stage). This electron flow continued until the maximum separation between the tribopositive and tribonegative sides was reached. When pressed again, the dipole moment vanished, resulting in a reverse flow of electrons between the electrodes (fourth stage). This electrostatic induction process generated output signals until both surfaces were fully in contact again. The output performances of the PSP-TENG were comprehensively investigated. An energy harvesting assessment was conducted utilizing a linear motor to replicate the cyclic contact separation between two opposing triboelectric surfaces by modifying operational factors like frequency and distance. The voltage output of the TENG was directly gauged using an oscilloscope (DSOX3014T) with an internal load resistance of 10 MΩ. Additionally, the output current was assessed by employing a source measurement unit (Keysight B2911A). The device achieved a peak open circuit voltage of 160 V at a working frequency of 4 Hz and a separation distance of 1 cm at room temperature, as depicted in Figure 4 c. Additionally, it generated a short circuit current of 0.2 A, as shown in Figure 4 d. The relationship between the output voltage of the PSP-TENG and the loading resistance was explored by measuring the output voltage of the PSP-TENG across six different loading resistances, spanning from 100 Ω to 150 MΩ. Figure 4 e shows the output voltage waveforms of the developed device at different loading resistances with the conclusion that the output voltage of the PSP-TENG increased as the loading resistance increased. Figure 4 f shows the dependence of the output voltage of the PSP-TENG on loading resistance. The instantaneous power of the device was calculated by the given Equation (1).\n (1) P = V 2 R Figure 4 g shows the instantaneous power of the PSP-TENG at different loading resistances, with the conclusion that a maximum power of 2.2 mW at a loading resistance of 20 MΩ could be generated. Figure 4 h displays the rectified output voltage waveforms of the PSP-TENG at loading resistances of 4 MΩ and 5 MΩ, showing the positive and negative voltage profiles, respectively. A full-wave rectifier circuit was used to convert the obtained AC waveform of the PSP-TENG into DC voltage for using the developed device in practical applications like charging capacitors and powering different loads, as shown in Figure 5 a. Four different capacitors of 10 µF, 1 µF, 340 nF, and 8 nF capacitance, respectively, were charged by using the developed PSP-TENG. Figure 5 b shows the charging of these four capacitors. The obtained results showed that the 10 µF capacitor took 20 s to charge up to 1 V, the 1 µF capacitor took 15 s to charge up to 3 V, the 340 nF capacitor took 10 s to charge up to 4.3 V, and the 8 nF capacitor took only 2 s to charge up to 4.5 V. Figure 5 c shows the charging/discharging waveform of the 340 nF capacitor. The developed PSP-TENG was capable of charging a 340 nF capacitor to 6 V within a duration of 12 s. Additionally, the PSP-TENG was tested for powering 24 red LEDs, successfully illuminating them using its rectified output voltage. Figure 5 d visually depicts the 24 LEDs powered by the PSP-TENG. Supplementary Material S1 provides a video demonstration of the LEDs being powered by the PSP-TENG. The humidity sensor schematic is shown in Figure 1 b, where the active film of the PSP was sandwiched between the top and bottom electrodes. The resistance of this resistive humidity sensor was proportional to the environmental relative humidity (%RH). A measurement setup was constructed to investigate the humidity sensing properties of the proposed humidity sensor. The schematic diagram of the setup is shown in Figure 1 c. Different standard saturated salt solutions were used to achieve the different levels of %RH. The PSP-TENG acted as the power source of the humidity sensor, and the PSP-SPHS was placed in various saturated salt solution jars to characterize it electrically at different %RH levels. The output voltage across the humidity sensor was measured by an oscilloscope. As the relative humidity (%RH) increased, there was a corresponding decrease in the resistance of the self-powered humidity sensor. This decrease in resistance led to a monotonous decrease in the output voltage across the humidity sensor, in accordance with the voltage divider principle. The relationship between %RH and resistance is such that higher humidity levels result in lower resistance, which, in turn, leads to a decrease in the output voltage of the humidity sensor. This behavior can be explained by the moisture-dependent electrical properties of the sensor material, which exhibits a sensitivity to changes in humidity levels. Figure 6 a depicts the output voltage across the humidity sensor that was placed in different %RH jars. The SPHS device was successfully tested at five different %RH levels (10%, 30%, 60%, 80%, and 90% RH), and the corresponding measured output peak voltages were 70 V, 63 V, 51 V, 33 V, and 7 V, respectively. Figure 6 b shows the fitting line of the average voltage across the humidity sensor at different %RH levels. The response time of the device was tested by placing it in a 10%RH jar, and the voltage across it was constantly measured. It was then transferred to a 90%RH jar, and the time taken for the voltage to reach 90%RH was measured. Similarly, the recovery time was determined by moving the sensor from 90%RH to 10%RH. The proposed SPHS demonstrated a rapid response time of 4 s and a recovery time of 10 s. Figure 6 c illustrates the response and recovery time graphs of the SPHS. Stability is a crucial characteristic of humidity sensors, and thus the stability of the fabricated device was assessed at different %RH levels. Figure 6 d displays the stability test results, indicating that the SPHS maintained a consistent output voltage over an extended period at all %RH levels. The SPHS exhibited favorable performance in terms of response and recovery times while demonstrating excellent long-term stability. Table 1 summarizes the humidity sensing performance of this demonstrated PSP-SPHS with previously reported works. The response time, recovery time, sensitivity, and humidity range of PSP-SPHS were compared with those of previously reported SPHS devices made of tin disulfide nanoflowers/reduced graphene oxide [ 15 ] chitosan and activated carbon [ 27 ], graphene oxide [ 48 ], and Nb 2 CT x /sodium alginate [ 49 ]. The comparison showed that the PSP-SPHS had a fast response/recovery time with a greater sensitivity over a wide humidity range."
} | 5,420 |
23483665 | null | s2 | 1,528 | {
"abstract": "The mechanical holdfast of the mussel, the byssus, is processed at acidic pH yet functions at alkaline pH. Byssi are enriched in Fe"
} | 32 |
35938723 | PMC9426452 | pmc | 1,529 | {
"abstract": "ABSTRACT Bacteria display a remarkable capacity to organize themselves in space and time within biofilms. Traditionally, the spatial organization of biofilms has been dissected vertically; however, biofilms can exhibit complex, temporally structured, two-dimensional radial patterns while spreading on a surface. Kahl and colleagues report a ring pattern that indicates the alternating redox metabolism of P. aeruginosa biofilms under light/dark cycles. Does the presence of a rhythmic, daily phenotype imply a circadian rhythm? Here, we highlight several examples of rhythmic patterns reported in the literature for surface-colonizing multicellular assemblies and discuss the conceptual requirements for proving the presence of a prokaryotic circadian clock behind pattern formation."
} | 196 |
28952565 | PMC5597161 | pmc | 1,530 | {
"abstract": "Metabolic engineering of various industrial microorganisms to produce chemicals, fuels, and drugs has raised interest since it is environmentally friendly, sustainable, and independent of nonrenewable resources. However, microbial metabolism is so complex that only a few metabolic engineering efforts have been able to achieve a satisfactory yield, titer or productivity of the target chemicals for industrial commercialization. In order to overcome this challenge, 13 C Metabolic Flux Analysis ( 13 C-MFA) has been continuously developed and widely applied to rigorously investigate cell metabolism and quantify the carbon flux distribution in central metabolic pathways. In the past decade, many 13 C-MFA studies have been performed in academic labs and biotechnology industries to pinpoint key issues related to microbe-based chemical production. Insightful information about the metabolic rewiring has been provided to guide the development of the appropriate metabolic engineering strategies for improving the biochemical production. In this review, we will introduce the basics of 13 C-MFA and illustrate how 13 C-MFA has been applied via integration with metabolic engineering to identify and tackle the rate-limiting steps in biochemical production for various host microorganisms",
"conclusion": "5. Conclusions Metabolic engineering has been rapidly developed for various industrial microorganisms to produce bulk chemicals, fuels, and drugs from renewable feedstock, in order to free the modern society from the depleting fossil fuel feedstock. However, the complex microbial metabolism is one of the most challenging obstacles for metabolically engineered microorganisms to reach an industrially satisfactory yield, titer or productivity. In order to overcome this challenge, 13 C-MFA has been continuously developed and successfully applied to assist the rational design of metabolic strategies for both model and non-model microorganisms by rigorously quantifying the carbon flux distribution in central metabolism. As shown in this review, multiple issues in biochemical production, such as bottleneck pathways in biochemical synthesis, cofactor imbalance issue in host cells, and energetic requirements in cell maintenance, have been revealed by 13 C-MFA, which guides the development of the appropriate metabolic engineering strategies for successful improvements of the target chemicals to different extents. To further advance 13 C-MFA, several techniques have been recently developed to demystify the secondary metabolism, improve the accuracy and resolution of the flux distribution, and investigate the metabolism for autotrophic organisms as well as the microbial metabolism at the metabolic non-steady state. We believe that, by using 13 C-MFA for various cell factories, metabolic strategies will be more rationally designed and successfully applied to improve the biochemical production.",
"introduction": "1. Introduction Using microorganisms to produce various chemicals from renewable resources could be environmentally friendly and reduce strong dependence on petroleum. Recently, with the rapid development of metabolic engineering and synthetic biology, a wide range of bulk chemicals [ 1 , 2 , 3 , 4 ], biofuels [ 5 , 6 , 7 , 8 , 9 ], and drugs [ 10 , 11 , 12 , 13 , 14 , 15 , 16 ] from renewable feedstock have been produced by many industrial microorganisms such as Escherichia coli [ 17 , 18 , 19 , 20 , 21 , 22 ] and Saccharomyces cerevisiae [ 23 , 24 , 25 , 26 ] . However, only a few of these biosynthesized chemicals are able to be industrially commercialized due to low production levels with unsatisfactory titers, yields, and productivities [ 27 , 28 ]. Therefore, it is pivotal to develop novel strategies in metabolic engineering to improve microbe-based chemical production. One of the main reasons for the low production level of engineered microorganisms is the high complexity of cell metabolism[ 28 ]. Microbial production of chemicals is more than the enzymatic conversion of the precursors to the products. Instead, to achieve the production of target chemicals at a high level, controls over microbial metabolism must coordinate the carbon flux [ 29 , 30 ], cofactor supply [ 31 , 32 , 33 ], cell maintenance [ 10 , 34 , 35 ], as well as other factors [ 36 , 37 , 38 , 39 ]. In general, many of the metabolic engineering strategies adopted to manipulate microbial metabolism for biochemical production only focus on a few known challenges (e.g., poor gene expression). However, attempting to overcome these challenges could result in new problems in host cells (e.g., metabolic burden) and hence prevent the microorganisms from achieving high-level chemical production. The lack of knowledge on such complex behavior of microbial physiology presents one of the most significant issues in improving the microbe-based chemical production. To demystify the complex metabolic rewiring of engineered microorganisms and more importantly, derive the appropriate strategies to engineer microorganisms for better biochemical production, a technology named 13 C-Metabolic Flux Analysis ( 13 C-MFA) has been in development since the 1990s [ 40 , 41 , 42 , 43 , 44 , 45 ]. Basically, in 13 C-MFA, carbon isotopes have been used to trace the cell metabolism. The carbon flux distributions in metabolic network of microorganisms can be determined using computational algorithms with the development of a metabolic model and the measurement of 13 C-labeling patterns of the key metabolites [ 40 , 41 , 46 , 47 , 48 , 49 ]. By comparing variations of metabolic fluxes among different engineered microorganisms, the key issues, such as the bottleneck pathway, could often be revealed and hence guide the metabolic engineers to develop more appropriate strategies [ 35 , 50 , 51 , 52 , 53 , 54 , 55 ] for further improvement of chemical production. In the past decade, we have witnessed that numerous valuable biological insights pinpointed by 13 C-MFA successfully helped to enhance the microbial production of chemicals [ 30 , 34 , 51 , 53 , 56 , 57 ]. Therefore, 13 C-MFA has been widely considered as one of the most important tools to diagnose complex microbial metabolism and develop novel metabolic engineering strategies [ 29 , 30 , 32 , 34 , 35 , 53 ]. In this review, we aim to summarize the integrated tactics of 13 C-MFA and metabolic engineering from cases that attempted to improve microbe-based chemical production in the past decade. We will first briefly introduce the techniques of 13 C-MFA, and then categorize studies that integrated 13 C-MFA with metabolic engineering in different industrial microorganisms, including (1) Saccharomyces cerevisiae ; (2) Escherichia coli ; (3) Bacillus subtilis ; (4) Corynebacterium glutamicum ; and (5) other microorganisms. Finally, we envision the emerging areas where breakthroughs of 13 C-MFA could potentially promote rational metabolic engineering for improved microbe-based chemical production in future."
} | 1,746 |
31601819 | PMC6787259 | pmc | 1,531 | {
"abstract": "Bacterial diversity associated with corals has been studied extensively, however, localization of bacterial associations within the holobiont is still poorly resolved. Here we provide novel insight into the localization of coral-associated microbial aggregates (CAMAs) within tissues of the coral Acropora hyacinthus. In total, 318 and 308 CAMAs were characterized via histological and fluorescent in situ hybridization (FISH) approaches respectively, and shown to be distributed extensively throughout coral tissues collected from five sites in Japan and Australia. The densities of CAMAs within the tissues were negatively correlated with the distance from the coastline (i.e. lowest densities at offshore sites). CAMAs were randomly distributed across the six coral tissue regions investigated. Within each CAMA, bacterial cells had similar morphological characteristics, but bacterial morphologies varied among CAMAs, with at least five distinct types identified. Identifying the location of microorganisms associated with the coral host is a prerequisite for understanding their contributions to fitness. Localization of tissue-specific communities housed within CAMAs is particularly important, as these communities are potentially important contributors to vital metabolic functions of the holobiont.",
"conclusion": "Conclusion Localization of microorganisms associated with corals is vital to understand their symbiotic relationships and reveal their function(s) within the holobiont. Here we provide novel insight into the distributions and densities of bacteria within tissues of the coral Acropora hyacinthus sampled from five different locations. CAMAs were common in coral tissues sampled, although their prevalence were higher in colonies sampled close to shore compared to offshore geographic sites, potentially linked to water quality parameters. While each CAMA appeared to be dominated by a single bacterial morphological type, different CAMA hosted different bacterial morphotypes. CAMAs have been defined as facultative symbionts, not necessary for host fitness 33 , however their high prevalence and abundance in coral tissues may indicate they are integrated into shared metabolic pathways and central to maintaining coral fitness through provisioning benefits. For example, findings that CAMAs are often co-localized near Symbiodiniaceae cells highlights the potential metabolic integrated links between bacterial and Symbiodiniaceae symbionts. Such propositions require testing by tracing metabolic pathways, as well as improved taxonomic and functional assessment of the microorganisms housed within these CAMAs.",
"introduction": "Introduction Scleractinian corals associate with broad consortium of microorganisms, including endosymbiont dinoflagellates (Symbiodiniaceae), protozoa, fungi, bacteria, archaea and viruses, which collectively are termed the coral holobiont 1 – 3 . The importance of symbiotic dinoflagellates in provisioning the coral host with essential nutrients through translocated photosynthates has been established e.g. 4 , 5 , however the roles of other microorganisms within the holobiont are less well understood (reviewed in 3 ). Some of the functions attributed to coral-associated microbiota include supply of essential nutrients and vitamins through processes such as nitrogen fixation 6 – 9 and metabolizing dimethylsulfoniopropionate (DMSP) to produce biologically important byproducts like dimethylsulfide 10 . The coral microbiota is also likely important for directly facilitating disease resistance through production of antimicrobials 11 , 12 or indirectly by preventing colonization of opportunistic or pathogenic organisms 13 , 14 . Corals are considered simple metazoans, but despite their basal phylogenetic position, they nevertheless form complex three-dimensional structures. Anatomically, the coral consists of (1) a surface mucus layer, (2) polyps consisting of feeding tentacles, actinopharynx, mesenteries (including their filaments), and walls (surface and basal), (3) a gastrovascular system that includes gastrovascular cavity (formerly coelenteron) and connecting canals and (4) an external calcium carbonate skeleton 15 , 16 . The coral tissue layers are composed of epithelia, calicodermis (formerly calicoblastic epithelium) and the gastrodermis which also contains the Symbiodinaceae. Within all these microhabitat niches, bacteria can reside as either transient communities or established symbionts with functional roles that may be positive, neutral or negative to the coral holobiont 17 . A multitude of studies have reported on the diversity of the coral-associated microbial communities, in some cases finding conserved microbial communities associated with some coral species, and in others finding shifting microbial profiles that reflect varying geographic, temporal or health status patterns 18 – 21 . To understand the significance of coral microbial associations, care must be exercised so that diversity patterns reflect the specific ecological niche that these communities inhabit, such as the surface mucus layer, tissue layers, and/or the skeleton 22 – 29 . Defining the locations of specific microorganisms is essential for elucidating the importance of their role within the holobiont. For example, mucus bacteria are more likely to have a loose association with the coral host, being sloughed off as the mucus is exuded from the corals 30 . Conversely tissue-associated microorganisms are potentially more integrated in shared metabolic pathways and may reside in specific associations with host coral as a consequence of potential host selection 6 , 21 , 31 . To date, few studies have precisely localized bacterial communities within coral tissues. Studies that have focused on localization often find that bacterial communities within coral cell layers (i.e. epidermal and gastrodermal epithelia) form aggregations termed coral-associated microbial aggregates (CAMAs) 32 – 35 . CAMAs were first reported as potential pathogens when observed within healthy tissues of Caribbean corals displaying signs of white-band disease 36 , 37 . Further studies subsequently reported that CAMAs are widespread in tissues of healthy corals sampled from geographically dispersed areas 33 , 37 , 38 . To date, CAMAs have been reported from 5 species of corals in the Caribbean 36 , 37 and 24 species from the Indo-Pacific 33 , although their frequency varies among coral genera, with the genera Acropora , Porites , and Pocillopora most commonly hosting bacterial aggregates 33 . Identification of the microorganisms that constitute these CAMAs has been poorly resolved. Neave et al . (2016) visualized aggregates (i.e., “cyst-like aggregations”) of Endozoicomonas within tissues of Stylophora pistillata at the interface of the epidermis and gastrodermis and confirmed their distribution in samples taken from widely-separated biogeographic regions 39 , 40 . However, the fine-scale spatial distributions of microorganisms and potential microhabitat-associated structure of the microbiome still have not been clarified. To facilitate a more comprehensive understanding of the roles bacterial communities may play in the coral holobiont, an improved understanding of the localization of these microbial communities is essential, particularly tissue-associated bacterial communities housed within structures termed coral-associated microbial aggregates (CAMAs). Here, we visualize the localization, distribution and morphology of CAMAs associated with the coral Acropora hyacinthus sampled from Sesoko Island, Okinawa, Japan and sites in the northern and central regions of the Great Barrier Reef (GBR) Australia, located along an inshore to offshore gradient.",
"discussion": "Results and Discussion Comparison of CAMA prevalence among geographic locations At the time of field collection, all 48 colonies of A . hyacinthus sampled (one fragment from each colony), from the 5 geographic locations (see Fig. 1 ), appeared visually healthy. This was confirmed by subsequent histological analyses, which found that all tissues displayed normal cell morphology, including no signs of fragmentation, wound repair or necrosis (as per criteria in 41 ). In total, 318 CAMAs were characterized via histology (Fig. 2a ) with CAMAs detected in fragments from 27 of the 48 sampled colonies (~56%). The vast majority stained basophilic (95.9%), compared to only 4.1% staining eosinophilic with the hematoxylin and eosin staining procedure. At the Sesoko Island site, tissues derived from the fragments of all 10 sampled colonies contained CAMAs (Fig. 2c ), whereas at the other four sites, CAMAs were observed only in tissues of some of the coral fragments. For example, CAMAs were clearly visible in 80% of Inner Shelf samples, 30% of Lizard Island, 25% of Outer Shelf, and 40% of Orpheus Island samples (n = 10 samples at all sites, except at the Outer Shelf site where n = 8 samples). In general, the prevalence of CAMAs was significantly higher at the Sesoko Island site than at the Lizard Island, Outer Shelf, and Orpheus Island sites (Fisher’s exact test: p = 0.00099, Fig. 2c , and see more detail of the pairwise test in Suppl. Table S1 ). Figure 1 Map showing the locations of five study sites in two countries. ( a , b ) Sesoko Island (SI) in Okinawa, Japan. ( a , c ) Inner Shelf (IS), Lizard Island (LI) and Outer Shelf (OS) sites in the Northern Great Barrier Reef; and ( a , d ) Orpheus Island (OI) in the central Great Barrier Reef, Australia. Figure 2 Appearance and occurrence of CAMAs in the coral Acropora hyacinthus . ( a ) Numerous CAMAs (indicated by arrows) are visible in a histological section stained by hematoxylin and eosin of a colony from Sesoko, Japan. In the panel, small photo shows close-up of a CAMA located in a mesentery of a polyp. ( b ) Right panel shows two of CAMAs (red) distributed with Symbiodiniaceae (green) in the tentacle (coral tissue: blue) of a sample from the Inner Shelf, GBR using FISH. ( c , d ) Pie diagrams showing the proportion of colonies sampled that contained CAMAs at five sites in Japan and Australia based on the detections of ( c ) HE stain and ( d ) FISH. ( c ) A significantly higher proportion of Sesoko Is. samples contained CAMAs (including basophilic and eosinophilic) than samples collected from other sites (Lizard Is., Outer Shelf northern GBR site, and Orpheus Is). ( d ) In the FISH experiment, both proportions of CAMA in samples from Sesoko Is. and Inner Shelf were also significantly higher than in samples from other three sites (** p < 0001; Fisher’s exact test followed by a Benjamini–Hochberg false discovery rate correction). Scale bars indicate 600 µm ( a ), 50 µm (small panel in a ) and 100 µm ( b ). A total of 308 CAMAs were also identified by FISH in the serial-sectioned tissues derived from the sampled colonies (again observed in 27 fragments from 48 sampled colonies) (Fig. 2b ). CAMAs were observed from all samples derived from Sesoko Island and the Inner Shelf (Fig. 2d ). In contrast CAMAs were observed in 30% of Lizard Island, 25% of Outer Shelf, and 20% of Orpheus Island samples. The prevalence of CAMAs detected by FISH was significantly higher at both Sesoko Island and Inner Shelf than at the other sites (Lizard Island, Outer Shelf, and Orpheus Island sites; Fisher’s exact test: p = 0.0000019, Fig. 2d and see more detail of the pairwise test in Suppl. Table S1 ). The density of basophilic (i.e. stained with hematoxylin) CAMAs in tissues was highest in the Sesoko Island samples (n = 10) with 20.13 ± 17.1 per cm 2 (average and S.E.), compared to 6.78 ± 9.3 per cm 2 for the Inner Shelf (n = 8) samples, 0.92 ± 0.1 per cm 2 at Lizard Island (n = 3) and 3.55 ± 3.3 per cm 2 in Orpheus Island (n = 3) samples (see more detail in Suppl. Table S2 ). Basophilic CAMAs were not detected in samples of the Outer Shelf corals. Eosinophilic (i.e. stained by eosin) CAMAs were only detected in coral fragments derived from three sites, with 2.32, 5.39 ± 6.7, and 1.28 ± 2.8 per cm 2 in Inner Shelf (n = 1), Outer Shelf (n = 3) and Orpheus Island (n = 2) samples, respectively (see more detail in Suppl. Table S2 ). The densities of CAMAs detected by FISH were reflective of the histological patterns with 18.72 ± 12.7 per cm 2 in Sesoko Island (n = 10), 7.90 ± 8.2 per cm 2 in Inner Shelf (n = 10), 1.26 ± 0.6 per cm 2 in Lizard Island (n = 3), 0.90 ± 0.5 per cm 2 in Outer Shelf (n = 2), and 4.48 ± 4.2 per cm 2 in Orpheus Island (n = 2) samples (see more detail in Suppl. Table S2 ). In one sample from Sesoko Island, 48.9 basophilic CAMAs were detected per cm 2 of tissue (41.1 detected by FISH in the serial section), the greatest density of CAMAs observed across all 48 samples investigated. The maximum number observed in a sample from the Outer Shelf region was 10.10 eosinophilic CAMAs per cm 2 . The patterns of CAMAs prevalence derived from histological and FISH analysis confirm that CAMAs commonly occur in healthy tissues of the coral Acropora hyacinthus collected from sites in Japan and Australia separated by more than 40 degrees of latitude. Although CAMAs were not detected in all tissue samples collected, this may reflect limitations in the area of coral tissue that can be surveyed via histological approaches. The presence of CAMAs in a histological section will depend on the tissue sectioned (i.e. the location of the fragment on the colony and on the section from that fragment), and on the scale and orientation of the section. Indeed, our results showed variation of both presence and density of CAMAs, even from samples derived from the same site (see more detail in Suppl. Table S2 ). Therefore, although not all coral tissue samples (and therefore not all colonies) were found to host CAMAs, we cannot exclude the possibility that other tissue areas of the same colony had CAMAs present. Other studies have also reported that CAMAs are common in tissues of many coral species in the Caribbean 36 , 37 , Indo-Pacific and Red Sea 33 , 34 , 38 , 39 . In particular, CAMAs were common in species of Acropora , Porites , and Pocillopora , although often their presence was patchy within a population sample 33 . Interestingly the distance from the coast negatively correlated with the proportion and density of CAMAs within coral tissues across the sampling sites (Table 1 ). A predictive model was constructed which yielded an ordinary negative binomial regression which was found preferable to a zero-inflated model (Vuong test; Voung z-statistic = 22.57413, p -value < 0.001 in the basophilic CAMAs, and Voung z-statistic = 3.120e-6, p < 0.001 in CAMAs detected by FISH). Therefore an increased distance from the coastline influenced negatively the proportion and density of both basophilic CAMAs and CAMAs detected by FISH (Fig. 3a,c ). The density of eosinophilic CAMAs did not follow the negative binominal regression model (Table 1 and Fig. 3b ). Although our study provides only a snapshot of five sites sampled at one time point, it suggests that inshore reef environments may promote the development of CAMAs within coral tissues. The inshore site at Sesoko Island has high nutrient influxes, especially phosphates 42 , 43 . Similarly, nearshore GBR sites are influenced by influxes of dissolved nutrients from terrestrial runoff, with higher concentrations typically found at inshore compared with offshore sites 44 . Temperature fluctuations may also influence CAMA abundance, although differences in seasonal temperature fluctuations are minimal across the northern GBR sites 45 . The coral microbiome community has been shown to shift in response to environmental stressors 30 , 46 – 50 , thus water quality parameters influencing microbiological composition and function 51 could stimulate CAMA abundance. Further studies, particularly of potential links between nutrient levels and CAMA development, are needed to understand what might drive the increased prevalence and density of CAMAs in coral tissues at inshore sites, and to determine if hosting more CAMAs is beneficial to corals or is an indicator of negative impacts on the coral holobiont. Table 1 Results from the generalized linear regression models with negative binominal distribution testing the effect of the distance from coastline on the densities of CAMAs. Estimate SE Wald chi-square (z-value) p -value (z) AIC LR ( X 2 ) p -value (LR) Basophilic CAMAs Intercept 3.53176 0.399923 8.846 <2e-16*** 187.52 67.427 2.2e-16*** Distance † −0.17477 0.02769 −6.312 2.76e-10*** Eosinophilic CAMAs Intercept −3.508 1.3253 −2.647 0.00812** 46.991 Distance † 0.077 0.0459 1.678 0.09643 CAMAs detected by FISH Intercept 3.19428 0.35556 8.984 <2e-16*** 212.09 51.662 6.594e-13*** Distance † −0.12714 0.02092 −6.077 1.22e-09*** † The distance indicates distance to closest coastline. p- values are ** p < 0.01 and *** p < 0.001. AIC and LR indicates Akaike’s Information Criterion and likelihood-ratio. Figure 3 Relationship between densities of CAMAs and distance to the coast based on a negative binominal regression model. ( a ) Density of basophilic CAMAs was negatively correlated with the distance from the coast to the offshore sampling sites. ( b ) No correlation of eosinophilic CAMAs were detected in relation to the distance to the coast, although higher density (max. 10.1 cm 2 ) was found in the Outer Shelf (offshore). ( c ) Density of the CAMAs detected by FISH was also negatively associated with the distance. Model prediction and associated 95% confidence intervals were obtained from a negative binominal regression model between the densities of CAMAs and the distance to the coastline ( a , c ). Scale bars indicate 10 µm. Distribution of CAMAs within anatomical regions of the coral polyp CAMAs were localized (by FISH) within the six anatomical regions of the coral polyp: the tentacle, actinopharynx, mesentery, mesenterial filament, coenenchyme, and calicodermis (see Fig. 4a ). The CAMAs appeared to be randomly distributed across the anatomical regions investigated, though because the numbers of CAMAs characterized for the Lizard Island, Outer Shelf and Orpheus Island samples were low (Fig. 4b ), meaningful comparisons can only be made between the Sesoko Island and Northern Inner Shelf samples. For the Sesoko Island samples, CAMAs were found predominantly in the tentacles (36.5%), mesenterial filaments (34.7%) and coenenchyme (16.5%) regions; in Inner Shelf samples from Northern GBR sites, CAMAs were mostly located in the calicodermis (46.8%), tentacles (17.5%) and coenenchyme (16.7%) (Fig. 4b ). An additional 31 CAMA-like shaped structures were observed from three samples derived from Sesoko Island, but these were discounted as bacterial aggregates due to non-probe binding signals and excluded from the analysis. Figure 4 Distribution of CAMAs within six anatomical regions of the coral Acropora hyacinthus collected from five sites in Japan and Australia. ( a ) Schematic drawing of a coral polyp showing the six anatomical regions examined microscopically (same color coding used in ( b , c ). ( b ) Pie charts showing the distribution of CAMAs among six anatomical regions in samples from the five sites. ( c ) Dot plots comparing the size of CAMAs among anatomical regions for each location. N = 308 CAMAs detected in tissues treated with FISH. In the dot plot, the lines indicate median (thick line), 3 rd (upper) and 1 st (lower) quartiles (long thin lines), whiskers (short thin lines; upper value indicates largest value within 1.5* interquartile range from 3 rd quartile, lower: min value). CAMAs spanned a wide size range, from 23 to 6,761 µm 2 across tissue samples from all sites (Fig. 4c ). Measurements likely underestimated the size of CAMAs, given they are dependent on the orientation of sectioning and it is unlikely that most CAMAs were sectioned through their greatest diameter. Acknowledging constraints associated with sectioning, the average size of CAMAs was 1,304 µm 2 . Sesoko Island samples contained generally larger average and median-sized aggregates (1,507.7 ± 1,522.4 µm 2 , 921.3 µm 2 respectively) than samples from the GBR region (Inner Shelf: 637.4 ± 734 µm 2 , 321.6 µm 2 ). However, given the large range in sizes measured, the fact that only two locations had sufficient sample sizes for comparison, the issue of sectioning orientation potentially biasing size measurements, and high variation among the individual colony level (Suppl. Fig. S1 ), such patterns require further validation. No patterns in the size of CAMAs across different anatomical regions were detected (Fig. 4c ). Morphology of CAMAs within coral tissues High resolution imaging was used to partially characterize and compare the morphology of bacteria across the CAMAs detected (Fig. 5 ). Interestingly, each CAMA appeared to be composed of a single morphological type of bacteria, although morphological types varied among CAMAs. Overall, five different morphological types of bacteria were identified: rod-shaped (length 2.5 ± 0.1 µm, width 0.6 ± 0.0 µm, Fig. 5a,e ), an atypical coccus (length 4.8 ± 0.3 µm, width 3.1 ± 0.2 µm, Fig. 5b,f ), a longer rod morphology (length 8.0 ± 0.0 µm, width 0.8 ± 0.0 µm, Fig. 5c,g ), filamentous-like bacteria (length N.D., width 0.4 ± 0.0 µm, Fig. 5d,h ), and a rod-shaped morphology but with spore-like structures (Fig. 5i,j,m,n ). While the consistency of morphological characteristics within each CAMA may indicate that CAMAs are hosting single bacterial types, it is also possible that they host multiple bacterial species with similar morphologies. Interestingly, the fluorescent signal detected for some CAMAs was not uniform over the entire aggregation. The lack of signal within some CAMAs (see Fig. 5d for example) may be due to the probe not targeting microbial cells that inhabit that space. Defining a specific morphological shape for the bacterial cells observed was sometimes difficult, with patchy probe hybridization producing images of structures that potentially protruded from the tissue sections and seemed amorphous (Fig. 5k,l,o,p ). However, this again could be the result of mixed microbial communities within the CAMAs, with some cells targeted by the probes, but others not hybridizing to the probe-fluorochrome conjugate. Figure 5 Morphological variation of bacteria housed within different CAMAs. Note that bacteria within each CAMA are morphologically similar. Dotted lines (in a – d , i – l ) delineate regions magnified in close-up images ( e – h , m – p ). Overall, five bacterial morphologies were detected: rod-like ( a , e ), pleomorphic ( b , f ), long rods ( c , g ), filamentous-like ( d , h ), rod-shaped with spore-like structures ( i , j , m , n ), and putative amorphous masses ( k , l , o , p ). Scale bars indicate 10 µm ( a – d , i – l ) and 5 µm ( e – h , m – p ). The variation in bacterial cell morphology among CAMAs may contribute to the different HE staining properties observed, with the majority being basophilic (95.9%), but 4.1% staining eosinophilic (n = 318 CAMAs detected in total). Variability in the staining of CAMAs has been reported previously 33 , 52 , and attributed to varying degrees of protein and DNA production or local tissue pH conditions 33 . However, these staining patterns may also be influenced by different bacteria housed within the CAMAs. Previous studies have reported that the abundant coral-associated bacterial genera Endozoicomonas forms aggregates within tissues of the corals Stylophora pistillata and Pocillopora verrucosa 39 , 40 . Indeed, extensive sequence-based phylogenetic surveys of coral microbiomes have revealed that several dominant bacterial groups are common 21 , including the Proteobacteria (particularly Alpha - and Gammaproteobacteria inclusive of Endozoicomonas ), as well as Actinobacteria , Bacteroidetes (especially Flavobacteria ), and Cyanobacteria 3 . Hence some of the different cellular morphologies we detected within CAMAs may represent common coral-associated bacterial groups profiled in microbiome diversity studies. We note, however, that bacterial morphology can be plastic and dependent on numerous variables, such as division stage, colonization, chemical environment, physical constraints and nutrient availability 53 . Future work targeting CAMAs with taxa-specific probes and gene-sequencing approaches would help to resolve both the taxonomy, composition and potential functional roles of bacteria within these structures of the coral holobiont. FISH imaging showed that for some of the CAMAs, patchy probe-specific labeling occurred, resulting in dull or dark areas within the structures. There are a number of potential methodological reasons for this observation, including: (1) insufficient probe sensitivity due to low ribosomal rRNA content in target cells 54 , 55 , (2) methodological and environmental factors that prevent probes from accessing target cellular rRNA at these sites 56 , (3) the bacterial community penetrating and proliferating within the epidermis of coral tissues 38 , or (4) lipid or fat solvents deposited through dehydration and dewaxing steps showing empty spaces in the tissues 57 . Alternatively, little to no signal in central regions of some of the CAMAs could indicate that the probe EUB338mix did not target the taxonomic group of microbes within these regions. The EUB338mix is estimated to cover 96% of the Eubacteria domain, but taxa outside this coverage, including archaeal lineages, may be present 58 . We speculate that some of the CAMAs may be mixed communities containing bacteria not targeted by the probes or even Archaea , which have been identified to associate with corals in microbiome diversity studies 59 . A recent coral metagenomic study recovered Thermarchaeota genome bins from a Porites sp. that was potentially metabolically linked through nitrogen cycling to other coral microbial-associated taxa, including Nitrospira 6 . Co-aggregation of ammonia-oxidizing archaea with nitrite-oxidizing bacteria is common in other organisms, such as sponges 60 , 61 . Further three-dimensional reconstructions of z-stacked FISH images of select 100 µm stained sections of tentacles visualized the CAMAs as typically spheroid or ellipsoid-shaped structures (Fig. 6a ). In one example, a single CAMA was located independently in the epidermis of a tentacle (Fig. 6b ). In a different tentacle, multiple smaller CAMAs were localized close to Symbiodiniaceae cells in the gastrodermis (Fig. 6c ). Sizes of the large single CAMA and the multiple smaller CAMAs were 33,400 µm 3 (Fig. 6d , surface area 7,729 µm 2 ) and 1,978 ± 141.2 µm 3 (Fig. 6e , 809.5 ± 59.9 µm 2 , n = 4 CAMAs), respectively. Even though the CAMAs were located in the same polyp, the size of the single large CAMA was approximately 40-fold greater than the multiple smaller CAMAs, demonstrating inherent size variability for these structures. The bacteria within these CAMAs displayed a similar rod-shaped morphology (see Fig. 5a,e ), with the average cross-sectional area of each bacterium being 175.7 µm 2 and 22.6 ± 4.2 µm 2 , respectively (Suppl. Fig. S2 ). Based on cell size, the number of individual rod-shaped bacterial cells within the 3D rendered images of these CAMAs was estimated as ~47,275 cells for the large single CAMA located in the epithelium of the tentacle (Figs 6d and S2a ), and 2,799 ± 200 cells (Figs 6e and S2b ) for each of the smaller CAMAs localized close to Symbiodiniaceae cells. Figure 6 Three-dimensional (3D) images of CAMAs (red) within a tentacle of the coral Acropora hyacinthus as visualized using FISH. ( a ) Section of tentacle showing localization of two types of CAMAs (10x magnification). ( b ) Single aggregation of bacteria in a large structure within the epithelium (40x magnification; composed of 92 z-stack images). ( c ) Multiple aggregations of bacteria in smaller structures within the gastrodermis (40x magnification; composed of 56 z-stack images). 3D rendering of CAMAs ( d , e ) reconstructed from 3D images in ( b , c) . Coral tissue and Symbiodiniaceae appear as blue and green structures, respectively. Using these estimates of cell numbers within CAMAs we further extrapolated that bacterial densities in tissues from Sesoko Island and Inner Shelf GBR corals are approximately 5.6 × 10 4 and 1.5 × 10 4 cells per cm 2 (along a linear cross-section), respectively. Very few studies have accurately determined bacterial cell densities associated with corals. Counts for bacteria in the coral mucus layer can be as high as 10 7 cells ml −1 62 . Estimates of bacteria on tissue surfaces range from 1 × 10 5 to 10 6 cells per cm 2 for the coral Pocillopora damicornis 63 and 8.3 × 10 6 to 6.2 × 10 7 cells per cm 2 for the coral Oculina patagonica 25 . However, these counts are based on bacteria external to coral tissues and therefore not directly comparable to estimates from our 3D reconstructions of CAMAs visualized within coral tissues. Our results highlight that tissue-associated microbial communities are present across the different anatomical regions of the coral polyp and that differentiating these communities from external and mucus-associated microbial communities is important for accurate appraisals of the coral microbiome and for identifying their role(s) within the coral holobiont."
} | 7,361 |
20233076 | PMC3061588 | pmc | 1,532 | {
"abstract": "Interactions among individuals in social groups lead to the emergence of collective behaviour at large scales by means of multiplicative non-linear effects. Group foraging, nest building and task allocation are just some well-known examples present in social insects. However the precise mechanisms at the individual level that trigger and amplify social phenomena are not fully understood. Here we show evidence of complex dynamics in groups of the termite, Cornitermes cumulans (Kollar) (Isoptera: Termitidae), of different sizes and qualitatively compare the behaviour observed with that exhibited by agent-based computer models. It is then concluded that certain aspects of social behaviour in insects have a universal basis common to interconnected systems and that this may be useful for understanding the temporal dynamics of systems displaying social behaviour in general.",
"introduction": "Introduction Is there a minimal number of individuals necessary for a given trait of social behaviour to appear? Why do individuals present apparently chaotic and unpredictable activity, while the groups they belong to present ordered patterns? How do interactions turn uncorrelated and disordered individual behaviour into more ordered and coherent collective behaviour? Why does temporal activity in social groups seem to have self-similar patterns? In recent years, there is a renewed interest for addressing these kind of questions coming out from a more basic yet unanswered problem in evolutionary biology: how does collective social behaviour emerge out of cooperating individuals interacting locally? While these questions are general for all social organisms, termites present a particular set of well known behaviours, ranging from very simple self-grooming to extremely elaborate collective nest building. How such behaviours are triggered and controlled is not yet fully understood, but there is growing evidence that common general principles are in action: termite collective patterns emerge spontaneously from the concurrent action of simple individuals, performing simple local tasks with no information whatsoever on the global pattern to be achieved. When an insect colony is viewed as a complex sytem, “[...] common principles exist at the organismic and superorganismic levels, thus between individual insects and the tightly integrated colonies they compose” ( Wilson & Holldobler 2005 ). Several previous works have pointed to the emergence of collective behaviours in several organisms including termites. Recent examples range from behaviour amplification and mass recruitment in ants ( Jeanson et al. 2004 ; Dussutour et al. 2005 ), self-organised temporal synchronisation in ants ( Miramontes et al. 2001 ), colony-size influence on individual performance in wasps ( Karsai et al. 1998 ), nest construction in spiders ( Bourjot et al. 2003 ), self-organised aggregation in cockroaches ( Jeanson et al. 2005 ), synchronization in fireflies ( Camazine et al., 2001 ), among many others. In termites there are a number of complex behaviours that have been studied using agent-based models. Examples include self-organised nest construction ( Deneubourg 1977 ; Courtois et al. 1991 ; Bonabeau et al. 1997 ; O'Toole et al. 1999 , 2003 ), social facilitated survival ( Miramontes and DeSouza 1996 ), disease transmission ( Pie et al. 2004 ) and individual recognition ( Copren et al. 2005 ). Agent-based models have proved to be powerful tools for exploring aspects of these questions, because they incorporate basic rules of individual behaviour capturing the essence of the problem being explored. Global behaviour on these models are then accurate descriptions of their biological counterpart. So, by constructing simulations with very large number of parallel interacting agents, social biologists are now in better position to obtain relevant answers. Mobile Cellular Automata (MCA) is a class of agent-based models very suited to study the emergence of collective behaviours out of individual interactions ( Miramontes 1993 ). That is, to test whether a given global pattern may emerge from interindividual contacts among termite workers, a MCA model is build out of real-world parameters and compared to live termites. Convergence of results is interpreted as an evidence that the basic factors important to the production of the social behaviour have, in fact, been isolated ( Cole and Cheshire 1996 ). In this paper, we explore and compare the arising of coherent patterns of temporal activity in groups of termites and MCA of different sizes, arguing that such a convergence of results may undercover an universal property of social systems. Let us remember that MCA models are complex systems by their own right. The experiment was aimed to observe the essential traits of interacting behaviour among individuals that would lead to the emergence of social behaviour once the groups are sizable. As interactions between individuals increase in frequency, collective behaviour arises, and that is of course expected to be favoured by group size. Spontaneous activations should, therefore, tend to be dumped by induced activations in such a way that the number of spontaneous activations would decrease as the number of induced activations increases, both as a function of group size. Moreover, if collective behaviours are not a simple function of the superposition of the individual contributions (that is, collective patterns ‘emerge’ from the interindividual interactions), then decreasing spontaneous activations and increasing induced activations should obey a non-linear pattern. We, therefore, expect to find a non-linear increase in the number of induced activations among individuals as group size increases, a non-linear decrease in the number of spontaneous interactions, a non-linear initial increase in the Kolmogorov-Shannon entropy and a subtle self-similar structure in the temporal dynamics of activity ( Miramontes 1995 ; Solé et al. 1995 ).",
"discussion": "Discussion Complex behaviour has been identified in social insects. It is now firmly established that isolated ant workers, for instance, may show temporal patterns of activations and locomotion characterised by deterministic chaos ( Cole 1991a ). While these patterns of disordered behaviour are present at the individual level, their complete colonies exhibit periodic cycles of ordered activity. Furthermore, it has been shown that as the group size of workers vary, its temporal dynamics varies accordingly, from disordered states into periodic synchronisation ( Cole 1991b ; Cole 1992 ). Using these behavioural traits, MCA computer models have helped to elucidate that such changes in the collective behaviour are the outcome of interactions alone in ants ( Miramontes 1995 ; Miramontes & Rohani 2001 ; Solé et al. 1993 ) and in termites ( Miramontes and DeSouza 1996 ), and that the transition from disorder into order is indeed what physicists call a phase transition characterised by critical fluctuations (1/ f dynamics), where a number of information measures such as system entropy are maximised ( Miramontes 1995 ). The phase transition has the density of the nest as its control parameter. Lately it has been shown that real ants self-organise to attain density values that seems to correspond to those in the range predicted by computer models. That would pose the nest as the verge of such a phase transition ( Miramontes 1995 ; Solé et al. 1995 ) In termites there is an effort to study how group size influences and initiate certain traits of collective behaviour, mainly by addressing the role of social facilitation in survival, poisoning resistance, disease transmission and reproduction ( Miramonteset al.1996 ; DeSouza et al. 2001 ; DeSouza and Miramontes 2004 ; Santos et al. 2004 ; Muradian et al. 1999 ; Rosengaus al. 1998 ; Rosengaus al.2000 ; Pie el al.2004 ; Costa-Leonardo et al . 1999; Brent and Trainello 2001 ). Here we explored the interplay between group size and the temporal dynamics of spontaneous and induced activations, along with some complexity measures such as the group entropy and self-similarity in temporal patterns. In our laboratory experiment it was observed that the increase in group size translates into a non-linear decrease of the spontaneous activations, the same qualitative pattern that was observed in the computer simulations. The increase in group size had the effect of increasing the number of induced activations, the same pattern in the model and finally, the increase of the group size translated into an increase of entropy, both in the experiment and the model. We did not try to parametrize the model, that is to evaluate precise values for the activation probaility pa or the gain porameter g for two reasons: first, it is known that the MCA model is quite roboust in its dynamic behaviour for a range of the values of these two parameters and, second, we are interested in a qualitative comparison of generic dynamical patterns between experiments, observations and models. While we acknowledge that evaluating exact parameter values may be of interest, it certainly goes beyond the scope of the present work. Figure 4. Induced activations increase along with the increase of the group size. This result is in agreement with the MCA model and is the result of a process of increasing social contacts. Notice the non-linear increasing pattern. The line is a guide to the eye only. Figure 5. Entropy increase non-linearly along with the increase of the group size. This result is in agreement with the MCA model and is the result of a process of increasing social contacts. The line is a guide to the eye only. Figure 6. Iterated Function Analysis (IFS) of the dynamics of temporal activations. The IFS analysis works in the following way: the data set is sorted from the minimum to the maximum value and then subdivided into four segments such that each segment contains the same number of points (notice that the segments could be of different lengths). The original unsorted data set is then normalised and coarse-grained into four values, say 1,2,3 and 4, representing the quartile to where the data belong. The representation space is a square where the four corners are labelled 1,3,2,4 in a clockwise direction (starting in the lower left corner). Each value of the coarse-grained series is associated with the corner having the same number. A point is plotted half the way between the centre of the square and the first point of the series. A second point is plotted halfway between the first plotted point and the second point in the series, and so on. Results are shown of analysing the temporal patterns of group sizes of 8 (a), 16(b), 20(c) and 30(d). The time series are not large enough to form full-developed self-similar patterns in the IFS, nevertheless a comparison with MCA long series (e,f) shows that the patterns have similarities and so termite social activity have subtle temporal structure worth exploring in future research. MCA series were produced with 2000 long time-series. Density was low in (e) and high in (f). Our results seem to point towards the existence of the same non-linear trends arising in other social insects. The use of agent-based models provide evidence that this may be the case. More detailed termite experiments are needed to establish this apparently universal property of social systems: non-linear trends in the emergence of sociality and the potential existence of order-disorder phase transitions with the group-size as the order parameter. It seems that our data is consistent with self-similar 1/ f dynamics because the observed patterns are similar to those obtained with the MCA ( Figure 6 ) and are also strikingly similar with those found in ants ( Miramontes et al. 2001 ). However we do not conclude anything further because more data are needed. We hope that this work may serve as inspiration for a more complete exploration of this intriguing phenomenon."
} | 2,996 |
21932253 | null | s2 | 1,533 | {
"abstract": "Succinate has been recognized as an important platform chemical that can be produced from biomass. While a number of organisms are capable of succinate production naturally, this review focuses on the engineering of Escherichia coli for the production of four-carbon dicarboxylic acid. Important features of a succinate production system are to achieve an optimal balance of reducing equivalents generated by consumption of the feedstock, while maximizing the amount of carbon channeled into the product. Aerobic and anaerobic production strains have been developed and applied to production from glucose and other abundant carbon sources. Metabolic engineering methods and strain evolution have been used and supplemented by the recent application of systems biology and in silico modeling tools to construct optimal production strains. The metabolic capacity of the production strain, the requirement for efficient recovery of succinate, and the reliability of the performance under scaleup are important in the overall process. The costs of the overall biorefinery-compatible process will determine the economic commercialization of succinate and its impact in larger chemical markets."
} | 297 |
36501708 | PMC9739927 | pmc | 1,534 | {
"abstract": "Ionic conductive hydrogels used as flexible wearable sensor devices have attracted considerable attention because of their easy preparation, biocompatibility, and macro/micro mechanosensitive properties. However, developing an integrated conductive hydrogel that combines high mechanical stability, strong adhesion, and excellent mechanosensitive properties to meet practical requirements remains a great challenge owing to the incompatibility of properties. Herein, we prepare a multifunctional ionic conductive hydrogel by introducing high-modulus bacterial cellulose (BC) to form the skeleton of double networks, which exhibit great mechanical properties in both tensile (83.4 kPa, 1235.9% strain) and compressive (207.2 kPa, 79.9% strain) stress–strain tests. Besides, the fabricated hydrogels containing high-concentration Ca 2+ show excellent anti-freezing (high ionic conductivities of 1.92 and 0.36 S/m at room temperature and −35 °C, respectively) properties. Furthermore, the sensing mechanism based on the conductive units and applied voltage are investigated to the benefit of the practical applications of prepared hydrogels. Therefore, the designed and fabricated hydrogels provide a novel strategy and can serve as candidates in the fields of sensors, ionic skins, and soft robots.",
"conclusion": "4. Conclusions In summary, a facile strategy was employed to prepare the multiple-stimuli-responsive hydrogels, specifically employing acrylic acid (AA), 2-acrylamide-2-methylpropane sulfonic acid (AMPS), and high-modulus bacterial cellulose (BC) as a skeleton of a double network, and introducing tannic acid (TA) as adhesion units and anhydrous calcium chloride (CaCl 2 ) as anti-freezing components. Under the UV initiation, a dual-network ionic conductive hydrogel with excellent adhesion, anti-freezing, and moisturizing functions and high stretchability was prepared by radical polymerization. The hydrogel showed excellent mechanical properties (83.4 kPa and 1236%) because of the metal-carboxyl coordination bonds and the double network structure. Moreover, by introducing the high concentration of Ca 2+ ions, the hydrogel exhibited great anti-freezing performance from −33 °C to 0 °C, excellent moisture-retention performance at room temperature (70 wt%, 120 h), and stable ionic conductivity (1.92 S/m). In addition, the hydrogel can be tightly adhered to the surface of the human body because of the unique structure of TA, and can sensitively detect human motion, physiological signals, and the strains in the range of 2.5–900%. Interestingly, the RRC signals of the hydrogel showed a linear correlation with the tensile strain and bending angle during both continuous stretching and bending. Besides, the experiment on the sensing performance under different concentrations of Ca 2+ and different voltages further demonstrated that the strain-sensing ability of ionic conductive hydrogels was mainly related to the ion migration rate, rather than the conductivity or ion concentration. To summarize, the hydrogel exhibits robust and repeatable adhesion, excellent anti-freezing and moisturizing properties, and stable mechanical sensing properties, which are expected to be applied in the fields of a new generation of electronic-skin soft robots and human motion detection.",
"introduction": "1. Introduction As a soft material composed of liquid and solid phases, hydrogel has excellent flexibility, biocompatibility, and sensing capability [ 1 , 2 , 3 , 4 ]. Over the last few decades, using hydrogel as smart wearable devices in the field of intelligent medical and health monitoring has attracted considerable attention [ 5 , 6 , 7 , 8 ]. In general, traditional conductive hydrogels always contain conductive fillers such as metal-based nanowires, conductive polymers, and carbon-based nanofillers [ 9 , 10 , 11 ]. However, those hydrogels are complex to prepare, and the incompatibility between gels and fillers severely limits their properties and practical applications. Compared to traditional electronically conductive hydrogels, the conduction of ionic conductive hydrogels is mainly due to free-moving ions, which avoid the problems mentioned above [ 12 , 13 , 14 , 15 ]. In addition, they are usually transparent and have a similar conductive mechanism to that of the human tissues, giving them potential applications in the field of smart wearable devices such as sensory skins, human motion sensors, and personal healthcare diagnoses [ 16 , 17 , 18 ]. Zhang et al. [ 19 ] prepared ionic conductive hydrogels containing bacterial cellulose (BC), and the novel ion channels endow the hydrogels with stable and excellent sensing properties. Gupta et al. [ 20 ] fabricated silver-nanoparticle-loaded bacterial cellulose hydrogels for a potential wound-dressing application that can be directly attached to the skin surface. However, as a smart wearable material that can exhibit an effective strain-sensing and convert it into electrical signals, the ability of the hydrogels to deform with human skin in real time is essential. Tannic acid (TA) can be an efficient gelation binder for hydrogels since it has a molecular structure similar to that of mussels. For instance, Fan et al. [ 21 ] prepared dual-cross-linked single-network hydrogels with versatile adhesiveness by introducing TA as binders. Shao et al. [ 22 ] fabricated cellulose nanocomposite tough hydrogels with durable and repeatable adhesiveness and the authors ascribed it to the presence of catechol groups from the incorporated TA. Besides, in some harsh environments, such as sub-zero and arid ones, hydrogels are usually damaged by dehydration or freezing, resulting in the loss of their functionality [ 23 ]. Therefore, it is also necessary to enhance the anti-freezing and moisturizing properties of hydrogels [ 24 , 25 ]. Lately, numerous studies focused on the preparation of hydrogels to resist freezing and moisturizing have been reported. He et al. [ 26 ] prepared conductive hydrogels that exhibited excellent anti-freezing and moisturizing properties by introducing a water–glycerol dispersion medium. Morelle et al. [ 27 ] reported anti-freezing hydrogels with tough polyacrylamide–alginate double networks containing Ca 2+ ions. Nevertheless, among the studies of transparent conductive hydrogels that have been reported, most of them paid tremendous attention to the response and recognition of various mechanical signals, as well as to the exploration of multifunctions. Besides, those studies only concentrated on the stable conductivity, high sensitivity, and microstructure of ionic conductive hydrogels, without focusing on the specific factors related to their sensing performance and the mechanism of strain sensing. Hence, it is urgently needed to design and prepare transparent multifunctional/multi-response conductive hydrogels and deeply investigate the factors affecting their sensing performance to meet practical requirements. Herein, we prepared a new hydrogel with high stretchability, excellent adhesion, and stable mechanical sensing properties in freezing and arid environments, which consisted of TA@BC, acrylic acid (AA), and 2-acrylamide-2-methylpropane sulfonic acid (AMPS). TA-coated BC improved its uniform distribution in the hydrogel, and natural polyphenol TA formed hydrogen bonds with polymer that also toughen the polymer network [ 28 ]. Moreover, a large number of catechol groups supplied by TA can bind to both inorganic and organic surfaces through the formation of reversible noncovalent or irreversible covalent interactions [ 21 , 29 , 30 ]. The ionic conductive hydrogels can adapt to different deformations (including stretching and compression) and present a stable response to various stimuli through relative resistance changes (RRCs) and the hydrogel can maintain high ionic conductivities of 0.36 S/m at −35 °C. When applicated as a biosensor, the prepared hydrogel not only exhibits an excellent human-motion detection function but also has a sensitive response to micro-deformations such as pulse and writing. Additionally, to promote practical applications of hydrogels, we analyzed the sensing performance of the hydrogels under different tensile strains and bending strains, and explored the effects of two main factors (metal ion concentration and applied voltage). In summary, the prepared hydrogels exhibited excellent mechanical properties, great sensing performance, and strong adhesion, providing a novel strategy for the development of next-generation smart materials, including ionic skin, health monitoring, soft robots, etc. The exploration also provides a new prospect for the practical applications of ionic conductive hydrogels.",
"discussion": "3. Results and Discussion 3.1. Design and Fabrication of Hydrogel With good biocompatibility, high sensitivity, and stable ionic conductivity, ionic conductive hydrogels are considered as potential candidates for smart wearable devices, such as sensory skins, human motion sensors, and personal healthcare diagnoses. However, poor mechanical properties limit their applications. To solve this problem, in this study, BC and P(AA-AMPS) polymer chains with excellent biocompatibility were used to form a double network hydrogel, and the hydrogel with physical and chemical cross-linking can be easily prepared using a one-pot method ( Figure 1 a). Briefly, AA, AMPS, N, N -methylene bisacrylamide (BIS), and CaCl 2 were added to the prepared TA@BC suspension. Under UV irradiation (365 nm, 20 W) for 0.5 h, the hydrogel was then fabricated through radical polymerization, where the name of the gel is defined as P(AA-AMPS)-TA@BA x -Ca 2+ and x represents the solid content of TA@BC. The specific components of the hydrogel are shown in Table S1 . In this work, BC and P(AA-AMPS) polymer chains formed a double network structure to enhance the mechanical properties of gels. TA not only acted as adhesion units to endow the material with good adhesion properties but also coated the BC surface to improve the dispersion [ 22 ]. The high concentration of Ca 2+ ions enhanced the hydrogels’ strength through the formation of metal-carboxyl coordination bonds and also endowed the hydrogels with stable ionic conductivity. Moreover, the hydration between the Ca 2+ ions and water could enhance the moisture retention and freezing resistance of the gels [ 31 , 32 ]. In the AFM and SEM images ( Figure 1 b,c), the fiber of the TA@BC surface became rougher and appeared to have a smaller aspect ratio after TA coating [ 22 ], and the thickness of the TA@BC reached 31.9 nm in the thickness curve, which was 2.5 times that of the uncoated BC (12.9 nm). Besides, a large number of slightly smaller and uniform porous structures could be observed in P(AA-AMPS) after freeze drying ( Figure 1 b). After adding TA@BC, the double network structure was formed by TA@BC and P(AA-AMPS), which enlarged the pores inside of hydrogel. Meanwhile, flocculent fibers can be clearly observed in the pores. However, after adding a high concentration of Ca 2+ , there is only a small and sparse porous structure on the surface of P(AA-AMPS)-TA@BC-Ca 2+ . The reason was that the high concentration of Ca 2+ ions reduced the freezing point of water in the hydrogel during the freezing process since it formed smaller ice crystals. The uniform distribution of S and N elements in the prepared hydrogels can be observed in EDS ( Figure S1 ), which showed the uniform distribution of P(AA-AMPS) polymer chains in the hydrogel with excellent physical properties. In addition, the uniform distribution of the Ca element indicated that Ca 2+ ions could easily pass through the porous structure of the hydrogels, ensuring the stable conductivity of the gels. The Fourier transform infrared spectrometer (FT-IR) spectra of AA, AMPS, BC, TA, TA@BC, P(AA-AMPS), P(AA-AMPS)-TA@BC, and P(AA-AMPS)-TA@BC-Ca 2+ are presented in Figure 2 . Compared with unmodified BC fiber, the spectrum of TA@BC showed some remarkable characteristic peaks of TA, such as C=O(1716 cm −1 ), C=C(1606 cm −1 , 1535 cm −1 , 1448 cm −1 ), and O-H(1322 cm −1 , 1203 cm −1 ) bonds, [ 22 , 33 ] which represented the carboxyl group, benzene ring, and phenolic hydroxyl group in TA, respectively. In addition, after TA is coated on the surface of BC, the absorption peaks of -OH in TA (3425 cm −1 ) and BC (3349 cm −1 ) shifted to 3114 cm −1 of TA@BC, indicating that a strong hydrogen bond force was formed between TA and BC. Moreover, the changes of TA@BC in the AFM and SEM images ( Figure 1 b,c) and the thickness distribution curve ( Figure 1 d) confirmed that TA was successfully coated on the surface of the BC. For the hydrogel of P(AA-AMPS), the absorption peaks at 2976 and 1452 cm −1 represented the O-H [ 34 ] and C–N [ 35 ] bonds of the AA and AMPS, respectively, which showed the successful polymerization of the P(AA-AMPS) hydrogel, and the C=O at 1704 cm −1 was significantly enhanced when the fiber of TA@BC was added into the P(AA-AMPS) hydrogels. After adding Ca 2+ , the absorption peak of C=O shifted from 1704 cm −1 to 1700 cm −1 , demonstrating the form of metal-carboxyl coordination bonds between Ca 2+ ions and the carboxyl group of AA and AMPS [ 31 , 34 ]. Besides, the absorption peaks at 3297 cm −1 in the P(AA-AMPS)-TA@BC and P(AA-AMPS)-TA@BC-Ca 2+ hydrogels were associated with the hydrogen bonds between TA@BC and P(AA-AMPS), respectively. 3.2. Mechanical Properties of Hydrogel Owing to the uniform intermolecular and intramolecular interactions in the P(AA-AMPS)-TA@BC-Ca 2+ gel and the double network structure formed by BC and P(AA-AMPS), the hydrogel performed excellent mechanical properties. As a fiber with a large aspect ratio, BC showed high elastic modulus and strength [ 19 ], and its content has a significant effect on the strength of the hydrogel. With the increase in TA@BC content from 0.25 wt% to 0.5 wt%, the tensile strength of the P(AA-AMPS)-TA@BC-Ca 2+ hydrogel increased from 24.1 kPa to 83.4 kPa, and the elongation at break of the hydrogel increased from 756% to 1236% ( Figure 3 a). However, when the content of TA@BC increased to 1 wt%, the strength and strain of the hydrogel significantly decreased to 60.1 kPa and 1000%, respectively, which might be attributed to the large amount of BC and TA reducing the transparency of the hydrogel and hindering the UV-induced radical polymerization of AA and AMPS. Thus, the mechanical properties of the gel were reduced. The P(AA-AMPS)-TA@BC 0.5 -Ca 2+ gel was subjected to multiple cyclic tensile tests, and the results showed the stress–strain curves of the gel in the cycle (600% strain, 5 cycles) basically overlapped and had very small hysteresis, which indicated the good elasticity of the hydrogel. When TA@BC and Ca 2+ were gradually added into the P(AA-AMPS) system, the maximum stress and strain of the hydrogel were simultaneously increased because it was affected by the double network structure (formed by TA@BC fibers and polymer network) and the dynamic coordination bonds (formed by Ca 2+ and carboxyl groups). Moreover, the tensile strength and maximum elongation of P(AA-AMPS)-TA@BC 0.5 -Ca 2+ were 6.7 and 3.4 times higher than that of P(AA-AMPS) hydrogel, respectively ( Figure 3 c). In addition, the tensile properties of P(AA-AMPS)-TA@BC-Ca 2+ hydrogels with different Ca 2+ concentrations were tested. Figure 3 d shows that the hydrogels with a Ca 2+ concentration of 3M exhibited the best mechanical properties (stress and strain), which can be attributed to the more metal-carboxyl coordination bonds in the hydrogel, and the toughness of the P(AA-AMPS)-TA@BC-Ca 2+ hydrogels with different concentrations of TA@BC is shown in Figure S2 . The prepared hydrogels also showed excellent compression properties. Here, the compression properties of hydrogels with different TA@BC contents were tested, and the compression process is shown in Figure S3e . As the TA@BC content increased to 0.5 wt%, the strength of the gel increased from 105 kPa to 207 kPa, and the gel with 0.5 wt% content was twice as strong as the pure gel at 80% compressive strain ( Figure S3 ). However, when the content of TA@BC increased to 1 wt%, the compressive strength of the hydrogel dropped to 153 kPa, which was consistent with the tensile properties of the gel. The continuous compression-release performance of the hydrogel was tested by 80% strain, 5 cycles ( Figure S3 ). During compression-release cycles, the stress–strain curve and hysteresis loop of the hydrogel completely coincide. Therefore, the prepared hydrogel exhibited sufficient toughness to withstand an 80% continuous cyclic compressive deformation. When the external force was removed, the hydrogel could recover to its original state, indicating that it had good elasticity and fatigue resistance. Then, the compression properties of P(AA-AMPS), P(AA-AMMPS)-TA@BC 0.5 , and P(AA-AMPS)-TA@BC 0.5 -Ca 2+ were all tested ( Figure S3 ). Similar to the results for tensile properties, the addition of TA@BC and Ca 2+ significantly increased the compressive properties of the hydrogel. In view of the fact that the hydrogel system showed the best excellent mechanical properties when the content of TA@BC was 0.5 wt%, the P(AA-AMPS)-TA@BC 0.5 -Ca 2+ gel material was preferred as the main research object in the following studies. 3.3. Mechanical and Conductive Stability of the Hydrogel in Freezing and Arid Environments As a flexible material containing a large amount of water, hydrogels are prone to freezing or dehydration in cold or dry environments, resulting in the loss of their functionality. For example, in a cold environment, the hydrogel will be brittle because of the crystallization of water inside the gel, which then affects the mechanical properties and the ionic conductivity of the ionic conductive hydrogel. Therefore, maintaining the properties of the materials in multiple environments is a challenge for the development of flexible hydrogels. One of the ways to effectively resist freezing and moisturizing is to introduce anti-freezing components, such as high-concentration Ca 2+ /Li + , ethylene glycol, dimethyl sulfoxide [ 36 , 37 , 38 ], etc. In this work, the addition of 3 mol/L Ca 2+ endowed the hydrogel with excellent anti-freezing properties. To measure the freezing resistance of the material, the freezing points of P(AA-AMPS)-TA@BC 0.5 -Ca 2+ hydrogels with different Ca 2+ concentrations were recorded by DSC during heating from −50 °C to 20 °C. We define the peak of the ice crystals dissolving and absorbing heat as the freezing point of the gel. During the heating process of the DSC curve, the pure gel P(AA-AMPS) showed a large endothermic peak at 0 °C (freezing point of P(AA-AMPS) hydrogel), which was owing to the melting of the ice crystals in the sample. However, when Ca 2+ ions were added to the gel, the freezing point of the P(AA-AMPS)-TA@BC 0.5 -Ca 2+ hydrogel decreased significantly. As the concentration of Ca 2+ ions in the hydrogel increased from 1M to 3M, the freezing point of the P(AA-AMPS)-TA@BC 0.5 -Ca 2+ gel decreased from −11 °C to −33 °C ( Figure 4 a), showing a significant anti-freezing effect. The lower freezing point of the hydrogel was due to the fact that the high concentration of the salt solution will increase the osmotic pressure and reduce the freezing point based on the colligative properties of the solution. Additionally, the hydration of Ca 2+ will affect the hydrogen bonds of water molecules. These factors all affect the formation of ice crystals inside the gel ( Figure 4 a) [ 33 , 39 ]. As shown in Figure 4 b, the P(AA-AMPS) and P(AA-AMPS)-TA@BC 0.5 -Ca 2+ gels were simultaneously frozen at −20 °C for 24 h. The moisture inside the pure P(AA-AMPS) gel crystallized during the cold processes, and the gel changed from transparent to white. Meanwhile, the P(AA-AMPS) gel became fragile and easily broke during the stretching process. By contrast, the P(AA-AMPS)-TA@BC 0.5 -Ca 2+ gel remained transparent and flexible after freezing, and could be easily twisted and stretched. To clearly characterize the mechanical stability of the hydrogel at low temperatures, we tested the tensile properties of the hydrogel in the frozen state and normal state ( Figure 4 c). The results showed that the fracture stress and strain of the gel stored at −20 °C for 24 h can still reach 68.2 kPa and 1013.8%, which are 81.7% and 82.0% of the normal state, respectively. The performance proves that the hydrogel material has excellent anti-freezing performance and stability. Besides, the P(AA-AMPS)-TA@BC 0.5 -Ca 2+ hydrogel exhibited not only excellent anti-freezing properties but also sufficient moisturizing properties. Within 120 h in an open environment at room temperature, the hydrogel P(AA-AMPS) retained only 14.3% of its original weight, showing obvious shrinkage and dehydration ( Figure 4 e). However, the P(AA-AMPS)-TA@BC 0.5 -Ca 2+ gel still retained 70% of its initial weight, and is less deformed compared to the P(AA-AMPS) gel ( Figure 4 d). Meanwhile, after being placed in the open environment for 120 h, the elongation of the prepared hydrogel was still 1025.3%, and the stress was 75.2 kPa, which were 83% and 91% of their initial state, respectively ( Figure 4 f). The excellent moisture-retention performance of the P(AA-AMPS)-TA@BC 0.5 -Ca 2+ hydrogel was due to the fact that the addition of CaCl 2 could reduce the vapor pressure of water and effectively inhibit the evaporation of water in the hydrogel. Specifically, the hydration between Ca 2+ /Cl − and water molecules effectively inhibited the evaporation of water molecules. Therefore, the P(AA-AMPS)-TA@BC 0.5 -Ca 2+ hydrogel exhibited good moisturizing properties ( Figure S4a ). The high concentration of Ca 2+ ions in the hydrogel not only endowed the material with great anti-freezing and moisturizing properties but also with excellent ionic conductivity. The ionic conductivity of the P(AA-AMPS) and P(AA-AMPS)-TA@BC 0.5 -Ca 2+ hydrogels at different temperatures was tested by a comprehensive physical property measurement system ( Figure S4b ). Compared with the P(AA-AMPS) gel, the ionic conductivity of the P(AA-AMPS)-TA@BC 0.5 -Ca 2+ hydrogel reached 1.92 S/m at room temperature, which was 80 times higher than that of the P(AA-AMPS) gel (0.024 S/m), and the conductivity of the P(AA-AMPS) gel dropped by an order of magnitude (from 0.024 S/m to 0.001 S/m) as the temperature decreased from 20 °C to −35 °C. Since the freezing point of the P(AA-AMPS) gel is 0 °C, while the ion migration rate decreased with the change of temperature, the formation of ice crystals also hindered the migration of ions, thereby reducing the conductivity of the hydrogel. By contrast, the P(AA-AMPS)-TA@BC 0.5 -Ca 2+ hydrogels still exhibited an ionic conductivity of 0.36 S/m at −35 °C, which was 360 times higher than that of the P(AA-AMPS) gel. The high conductivity indicated that the material showed great anti-freezing properties and excellent ionic conductivity above the freezing point. 3.4. Adhesion Properties of Hydrogel As a wearable device in direct contact with the human body, it is highly desirable to improve the adhesion of the hydrogel to human skin or a prosthesis. Strong and repeatable adhesion is conducive to the conformal attachment of the strain sensor and avoids the occurrence of interface delamination under large deformation to ensure the accuracy of monitoring. The presence of catechol in TA is believed to fulfill the dual role of interfacial binding and the solidification of the adhesive proteins [ 40 ]. Catechol is capable of diverse chemistries, which enable it to bind to both organic and inorganic surfaces through the formation of reversible non-covalent or irreversible covalent interactions [ 29 , 30 ]. Based on the catechol groups of oxidized polyphenols mimicking the mussel adhesion mechanism, the TA@BC hydrogels exhibited unique adhesiveness to a wide range of surfaces [ 22 ]. The P(AA-AMPS)-TA@BC 0.5 -Ca 2+ hydrogels showed robust adhesion to various substrates such as plastic, glass, PTFE, ceramic, rubber, and iron ( Figure 5 a). These results indicated that the hydrogel could be adhered directly on prosthetic or robotic surfaces made of various materials without any other additional processing. Moreover, we can evaluate the adhesion properties of hydrogels by measuring the tensile adhesion strength between the hydrogel and the matrix. Among the matrix materials, the P(AA-AMPS)-TA@BC 0.5 -Ca 2+ hydrogel showed the highest adhesion strength and adhesion force to wood, reaching 27 kPa and 204 N/m, respectively ( Figure 5 c,d). The main adhesion units of the hydrogel were TA containing abundant of o-phenolic hydroxyl groups, which can be tightly combined with the surface of the material through a large number of hydrogen bonds, especially a matrix with a large number of hydroxyl groups such as wood and glass [ 22 , 33 ]. More importantly, dynamic hydrogen-bonding interactions endowed TA@BC hydrogels with biocompatibility and repeatable adhesion without sacrificing other desirable properties, which were attractive for practical applications as wearable strain sensors. 3.5. Electrical Sensing Properties of Hydrogel Owing to the dual network structure and uniform conductive ions, the hydrogel not only exhibited excellent mechanical properties but could also serve as a sensing medium to transmit the relative resistance changes (RRCs) caused by the change of the conductive pathway. The conductivity of the P(AA-AMPS)-TA@BC 0.5 -Ca 2+ hydrogel reached 1.92 S·m −1 because of the existence of a large number of free Ca 2+ ions ( Figure S4b ). To visually display the sensing ability of the P(AA-AMPS)-TA@BC 0.5 -Ca 2+ hydrogel, we connect the gel strip in the closed circuit with a LED lamp under a direct current voltage of 4.5 V ( Figure 6 a). The brightness of the LED lamp was gradually decreased when the hydrogel was stretched, which indicated the increase in resistance. To show the real-time mechanical response properties of the hydrogels more clearly, the RRC vibration of the hydrogels was measured under diverse degrees of tensile strains. As shown in Figure 6 b,c, both in small and large strains, the RRCs of the hydrogel changed obviously without delay. For a single stretching and releasing process, the RRCs value drops to the original state without obvious hysteresis after removing the tensile load. To verify the effect of different concentrations of Ca 2+ ions on the sensing performance of hydrogels, the RRCs of P(AA-AMPS)-TA@BC 0.5 -Ca 2+ with 1M, 2M, and 3M Ca 2+ concentrations under diverse degrees of tensile strains are shown in Figure 6 d–f. When the ion concentration increased from 1 M to 3 M, the conductivity of the gel increased from 0.42 to 1.92 S/m ( Figure S5 ). However, at the same strain, hydrogels with higher conductivity had no significant effect on the corresponding change in RRC values. The reason was that the ion concentration in the hydrogel did not change the number of charges during stretching. Therefore, as for the same ions, higher conductivity was not a relevant factor for the sensitivity of the gel sensor. Whereas, when the voltage gradually increased from 0.5 V to 1 V, the RRC value at the same strain increased, and the gauge factor (GF) of the hydrogel at 500% strain increased from 1.49 to 2.18. Due to this, the initial current of the gel increased as the voltage increased, resulting in an expansion of the current variation range. Additionally, the migration rate of Ca 2+ ions in the hydrogel was affected by the voltage, thereby enhancing the sensitivity of the hydrogel to the strain responses ( Figure 6 g–i). Hence, as for ionic conductive hydrogels, within a specific range, the migration rate of ions is the main factor related to the sensing performance of the hydrogel. The strain-sensing ability of 3M P(AA-AMPS)-TA@BC 0.5 -Ca 2+ hydrogel was calculated using the value of GF. When the tensile strain changed from 2.5% to 900%, the GF value of the gel increased from 0.91 to 2.91 ( Figure S6 ), which indicated that in the range of 2.5% to 900%, the P(AA-AMPS)-TA@BC 0.5 -Ca 2+ hydrogel exhibited reliable, stable and sensitive strain-sensing performance. In addition, the hydrogel also presented a sensitive and stable response to the stretching frequency from 0.16 Hz to 1 Hz ( Figure S7a ), and under 400% strain, the sensing response time and recovery time of the hydrogel were 460 and 500 ms, respectively, which satisfied the requirements of human motion detection ( Figure S7c ). For the long-term use of hydrogel sensors, the ability to stably transmit real-time electrical signals during repeated stretch–release cycles is critical. As for the P(AA-AMPS)-TA@BC 0.5 -Ca 2+ hydrogels, stable mechanical response properties could be maintained under 400% strain after 300 stretch–release cycles ( Figure S7b ), and during the test, there was no obvious shift in the RRCs of the hydrogel, indicating that the P(AA-AMPS)-TA@BC 0.5 -Ca 2+ hydrogel had excellent mechanical response properties and long-term reusability. 3.6. Behavioral Monitoring of Human Movement and Physiological Signals by Hydrogel With high mechanical stability, strong adhesion, great ionic conductivity, and excellent sensitivity, the P(AA-AMPS)-TA@BC 0.5 -Ca 2+ hydrogel is a potential candidate for human motion detection and physiological signal monitoring. As shown in Figure 7 c, the adhered P(AA-AMPS)-TA@BC 0.5 -Ca 2+ hydrogel on the finger joint rapidly responded to each bending (0° to 90°) angle by RRC values. Additionally, the stable current signal from the ionic conductive hydrogel could monitor the process of instantaneous change of different bending angles. When the bending angle of the finger abruptly varied from 30° to 60°, the RRC value of the gel increased from 0.968 to 1.912 without delay, and when it changed from 90° to 0°, the RRC value quickly recovered to the original state, which indicated that the hydrogel could maintain stable and highly sensitive response characteristics during continuous motion detection ( Figure 7 a). In Figure 7 b, the finger’s slow and continuous bending process was monitored. After fitting, the results showed an obvious linear function between the output RRC value and the angle (R 2 = 0.9959). Hence, the P(AA-AMPS)-TA@BC 0.5 -Ca 2+ hydrogel can be considered as an ideal candidate for detecting human motion. In addition to macroscopic deformations related to human motion detection, the P(AA-AMPS)-TA@BC 0.5 -Ca 2+ hydrogel could also respond to various physiological signals. When writing on the surface of the hydrogel, the specific characteristics of the handwriting, like writing order and writing strength, can be translated into different electrical signals, which means that the hydrogel can distinguish different handwritten letters. As shown in Figure 7 e, we connected the copper electrodes on the left and right sides of the hydrogel, and then covered the upper and lower sides with PET film to prevent water loss and scratched the gel. Three numbers were written on the hydrogel surface, and the corresponding RRC signals differed and had their own characteristics ( Figure 7 f). Moreover, the hydrogel could also be used to monitor subtle changes of the human pulse. In Figure 7 g, the P(AA-AMPS)-TA@BC 0.5 -Ca 2+ hydrogel showed the ability to clearly identify the pulse before and after exercise. We took the ionic conductive hydrogel adhered to the wrist to monitor the subtle changes of the volunteers’ pulse before and after running for 5 min. We found that the corresponding RRC value increased significantly after running, indicating that the pulse beat more vigorously after exercise. The experiment also proved the high sensitivity of the gel for micro-motion detection. Therefore, the P(AA-AMPS)-TA@BC 0.5 -Ca 2+ hydrogel can be used to detect the physiological signals of the human body, and return real-time electric signals in the form of RRC values."
} | 8,020 |
35451146 | PMC9325067 | pmc | 1,535 | {
"abstract": "Abstract Mycorrhizal fungi can colonize multiple trees of a single or multiple taxa, facilitating bidirectional exchange of carbon between trees. Mycorrhiza‐induced carbon transfer was shown in the forest, but it is unknown whether carbon is shared symmetrically among tree species, and if not, which tree species are better donors and which are better recipients. Here, we test this question by investigating carbon transfer dynamics among five Mediterranean tree species in a microcosm system, including both ectomycorrhizal (EM) and arbuscular (AM) plants. Trees were planted together in “community boxes” using natural soil from a mixed forest plot that serves as a habitat for all five tree species and their native mycorrhizal fungi. In each box, only the trees of a single species were pulse‐labelled with 13 CO 2 . We found that carbon transfer was asymmetric, with oak being a better donor, and pistacia and cypress better recipients. Shared mycorrhizal species may have facilitated carbon transfer, but their diversity did not affect the amount, nor timing, of the transfer. Overall, our findings in a microcosm system expose rich, but hidden, belowground interactions in a diverse population of trees and mycorrhizal fungi. The asymmetric carbon exchange among cohabiting tree species could potentially contribute to forest resilience in an uncertain future.",
"conclusion": "5 CONCLUSION Carbon transfer between trees of different species has so far been demonstrated mainly in temperate and boreal forests (Fitter et al., 1998 ; Francis & Read, 1984 ; Högberg et al., 2008 ; Klein et al., 2016 ; Simard, Perry, et al., 1997 ). Here, it was shown that carbon was moving between tree species of the Mediterranean forest whether by CMNs or by diffusion through the soil. Considering that our microcosm experiment simulated a productive mixed Mediterranean forest, it is possible that asymmetric resource distribution is part of a healthy forest ecosystem. Carbon transferred among the trees from a donor species to a recipient species, and sometimes vice versa. Overall, among our five species, Quercus served only as a carbon donor; Cupressus , Ceratonia and Pinus had both donor and recipient functions, and Pistacia was the most dominant carbon recipient (we did not test it as a donor). The function of carbon trade induced by mycorrhizal networks is still unknown. This mechanism can have large effects on trees dynamics in the forest, whether it underlies a competitive symbiosis in which highly connected trees have an advantage over not well‐connected trees or a mutual symbiosis allowing trees to support each other in times of need. Further research needs to be done to examine the effect of carbon trading on various conditions such as establishment of young seedlings, deficiency in nutrients such as sugars, N and P, and drought.",
"introduction": "1 INTRODUCTION Roots of vascular plants interact with mycorrhizal fungi along with other components of the soil microbiome including nonmycorrhizal fungi, archaea, and bacteria (Högberg et al., 2008 ). The symbiotic interaction between the fungus and the root relies on transmission of soil‐derived nutrients from the fungus to the host tree (Collins Johnson et al., 2010 ), increasing root absorption of water by the fungal hyphae and mediating the interaction of the root with other microbes in the soil (Aroca et al., 2007 ; Hestrin et al., 2019 ). The heterotrophic fungus benefits from the interaction by receiving carbon from the autotrophic host tree (Högberg et al., 2008 ). There are two main functional groups of mycorrhizae: ectomycorrhiza (EM) which do not penetrate the root cortex of the host and interact mainly with trees that are located in seasonally cold and dry climates, and arbuscular mycorrhiza (AM) whose hyphae penetrate the root cortex and interact mainly with plants that are located in seasonally warm and wet climates (Steidinger et al., 2019 ). Each individual tree may interact with tens to hundreds of different mycorrhizal taxa at the same time (Bahram et al., 2011 ). It is generally thought that plants associate exclusive with a single mycorrhizal type, but it has also been shown that some plants can be colonized by fungi of several mycorrhizal types, including both arbuscular mycorrhizal fungi (AMF) and ectomycorrhizal fungi (EMF) (Teste et al., 2020 ). The fungal affinity to their hosts is also complex since fungi have a wide range of specificity with some species that are specialist to their host trees and others that are generalist, having multiple partners (Massicotte et al., 1994 ; Heijden et al., 2015 ). The formation of common mycelial networks among trees depends on both the neighbouring trees specificity to compatible fungal species and on the fungal specificity to the host trees. Trees in the forest compete over limited resources such as light and nutrients (Lindenmayer & Laurance, 2017 ). However, some studies demonstrated carbon exchange among trees which might serve as a mutualistic interaction between them. Studies showing asymmetrical carbon transfer to young seedlings and supporting their establishment (Van Der Heijden & Horton, 2009 ) or studies presenting increased carbon transfer to trees under starvation (Simard, et al., 1997 ) certainly point that way. Several studies showed different benefits from CMNs among tree species. In a microcosm with the EM trees Pinus and Larix , Pinus got more carbon when having Suillus bovinus fungi and Larix got more carbon when ectomycorrhizal fungus was Suillus grevillei or Boletinus cavipes (Finlay, 1989 ). In AM plants it was shown that AMF favoured legume over grasses (Scheublin et al., 2007 ). Carbon transfer has been shown between legumes of the same species and between trees of different, usually related, species in the laboratory and later in the forest (Fitter et al., 1998 ; Francis & Read, 1984 ; Högberg et al., 2008 ; Simard, et al., 1997 ). Recently it has been shown that carbon is bidirectionally transferred among trees of different taxa in a mature forest (Klein et al., 2016 ). There are several possible mechanisms underlying the transfer of carbon between trees: Roots of different trees form fusions (natural root grafts) and can exchange carbon, water, or nutrients among them (Fraser et al., 2006 ; Graham & Bormann, 1966 ; Nara, 2006 ). However, root grafts occur mostly between trees of the same species (Fraser et al., 2006 ; Graham & Bormann, 1966 ; Woods & Brock, 1964 ) and therefore it does not provide a good explanation for carbon transfer between trees of different taxa. Another possible explanation is that carbon of one tree is secreted to the soil and is being absorbed by a root of a different tree (Pérez‐Pazos et al., 2021 ). A third possibility is that trees exchange carbon through mycorrhizal networks (Fitter et al., 1998 ; Francis & Read, 1984 ; Högberg et al., 2008 ; Simard, et al., 1997 ). Some evidence have been found to support this theory in mature trees (Klein et al., 2016 ). First, fruit bodies of mycorrhizal fungi that interacted with carbon‐labelled trees contained labelled carbon in contrast to identical fungi located far from the carbon‐labelled trees and to fungi that did not interact with the labelled trees (Klein et al., 2016 ). In addition, trees that shared more mycorrhizal fungi tended to exchange more carbon (Rog et al., 2020 ). The temporal and quantitative dynamics of carbon transfer between trees is still unclear. Once carbon is assimilated in the leaf of a tree it can reach three major sinks: respiration, biomass, and exudation from the root (Epron et al., 2012 ; Klein & Hoch, 2015 ). Carbon reaching the phloem moves in both directions and reaches the roots according to the source‐sink gradient from the leaf to the root (Liesche et al., 2015 ). Studies have shown that carbon flux in the phloem is faster in angiosperms compared to gymnosperms with the rate of 0.2–6 m/h and 0.1–0.2 m/h, respectively (Epron et al., 2012 ). This is probably due to the anatomical differences in their transport system (Liesche et al., 2015 ). A pulse labelling experiment on individual Mediterranean saplings showed carbon allocation to roots three days post labelling in saplings of Pinus halepensis , Cupressus sempervirens , Quercus calliprinos , Ceratonia siliqua and Pistacia lentiscus ( Rog, Jakoby, et al., \n 2021 \n ) . The partitioning of carbon to belowground compartments was 28%–38% in gymnosperms ( Pinus halepensis and Cupressus sempervirens ), and 5%–10% in angiosperms ( Quercus calliprinos , Ceratonia siliqua and Pistacia lentiscus ) (Rog, Jakoby, et al., 2021 ). These quantities refer to allocation of carbon from the leaf to the root of the same tree. After reaching the roots, carbon can transfer to the rhizosphere, which is composed of the soil and the rhizosphere microbiota, including mycorrhizal fungi. Carbon is expected to further decrease while transferring to mycorrhizal fungi and to the roots of a neighbouring tree 2021 . Indeed, in an experiment studying carbon transfer between paper birch and Douglas fir saplings it was shown that small amounts of 4.7% of the carbon fixed by paper birch were transferred to Douglas fir (Simard, et al., 1997 ). While most common mycorrhizal networks (CMNs) were studied in temperate ecosystems with rich organic soil and wet climates, less in known on CMNs in water limited environments such as the Mediterranean forest. In such environments the trees must deal with frequent droughts and often limit their activity to specific seasons or times of the day. This can affect the dependency of trees on the mycorrhizal network for resource uptake. It can also possibly change the specificity level of mycorrhizal fungi, connecting to additional hosts to reduce risk and have an advantage in a harsher environment. Considering that drought periods are becoming longer and harsher also in temperate forests (Klein et al., 2022 ), ecosystems such as the Mediterranean forest can help us predict how mycorrhizal networks will look like in the future. Here, saplings of five Mediterranean tree species, EM: Pinus halepensis and Quercus calliprinos (Torres & Honrubia, 1994 ; Trocha et al., 2012 ), and AM: Cupressus sempervirens (Zarik et al., 2016 ), Ceratonia siliqua (Essahibi et al., 2018 ; Lahcen et al., 2012 ) and Pistacia lentiscus (Caravaca et al., 2002 ; Green et al., 2005 ) were labelled with 13 CO 2 . We measured the amount and direction of belowground carbon transfer to other saplings of the same species and of different species. We aimed to identify which tree species serve as carbon donors and which serve as carbon recipients in this microforest community, and to map the mycorrhizal networks connecting them. To do so, we planted trees in microcosm system and used 13 CO 2 pulse labelling to track carbon allocation within and between trees. Then we used high throughput DNA sequencing to identify the mycorrhizal fungi colonizing each tree. We hypothesized that carbon would transfer: (i) in a bidirectional way in an asymmetrical manner, (ii) among tree species saplings that share mycorrhizal species, and (iii) among EMF hosts and among AMF hosts but not between these two guilds of plants.",
"discussion": "4 DISCUSSION Here, we studied belowground carbon allocation in a microcosm system for the mixed Mediterranean forest. We identified carbon donor and recipient trees in this community and found overlapping mycorrhizal fungi species among them. We hypothesized that carbon would transfer in a bidirectional and asymmetrical way. Indeed, most species had the positions of both carbon recipients and carbon donors (Figure 2 , Figures S13 and S14 ). This aligns with the results of previous studies and with our perception of the mycorrhizal network as a pathway that transports materials in both directions (Simard, et al., 1997 ). Surprisingly, Quercus trees demonstrated a pattern of unidirectional carbon transfer. They transferred carbon to Pistacia , Cupressus and Pinus , while not recicarbon from any other tree, including other Quercus trees (Figure 2 , Figure S14 ). Previous studies have suggested that carbon transfers among trees according to a carbon gradient (Francis & Read, 1984 ; Simard, et al., 1997 ). Indeed, we found high carbon percentage in Quercus roots, making the Quercus trees a potential physiological source relative to their neighbours (Figure S16 ). This result could also be explained by the small sample size in this study together with low quantities of transferred carbon, making it hard to detect. In other experiments, Quercus was shown to receive carbon from both other Quercus trees and pines, implying bidirectional transfer could happen also in Quercus trees (Cahanovitc et al., 2022 ). There were some differences between the studies that can explain why we did not see carbon transfer to Quercus in the current study. First, the mycorrhizal community differed between the two experiments, with the previous study dominated by Tomentella fungi and the current study containing mainly Scleroderma fungi. Second, in this current study, we had many trees of several species in each box creating a highly competitive environment. We found that carbon transfer was asymmetrical as expected, in line with previous studies (Finlay, 1989 ; Scheublin et al., 2007 ; Simard, et al., 1997 ). A mixing model indicated that 4–29% of fine root carbon in recipients originated from donor trees (Table 2 ). However, the utility of this approach is limited in the case of pulse labelling with restricted root sampling. For example, in the box with labelled Cupressus we collected a Cupressus root that according to the calculation was composed entirely by carbon from its neighbour, and a Pistacia root with ~63% carbon from its neighbour (Table 2 ). To get these values we used the peak in the donor tree, however, there were large variations in the roots of labelled trees and perhaps the root that served as the source for the labelled carbon in the Pistacia had much higher levels, and thus transferred lower quantities of labelled carbon than estimated. The movement of carbon can be either symmetrical or asymmetrical, forming a platform for competition when specific species exploit the network to their advantage or for collaboration when stronger species can support the weaker ones. It also fits the results of previous studies showing different benefits from CMNs among tree species. In a microcosm with the EM trees Pinus and Larix , Pinus got more carbon when interacting with Suillus bovinus fungi and Larix got more carbon when associating with the ectomycorrhizal fungus Suillus grevillei or Boletinus cavipes (Finlay, 1989 ). In AM plants it was shown that AMF favoured legumes over grasses (Scheublin et al., 2007 ). TABLE 2 Carbon transfer between donor and recipient trees. Carbon excess was calculated using Equations (2) , (3) , (4) , (5) , (6) , (7) , (8) , and the % of donor carbon in recipients was calculated by the mixing model in Equation 9 \n Donor Recipient Mycorrhizal group Days post labelling Shared mycorrhizal ASVs δ 13 C in recipient (‰) Carbon excess (mg) Donor carbon (%) \n Quercus \n \n Pinus \n EM 42 137 –16.66 0.22 6.51 \n Quercus \n \n Cupressus \n EM +AM 240 57 –19.836 0.19 15.09 \n Quercus \n \n Pistacia \n EM +AM 15 103 –21.35 0.07 27.12 \n Ceratonia \n \n Cupressus \n AM 240 73 –20.75 0.28 4.41 \n Ceratonia \n \n Ceratonia \n AM 240 130 –19.419 0.01 6.35 \n Ceratonia \n \n Pistacia \n AM 56 45 –1.818 0.21 29.01 \n Cupressus \n \n Cupressus \n AM 2 52 35.195 0.37 118.88 \n Cupressus \n \n Pistacia \n AM 1 40 6.147 1.28 62.87 \n Pinus \n \n Pinus \n EM 56 300 16.762 0.04 13.38 Abbreviations: AM, arbuscular mycorrhizal; EM, ectomycorrhizal. John Wiley & Sons, Ltd We further hypothesized that carbon will be shared among tree species that host the same mycorrhizal species. Here, various fungal species were colonizing roots of all the five dominant tree species in the Mediterranean forest. These mycorrhizal fungi could form belowground networks among trees and serve as a platform for the carbon transfer among trees. Our last hypothesis was that carbon will move among EMF hosts and among AMF hosts but not between these two guilds of plants. Here, most cases of carbon transfer indeed occurred within the same mycorrhizal groups. In the fourth box having Quercus trees as donors, trees of different guilds served as recipients: Pinus (EM), Pistacia and Cupressus (AM). This could be explained by carbon movement in the soil and not by direct CMNs connecting the trees. Several studies have shown that Quercus trees can have both EM and AM fungi, allowing it to move carbon to both groups of trees (Egerton‐Warburton & Allen, 2001 ; Rothwell et al., 1983 ). Here, we have not verified and identified the mycorrhizal association between trees and fungi and therefore this explanation is merely speculation. Carbon had transferred among saplings that share the same phylogenetic groups (gymnosperms vs. angiosperms) but also among trees from other groups (e.g., Cupressus to Pistacia ). These results suggest that the similarity of the fungal community (guild) of the trees is more relevant to carbon transfer than the phylogenetic proximity among the trees. 4.1 Different dynamics of carbon transfer suggest multiple mechanisms We found several cases in which carbon was found in roots of nonlabelled trees. This carbon was found over a broad time range, between 1 day and 240 days post labelling. We have focused on the dynamics among several species and did not have many replicates of each donor‐recipient couple. Together with the complex dynamics of carbon within the compartments of each tree it reduced our ability to point out significant differences of carbon transfer timing across species. However, since natural levels of root δ 13 C are well studied, we can say with confidence that carbon transfer had occurred in the cases mentioned. In our experiment we did not include a separation between carbon exchange by roots, soil, or CMNs. However, trees were taken apart after the termination of the experiment, and no root grafting that could explain our results were found (Figure S2 ). A plausible mechanism underlying carbon transfer belowground is diffusion in soil. In another experiment using the same soil, diffusion rate was calculated at 0.2–0.3 cm day −1 (Cahanovitc et al., 2022 ). Saplings were planted 10 cm apart from each other (Figure 1 ), and their root systems were ~5 cm apart (Figure S2 ), and hence C transfer through soil was possible after ~17 days. Here, levels of δ 13 C in the soil increased 2 days post labelling then decreased and had another peak after 230 days (Figure S9 ). Several cases of carbon transfer demonstrated in our results certainly fit these dynamics, for example carbon transfer between Quercus and Cupressus 240 days post labelling. In other cases, such as the transition between Quercus and Pinus this rate did not correspond to the level of labelled carbon in the soil, making the CMN more plausible than simple diffusion. 4.2 Generalist mycorrhiza as a possible mechanism underlying carbon exchange Most studied forests are dominated by either EMF or AMF and accordingly present each tree as either EM host or AM host. In the Mediterranean mixed forest, however, both AMF and EMF naturally cohabit, allowing trees to host both types simultaneously. There is prior evidence for trees hosting AMF and EMF within the same root system in seedlings of Quercus in different ages (Egerton‐Warburton & Allen, 2001 ) and next to different host plants (Dickie et al., 2001 ). Also, in a temperate forest in Japan, where AM and EM trees co‐occur, it was shown that the fungal composition of the AMF host Chamaecyparis obtuse contained EMF in addition to AMF (Toju & Sato, 2018 ). Here, we have found multiple overlapping mycorrhizal fungi among our trees. These fungi could potentially form networks connecting tree roots and allowing carbon transfer. While many of the fungi were host specific, most of the highly abundant fungi had multiple hosts and were found on several species (Figure 3 ). In this experiment we had Quercus and Pinus , which are considered as EMF hosts (Torres & Honrubia, 1994 ; Trocha et al., 2012 ), alongside the AMF hosts Pistacia (Caravaca et al., 2002 ; Green et al., 2005 ), Cupressus (Zarik et al., 2016 ) and Ceratonia (Essahibi et al., 2018 ; Lahcen et al., 2012 ). Most cases of transferred carbon occurred within the same guild, suggesting that the mycorrhizal fungi were active only in their traditional hosts. There was no correlation between the number of shared mycorrhizal fungi and the amount of shared carbon among trees, in contrast to other studies (Rog et al., 2020 ), raising the possibility that the identity of the shared fungus is more relevant than its abundance. Here, Quercus and Pinus shared carbon and were colonized with Tomentella eliisi and Suillus collintus , that were shown to mediate carbon transition between these two tree species (Cahanovitc et al., 2022 ). In addition, we identified two abundant AMFs‐ Rhizophagus irregularis and Rhizophagus fasciculatus , mostly on the AMF hosts Cupressus and Pistacia . This finding increases the likelihood that they have mycorrhizal association with the trees, especially due to the high and fast carbon exchange that was documented between Cupressus and Pistacia saplings. However, in the fourth box, we did document these two AMF species on the EM tree species Pinus and Quercus (Figure 4 ). Unexpectedly, we also found seven abundant EMF that colonized all tree species, including the AM tree species Cupressus and Pistacia . The fact that we detected EMF species on non‐EMF hosts does not necessarily mean that these fungi functioned as mycorrhizal symbionts. Since we did not verify the mycorrhizal colonization status using microscopy approaches, we cannot rule out the possibility that these are simply hyphae growing on the surface of roots. However, we can neither rule out that tree species in our experiment were dual‐mycorrhizal (Teste et al., 2020 ). It can stem from the conditions of our experiment, having trees of several species growing densely by each other (Figure S2 ), as they grow in the Mediterranean mixed forest. 4.3 Effects on mycorrhizal community composition in roots The young saplings were planted in microcosm boxes along with their primary fungal community. Prior studies have shown that initial fungal colonization of roots can affect the fungal community in a system through priority effect (Hausmann & Hawkes, 2010 ). Here, the mycorrhizal fungi in the individual trees were identified before the start of the experiment and contained a small and specific community of mycorrhizal fungi on each tree species (Figure S10 ). The separation between the functional groups of AMF and EMF hosts was clear in the individual trees, however trees grown in mixed microcosm presented a more diverse and mixed community of mycorrhizal fungi. Only two mycorrhizal fungi were overlapping among fungi in individual trees and in microcosms. These are the Tomentella ellisi that was found only in individual Pinus trees, and Sphaerosporella brunnea found in Pinus and Quercus trees. In the microcosm on the other hand, these two fungi were observed on all the tree species that participated in the experiment. One of the most abundant EMF taxa we found was Suillus , known to be restricted to members of the pinoid clade of Pinaceae (Kretzer et al., 1996 ). Yet, there are new reports of various Suillus species sporocarps from forests in which their Pinaceae hosts are absent. In bioassays, Suillus was found to form mycorrhiza with Abies and Tsuga that belong to the abietoid clade of Pinaceae, indicating that Suillus host specificity is more flexible than previously thought (Pérez‐Pazos et al., 2021 ). Here, Suillus collinitus was indeed most abundant on Pinus saplings, yet we found it on roots of all species. Accordingly, in other Mediterranean forest species, Suillus collinitus was found on the classic Pinus host but also on Quercus roots (Cahanovitc et al., 2022 ). Root tips with Scleroderma were found only on Quercus saplings and on Pistacia saplings that were planted in the fourth box (with many Quercus trees), but not in other boxes even though sharing the same soil. Both results can be explained by a host neighbour effect. It was previously shown that the presence and composition of neighbouring plants can affect the ability of some mycorrhizal fungi to develop mycorrhizae with hosts (Hausmann & Hawkes, 2010 ; Kohout & Sýkorová, 2011 ; Massicotte et al., 1994 ; Molina et al., 1997 ). Specifically, there are prior examples of specialist mycorrhizal fungi that formed symbiosis with other trees when growing in a mixed community near to their main host (Pringle, 2009 ). For example, the Larix specialist Suillus larcinus was detected on a Betula sapling when growing next to a Larix sapling (Nara, 2006 ). In another study it was shown that Suillus subaureus can germinate and associate with both Pinus and Quercus hosts, both in the laboratory and in the forest. Two other fungi, Suillus americanus and Suillus clintonianus , germinated by spores only in the presence of their primary Pinus hosts but could also form mycorrhizal association with Quercus and Larix trees when colonizing via mycelial networks (Lofgren et al., 2018 ). 4.4 Similarity of fungal composition in the community boxes and in the Mediterranean forest The current experimental setup serves as a microcosm for the water limited Mediterranean forest. It is important to understand CMNs composition and function in such environments due to the effects of drought on the fungal community abundance and diversity in soil (Hawkes et al., 2011 ; Yang et al., 2010 ). In addition, tree species vary in their spatial root distribution. In our system, Quercus is more deep rooted than Pinus (Rog, Tauge, et al., 2021 ). This variation can also affect resource sharing by CMNs. We compared the fungal communities in our microcosms with fungal communities of roots from the Mediterranean mixed forest tree species (Cahanovitc et al., 2022 ). Among the most abundant mycorrhizal fungi, four species were found in both communities: Suillus collinitus , Tomentella ellisii , Terefezia pini , Inocybe ( roseipes vs. rhodiola ). In other studies on mycorrhizal colonization of Pinus halepensis saplings in natural Mediterranean soil and in the forest itself, Suillus collintus and Inocybe were highly abundant, as was shown here (Livne‐luzon et al., 2017 ; Querejeta et al., 1998 ). The composition similarity between fungal communities in the natural forest and in the community boxes in this experiment reinforces our results. Yet, the ability of plants and fungi to create mycorrhizal association and trade resources may be different from what was demonstrated by experimental syntheses. Therefore, it is important to study the dynamics of carbon exchange and quantify its extent under different conditions in the mixed forest itself."
} | 6,846 |
34349920 | PMC8278901 | pmc | 1,536 | {
"abstract": "The biomineralization of intracellular magnetite in magnetotactic bacteria (MTB) is an area of active investigation. Previous work has provided evidence that magnetite biomineralization begins with the formation of an amorphous phosphate-rich ferric hydroxide precursor phase followed by the eventual formation of magnetite within specialized vesicles (magnetosomes) through redox chemical reactions. Although important progress has been made in elucidating the different steps and possible precursor phases involved in the biomineralization process, many questions still remain. Here, we present a novel in vitro method to form magnetite directly from a mixed valence iron phosphate precursor, without the involvement of other known iron hydroxide precursors such as ferrihydrite. Our results corroborate the idea that phosphate containing phases likely play an iron storage role during magnetite biomineralization. Further, our results help elucidate the influence of phosphate ions on iron chemistry in groundwater and wastewater treatment.",
"conclusion": "Conclusions Here, we present a new in vitro method of magnetite synthesis through an iron phosphate precursor phase that was initially inspired by prior observations of disordered phosphate-rich ferric hydroxides within MTB. Phosphate ions played a crucial role in the process by aiding in the co-localization of Fe 2+ and Fe 3+ ions while inhibiting the precipitation of iron oxides other than magnetite. The demonstrated utility of phosphate in favoring magnetite formation contributes to the understanding of where, how, and why the disordered phosphate-rich ferric hydroxide precursor phase in some strains of MTB may be formed and converted into magnetite. In particular, our synthesis suggests the importance of this phase as a potential storage mechanism for iron within MTB prior to magnetite biomineralization. More broadly, this study contributes to the understanding of iron chemistry in the presence of phosphate ions in alkaline conditions. We believe this work will encourage the development of new alternative approaches for the experimental design of magnetic nanoparticles with defined shapes and sizes that will have many useful commercial applications. Further, this work has implications in the fields of groundwater and wastewater treatment, and could potentially contribute to more efficient and cost-effective water purification strategies in these settings.",
"introduction": "Introduction Magnetite (Fe 2+ Fe 2 3+ O 4 ), a magnetic mineral found in both geological and biomineralization contexts, has magnetic and biocompatible properties that allow for a wide range of applications. 1–3 While there are many ways to synthesize magnetite, producing it with control over crystal habit, shape, and size usually requires high temperatures and environmentally harmful solvents. 4,5 Magnetotactic bacteria (MTB), on the other hand, are able to form crystals of magnetite at room temperature and in aqueous media with exquisite control over their size, shape and organization, and thereby over their magnetic properties. 6 This biological capability drives interest in developing biomimetic pathways, using peptides and proteins similar to those found within MTB, in an attempt to unravel and ultimately exploit their magnetite formation strategies. 7–10 MTB produce magnetite nanocrystals in specialized vesicles called magnetosomes. The mineralization process is regulated by a large number of proteins with specific functions that only now are beginning to be resolved. 11–13 In particular, many questions remain regarding the role of precursor phases such as ferrihydrite, 14,15 hematite, 16–18 and ε-Fe 2 O 3 . 19 Previous studies on the AMB-1 strain of Magnetospirillum magneticum 20 and the MSR-1 strain of Magnetospirillum gryphiswaldense 21 present evidence of a multi-step mechanism in which a disordered phosphate-rich ferric hydroxide precursor phase is formed first and subsequently converted into magnetite. Given its similarity to phosphate-rich ferritin, which is known to exist in many prokaryotes, 22 this precursor phase is thought to act as storage for iron before the initiation of the magnetite formation process. A similar amorphous, hydrated ferric phosphate phase has also been described in the dermal granules of Molpadia intermedia , 23 and ferric phosphates have been readily observed in many marine invertebrates. 24 While the interplay of iron and phosphate clearly plays a significant role in many biomineralization contexts, exploring the precipitation chemistry of ferrous and ferric ions in the presence of phosphate is also crucial for understanding the fate of nutrients ( e.g. phosphate) 25 in groundwater and wastewater treatment. 26,27 It has been proposed that inorganic phosphate (P i ) interacts with iron oxy-hydroxides by binding strongly to their surfaces, thereby stabilizing and favoring the formation of poorly crystalline iron phosphate precipitates. 28–32 Thus, during in vivo magnetite formation, the role of P i could be the sequestration of iron ions into precursor phases, which prevents the precipitation of iron oxides in neutral pH conditions. A similar strategy has been demonstrated in the case of crayfish gastroliths, where P i plays an important role in stabilizing biogenic amorphous calcium carbonate, 33–36 a mineral phase which stores CaCO 3 during the skeleton formation process. To help investigate the role that phosphate ions play in magnetite formation in MTB, we designed a method to form magnetite nanoparticles through the controlled formation and transformation of a phosphate precursor using a titration setup. We characterized the products formed at several stages of the reaction using cryogenic transmission electron microscopy (cryoTEM), selected area electron diffraction (SAED), X-ray absorption near-edge spectroscopy (XANES), and Raman spectroscopy. We demonstrate, for the first time, the transformation of an amorphous iron phosphate into magnetite at room temperature and in aqueous solution without evidence of other intermediate iron oxy-hydroxides.",
"discussion": "Discussion Here we have demonstrated, for the first time, an in vitro method to synthesize magnetite through an amorphous mixed valence iron phosphate precursor phase. The formation of this precursor phase was crucial to allow for the controlled formation of magnetite in our synthetic approach. When the reaction was performed with the same titration protocol in the absence of phosphate, the initial product was ferrihydrite, 44 which was converted to magnetite already at pH ∼ 8, while in the presence of phosphate magnetite was formed above ∼pH 11 ( Fig. 3 ). This shift to a higher pH value indicates that phosphate forms a stable precursor phase that inhibits the formation of all other iron oxides, which under many synthetic conditions compete with magnetite formation. 32,45 Indeed, prior research has demonstrated that ferrous iron ions aerated in the presence of phosphate at near-neutral pH leads to the predominant formation of amorphous ferric phosphate, and even small amounts of phosphate in the system can prevent the formation of iron oxides such as goethite. 29 In our system, the formation of magnetite was then triggered by the increase in pH, which favored the release of phosphate ions into solution due to competition with the hydroxide ions. 46–48 Thus, with a rapid pH increase in our synthesis, we hypothesize that hydroxy ions replaced the phosphate ions and that magnetite was formed after the release of any associated water molecules. However, additional experiments are needed to verify this hypothesis by performing in situ analysis to follow the evolution of the sample during rapid pH increase. The fact that none of the other common iron oxides were observed in our synthesis suggests that by coordinating Fe 2+ and Fe 3+ in one mixed phase, it is possible to drive the conversion directly to magnetite upon removal of phosphate ions. We suggest that the small size of the magnetite crystals formed in our bioinspired experiment ( Fig. 3 ) could have been caused by the presence of the phosphate ions that, despite the high pH, may still interact with developing crystals and inhibit their growth. Thus, if we had removed P i as we increased pH in our experimental system, formation of larger magnetite crystals may have been possible. Due to its presence in many fertilizers, phosphate is ubiquitous in the natural environment and present in high levels in groundwater and wastewater, contributing to eutrophication in many rivers and lakes. 27,49 Iron oxide species have been used for many years in attempts to recover phosphate from wastewater given their propensity for phosphate adsorption, 27,49 and iron has otherwise been used in groundwater systems in attempts to remove contaminants. 26 The insights we have presented here regarding the interactions between iron and phosphate in an aqueous system contributes to the understanding of the interplay of these ions in complex groundwater and wastewater systems. Further, our work suggests a potential mechanism through which phosphorus existing in these systems can be leveraged to produce size-controlled magnetite nanoparticles. This process could be particularly useful in places where groundwater comes from iron-rich aquifers and iron needs to be removed before the water can be used as drinking water. 50 Currently, iron is usually removed from groundwater using aeration followed by rapid sand filtration. This process causes oxidation and ultimately hydrolysis of iron ions, resulting in a highly-hydrated ferric hydroxide sludge which has limited market value. 50,51 If magnetite nanoparticles, which have many commercial applications, 5,52,53 were to be formed instead of this sludge, there would be considerable economic rewards. 54 Indeed, among many other uses, magnetite nanoparticles have been proposed for use in anti-cancer drug delivery systems, 55,56 cancer-treatment processes including magnetic hyperthermia, 57 and MRI contrast materials. 58,59 Notably, many of the existing methods to produce size- and shape-controlled magnetite nanoparticles with specific magnetitic properties require high temperatures, high pressures, or organic solvents. Although we have not developed a mechanism for precise control over magnetite crystal shape and size, we have shown that it is possible to achieve a very narrow size distribution of magnetite nanoparticles through an amorphous phosphate precursor in an aqueous solution. Future iterations of this synthetic pathway could yield new, more environmentally friendly methods to produce magnetite for its many commercial applications. In addition to serving as inspiration for new methods to produce magnetite, this synthesis also contributes to the understanding of the purpose of phosphate-rich precursors observed in some strains of MTB. Time-resolved studies have demonstrated phosphate-rich ferric oxy-hydroxide precursor phases in vivo early in the magnetite formation process in both the AMB-1 strain of Magnetospirillum magneticum 20 and the MSR-1 strain of Magnetospirillum gryphiswaldense . 21 These poorly ordered species are thought to be similar to bacterioferritin, which is an iron storage protein observed in other prokaryotes. 22 In the RS-1 strain of Desulfovibrio magneticus , a most likely mixed valence (but predominantly ferrous) amorphous iron phosphate precursor phase has been observed. 60 However, the exact role of these amorphous iron phosphate precursors in magnetite biomineralization remains unclear. Of note, a recent study has suggested that since the absence of two ferritin-like proteins in MSR-1 strains did not affect magnetite formation, bacterioferritin-like species are not directly involved in magnetite biomineralization. 61 However, the authors admit that other ferritin-like proteins that have yet to be characterized could still be involved directly in magnetite biomineralization. Further, they did not perform time-resolved studies to prove that no precursor phase was formed at any point in the absence of the two studied ferritin-like proteins. Even so, they showed that the two ferritin-like proteins they studied were useful in resisting oxidative stress. 61 This finding suggests that sequestering iron in a stable precursor phase prior to magnetite biomineralization may help prevent the toxic effects of free intracellular iron. Our synthesis supports the hypothesis that iron phosphate precursor phases function as a stable iron storage phase prior to magnetite biomineralization, as has been suggested in previous studies. 20,21,60,62 As we have shown, iron ions were very stable in the phosphate containing phases during our synthesis. Thus, phosphate ions may act as a control agent that allows for accumulation and co-localization of Fe 2+ and Fe 3+ ions, while preventing the precipitation of unwanted iron oxide phases. 32 The idea that iron is stored in a stable phase prior to magnetite formation is corroborated by a recent study looking at in vivo iron isotope measurements in AMB-1. 63 In that study, bacterial lysates (representing everything in the cell besides the magnetosome) contained at least 50% of the total cellular iron when grown in media containing relatively high iron concentrations, which supports an iron reservoir besides magnetite in these cells. Further, these lysates showed enrichment in heavy isotopes, suggesting the predominant presence of Fe 3+ . 63 Another study has found that magnetite crystals contain at most 30% of the total intracellular iron, and that another large pool of iron exists in MTB. 64 These findings are consistent with a ferric iron containing precursor storage phase. Further, magnetite formed in the iron isotope experiments showed depletion in heavy isotopes, suggesting that magnetite was formed following partial reduction of Fe 3+ to Fe 2+ . Thus, the authors presented a model in which iron accumulates intracellularly as Fe 3+ and is later reduced to Fe 2+ prior to transport into the magnetosome. Similarly, another study has found evidence of reduced Fe 2+ both within the magnetosome and in the cytoplasm. 64 Whether this reduction happens at the magnetosome membrane or in another intracellular compartment remains unclear. As we have shown, iron ions can stably exist as part of a mixed-valence phosphate containing phase under pH conditions typically found intracellularly in MTB (7.0–7.6), 65 which suggests that the reduction of Fe 3+ to Fe 2+ could occur outside of the magnetosome while the iron ions are stabilized within an iron phosphate precursor phase. We have shown that magnetite can be formed directly from an iron phosphate species, which raises the question of whether magnetite could be formed directly from a phosphate containing precursor under appropriate conditions in MTB. Intriguingly, electron microscopy experiments in the RS-1 strain of Desulfovibrio magneticus have suggested that amorphous iron phosphate containing granules, which rapidly accumulate iron early in the biomineralization process, can convert directly to magnetite. 60 However, there was evidence to suggest that iron was mostly transferred from this precursor phase to other iron reservoirs prior to magnetite formation. In addition, another study in RS-1 found evidence of amorphous, iron phosphate granules that are likely separate from magnetite and within separate bacterial organelles. 62 Further, in our experiments, the high pH values necessary for the direct conversion of the iron phosphate precursor to magnetite (>11) is inconsistent with magnetosomal pH values (7.0–7.4). 65 Thus, while it is worth considering the possibility of a direct conversion of iron phosphate precursors to magnetite, an iron storage function of the precursors is more likely."
} | 3,973 |
35264607 | PMC8907315 | pmc | 1,537 | {
"abstract": "The development of highly durable, stretchable, and steady triboelectric nanogenerators (TENGs) is highly desirable to satisfy the tight requirement of energy demand. Here, we presented a novel integrated polymeric membrane that is designed by PEDOT: PSSa-naphthalene sulfonated polyimide (PPNSP)-EMI.BF 4 Electronic skin (e-skin) for potential TENG applications. The proposed TENG e-skin is fabricated by an interconnected architecture with push–pull ionic electrets that can threshold the transfer of charges through an ion-hopping mechanism for the generation of a higher output voltage (Voc) and currents (Jsc) against an electronegative PTFE film. PPNSP was synthesized from the condensation of naphthalene-tetracarboxylic dianhydride, 2,2′-benzidine sulfonic acid, and 4,4′diaminodiphenyl ether through an addition copolymerization protocol, and PEDOT: PSSa was subsequently deposited using the dip-coating method. Porous networked PPNSP e-skin with continuous ion transport nano-channels is synthesized by introducing simple and strong molecular push–pull interactions via intrinsic ions. In addition, EMI.BF 4 ionic liquid (IL) is doped inside the PPNSP skin to interexchange ions to enhance the potential window for higher output Voc and Iscs. In this article, we investigated the push–pull dynamic interactions between PPNSP-EMI.BF 4 e-skin and PTFE and tolerable output performance. The novel PPNSP- EMI.BF 4 e-skin TENG produced upto 49.1 V and 1.03 µA at 1 Hz, 74 V and 1.45 µA at 2 Hz, 122.3 V and 2.21 µA at 3 Hz and 171 V and 3.6 µA at 4 Hz, and 195 V and 4.43 µA at 5 Hz, respectively. The proposed novel TENG device was shown to be highly flexible, highly durable, commercially viable, and a prospective candidate to produce higher electrical charge outputs at various applied frequencies.",
"conclusion": "Conclusion In this research, we have demonstrated an ionic networked polymer TENG based on PPNSP.EMI.BF 4 -PTFE with ion-ion hopping alternates hydrophilic nano-channels integrated with sulfonic acid groups, resulting in superior electro-chemo-mechanical properties. Well-arranged hydrophilic sulfonic acid groups have good compatibility with both PEDOT: PSSa, and EMI.BF 4 to penetrate at atom–atom interactions owing to PPNSP.EMI.BF 4 composite film thereby provided constant interconnected ion-ion interactions to transport charges through ionic nano-channels to enhance Voc, and Jsc performance and robustness. The FE-SEM analysis clearly revealed the ion-ion nano-channels within the polymer network that trigger the charges very fast when PPNSP.EMI.BF 4 e-skin interacted with the PTFE surface. The SS curves of the NSP.H + , PPNSP, and PPNSP.EMI.BF 4 noticed the strong flexibility and resilience of the membranes to be compatible with the contact-separation mode TENG without damaging the membranes even after several cycles of contact-separation mode TENGs. Mainly, the PPNSP.EMI.BF 4 e-skin displayed a dramatic increase in the tensile modulus, up to 131%, tensile strength, up to 174%, and elongation at break values up to 277% compared to those of its starting membranes, such as the NSP.H + and PPNSP membranes. These synergistic effects were beneficial after incorporation of EMI.BF 4 IL in the PPNSP membrane in designing a high-performance ion-mediated e-skin-based TENG that shows large affective charge induction to provide higher Voc, and Jsc. In particular, the NSP-H + membrane was sandwiched by the PEDOT: PSSa conducting polymer, which enhanced superior conductivities. The Voc and Jsc of the NSP.H + at 5 Hz surged upto − 45.2 to 51.7 V and − 0.62 µA to 0.72 µA, respectively. Next, the PPNSP TENG showed an output voltage of at 5 Hz, − 76.5 V to 88 V and − 1.44 to 1.6 µA, respectively. On the other hand, PPNSP-EMI.BF 4 e-skin was shown upto − 86.5 V to 109 V and − 2.08 to 2.35 µA at 5 Hz owing to EMI.BF 4 IL interpenetrated within the polymer network, which enhanced the accumulation of high power densities. Significant enhancement was observed in both voltage and current after doping with EMI.BF 4 IL into the PPNSP film by dip coating method. The homogeneous deposition EMI.BF 4 into PPNSP, quantum jump enhancement was observed, and the quantitative output performance of PPNSP-EMI.BF 4 was determined due to active mobility of EMI.BF 4 ions within the host PPNSP polymer film as well as the surface. In specific, it can be seen that the PPNSP-EMI.BF 4 -PTFE TENG showed higher electrical output than the other two systems (i.e., NSP.H + -PTFE, and PPNSP-PTFE TENG) owing to the availability of abundant mobile BF 4 − ions from EMI.BF 4 IL. Next, performance characteristics of PPNSP.EMI.BF 4 -PTFE TENG showed the load resistance analysis, power density calculations, and device stability at 5 Hz applied frequency, and load resistance was increased from 100 ohm Ω to 570 MΩ. The areal power density of the device was 33.2 mW, and the load resistance was 570 MΩ. Additionally, the unidirectional output was stored in energy storage systems such as capacitors and batteries and the rectified Voc of PPNSP.EMI.BF 4 -PTFE e-skin TENG device. The performance stability of Voc and Jsc of PPNSP.EMI.BF 4 -PTFE TENG system was checked at a 5 Hz applied frequency, and it was stable for ~ 10,000 cycles without any fluctuations. The durability and stability of the proposed system showed excellent harvesting performance and superior mechanical strength without any surface damage. The present results have suggested that the controlled self-assembly process for strong ion-ion connections and ion transport nanochannels can be used for tailoring superior TENG applications, which are potentially required for next-generation electronic products such as wearable soft electronics, flexible displays, and smart mobile phones.",
"introduction": "Introduction For two decades, fossil fuel reservoirs are gradually decreasing owed to abundant consumption of human society. Alternatively, scientists and engineers are focused to develop clean, and renewable energy techniques which includes piezoelectric, photoelectric, thermoelectric, pyroelectric, electrostatic, and electromagnetic devices to reduce the dependency on fossil fuels 1 – 5 . They can work as portable energy harvesters to drive low-powered electronic devices and work as self-powered sensors 6 , 7 . In addition, numerous technologies suffer from issues concerning device design limitations, the development of highly efficient composite films, long processing periods, power control circuits, unsteady production performance, packaging, and shelf-life issues. In recent times, triboelectric nanogenerators (TENGs) have established global commitment for the collecting of viable green energy from ambient resources 8 . TENGs were technologically advanced based on an amalgamation of contact separation electrification and electrostatic stimulation for scavenging attenuated mechanical energy via triboelectric resources 9 , 10 . The appropriate selection of triboelectric paired materials and their coherent design can upsurge the rate of energy collection and conversion efficiency 11 , 12 . At regular intermissions of TENG resources with oppositely charged electrets, ions or electrons can be driven to movement through the external load and produce a continuous current 13 .\n In recent years, sulfonated polyimide block copolymers (SPIs), and PEDOT: PSSa conducting polymers are being used as the most promising organic polymers that contain regular porous nanochannels and present numerous manufacturing qualities, such as film forming ability with resilience, elasticity, bendability, stretching ability, long shelf life, and electrochemomechanical properties 14 – 19 . Although several groups have used a series of SPIs as ionomers for high-performance fuel cell applications and actuators but not applied for TENG applications 20 , 21 . PEDOT: PSSa is a combination of polymer mixture of two conductive ionomers. One of the components in this mixture is made up of sodium polystyrene sulfonate and some of the sulfonyl groups are deprotonated and carry a negative charges 22 – 24 . Recently, Jang have designed conductivity enhancement experiments by mixing of ionic liquids (ILs) into PEDOT: PSSa conducting polymer for a noteworthy conductivity improvement 25 . The ions exchange between PEDOT: PSSa, and IL can assist PEDOT to decouple from PSSa, and harvest bulk-scaled loosely bound conducting domains 26 . The spontaneous ion exchange followed by nano channeled phase between PEDOT and PSSa chains, with formation of a regular π–π stacked PEDOT cations were intercalated by IL anions, is further sustained by molecular dynamics performed on bulk PEDOT: PSSa models in solution 27 – 33 . Furthermore, the IL can act as a bridge electrolyte between the core SPI polymer and PEDOT: PSSa electrodes, it can provide a further significant intensification of charge distribution performance and extended durability. In addition, the EMI.BF 4 is evaluated as a low temperature electrolyte additive, and enhance the cycle stability 34 . Yan et al. have reported porous g-C3N4, and Mxene dual confined FeOOH Quantum Dots for superior energy storage in an EMI.BF 4 IL for high efficiency supercapacitors 35 . From their exploration, we inspired to use an EMI.BF 4 for emerging high performance TENG by producing fast ion-hopping rate between cations and hydrophilic NSP.H + , and PEDOT: PSSa conduction layers by ion-ion interactions through porous nano channels 36 . To develop an economically viable, highly endurable, and exchange of interionic ionic electrets between electropositive, and electronegative porous next-worked NSP.H + , and PPNSP membranes were used to generated high output voltage and currents through TENG process. Additionally, we presented a simple but ultrafast two-step synthesis including dry casting and drop casting for a high-performance TENG. The ionic electrets networked, hydrogen ion (H + ) rich naphthalene sulfonated polyimide (NSP.H + ) ionic membrane, polyethylenedioxythiophene (PEDOT): polystyrene sulfonate (PSSa) as a conducting electrode layer, and 1-ethyl-3-methylimidazolium tetra fluoroborate [EMI.BF 4 ] ionic liquid (IL) as the mobile electrolyte 37 . Molecular-level region-specific interaction of cations and anions in IL with hydrophilic-hydrophobic co-blocks of NSP mediocre is utilized for building a self-assembled ionic networked polymer with uninterrupted and intersected ion transport nano channels for high-performance TENG 38 . Besides, the determination of the present work is to fabricate an ultrafast solvent drop-casting method to produce a combined networked electropositive ionic layer of PEDOT: PSSa—EMI.BF 4 -NSP (PPNSP) e-Skin strongly follows an ion-ion hopping mechanism (stages 1 and 2) using ionic electrets. The ionic conductivity, and ion exchange capacity of PPNSP are increased up to 3.3 times and 3.5 times through ionic electrets by an ion hopping mechanism that established that the higher density of excess protons (H + ions) on the active surface can activate polarized charges to produce a higher TENG output voltage (Voc) and output currents (Isc) when interacting with the electronegative PTFE surface by contact separation mode (stages 3 and 4). The developed PPNSP-PTFE-TENG system is a virtuous candidate for the generation of higher Voc and Isc through a ion-ion hopping mechanism due to its significant benefits, such as a π–π stacked layer that helps to push and pull quick response to travel the ions via interconnected neural networked knots when undergoing contact-separation time. As a result, the novel PPNSP.EMI.BF 4 -PTFE TENG produced the Voc and Jsc values from 49.1 V and 1.03 µA at 1 Hz, 74 V and 1.45 µA at 2 Hz, 122.3 V and 2.23 µA at 3 Hz, and 171 V and 3.6 at 4 Hz, and 195.5 V and 4.43 µA at 5 Hz, respectively, 92 V and 2 mA, 74 V and 1.4 mA, 52 V and 0.93 mA, 32 V and 0.3 mA open-circuit voltages Voc and Isc at 5 Hz, 4 Hz, 3 Hz, 2 Hz, and 1 Hz, respectively, as shown in Fig. 1 . Figure 1 Schematic representation of PPNSP-EMI.BF 4 -PTFE TENG for harvesting energy through ionic electrets and ion-hopping mechanisms. Stage 1. Synthesis of NSP.H + s: PEDOT: PSSa composite film, Stage 2 Fabrication of ionic net-worked conductive composite electronic skin (e-Skin); Stage 3. Size of (4 cm × 4 cm) PPNSP.EMI.BF 4 e-skin and PTFE, and vertical contact separation mode TENG; Stage 4. The Voc and Jsc of conductive PPNSP-EMI.BF 4 -PTFE e-Skin TENG.",
"discussion": "Results and discussion Preparation of naphthalene SPI (NSP.H + ), NSP.H + -PEDOT: PSSa (PPNSP), PPNSP-EMI.BF 4 composite e-skins Figure 2 a shows the synthesis of NSP - . H + oligomer film by the sequential addition of monomers of BDSA. TEA, NTDA, and ODA in m-cresol using a literature method by random copolymerization through the addition of copolymers in a one-pot method. Next, NSP.H + . TEA was subjected to a proton exchange reaction in 3 N HCl to generate NSP - . H + oligomer membrane, Fig. 2 b. Next, the PPNSP film was fabricated using the solvent dip-coating method by mixing NSP.H + oligomer film in 20 wt% PEDOT: PSSa in DMSO, Fig. 2 c. Later, ionic networked PPNSP-EMI.BF 4 composite film was synthesized using the dip-coating method by immersing the PPNSP film prepared in 20 wt% EMI.BF 4 in DMSO for 12 h, respectively. During dip-coating process, EMI.BF 4 is permeated into the host PPNSP through an ion hopping mechanism through a solvent sorption method. Additionally, long chain oligomers can show inter- and intramolecular charge transfer complex (CTC) singularities intercalated between SO 3 H groups with EMI.BF 4 . In addition, the hydrophilic SO 3 H functional groups were strongly intercalated with each other between PPNSP and EMI.BF 4 through hydrophilic-hydrophilic interactions to generate higher Voc and Jsc 39 , 40 .\n Polymerization of PEDOT: PSSa, and interaction of PPNSP, and EMI.BF 4 Figure 3 shows the mechanistic approach of the formation of PPNSP: EMI.BF 4 e-skin TENG. From oxidation and dimerization of ethylene-dioxythiophene (EDOT) through an oxidative polymerization procedure catalyzed by metal cations, and the universal oxidant is present in various organic solvent systems. The stage 1, showed the oxidation of EDOT (1) monomers into unstable cationic radical structures (2). The unstable radicals were rearranged into a stable dimer (3) by reaction steps that involved combination and deprotonation, as shown in stage 2. In addition, a neutral PEDOT chain itself has a conjugated network with alternate double bonds with a low energy band gap and low oxidation potential. Ionic interactions were occurred with NSP.H + , and PEDOT: PSSa to generate the PPNSP e-skin, stage 3. Next, the PPNSP was soaked in 20% EMI.BF 4 in DMSO to generate the PPNSP: EMI.BF 4 composite e-skin, stage 4. Continuous recurrence of those steps resulted in the formation of well-dispersed, and doped PEDOT: PSSa solution showed the chemical structures of 3,4-ethylenedioxythiophene (EDOT), and poly(3,4-ethylenedioxythiophene) (PEDOT). The structure of the as-prepared PEDOT is made up of benzoid, and quinoid forms. The benzoid structure possesses a π-electron localized, conjugated structure that remains largely unaffected by external sources. In contrast, the quinoid form of PEDOT owns a delocalized state of π-electrons, which can be strongly exaggerated by solvent treatment 41 . In the electrically active, oxidized state, there remains a positive charge on every PEDOT polymer chain. The charges on the backbone were balanced with an anions are from small molecules or a macromolecular anions such as poly (4-styrene sulfonic acid) (PSSa). This higher charge transfer performance of the newly developed PPNSP: EMI.BF 4 e-skin TENG through an ion-hopping mechanism that induces high ionic conductivity and tuned the mechanical properties, resulting from strong ionic interactions among the NSP.H + , EMI.BF 4 , and PEDOT: PSSa, and interconnected networked polymer matrix 42 , 43 . Figure 3 Mechanistic approach of the formation of oxidative and dimerization of EDOT, stage 1; polymerization and isomerization of EDOT, stage 2; Ionic interactions of EMI.BF 4 with NSP - . H + , and PEDOT: PSSa, stage-3; and ion hopping mechanism between PPNSP and EMI.BF 4 , stage 4. SEM analysis of NSP.H + , PPNSP, and PPNSP-EMI.BF 4 composite films Scanning Electron Microscopy (SEM) images of the NSP.H + and PPNSP films are shown in Fig. 4 and reveal the surface morphologies. In particular, Fig. 4 a and b show the wrinkle-free plain surface with no obvious changes. However, the inbuilt hydrophilic –SO 3 H groups were attached to the host polyimide co-blocks that could intercalate and generate nano level distances between the oligomeric networked chains. Significant submicron-sized porous grooves and micron-sized crests with tightly packed networks were found on the surface, as shown in Fig. 4 c. The hollow groves and crests were created due to a significant ionic network on the surface morphology. In the cross-sectional view, hydrophilic and hydrophilic interactions take place between NSP.H + , and PEDOT: PSSa through hydrogen bonding that created a loosely bound network between them, Fig. 4 d 21 . The hydrophilic-hydrophilic ionic system can establish a flexible interpenetration complex between NSP – H + , and PEDOT: PSSa to improve the charge densities on the polymeric surface. The ionic networked morphology can exchange ions through strong ionic knots, which can travel through hydrophilic ionic channels and strongly support the ion hopping mechanism. After impregnation of EMI.BF 4 IL, the thickness of PPNSP e-Skin was increased due to swelling of ionic liquid. Figure 4 e shows the strong spherical aggregations on the surface by ionic clusters 44 . In addition, the inset image clearly indicates the ionic clusters at nanoscale levels. In the cross-sectional view, a loosely bounded net-like channeled network appears due to the strong intercalation between PEPDOT: PSSa and EMI.BF 4 IL. Additionally, the high-resolution surface and cross-sectional images display the formation of zigzag net-like channels. In particular, EMI + cations were deposited on the hydrophilic regions as bright spots, which clearly indicate the formation of PPNSP-EMI + by an ion hopping mechanism, as shown in Fig. 4 f. The homogeneity of the blend membrane confirmed the strong ionic interactions, which included ionic cross-linking and hydrogen bonding between EMI.BF 4 and the PPNSP composite membrane, resulting in the enhanced interfacial compatibility and mechanical stiffness of the composite membrane. In addition, ionic cross-linking and hydrogen bonding can generate higher charges on the surface. These charges can enhance the open circuit voltage and short circuit currents. The overall morphologies were strongly justified to accelerate the charges within the charge transfer complexes when a contact-separation mode TENG was performed 39 . The FT-IR, XRD, Stress–strain (SS) curves, and TGA analyses of NSP.H + , PPNSP, and PPNSP.EMI.BF 4 were reported in Supporting information . Figure 4 FE-SEM images of the surface morphologies of NSP.H + , PPNSP, and PPNSP-EMI.BF 4 composite films; ( a , b ) regular plain surface; ionic networked submicron to micron hallow grooves and knots on the surface, ( c , d ); ionic clusters on the surface of PPNSP film by aggregations of EMI.BF 4, (inset image, strongly glued PPNSP-EMI.BF 4 ionic clusters, which can help to promote the ion hopping mechanism, at 1 μm magnification) ( e ), and loosely bonded ionic networks in the cross-sectional view of PPNSP-EMI.BF 4 composite film ( f )."
} | 4,917 |
26692424 | PMC4686896 | pmc | 1,538 | {
"abstract": "In the current global climate change scenario, stressors overlap in space and time, and knowledge on the effects of their interaction is highly needed to understand and predict the response and resilience of organisms. Corals, among many other benthic organisms, are affected by an increasing number of global change-related stressors including warming and invasive species. In this study, the cumulative effects between warming and invasive algae were experimentally assessed on the temperate reef-builder coral Cladocora caespitosa . We first investigated the potential local adaptation to thermal stress in two distant populations subjected to contrasting thermal and necrosis histories. No significant differences were found between populations. Colonies from both populations suffered no necrosis after long-term exposure to temperatures up to 29 °C. Second, we tested the effects of the interaction of both warming and the presence of invasive algae. The combined exposure triggered critical synergistic effects on photosynthetic efficiency and tissue necrosis. At the end of the experiment, over 90% of the colonies subjected to warming and invasive algae showed signs of necrosis. The results are of particular concern when considering the predicted increase of extreme climatic events and the spread of invasive species in the Mediterranean and other seas in the future.",
"discussion": "Discussion Understanding how global change-related disturbances interact is critical in the current context of rapid changes and the enlargement of the geographical overlap of stressors. Using an experimental approach, we showed that the interaction of two of the most widespread global change stressors (warming and invasive species) trigger drastic additive effects on a temperate reef-builder coral that shows uniform thermotolerance despite contrasting temperature regimes inherent to its geographical distribution. While mortalities of Cladocora caespitosa , related to abnormally high summer temperatures, have been reported in many Mediterranean coastal zones exposed to differing thermal regimes (La Spezia, NW Italy 13 ; Columbretes, E Spain 5 ; Mljet and Piran, Croatia and Slovenia 15 ; Cyprus 14 ), information on its thermotolerance in relation to geographical differences in temperature is lacking. Strikingly, our results showed that C. caespitosa colonies from a population recurrently affected by positive thermal anomalies and mortality events (Columbretes Islands 5 ) and colonies thriving in much colder waters and at least not recently impacted by mortality events (Medes Islands: this study), did not exhibit signs of necrosis after being exposed over 68 days to thermal stress conditions. In the present study, temperatures in the aquaria experiment were similar or even higher to those recorded in summers triggering necrosis in the Columbretes Islands 5 , and the maxima were over 4 °C higher than those registered in the Medes Islands. However, in contrast to what has been described in situ 5 or in previous experimental studies 29 30 31 , no significant effects on necrosis and photosynthetic efficiency were recorded. Previous studies on temperate octocorals determined similar discrepancies between experimental and in situ studies 23 24 26 . These contrasting results support the hypothesis of the multi-factorial cause of the necrosis events affecting C. caespitosa 5 or similar species 32 . This shows that this species, despite its potential broad thermal tolerance in the aquaria, is subjected to unstudied factors and synergies that need further investigation. Overall, the results of the thermotolerance experiment are noteworthy, considering that C. caespitosa is an endemic species that has been thriving in the Mediterranean Sea for the last 2.5 million years (Late Pliocene 33 ) and with populations showing a certain grade of isolation and genetic differentiation 34 . Therefore, hypothetically, local thermal adaptation could be expected as described in other Mediterranean and tropical benthic species 24 25 35 36 . However, our results are similar to those obtained for the Mediterranean sleractinian coral Oculina patagonica 32 , which belongs to the same family as C. caespitosa 37 . These contrasting results raise further questions on the potential for thermal adaptation in temperate corals and further studies are needed to understand these processes. While neither necrosis nor a strong reduction of photosynthetic efficiency was registered during the thermotolerance experiment, both variables were significantly affected when subjected to both a temperature increase and invasive algae ( Figs 4 and 5 ). However, it has to be noted that photosynthetic efficiency showed a partial recovery at the end of the experiment in the covered treatments, both under control and enhanced temperature. This recovery could be associated to a potential low-light photo-acclimation mechanism of the zooxanthellae in symbiosis with C. caespitosa , as has been reported to occur after some days in tropical corals 38 and in agreement with the capability of C. caespitosa of dwelling in contrasting light conditions 39 . Both the invasive algae and the artificial material treatments had negative effects on the coral, which could point to a primary influence of a physical effect of overgrowth. However, although secondary metabolites of W. setacea have not been studied to date, chemical interactions should not be disregarded, taking into account that comparatively the alga treatment had increased negative effects, both in tissue necrosis and photosynthetic activity. These types of interactions have been suggested in tropical reefs between macroalgae and corals 40 or related to algae-associated pathogens 41 . It should be noted that turf algae have been reported to notably overgrow and damage coral tissue in tropical corals. For example, the filamentous red alga Anotrichium tenue was reported to damage Porites spp . and outcompete coral colonies through physical and chemical mechanisms 42 . In our study, besides potential chemical interactions, thermal stress was exacerbated by the physical shield effect of W. setacea directly covering the colonies. As mentioned in previous studies on sponges, this effect reduces or even impairs water exchange with the associated impact on respiratory and nutritional needs 43 . Hence, similar effects on the physiological status of C. caespitosa colonies can also be expected because its nutrition relies on both autotrophy and heterotrophy, thus being highly dependent of water exchange and light availability 44 . Recently, several studies have underlined the crucial need to account for multiple-stressor interactions in ecological studies and conservation planning 7 9 10 . The results reported in our study on the synergistic effects between warming and invasive species are worrying, since impacts on C. caespitosa , in particular, and other macroinvertebrates, in general, could notably increase when considering future warming and the spread of invasive species in the Mediterranean region."
} | 1,774 |
30774969 | PMC6370796 | pmc | 1,539 | {
"abstract": "Disturbance is known to affect the ecosystem structure, but predicting its outcomes remains elusive. Similarly, community diversity is believed to relate to ecosystem functions, yet the underlying mechanisms are poorly understood. Here, we tested the effect of disturbance on the structure, assembly, and ecosystem function of complex microbial communities within an engineered system. We carried out a microcosm experiment where activated sludge bioreactors operated in daily cycles were subjected to eight different frequency levels of augmentation with a toxic pollutant, from never (undisturbed) to every day (press-disturbed), for 35 days. Microbial communities were assessed by combining distance-based methods, general linear multivariate models, α-diversity indices, and null model analyses on metagenomics and 16S rRNA gene amplicon data. A stronger temporal decrease in α-diversity at the extreme, undisturbed and press-disturbed, ends of the disturbance range led to a hump-backed pattern, with the highest diversity found at intermediate levels of disturbance. Undisturbed and press-disturbed levels displayed the highest community and functional similarity across replicates, suggesting deterministic processes were dominating. The opposite was observed amongst intermediately disturbed levels, indicating stronger stochastic assembly mechanisms. Trade-offs were observed in the ecosystem function between organic carbon removal and both nitrification and biomass productivity, as well as between diversity and these functions. Hence, not every ecosystem function was favoured by higher community diversity. Our results show that the assessment of changes in diversity, along with the underlying stochastic–deterministic assembly processes, is essential to understanding the impact of disturbance in complex microbial communities.",
"conclusion": "Implications and concluding remarks The implications of this study relate to both process engineering and environmental management. Sludge communities within wastewater treatment are not only model systems in microbial ecology, 65 but also a key driver for water sanitation and the environmental impact of anthropogenic water discharges. 66 Disturbances could promote stochastic assemblages of the sludge communities, which despite harbouring higher diversity could lead to variable overall ecosystem function. This could be the reason why after similar perturbations the process outcome differs, causing operational problems for water utilities. 67 Furthermore, cases where disturbance temporally favours stochastic assembly could lead to a different final community after the perturbation, 27 which could compromise the expected ecosystem function. More research is needed to identify such scenarios in practice. We described how different frequencies of disturbance affected ecosystem function and bacterial community diversity and assembly in a closed microcosm bioreactors system. Communities were assessed through different molecular methods that nonetheless yielded very similar patterns. Furthermore, besides the wastewater treatment microbial community, other complex microbial systems (e.g., the gut microbiome) might display similar responses to disturbance. We argue that changes not only in diversity but also in the underlying deterministic–stochastic assembly mechanisms should be evaluated in studies of the effects of disturbance on such systems. For such an assessment, both replication and wide-enough disturbance ranges are key. Additionally, the ISH could be evaluated within open systems to include the effect of dispersal processes. This calls for more studies in microcosm 45 , 68 and mesocosm settings, as well as meta-analysis from full-scale application studies.",
"introduction": "Introduction Understanding what drives patterns of community succession and structure remains a central goal in ecology 1 , 2 and microbial ecology, 3 especially since community diversity and assembly are thought to regulate the ecosystem function. 4 , 5 Assembly processes can be either stochastic, assuming that all species have equal fitness and that changes in structure arise from random events of ecological drift, 6 or deterministic, when communities form as a result of niche diversity shaped by abiotic and biotic factors. 7 Deterministic and stochastic assembly dynamics have been proposed to simultaneously act in driving assembly patterns observed in nature. 8 – 12 This has stimulated scientific discourse including modelling of experimental data 13 – 16 and both observational and manipulative experimentation in a variety of ecosystems, like deserts on a global scale, 17 groundwater, 18 subsurface environments, 2 , 19 , 20 soil plant–fungi associations, 21 rock pools, 22 water ponds, 23 and sludge bioreactors. 15 , 24 , 25 These prior studies emphasized the need to understand what governs the relative balance between stochastic and deterministic processes and what conditions would lead to stochastic processes overwhelming deterministic processes, particularly under disturbance. 20 To investigate their roles, well-replicated time series experiments are needed. 18 , 25 Disturbance is defined in ecology as an event that physically inhibits, injures, or kills some individuals in a community, creating opportunities for other individuals to grow or reproduce. 26 When disturbance is long-term or continuous, it is classified as press-disturbance. 27 Disturbance is deemed the main factor influencing variations in species diversity 28 and structuring of ecosystems, 27 , 29 but a clear understanding of its outcomes is lacking. 30 Particularly, the intermediate disturbance hypothesis (IDH) 31 predicts that diversity should peak at intermediate levels of disturbance due to trade-offs between species’ ability to compete, colonize ecological niches, and tolerate disturbance. The IDH has been influential in ecological theory, as well as in management and conservation, 32 but its predictions do not always hold true. 28 , 33 For example, in soil and freshwater bacterial communities, different patterns of diversity were observed with increasing disturbance frequency with biomass destruction 34 and removal 35 as disturbance type, respectively. Meanwhile, the effect of varying frequencies of non-destructive disturbances on bacterial diversity remains unknown. Furthermore, the IDH predicts a pattern but it is not a coexistence mechanism as it was originally purported to be. 36 Hence, its relevance is being debated 37 , 38 with multiple interpretations and simplicity as the main points of critique. To date, the mechanisms behind the observed patterns of diversity under disturbance remain to be elucidated. 39 , 40 The objective of this work was to test the effect of disturbance on the bacterial community structure, diversity, and ecosystem function of a complex bacterial system, with emphasis on the underlying assembly mechanisms. We employed sequencing batch bioreactors inoculated with activated sludge from an urban wastewater treatment plant, in a laboratory microcosm setup with eight different frequency levels of augmentation with toxic 3-chloroaniline (3-CA) as disturbance. Triplicate reactors received 3-CA either never (L0, undisturbed), every 7, 6, 5, 4, 3, and 2 days (L1–6, intermediately-disturbed), or every day (L7, press-disturbed) for 35 days. Chloroanilines are toxic and carcinogenic compounds and few bacteria encode the pathways to degrade 3-CA, 41 which is also known to inhibit both organic carbon removal and nitrification in sludge reactors. 42 Microcosm studies are useful models of natural systems, 43 can be coupled with theory development to stimulate further research, 44 and by permitting easier manipulation and replication can allow inference of causal relationships 45 and statistically significant results. 46 We analysed changes in the ecosystem function over time by measuring the removal of organic carbon, ammonia, and 3-CA, as well as biomass. Changes in community structure were examined at different levels of resolution using a combination of metagenomics sequencing and 16S rRNA gene fingerprinting techniques. Such changes were assessed by employing a combination of ordination tools, diversity indices, cluster analysis, univariate and multivariate statistical analyses. We also explored how diversity was related to function, focusing on trade-offs. Furthermore, the role of stochasticity in community assembly was investigated by employing null model techniques from ecology. We hypothesized that time would lead to a decrease in α-diversity at the extreme sides of the disturbance range due to deterministic adaptation to the environment, while less predictable conditions at intermediate disturbance levels would lead to higher α-diversity and stochastic assembly. Consequently, replicates at intermediately disturbed levels should display higher variability in terms of both community structure and function, compared with the ones at the extreme sides of the disturbance range where the opposite (that is, less variability) should occur.",
"discussion": "Results and discussion Overall community dynamics and differentiation of clusters Bacterial community structure displayed temporal changes and varied between disturbance levels, as assessed by 16S rRNA gene terminal restriction fragment length polymorphism (T-RFLP) (Fig. 1 ). Constrained ordination showed a defined cluster separation with 0% misclassification error of the outermost levels L0 and L7 from the remaining intermediate levels L1–6 (Fig. 1a ). Overall community structure differed over time with a dispersion effect after 14 days (Fig. 1b ). Levels across disturbance and time factors showed significant differences (PERMANOVA P = 0.003, Supplementary Table 1 ), with a non-significant interaction effect ( P = 0.15). Disturbance was the factor responsible for the observed clustering (Fig. 1a ) and not heteroscedasticity (PERDIMSP P = 0.35). Fig. 1 Microbial community dynamics across disturbance frequencies and time, as assessed by 16S rRNA gene terminal restriction fragment length polymorphism (T-RFLP) fingerprinting. a Canonical analysis of principal coordinates (CAP, constrained ordination) plot, with disturbance levels as differentiation criteria, shows cluster differentiation for L0 (CAP1 axis) and L7 (CAP2 axis) from intermediately disturbed levels (L1–6). Disturbance levels: L0 [light-green triangles], L1 [blue upside-down triangles], L2 [light-blue open squares], L3 [open red rhombuses], L4 [purple circles], L5 [black crosses], L6 [green x-symbols], and L7 [blue stars]. b Non-metric multidimensional scaling (NMDS, unconstrained ordination) shows temporal dispersion effect. Days: 14 [open triangles], 21 [light-grey upside-down triangles], 28 [dark-grey squares], and 35 [black rhombuses] Ecosystem function dynamics and trade-offs The undisturbed community (L0) was the only one with complete dissolved organic carbon (COD) removal and nitrate generation without nitrite residuals, while the press-disturbed community (L7) was the only one where nitrification products were never detected and also had the lowest biomass (Fig. 2 ). Initially, reactors at the disturbed levels showed an inability to remove all of the 3-CA (with the exception of L1). Such lack of 3-CA degradation was accompanied by a reduction in organic carbon removal in the first 3 weeks (Fig. 2a , Supplementary Figure 2a,c,e ), and a complete inhibition of nitrification with subsequent accumulation of ammonium (Fig. 2b , Supplementary Figure 2b,d,f ). Removal of 3-CA recovered and was above 95% for all disturbed levels after 28 days (Supplementary Figure 2g ), but COD removal was still not 100% despite complete 3-CA removal towards the end of the experiment (Fig. 2c ). Fig. 2 Process performance indicators across disturbance levels. Effects include temporal changes and trade-offs in community function. a , c Percentage of organic carbon as chemical oxygen demand (COD, black circles) and 3-CA (open purple rhombuses) removal for all levels (negative values represent accumulation). c Biomass as volatile suspended solids (VSS, open green squares). b , d Concentration of ammonium (black rhombuses), nitrite (open blue triangles), and nitrate (open red circles) as nitrogen for all levels. Data are from days 7 ( a , b ) and 35 ( c , d ) of the study (for all time points sampled, see Supplementary Figure 2 ). Mean ± s.d. ( n = 3) are shown. Undisturbed L0 replicates had consistent organic carbon removal and complete nitrification, whereas press-disturbed L7 never showed nitrification and had the lowest final biomass. Intermediate levels L1–6 displayed changing functionality with higher s.d. values that increased over time Nitrification was detected on day 21 for L1 and later for other disturbance levels, except for the press-disturbed L7. The dominant NO X component was nitrite, but some nitrate was also produced (Fig. 2d , Supplementary Figure 2h–j ). At the end of the study, there was a significant negative Spearman’s correlation between organic carbon removal and nitrification (Supplementary Figure 3a-b ) in terms of nitrite ( ρ = −0.901) and nitrate production ( ρ = −0.697). Biomass values on day 35 differed significantly among levels with the highest value at L1 and the lowest at L7 (Fig. 2c ). There was a significant positive correlation between biomass and nitrification in terms of nitrite ( ρ = 0.466) and nitrate production ( ρ = 0.656) (Supplementary Figure 3c-d ). Intermediate levels of disturbance displayed increased dissimilarity with time To distinguish the effect of disturbance from temporal community dynamics (Fig. 1 ), community dissimilarity was assessed on the T-RFLP dataset at each time point by ordination analysis using principal coordinates analysis (PCO) (Fig. 3a, b ), non-metric multidimensional scaling (NMDS), and canonical analysis of principal coordinates (CAP) with cluster similarity analysis (Supplementary Figure 4 ). The combination of constrained and unconstrained ordination methods allowed differentiating location from dispersion effects in community structure. 47 L0 was consistently different in all ordination plots and L7 differed after 21 days, both with 0% misclassification error at all time points for CAP plots. Dispersion effects within intermediate levels were evident in the unconstrained ordination plots with higher differentiation of biological replicates after 35 days (Fig. 3b ), coinciding with the production of nitrite and low levels of nitrate (Fig. 2d ). Community differentiation was statistically significant from day 21 onwards as supported by PERMANOVA and PERMDISP (Supplementary Table 1 ). Fig. 3 Community dissimilarity assessed by principal coordinates analysis (PCO) plots for all disturbance levels on T-RFLP datasets on days a 14 and b 35 of the study. Ovals with dashed lines represent 80% similarity calculated by group average clustering. Disturbance levels: L0 [light-green triangles], L1 [blue upside-down triangles], L2 [light-blue open squares], L3 [open red rhombuses], L4 [purple circles], L5 [black crosses], L6 [green x-symbols], and L7 [blue stars]. c Procrustes analysis on PCO at day 35 comparing metagenomics (circles) and T-RFLP (triangles) datasets. Lines unite data points from the same reactor ( n = 24). Same colour palette as for disturbance levels. Tests comparing both methods were statistically significant (Supplementary Table 2 ). Intermediate treatments’ (L1–6) within-treatment dissimilarity increased with time. L0 and L7 clusters consistently displayed higher similarity after 14 days Metagenomics community analysis validates observations from fingerprint dataset β-Diversity patterns observed from 16S rRNA gene amplicon T-RFLP data on day 35 were significantly similar to those from shotgun metagenomics data. A Mantel test on Bray–Curtis distance matrixes for both datasets ( n = 24) yielded significant similarity ( r = 0.73, P = 0.002). Procrustes tests of comparisons within ordination methods of PCO (Fig. 3c ) and NMDS also yielded significant similarities for both datasets ( P = 0.002, Supplementary Table 2 ). Multivariate PERMANOVA tests on the metagenomics dataset produced statistically significant results, but with significant heteroscedasticity as shown by PERMDISP (Supplementary Table 1 ). We resolved these mean–variance relationship concerns by running a general linear multivariate models (GLMMs) test to fit the data to a negative binomial distribution. Both residuals vs fitted and mean–variance plots supported the choice of a negative binomial distribution for the regression model (Supplementary Figure 5 ). The analysis of deviance of the regression rejected the null hypothesis of no difference between communities at different disturbance levels, independent of heteroscedasticity ( P = 0.0149). Higher α-diversity for intermediately disturbed treatments and diversity-function correlations The observed patterns in α-diversity were time-dependent, as diversity decreased over time with respect to the initial sludge inoculum (Fig. 4a , T-RFLP dataset). Such a temporal decrease in diversity was higher at the extreme ends of the disturbance range, resulting in a parabolic pattern on day 35 (Fig. 4b, c ). The final α-diversity pattern based on Hill number 2 D was similar for both T-RFLP and metagenomics methods (Fig. 4b ), although the latter showed higher variability. For the metagenomics dataset we also calculated the lower-order Hill numbers ( 0 D, 1 D) which give higher weight to less abundant operational taxonomic units (OTUs). They displayed the same parabolic pattern (Fig. 4c ). Welch’s ANOVA tests were statistically significant for all Hill numbers ( P < 0.01, P = 0.022 for 2 D metagenomics ). Additionally, there were strong significant correlations between α-diversity and ecosystem function (Supplementary Figure 6 ), focusing on the more robust estimators of microbial diversity 1 D and 2 D. 48 Both 1 D and 2 D correlated positively with ammonia removal and nitrite generation (Supplementary Figure 6a-b ), while 2 D had a positive correlation with biomass (Supplementary Figure 6c ) but a negative correlation with organic carbon removal (Supplementary Figure 6d ). Fig. 4 α-Diversity patterns. a Temporal dynamics of Hill number 2 D for abundant OTUs, calculated from T-RFLP data across disturbance levels. b Hill number 2 D calculated from T-RFLP (black dashed bars) and metagenomics (grey solid bars) data at days 0 (seed) and 35 (disturbance levels L0–L7). c Hill numbers 0 D (black solid bars) and 1 D (blue solid bars) from metagenomics data on days 0 (seed) and 35 (L0–L7). Values represent mean ± s.d. ( n = 3). Characters above bars indicate Games–Howell post-hoc grouping Null model analysis suggests different assembly mechanisms across disturbance frequencies To test if the observed changes in β-diversity (Figs 1a and 3 , Supplementary Figure 4 ) were due to variations in the underlying stochastic and deterministic mechanisms or due to changes in α- and γ-diversity ratios (α:γ) alone, 49 we employed a null model analysis from Kraft et al. 50 on the bacterial genus-level metagenomics datasets on day 0 and day 35. The model estimated null β-diversity values after randomizing the location of each individual within the three independent reactors for each of the eight disturbance treatment levels, while keeping the total quantity of individuals per reactor, the relative abundance of each OTU, and the γ-diversity constant over 10,000 iterations. Under this model, stochastic assembly mechanisms were found to be higher for some intermediately disturbed levels (L2–L5) in terms of stochastic intensity (SI) and standard effect size (SES) values, which corresponded to communities less deviant from the null expectation (Fig. 5 ). SI was also higher at d35 with respect to the sludge inoculum (d0). Fig. 5 Influence of stochastic assembly mechanisms in bacterial communities as assessed by a stochastic intensity and b standard effect size (SES). Both metrics were calculated through null model analysis on the metagenomics genus-level dataset at days 0 (seed) and 35 (disturbance levels L0–L7). Each calculation involved all replicates of each treatment ( n seed = 2, n L0–L7 = 3) evaluated over 10,000 null model iterations. SES values closer to zero represent communities less deviant from the null expectation, implying stronger stochastic assembly processes. Overall, stochasticity was stronger for intermediate disturbance levels L2–L5 and also increased with respect to the sludge inoculum Deterministic and stochastic patterns of assembly amongst different disturbance levels Niche-structuring at both ends of the disturbance frequency range was suggested by community structure patterns and ecosystem function. The undisturbed (L0) and press-disturbed (L7) levels were distinct from each other as well as from the remaining intermediate levels, as supported by multivariate tests (both distance-based and GLMMs). The ordination plots and cluster analyses showed a clear separate clustering for the independent replicates of these two disturbance levels along the experiment, and particularly the constrained ordination plots displayed this with 0% misclassification error. Furthermore, the ecosystem function was clearly differentiated between L0 and L7, as well as being consistent across replicates at each level. We contend that the observed clustering is an indication that both the undisturbed and press-disturbed levels favoured deterministic assembly mechanisms, where the selective pressure due to unaltered succession (L0) or sustained toxic-stress (L7) promoted species sorting, resulting in similar community structuring among biological replicates over the course of the experiment. Conversely, the communities from intermediately disturbed levels (L1–6) did not form distinct clusters for any particular level through the experiment. Within-treatment dissimilarity among replicates increased over time, with some replicates being more similar to those of other intermediate levels. Concurrently, ecosystem function parameters also displayed within-treatment variability for L1–6. For example, the conversion of ammonia to NO X products, which was initially hampered when communities were still adapting to degrade 3-CA, was not the same across all equally handled independent replicates. The observed divergence across independent replicates is considered here as a strong indicator of stochasticity in community assembly. Additionally, the lower deviation for L2–L5 from expected β-diversity values estimated via null model analysis indicates a higher role of stochasticity at intermediate disturbance levels. Several processes might be promoting stochastic assembly, like strong feedback processes 51 that are linked to density dependence and species interactions, 52 priority effects, 53 and ecological drift. 54 Reactors within this study were designed as closed systems, hence stochastic dispersal processes 55 could not affect community assembly. We argue that there were different underlying stochastic–deterministic mechanisms operating in the resulting community assembly along the disturbance range of our study. Similarly, a study on groundwater microbial communities 18 found through null model analysis that both deterministic and stochastic processes played important roles in controlling community assembly and succession, but their relative importance was time-dependent. The greater role of stochasticity we found on day 35 concurred with higher observed variability in the ecosystem function and structure among replicates for intermediately-disturbed levels. Likewise, previous work on freshwater ponds tracking changes in producers and animals 49 found β-diversity (in terms of dissimilarity) increasing with stochastic processes. These observed patterns are also in accordance with ecological studies proposing deterministic and stochastic processes balancing each other to allow coexistence, 10 with communities exhibiting variations in the strength of stabilization mechanisms and the degree of fitness equivalence among species. 9 Thus, it is not sufficient to ask whether communities mirror either stochastic or deterministic processes, 8 but also necessary to investigate the combination of such mechanisms that in turn explain the observed community structures along a continuum. 9 Diversity–disturbance patterns and trade-offs with function We observed the highest α-diversity at intermediate levels as predicted by the IDH, 31 both in terms of composition ( 0 D) and abundances ( 1 D, 2 D). This finding is non-trivial in two aspects. First, Svensson et al. 32 have shown that most studies find support for the IDH by using species richness ( 0 D) rather than evenness or other abundance-related indices (like 1 D and 2 D). They suggested that low evenness at high disturbance levels could be caused by the dominance of a few disturbance specialists. Second, the use of richness for microbial communities is not reliable 48 since it is heavily constrained by the method of measurement, 56 which makes it hard to compare results from different studies using this metric. Additionally, for complex communities there is often a huge difference between the abundance of rare and abundant taxa. Hence, for microbial systems, it is reasonable to assess diversity in terms of more robust compound indices rather than richness, the reason why we focused on 1 D and 2 D for diversity-function analyses. Importantly, the observed pattern in α-diversity was time-dependent and resulted in an IDH pattern after 35 days. Temporal dynamics were expected since the sludge community experienced an initial perturbation in all reactors after transfer from a wastewater treatment plant to our microcosm arrangement. For the sludge inoculum, this implied changes in reactor volume, frequency of feeding (continuous to batch), type of feeding (sewage to complex synthetic media), immigration rates (open to closed system), and mean cell residence time (low to high). This was a succession scenario in which communities had to adapt to such changes along with the designed disturbance array. For L0 and L7, 2 D decreased over time in agreement with deterministically-dominated processes, probably because such levels represented the most predictable environments within our disturbance range. In contrast, intermediate levels either increased or maintained the same 2 D over time (after an initial decrease within the first 2 weeks), seemingly a case where niche overlap promoted stochastic assembly. 8 The emergence of an IDH pattern after time is coherent with findings in previous microcosm studies using synthetic communities of protists 57 and freshwater enrichment microbial communities. 35 Yet, none of these studies evaluated the relative importance of the underlying assembly mechanisms for the observed diversity dynamics. Additionally, both 1 D and 2 D were positively correlated with nitrification and productivity, suggesting that higher community evenness favours functionality under selective pressure, 58 but were negatively correlated with organic carbon removal. Thus, we cannot affirm that more diverse communities have better functionality without considering trade-offs. This supports the notion that higher α-diversity does not necessarily imply a “better” or “healthier” system. 56 In addition to the observed changes in OTU diversity, there was an evident variation in ecosystem function along the disturbance range studied (Fig. 2c, d ), a similar finding to that of previous studies with simpler planktonic communities. 59 Functional trade-offs are expected under disturbance since organisms need to allocate resources normally used for other functions to recover after a disturbance. 60 In our study, communities with higher biomass had lower organic carbon removal efficiencies, which together with the trade-offs described for nitrification, suggest the adoption of different community life-history strategies depending on the frequency of disturbance. The results presented here were all taxonomy-independent since our focus was on diversity, function, and mechanisms of community assembly (phylum-level community changes are provided as supplemental material Supplementary Figure 7 ). Taxonomy-independent approaches continue to be useful to describe diversity patterns and mechanisms of community assembly. 2 , 61 However, it has been proposed that species’ traits can predict the effects of disturbance and productivity on diversity. 62 Hence, further analysis of the different taxa and their genetic potential paired with the observed trade-offs in ecosystem function will aid in the understanding of potential life-history strategies 60 and their relationship with community aggregated traits 63 in the near future. Merging mechanisms of community assembly and alpha-diversity patterns: an intermediate stochasticity hypothesis Knowing that the validity of the IDH is still under debate 37 , 38 and that many different diversity–disturbance patterns have been reported, 28 , 30 , 33 we asked whether there is a relationship between the peaked pattern in diversity observed and the underlying stochastic–deterministic processes of community assembly. Under purely stochastic processes, diversity should vary randomly as all species have equal fitness, 55 unless some other mechanism acts to prevent this. It is recognized that, beyond empirical pattern description, an understanding of the underlying mechanisms is necessary to comprehend the outcomes of intermediate disturbance regimes. 30 , 40 , 64 We hypothesize that higher α-diversity at intermediate disturbance frequencies is the result of weaker stabilizing mechanisms (niches), which are stronger at extreme ends of the disturbance range. Stochastic mechanisms will produce even assemblages (higher α-diversity) at intermediately disturbed levels, whilst infrequent or too-frequent disturbances will favour some species over others (lower α-diversity). We propose this idea as the intermediate stochasticity hypothesis (ISH, Fig. 6 ) and contend that it should hold particularly for compound α-diversity indices, 48 since the underlying assembly mechanisms would affect taxa abundance distributions. Fig. 6 Intermediate stochasticity hypothesis (ISH) for community assembly under varying disturbances. Conceptual representation of the classic relationship between α-diversity and disturbance, 31 including the effect of underlying stochastic and deterministic processes driving bacterial community assembly. When intermediate disturbance regimes result in less predictable environments, specialized traits would be less advantageous to taxa, and the stochastic equalization of competitive advantages would lead to higher α-diversity. On the contrary, extreme ends of the range where conditions are recurrent would select for adapted organisms whose dominance would result in a lower α-diversity The ISH can be further portrayed by noting a key reasoning behind the IDH, namely, that a competition–colonization trade-off would lead to higher diversity at intermediate levels of disturbance. 31 In the context of our study, which comprised a closed system, colonization would come from the low abundance taxa that have an opportunity to grow after different disturbance frequencies reduced the competitive ability of previously dominant taxa. Stochastic mechanisms of ecological drift could then play a critical role in shaping the emerging structure of microbial communities 3 where random processes of birth, death, and reproduction can have an effect on which of these low abundance taxa will be more benefited as a result of intermediate disturbance frequencies. Drift could also lead to historical contingency and priority effects that are also stochastic, 53 where taxa that occupy early the disturbance-opened niches could reduce the availability of resources to other taxa whose abundance will then be limited. Such reasoning could explain why, while higher α-diversity was found at intermediate levels of disturbance in our study, community structure and ecosystem function differed across identically treated replicates. Implications and concluding remarks The implications of this study relate to both process engineering and environmental management. Sludge communities within wastewater treatment are not only model systems in microbial ecology, 65 but also a key driver for water sanitation and the environmental impact of anthropogenic water discharges. 66 Disturbances could promote stochastic assemblages of the sludge communities, which despite harbouring higher diversity could lead to variable overall ecosystem function. This could be the reason why after similar perturbations the process outcome differs, causing operational problems for water utilities. 67 Furthermore, cases where disturbance temporally favours stochastic assembly could lead to a different final community after the perturbation, 27 which could compromise the expected ecosystem function. More research is needed to identify such scenarios in practice. We described how different frequencies of disturbance affected ecosystem function and bacterial community diversity and assembly in a closed microcosm bioreactors system. Communities were assessed through different molecular methods that nonetheless yielded very similar patterns. Furthermore, besides the wastewater treatment microbial community, other complex microbial systems (e.g., the gut microbiome) might display similar responses to disturbance. We argue that changes not only in diversity but also in the underlying deterministic–stochastic assembly mechanisms should be evaluated in studies of the effects of disturbance on such systems. For such an assessment, both replication and wide-enough disturbance ranges are key. Additionally, the ISH could be evaluated within open systems to include the effect of dispersal processes. This calls for more studies in microcosm 45 , 68 and mesocosm settings, as well as meta-analysis from full-scale application studies."
} | 8,580 |
32455248 | PMC7241024 | pmc | 1,540 | {
"abstract": "The ring stain phenomenon is a critical\nhindrance to the distribution\nof the solute during drying for biochemical assays and materials deposition.\nHerein, we developed a substrate, characterized with hydrophilic spots\nsurrounded by hydrophobic areas, to suppress the ring stain effect,\nand fabricated four kinds of patterned surfaces to investigate the\nrelationship between the surface free energy and ring-suppressing\nperformance. We found that during the evaporation process, a drop\nwas constrained on the hydrophilic spot with a pinned contact line,\nand the ring stain effect was suppressed significantly. The suppressing\nperformance of the ring stain effect increases with surface free energy\ndifferences between the hydrophilic and hydrophobic regions.",
"conclusion": "3 Conclusions In summary, our experimental results reveal that the ring deposits\ncan be suppressed significantly on hydrophilic/hydrophobic patterned\nsurfaces. The improvement of uniformity of spots depends on surface\nfree energy differences between the hydrophilic and hydrophobic regions.\nOur results provide ways to better control the distribution of the\nsolute during drying, which is important for biochemical assays and\nmaterials deposition.",
"introduction": "1 Introduction Ring-like\ndeposits are commonly observed along the perimeter of\nevaporating drops, known as the “coffee ring” phenomenon,\nwhich is undesirable in many cases, including DNA chips, 1 , 2 painting, and printing. 3 , 4 In 1997, Deegan and\nco-workers 5 first proposed that an outward\ncapillary flow in a drying drop of liquid carried dissolved solids\nto the periphery, forming ring-like deposits, and two conditions are\nnecessary for forming the capillary flow: contact line pinning and\nevaporation from the edge of the drop. Since then, efforts for\nsuppressing the ring stain effect were\nmainly focused on three strategies based on physical chemistry: (i)\nattenuating the pinning of the contact line, (ii) disturbing the outward\ndirection of the capillary flow, and (iii) preventing the nonvolatile\nsolutes to be transported to the edge of droplet. 6 In short, it can be concluded that all these methods took\ncontrol of the drying process by transition of the liquid property,\ntransition of substrate property, interactions at solid–liquid\nor liquid–gas interfaces, and altering environmental conditions.\nThe details are as follows: If a droplet was pure ethanol instead\nof water, the drying process\nbehaved in the manner with a constantly decreasing contact radius\nat an essentially constant contact angle. 7 Jin et al. 8 significantly reduced the\ncoffee ring and improved the film uniformity by controlling the amount\nof ethanol added in the MnO 2 droplet. Addition of surfactants\nor zwitterionic detergents 9 and heating\nof the liquid were the methods based on the Marangoni flow effect, 10 which needed to change the properties of solutes\nsometimes. The coffee ring effect suppression could also be demonstrated\nby addition of a hydrosoluble polymer, 11 cellulose nanofiber, 12 or a biocompatible,\nsurfactant-like polymer (PEG) 13 into the\nrelative droplets. Shimobayashi et al. 14 found that sweet coffee drops above a threshold sugar concentration\nleft a uniform rather than the ring-like pattern. For altering\nenvironmental conditions including humidity, temperature,\nand acoustic or electric fields, 15 , 16 complicated\ndevices were needed to control the drying process. Eales and Routh 17 showed that ring-shaped deposits could be removed\nthrough careful selection of the atmospheric conditions, and humidity\ncycling had potential for controlling the film shape of the volatile\ndroplet. Deegan et al. 18 restricted evaporation\nfrom the edge of drop by covering the drop with a lid that had only\na small hole over the center of the drop. Yen et al. 19 invented a methodology of laser-induced differential evaporation\nto remove the coffee ring effect. In order to investigate the\ntransition from the coffee ring deposition\nto the uniform coverage in drying pinned sessile droplets, Crivoi\nand Duan 20 , 21 developed a Monte Carlo model to comprehend\nthe relationship between ring stain inhibitions and interactions at\nsolid–liquid and liquid–gas interfaces, while Xu et\nal. 22 developed a discrete element model\nto comprehend the interparticulate activities. In the field of biochip\nand nanostructuring, Askounis et al. 23 observed\nsmoother ring stains with some nanostructuring when DNA self-assembly\nwas indicative of the importance of DNA length. Nanosheets, 24 nanorods, 25 or their\nhybridization could also break the formation of coffee rings and deposit\nuniform films after drying. Yunker et al. 26 carried the ellipsoids to the air–water interface by the\noutward flow that caused the coffee ring effect for spheres, but strong\nlong-ranged interparticle attractions between ellipsoids led to the\nformation of loosely packed or arrested structures on the air–water\ninterface. Superhydrophilic or superhydrophobic 27 , 28 surfaces could\nbe used to suppress ring deposition because on these two kinds of\nsurfaces, a water drop has a moving contact line, but it is difficult\nto define the area for solute distribution. Maran-Mirabal et al. 29 used a hydrophilic/hydrophobic mosaic surface\nto suppress ring deposition and found that the solute was concentrated\non the hydrophilic zone while the capillary flow was reversed on the\nhydrophobic zone. Das et al. 30 showed a\nsuppression of ring deposits when a droplet, deposited on a glass\nsubstrate coated with a thin layer of silicone oil, was evaporated.\nJi et al. 31 presented a suppressed coffee\nring system via a combination of a magnetically functionalized membrane\nand reciprocating magnetic field for highly reliable and ultrasensitive\nSERS detection. However, very little attention has been paid to a\npatterned substrate. Herein, we developed a substrate, characterized\nwith hydrophilic spots surrounded by hydrophobic areas, to suppress\nthe ring stain effect and fabricated four kinds of patterned surfaces\nto investigate the relationship between the surface free energy and\nring-suppressing performance.",
"discussion": "2 Results and Discussion According to Young’s equation, 32 when a drop is placed on a homogeneous surface, there is a three-phase\nequilibrium at the edge of drop 1 where γlg, γsl,\nand γsg denote the interfacial tensions of the liquid/gas, the\nsolid/liquid, and the solid/gas interface, respectively. On a hydrophilic\nsurface (θ < 90°), γsg > γsl, an aqueous\nsolution will spread, while on a hydrophobic surface (θ <\n90°), γsg > γsl, aqueous solution will bead up.\nSo,\nwhen a drop is placed on the interface between hydrophilic and hydrophobic\nregions, the drop will have a tendency to leave the hydrophobic region\nand stay in the hydrophilic region. The driving force for the solution\nto move and remain depends on the difference of (γgs –\nγls) between the hydrophilic and hydrophobic region: the larger\ndifference of (γgs – γls), the greater the driving\nforce. In another perspective, the driving force could prevent water\nmolecules escaping from liquid to air, which results in a slower evaporation\nat the edge of the drop and suppressed capillary flow. So, a substrate\nwith hydrophilic spots surrounded by a hydrophobic area may be good\nfor suppressing the ring stain effect. To illustrate the abovementioned\ndesignation, four kinds of patterned\nsurfaces were fabricated by removing fluoroalkylsilane (FAS) locally\nby O 2 plasma from the hydrophobic or superhydrophobic surface:\nhydrophilic (CA 35 ± 2°)/hydrophobic (CA 105 ± 2°),\nhydrophilic (CA 35 ± 2°)/superhydrophobic (CA 150 ±\n2°), superhydrophilic (CA 3 ± 1°)/hydrophobic (CA 105\n± 2°), and superhydrophilic (CA 3 ± 1°)/superhydrophobic\n(CA 150 ± 2°) surfaces. Drops of 1 μL aqueous solution\nwith different amounts of fluorescein isothiocyanate (FITC) (0.2,\n0.4, 0.6, 0.8, and 1.0 ng) were spotted on hydrophilic (or superhydrophilic)\nspots of patterned surfaces and then were observed by a fluorescence\nmicroscope after being dried ( Figure 1 a–e). It is obvious that spots on patterned\nsurfaces have higher uniformity compared to those on the homogeneous\nhydrophilic surface (CA 35 ± 2°), demonstrating that the\nring-like stain was significantly suppressed on a hydrophilic/hydrophobic\npatterned surface. To assess the uniformity of FITC deposition, the\npercentage standard deviation (PSD) values of pixel fluorescence intensities\nwithin the spots were calculated ( Figure 1 f). Based on the analysis of PSD values,\nwe can conclude that the ring stain effect can be suppressed most\nsignificantly on a superhydrophilic/superhydrophobic patterned surface,\nnext came hydrophilic/superhydrophobic, the third rank is given to\nsuperhydrophilic/hydrophobic, and then hydrophilic/hydrophobic patterned\nsurfaces. In our experiment, when the FITC concentration was 0.25\nng/mm 2 , the PSD values on the abovementioned four kinds\nof patterned surfaces are 9, 12, 18, and 22%, respectively. It can\nbe also noticed that the uniformity of drops on superhydrophilic spots\nis slightly better than that on hydrophilic ones and is much better\non spots with superhydrophobic surroundings than that on spots with\nhydrophobic surroundings. Figure 1 Fluorescence microscopy images of the deposition\nof FITC after\nbeing dried on patterned surfaces: (a) superhydrophilic/superhydrophobic,\n(b) hydrophilic/superhydrophobic, (c) superhydrophilic/hydrophobic,\n(d) hydrophilic/hydrophobic, and (e) hydrophilic surface. The concentrations\nof FITC solution from top to bottom in every photograph are 1.25,\n1.00, 0.75, 0.50, and 0.25 ng/mm 2 , respectively. (f) Percentage\nstandard deviation of the fluorescent intensities obtained from the\nanalyzed spots on surfaces (a)–(e). It was noteworthy that PSD values of drops on all kinds of surfaces\ndecreased significantly with the increasing of FITC concentration,\nimplying that the ring stain effect was closely related to the initial\nsolute concentration. The higher the concentration, the less ring\ndeposition is. For further explanation of this ring-suppressing\nperformance, the\ndroplets containing PS microsphere solutions (1uL, 19%wt) were dried\non the hydrophilic anodic aluminum oxide (AAO) surface and the superhydrophilic/superhydrophobic\npatterned surface. The SEM images of the residues on the edge of the\nring showed that the PS microspheres uniformly dispersed on the patterned\nsurface, while those close packed on the hydrophilic surface ( Figure 2 ). Figure 2 SEM images of the residues\non the edge of the ring formed on the\n(a) superhydrophilic/superhydrophobic patterned surface and (b) hydrophilic\nAAO surface. According to Deegan and co-workers’\ntheory, 5 , 18 an outward capillary flow carried the solute\nto the edge of the\ndrop, forming ring-like deposits. To detect whether the capillary\nflow in drops was altered on our hydrophilic/hydrophobic patterned\nsurface, drops of an aqueous solution containing CdTe quantum dots\nwere spotted on superhydrophilic/superhydrophobic patterned surfaces\n(CA 3/150°), and a hydrophilic surface (CA 35 ± 2°)\nwas used as a control. The images of the drying process were captured\nwith a camera from above down with a perpendicular angle to the surface\nin order to observe the residue dispersing process of a spotted droplet,\nespecially to take the images of the droplet edge. By recording fluorescence\nmicroscopy images of spotted drops during drying, we observed that\nfluorescence intensity kept invariant as time was prolonged on the\nsuperhydrophilic/superhydrophobic patterned surface ( Figure 3 a). However, on the control\nsurface, the fluorescence intensity at the edge increased with the\nevaporating time and decreased at the center of the drop simultaneously\n( Figure 3 b), showing\nthat quantum dots were carried to the edge by an outward capillary\nflow. In other words, the outward direction of the capillary flow\nwas suppressed on the hydrophilic/hydrophobic patterned surface. Figure 3 Fluorescence\nmicroscopy images of the edge of drying drops containing\n20 nm-sized CdTe nanoparticles on (a) superhydrophilic/superhydrophobic\npatterned and (b) hydrophilic surfaces. There are two necessary factors for the outward capillary flow:\na pinned contact line and larger evaporation rate at the edge of the\ndrop. Based on the video of the drying profile of the spotted drop\non the superhydrophilic/superhydrophobic patterned surface ( Figure 4 ), we found that\nthe drop was constrained on the hydrophilic spot with a pinned contact\nline. So, we presumed that alteration of capillary flow direction\ninside the drop on a hydrophilic/hydrophobic patterned surface may\nbe caused by the changing evaporation rate at the edge of the drop\nas proposed previously. In our case, water contact angles for the\nfour kinds of patterned surfaces are 3/150, 35/150, 3/105, and 35°/105°,\nand the differences of cosθ were 1.865, 1.685, 1.258, and 1.078\nrespectively. As shown in Figure 1 , the efficiency of ring stain suppressing decreased\nin the same sequential order. As well known, the water contact angle\ndepends on the solid surface free energy, so the efficiency of ring\nstain suppressing increases with the difference of the surface free\nenergy between hydrophilic and hydrophobic surfaces, and the superhydrophilic/superhydrophobic\npatterned surface is the best one. Figure 4 Evaporation behavior and model of a water\ndrop on the superhydrophilic/superhydrophobic\npatterned surface."
} | 3,357 |
34975995 | PMC8718876 | pmc | 1,541 | {
"abstract": "Plants and arbuscular mycorrhizal fungi (AMF) can form complex symbiotic networks based on functional trait selection, contributing to the maintenance of ecosystem biodiversity and stability. However, the selectivity of host plants on AMF and the characteristics of plant-AMF networks remain unclear in Tibetan alpine meadows. In this study, we studied the AMF communities in 69 root samples from 23 plant species in a Tibetan alpine meadow using Illumina-MiSeq sequencing of the 18S rRNA gene. The results showed a significant positive correlation between the phylogenetic distances of plant species and the taxonomic dissimilarity of their AMF community. The plant-AMF network was characterized by high connectance, high nestedness, anti-modularity, and anti-specialization, and the phylogenetic signal from plants was stronger than that from AMF. The high connected and nested plant-AMF network potentially promoted the interdependence and stability of the plant-AMF symbioses in Tibetan alpine meadows. This study emphasizes that plant phylogeny and plant-AMF networks play an important role in the coevolution of host plants and their mycorrhizal partners and enhance our understanding of the interactions between aboveground and belowground communities.",
"conclusion": "Conclusion This study assesses the relationship between the common plant phylogeny and AMF and the characteristics of the plant-AMF symbiotic network in a Tibetan alpine meadow. There was a significant positive correlation between the phylogenetic distance of the 23 herbs and their AMF community dissimilarity, indicating the selectivity of plant species on their AMF community. The network structure of the plant and AMF was characterized by high connectance, high nestedness, anti-modularity, and anti-specialization, which potentially promoted the stable symbiosis and interdependence between plants and AMF in Tibetan alpine meadows. The plant-AMF network was affected by the plant-AMF phylogenies, and the phylogenetic signal from plants was stronger than that from AMF. This study emphasizes that plant phylogeny and plant-AMF network play an important role in the complex plant-AMF community mechanism and enhance our understanding of the interaction between aboveground and belowground communities.",
"introduction": "Introduction Arbuscular mycorrhizal fungi (AMF) are an ancient root symbiotic group, whose origin is consistent with the first appearance of terrestrial plants, and generally considered to be the result of coevolution of fungi and plants ( Bonfante and Genre, 2008 ; Walder and van der Heijden, 2015 ; Genre et al., 2020 ). As a key component of the underground biological community, AMF can form a symbiotic relationship with most terrestrial plant species and can increase the absorption of phosphorus and nitrogen for plants ( Wang and Qiu, 2006 ; Smith and Read, 2008 ; Zhang et al., 2021 ). At the same time, plants provide photosynthetic carbon products to AMF ( Bever et al., 2009 ; Kiers et al., 2011 ; Wipf et al., 2019 ). The combination of AMF and plants improves the adaptability of plants and AMF symbionts ( Blackwell, 2000 ; Chen et al., 2018 ; Tedersoo et al., 2020 ). Furthermore, the plant-AMF symbioses are not formed stochastically but are dependent on their functional traits ( Bever et al., 2009 ; Kiers et al., 2011 ). Both plant hosts and AMF have been shown to preferentially allocate resources to higher-quality partners ( Ji and Bever, 2016 ), inducing reciprocal shifts in each assemblage, ultimately affecting the partner selection in the mycorrhizal symbiosis ( Werner and Kiers, 2015 ). Due to the simultaneous processes of coevolution, niche differentiation, and niche conservatism, the closely related plants tend to interact with similar AMF species (i.e., phylogenetic signal in the interaction) ( Gomez et al., 2010 ; Sargent et al., 2011 ). However, negative correlations between the phylogenetic distance of plants and AMF community distance were also reported in three locations of the northern Great Plains of the United States ( Reinhart and Anacker, 2014 ). In addition, the specificity of the interactions between the plant and AMF changed with community succession ( Montesinos-Navarro et al., 2012a ). These contrasting findings indicate the complexity of plant and AMF interactions. Many studies have shown that phylogenetic relatedness affects the mycorrhizal symbiotic network structure, which could answer the question of how coevolution explained species assemblages ( Barberan et al., 2012 ; Chagnon et al., 2012 ; Montesinos-Navarro et al., 2012b ; Chen et al., 2017 ; Goberna et al., 2019 ). The network is increasingly used to study co-occurrence patterns in plant and AMF symbioses, providing insights for biological interactions ( Barberan et al., 2012 ; Chagnon et al., 2012 ; Chen et al., 2017 ; Gao et al., 2019 ). The AMF communities of 166 root samples from 17 woody plants in subtropical forests of China showed that the woody plant-AMF network was highly interconnected and nested but in anti-modular and anti-specialized manners ( Chen et al., 2017 ). Most of the studies on the plant-fungus symbiotic network focused on temperate and tropical ecosystems, and the research objects were mainly woody plant-AMF symbiotic interactions. However, herbaceous plant-AMF network characteristics in the alpine ecosystems remain poorly understood. As the Third Pole on the earth, the Tibetan Plateau is a unique geographic unit with great biodiversity and critical ecosystem functions and is experiencing striking climate warming ( Bai-ping et al., 2002 ; Duan et al., 2006 ; Qiu, 2008 ). Alpine meadow is a typical vegetation type, accounting for approximately 40% of the total grassland area on the Tibetan Plateau ( Miehe et al., 2008 ; Wang et al., 2016 ). The importance of AMF in plant succession on the Tibetan plateau has been well documented ( Liu et al., 2011 ; Gai et al., 2012 ; Li et al., 2015 ). However, most of the studies focused on the effects of altitude gradients ( Liu et al., 2011 ; Gai et al., 2012 ; Li et al., 2014 ), precipitation ( Zhang et al., 2016 ), and vegetation types ( Gai et al., 2009 ; Gao and Guo, 2010 ; Xu et al., 2017 ) on AMF community composition. The selectivity of the interactions between the plant and AMF and their network properties in Tibetan alpine meadows remains unclear. In this study, to explore the selectivity in the plant-AMF interactions and clarify the maintenance mechanism of symbiosis, 23 species of common herbs in a Tibetan alpine meadow were selected. Illumina-MiSeq sequencing was used to characterize AMF communities living in each plant root to address the following questions: (1) How is the selectivity in the plant-AMF interactions in Tibetan alpine meadow? and (2) What are the characteristics of the plant-AMF network in Tibetan alpine meadows?",
"discussion": "Discussion There were 121 AMF molecular species detected from plant roots in this study, indicating diverse AMF species in Tibetan alpine meadow, in concordance with the previous studies ( Gai et al., 2009 , 2012 ; Liu et al., 2011 ; Zhang et al., 2016 ). At the same time, it was consistent with previous studies ( Gai et al., 2012 ; Li et al., 2014 ) that Cyperaceae plants which were considered unable or not easy to form mycorrhizal fungi also had high AMF colonization ( Figure 2A ; 40.4–60.2%) and diversity ( Figures 2B,C ; Shannon-Wiener: 2.31–2.60; richness: 86.67–90.67). These findings suggest the potential ecological importance of AMF for plant growth in the alpine meadow. To survive in the harsh environment of the alpine meadow, plants might choose AMF species to improve their ability to absorb nitrogen and phosphorus or the stress-resistance ability ( Bever, 2002 ). It has been found that some AMF species with strong cold resistance ability could survive, thus affecting the growth of host plants and community dynamics ( Deepika and Kothamasi, 2015 ; Bauer et al., 2017 ). Furthermore, this study found a significant positive correlation between plant phylogenetic distance and AMF community dissimilarity. Thus, the selectivity existed in the plant-AMF interactions, so that closely phylogenetic-related plant species select similar AMF communities ( Figure 5 ). This plant-AMF selectivity has been well documented in previous studies ( Vandenkoornhuyse et al., 2003 ; Horn et al., 2014 ; Veresoglou and Rillig, 2014 ; Chagnon et al., 2015 ; Deepika and Kothamasi, 2015 ). The different AMF communities in the rhizosphere of different plant species may be achieved by preferentially allocating plant carbon to the most beneficial fungal partners ( Bever et al., 2009 ; Kiers et al., 2011 ). The construction of the plant-AMF community was to maximize the functional matching between partners ( Thompson, 2005 ). Plants with similar traits preferentially hosted similar AMF and, likewise, phylogenetically related AMF (assumed to have similar functional traits) interacted with similar plants. No doubt that the plant phylogeny was used to map the mycorrhizal information to examine evolutionary patterns ( Wang and Qiu, 2006 ), but the complex symbiotic relationships between plants and AMF cannot be fully explained by plant phylogeny alone. The phylogenetic correlations between the interacting plant and AMF were modified by the competitive intransitivity ( Laird and Schamp, 2006 ) and AMF species competition within a plant host ( Roger et al., 2013 ). Therefore, future studies should involve factors, such as plant and AMF functional traits, species competition, climates, and soil nutrients, which can well explain the AMF community dynamics. The network structure of 23 common plants and AMF was characterized by high connectance, high nestedness, anti-modularity, and anti-specialization, which was in line with the characteristics of the woody plant-AMF network in a subtropical forest in China ( Chen et al., 2017 ). Generally, with the increase of connectance, the nestedness increased and the modularity decreased ( Põlme et al., 2018 ), while the nestedness decreased, the modularity and the specialization increased ( van der Heijden et al., 2015 ), as found in our network. High network connectance and nestedness and low network modularity indicated less segregation or sparse connections within subnetworks and dense connections between subnetworks for the plant-AMF network. These network properties enhanced the stability of the communities in the mutually beneficial network, which was considered to be an important property to promote the coexistence of species in the mutual system ( Thebault and Fontaine, 2010 ). Theoretical studies suggested that the common nestedness patterns in the species network can determine the feasibility, resilience, durability, and structural stability of ecological communities ( Toju et al., 2014 ), and the higher the nestedness of the interaction network, the better it can adapt to the effect of environmental disturbance. In the analysis of plant-AMF phylogenetic signals based on the association frequency matrix, the results showed that the phylogenetic signals of plants and AMF had significant effects on the structure of the plant-AMF network. The phylogenetic signal from plants was stronger than that from AMF ( Table 1 ). Close relatives of plants often interacted with AMF OTU species in the same group, but the interaction between AMF and plant species had a low correlation with AMF phylogeny, which was consistent with the study by Chen et al. (2017) . In contrast, findings by Montesinos-Navarro et al. (2012a , b) showed that AMF phylogenetic diversity might also explain that there were different plant-AMF network structures in different ecosystems. In the AMF community with a more diverse phylogeny, there were more opportunities for specific AMF to coexist with common host plants, thus forming a highly nested symbiotic network structure ( Chen et al., 2017 ). The phylogeny of plants and AMF played an important role in the non-random pattern of the AMF network in subtropical forest sites ( Chen et al., 2017 ). In addition, the plant root traits in subtropical forests also showed strong interspecific phylogenetic signals ( Kong et al., 2014 ). The modularity and nestedness in the plant-AMF network were caused by a variety of ecological processes, such as habitat heterogeneity ( Olesen et al., 2007 ), specific selection of plant and AMF combination, relative species abundance and phylogenetic diversity, or AMF competition within the root ( Verdu and Valiente-Banuet, 2011 ). The complex symbiotic interaction between plants and AMF has a potential effect on community assembly. Our study found that the characteristics of the plant-AMF network with high connectance and nestedness promote the stable symbiosis and interdependence between plants and AMF in Tibetan alpine meadows. Plants established network relationships with underground AMF through selection, which may affect the nutrient cycle of symbionts and even the succession of grassland community, so as to improve the adaptation to alpine ecological environment and stress. Therefore, further studies are required to elucidate the reciprocal symbiotic selection strategy, the coexistence pattern, community construction mechanism, and community dynamic rules of plant and AMF species. Furthermore, to enhance our understanding of the influence factors in the coevolution of plant-AMF is of great significance to study community structure, species evolution, and biodiversity."
} | 3,379 |
29241010 | null | s2 | 1,542 | {
"abstract": "Phytostabilization is a cost-effective long-term bioremediation technique for the immobilization of metalliferous mine tailings. However, the biogeochemical processes affecting metal(loid) molecular stabilization and mobility in the root zone remain poorly resolved. The roots of Prosopis juliflora grown for up to 36 months in compost-amended pyritic mine tailings from a federal Superfund site were investigated by microscale and bulk synchrotron X-ray absorption spectroscopy (XAS) and multiple energy micro-X-ray fluorescence imaging to determine iron, arsenic, and sulfur speciation, abundance, and spatial distribution. Whereas ferrihydrite-bound As(V) species predominated in the initial bulk mine tailings, the rhizosphere speciation of arsenic was distinctly different. Root-associated As(V) was immobilized on the root epidermis bound to ferric sulfate precipitates and within root vacuoles as trivalent As(III)-(SR)"
} | 231 |
29241010 | null | s2 | 1,543 | {
"abstract": "Phytostabilization is a cost-effective long-term bioremediation technique for the immobilization of metalliferous mine tailings. However, the biogeochemical processes affecting metal(loid) molecular stabilization and mobility in the root zone remain poorly resolved. The roots of Prosopis juliflora grown for up to 36 months in compost-amended pyritic mine tailings from a federal Superfund site were investigated by microscale and bulk synchrotron X-ray absorption spectroscopy (XAS) and multiple energy micro-X-ray fluorescence imaging to determine iron, arsenic, and sulfur speciation, abundance, and spatial distribution. Whereas ferrihydrite-bound As(V) species predominated in the initial bulk mine tailings, the rhizosphere speciation of arsenic was distinctly different. Root-associated As(V) was immobilized on the root epidermis bound to ferric sulfate precipitates and within root vacuoles as trivalent As(III)-(SR)"
} | 231 |
40148482 | PMC11950176 | pmc | 1,545 | {
"abstract": "Magnetotactic bacteria are ubiquitous aquatic prokaryotes that have the ability to biomineralize magnetite (Fe 3 O 4 ) and/or greigite (Fe 3 S 4 ) nanoparticles called magnetosomes. Magnetotactic cocci belonging to the class “ Ca . Magnetococcia” are most frequently identified in freshwater habitats, but remain uncultivated. Here, we report for the first time axenic cultivation of freshwater magnetotactic coccus FCR-1 isolated from Chichijima, Japan. Strain FCR-1 grows microaerophilically in a semi-solid gellan gum medium. We find that strain FCR-1 biomineralizes Fe 3 O 4 nanoparticles, which are not chained, into a cell. Based on phylogenomic analysis, compared with strains of the class “ Ca . Magnetococcia”, strain FCR-1 represents a novel genus of candidate family “ Ca . Magnetaquicoccaceae” within the class “ Ca . Magnetococcia” and we tentatively name this novel genus “ Ca . Magnetaquiglobus chichijimensis”. Our isolate provides a promising tool for elucidating the functions of unchained magnetosomes, the global distribution of magnetotactic bacteria and the origin of magnetotaxis.",
"introduction": "Introduction Chichijima; one of the Ogasawara (Bonin) Islands, which are volcanic islands formed during the Eocene time between 46 and 48 Ma, is located in the northwestern region of the Pacific Ocean, approximately 1000 km south from the Japanese mainland 1 . The Oceanic islands, which are remote and isolated from the continental landmass, have been considered a suitable model for evolutionary studies 2 . Because the Bonin Islands have never been connected to any continental landmass, quite a few endemic terrestrial species have been found in the islands 3 – 5 . Thus, it is still believed that the Bonin Islands could provide a clue to a better understanding of the distribution and evolution of organisms on earth. Magnetotactic bacteria (MTB) are phylogenetically a diverse group of prokaryotes that biomineralize intracellular magnetosomes, which are proteolipid membrane-enclosed magnetic nanoparticles (NPs) composed of magnetite (Fe 3 O 4 ) and/or greigite (Fe 3 S 4 ). Magnetosomes are generally characterized by a distinct species-specific crystal morphology and the arrangement in chains within the cells 6 , 7 . Thanks to the linear arrangement of magnetosomes, MTB have the ability to swim along geomagnetic field lines, a behaviour known as magnetotaxis 6 . MTB are microaerophiles or anaerobes that mainly inhabit the oxic-anoxic transition zones (OATZs) in the sediment or chemically stratified water column 8 . To our knowledge, known MTB species are affiliated with the classes Alphaproteobacteria , Gammaproteobacteria , Zetaproteobacteria , “ Ca . Magnetococcia” of the phylum Pseudomonadota , and with the phyla SAR324, Desulfobacterota , Bdellovibrionota , Nitrospirota, Nitrospinota, Fibrobacterota , Omnitrophota , Latescibacterota , Planctomycetota, Riflebacteria , Hydrogenedentota , UBA10199 and Elusimicrobiota in the GTDB taxonomy 9 – 12 . Magnetotactic cocci belonging to the class “ Ca . Magnetococcia” are most commonly observed morphotype among MTB and frequently detected from both freshwater and marine habitats 13 . However, only five magnetotactic cocci from marine habitats and hypersaline lagoon have been isolated in axenic cultures; i.e., Magnetococcus marinus MC-1 14 , 15 , “ Ca . Magnetococcus massalia” MO-1 16 , Magnetofaba australis IT-1 17 , and strain PR-3 and SS-1 18 . Besides the biomineralization of iron oxide crystals, it is known that magnetotactic cocci possess intracellular polyphosphate (Poly P) granules that frequently occupy most of the cell volume, whereas those also possess intracellular sulphur granules, which indicates the capability of utilizing reduced sulphur compounds 19 . The capability of forming the Poly P granules indicates that magnetotactic cocci utilize a large amount of phosphorus for supporting the “bacterial shuttle” around OATZ 20 – 22 . Three cultured strains; Mc. marinus MC-1 15 , “ Ca . Mc. massalia” MO-1 16 and Mf. australis IT-1 17 , which are microaerophiles, utilize thiosulphate for autotrophic growth. The metagenomic analysis indicated that uncultured freshwater magnetotactic coccus “ Ca . Magnetaquicoccus inordinatus” UR-1 was potentially capable of oxidizing sulphide, while it also appeared to have a potential to oxidize sulphite produced in dissimilatory sulphur oxidization 19 . Therefore, it is supposed that magnetotactic cocci play an important role in a biogeochemical cycling of iron, phosphorus and sulphur in natural environments 23 , 24 . Due to the difficulty of isolating freshwater magnetotactic cocci in an axenic culture, the diversity and microbial ecology of cocci and the characterization of magnetosomes biomineralized by cocci have been mainly investigated by cultivation-independent methods using fluorescent in situ hybridization (FISH) with specific 16S rRNA-based probes and FISH coupled with transmission electron microscopy (FISH-TEM) and scanning electron microscopy (FISH-SEM). These approaches have revealed that freshwater magnetotactic cocci are highly diverse in terms of the size, number, morphology and arrangement of magnetosome crystals 13 , 19 , 25 . Especially, the spatial arrangement of magnetosome crystals within the cells varies among strains of uncultured freshwater magnetotactic cocci. Strain YQC-1, BHC-1, DMHC-1, WYHC-1 and WYHC-3 biomineralize a single chain of magnetosome in each cell, while strain MYC-4, MYC-5, YQC-3, YQC-5 and XQGC-1 double chains of magnetosomes within a cell. Strain DMHC-2, DMHC-8 and MYC-9 biomineralize two double chains of magnetosomes in a cell, whereas “ Ca . Mq. inordinatus” UR-1, strain WYHC-2, THC-1, MYC-3, MYC-7, DMHC-6 and YQC-9 unchained magnetosomes 13 , 19 , 25 , 26 . Intriguingly, unchained magnetosomes have been frequently found in magnetotactic cocci from freshwater habitats 13 , 19 , 25 . Now, a question; “How can the cocci sense the geomagnetic field lines with unchained magnetosomes” arises. Freshwater magnetotactic cocci biomineralizing unchained magnetosomes have not yet been cultivated. The cultivation and isolation of these magnetotactic cocci are essential to understand the diversity and ecological functions of cocci in freshwater habitats and elucidate the functions and evolution of unchained magnetosomes. In this study, we for the first time cultivate novel magnetotactic coccus FCR-1 isolated from Renju Dam in Chichijima and investigate the characteristics of strain FCR-1. Magnetotactic coccus FCR-1 is cultivated using a semi-solid medium using gellan gum as a gelling agent. We find that the spatial arrangement and morphology of magnetosome crystals of strain FCR-1 are different from those of three cultured Mc. marinus MC-1, “ Ca . Mc. massalia” MO-1 and Mf. australis IT-1. We also conduct the whole genome sequencing to obtain functional annotations of genes related to the formation of the magnetosome. We also find, comparing the genome of strain FCR-1 with the available “ Ca . Magnetococcia” genomes, that strain FCR-1 represents a novel genus of candidate family “ Ca . Magnetaquicoccaceae” within the class “ Ca . Magnetococcia”, tentatively named “ Ca . Magnetaquiglobus chichijimensis”.",
"discussion": "Discussion Freshwater magnetotactic cocci were discovered more than 40 years ago and frequently found from various freshwater habitats 13 . However, the freshwater magnetotactic cocci remain uncultivated before this work. We have reported in the above the isolation of a novel freshwater magnetotactic coccus FCR-1 that biomineralizes unchained magnetosomes. Axenic culture of the freshwater magnetotactic coccus FCR-1 was successfully attained, verified by the genomic sequencing, and we found that almost all of the FCR-1 cells biomineralized unchained magnetosomes unlike commonly observed chained magnetosomes (Supplementary Fig. 5 in the Supplementary Information). Strain FCR-1 also had the ability to form Poly P granules containing either magnesium or calcium in individual cells (Fig. 3f, g , Supplementary Figs. 6 , 7 in the Supplementary Information, and Supplementary Data ). However, it still remains unknown why strain FCR-1 mainly biomineralized unchained magnetosomes and formed Poly P granules containing magnesium or calcium. We found that strain FCR-1 belongs to the family “ Ca . Magnetaquicoccaceae” within the class “ Ca . Magnetococcia” and represents a novel genus of the family “ Ca . Magnetaquicoccaceae” (Figs. 2 , 5 and Supplementary Fig. 11 in the Supplementary Information). Interestingly, the 16S rRNA gene sequence of strain FCR-1 was the same as the sequence of uncultured magnetotactic coccus clone HCH0515, which was detected from freshwater sediments of moat, Xi’an city, China 28 . On the basis of the above result and the fact that Chichijima has never been connected to the Eurasia continent, it can be inferred that strain FCR-1 might have migrated by carriers such as migratory birds and/or yellow sand. In the present study, we, for the first time, isolated and cultivated strain FCR-1 in a semi-solid medium, in which an oxygen concentration gradient was established, using gellan gum as a gelling agent. Note that gellan gum is an extracellular polysaccharide produced by Sphingomonas elodea and usually used as a gelling agent for plant tissue culture and cultivation of thermophilic microorganisms that grow at high temperature. In fact, some of the previous studies suggested that gellan gum could be used as a gelling agent for cultivation and isolation of mesophilic microorganisms from soil and freshwater environments 31 , 32 . Strain FCR-1 grew in a semi-solid medium, in which an oxygen concentration gradient was established, using Bacto agar as a gelling agent, but the growth rate was lower than that grown in the semi-solid gellan gum medium, whereas strain FCR-1 was unable to grow in an oxygen-gradient semi-solid medium using Noble agar as a gelling agent. Although agar is commonly used in microbiological studies, it could be a potential inhibitor of the growth of some of microorganisms 33 , 34 . It is therefore supposed that agar slowed the growth of FCR-1 cells. Thanks to gellan gum, which forms highly transparent gels, a very fine band of cells was easily observed by the naked eye. Therefore, semi-solid gellan gum media can be a useful tool for observation of the process of forming a band of cells, and isolation and cultivation of as-yet-uncultured MTB, including uncultured microaerophilic MTB, which form a very fine band of cells. In contrast to Mc. marinus MC-1 15 and Mf. australis IT-1 17 , strain FCR-1 was unable to oxidize thiosulphate when grown chemolithoautotrophically (Table 1 and Supplementary Fig. 3 in the Supplementary Information). The essential sox genes for oxidizing thiosulphate such as soxXYZAB operon and putative sox genes are present in the Mc. marinus MC-1 genome 15 . As in the case of several strains of the class “ Ca . Magnetococcia” 20 , strain FCR-1 contained the soxY (locus_tag: SIID45300_02975) and soxZ (locus_tag: SIID45300_02974) genes, while the soxA, soxB and soxX genes were not found in the genome of strain FCR-1. The absence of several essential sox genes is congruent with the experimental result; that is strain FCR-1 was not capable of oxidizing thiosulphate for growth. In addition, strain FCR-1 had the ability to oxidize sulphide, but the strain was not capable of reducing sulphate (Supplementary Fig. 3 in the Supplementary Information). Strain FCR-1 contained dsrABL (locus_tag: SIID45300_00713, SIID45300_00712, SIID45300_00711) genes, dsrMK (locus_tag: SIID45300_00870, SIID45300_00869) genes and dsrJOP (locus_tag: SIID45300_01599, SIID45300_01598, SIID45300_01597) genes, which indicates that the strain is potentially capable of reducing sulphate and oxidizing sulphide and/or sulphur. “ Ca . Mq. inordinatus” UR-1 also contained dsrABL genes, dsrC gene and dsrFEHJOP genes and the presence of the dsrEFH genes and dsrL gene indicated the reverse type of Dsr pathway 19 . Strain FCR-1 also had the ability to synthesize intracellular elemental sulphur (Supplementary Fig. 8 in the Supplementary Information) and contained fccA gene (locus_tag: SIID45300_02979) and fccB gene (locus_tag: SIID45300_02980), which are related to oxidization of sulphide to elemental sulphur. Formation of Poly P granules is commonly found among MTB. Strain FCR-1 formed 1 to 3 Poly P granules containing either magnesium or calcium (Fig. 3f, g , Supplementary Figs. 6 , 7 in the Supplementary Information, and Supplementary Data ). MTB also often produce intracellular calcium granules such as calcium carbonate (CaCO 3 ), calcium phosphate and calcium-rich phosphate 35 , 36 . Previous studies have shown that calcium phosphate and calcium-rich phosphate can be distinguished by the Ca Kα/P Kα ratio, in such a way that the former has higher Ca Kα/P Kα ratio (1.2–2.2) than the latter (0.22–0.26) 37 , 38 . The Ca Kα/P Kα ratio of the Poly P granules in FCR-1 cells showed 0.28 ± 0.02 (Fig. 3f , g and Supplementary Data ). Thus, strain FCR-1 produced calcium-rich phosphate granules. Recently, there are reports that freshwater magnetotactic cocci accumulated large Poly P granules under anoxic conditions and hydrolysed the Poly P granules under suboxic conditions, which indicates that the cocci utilized the Poly P for supporting their phosphorus metabolisms 20 , 21 . Polyphosphate kinase (PPK) is a main enzyme that catalyses the initial step in the formation of long-chain Poly P using ATP (PPK1) or GTP (PPK2) 39 , 40 , while exopolyphosphatase (PPX) has hydrolysis activity that releases the terminal phosphate from Poly P 41 . Two genes encoding PPK2 (locus_tag: SIID45300_00168) and exopolyphosphatase/guanosine pentaphosphate phosphohydrolase (PPX/GppA) (locus_tag: SIID45300_01972) were found in the FCR-1 genome. A possible explanation is that the formation or hydrolysis of Poly P granules in FCR-1 cells may require either magnesium or calcium ions for the PPK2 or the PPX/GppA activity. These also indicate that strain FCR-1 significantly contribute to the biogeochemical cycling of iron, sulphur, phosphorus, magnesium and calcium in freshwater environments. The morphology of magnetosome is correlated with MTB phylogenetic affiliations 42 . The morphology of magnetosomes of strain FCR-1 was clearly different from that of magnetosomes of three cultured strains of Mc. marinus MC-1 15 , “ Ca . Mc. massalia” MO-1 16 and Mf. australis IT-1 17 (Table 1 ). The morphology and spatial arrangement of magnetosome are known to be strictly controlled by the magnetosome-associated genes and therefore it is supposed that the organization of unchained magnetosomes in strain FCR-1 can be determined by the MGC identified in this study. It has been demonstrated that the mamK , mamJ , mamY, mcaA and mcaB genes, which are involved in the formation of chained magnetosomes, were found only in the genomes of Magnetospirillum species 43 – 46 . However, as in the case of other strains of the class “ Ca . Magnetococcia”, the mamJ , mamY, mcaA and mcaB genes were not found in the MGC of strain FCR-1, which indicates that some other mechanisms for forming chained magnetosomes are present in this group of MTB 19 , 26 . The mamK gene, which encodes an actin-like protein and organizes chained magnetosomes 43 , was found in the MGC of strain FCR-1, noting that the mamK gene is commonly present in MTB except for magnetotactic Elusimicrobiota 12 . As in the case of the “ Ca . Magnetococcia” strains containing multicopy mamK genes 19 , 26 , strain FCR-1 also possessed two copies of mamK genes located inside and outside of the MGC, which indicates that two MamK proteins are related to the specific spatial arrangement of magnetosomes; that is unchained magnetosomes, in the cell. The copy number of mamK genes and the similarity of multicopy mamK genes appeared to be associated with diverse magnetosome chain configuration, which indicated that the mamK genes of “ Ca . Magnetococcia” strains biomineralizing unchained magnetosomes had a relatively low similarity with each other (≤ 67%) 26 . In addition, the relationship between the similarity of the mamK genes and the shape factor of magnetosomes illustrated that the elongated shape of Fe 3 O 4 NPs might be also correlated with the unchained magnetosomes (Supplementary Fig. 13 in the Supplementary Information). Additionally, the MGC of strain FCR-1 contained several proteins with unknown functions, which may be attributed to the formation of unchained truncated hexagonal prism-like Fe 3 O 4 NPs. The Magnetaquicoccaceae -specific gene ( maq1 ), which is also found in the MGC of strain FCR-1, is located between mamE and mamK genes in strains belonging to the family “ Ca . Magnetaquicoccaceae” and similar to the position of mamJ gene in the genus Magnetospirillum 19 , which indicates that the organization of unchained magnetosomes could be attributed to the maq1 gene. Strain FCR-1, YQC-9 and UR-1, which belong to the family “ Ca . Magnetaquicoccaceae” and biomineralize unchained magnetosomes, possessed maq1 gene (Supplementary Fig. 14 and Supplementary Table 1 in the Supplementary Information). However, strain DMHC-6 and THC-1, which biomineralize unchained magnetosomes, did not possess the maq1 gene, which implies that the biomineralization of unchained magnetosomes is not solely attributed to the maq1 gene. Three hypothetical proteins (locus_tag: SIID45300_02215, SIID45300_02219, SIID45300_02220) of strain FCR-1 had a homology with those of “ Ca . Mq. inordinatus” UR-1, but had no homology with any of the proteins of cultured organisms in the GeneBank database. Therefore, these three proteins might be specifically encoded by the members of the family “ Ca . Magnetaquicoccaceae”, which biomineralize unchained truncated hexagonal prism-like Fe 3 O 4 NPs (Fig. 6a and Supplementary Data ). In future work, it would be interesting to create mutants without these three genes in order to elucidate the roles of these proteins. It is of great interest and importance to understand the mechanism of the biomineralization of unchained magnetosomes from both ecological and evolutionary points of view. From a thermodynamical point of view, the configuration of magnetosomes should be determined based on the free energy minimal conditions; i.e., the energy is lowered by the formation of chains, whereas the entropy is increased in the case of unchained (disassembled) magnetosomes. In other words, both configurations of magnetosomes; that is, chained or unchained ones, are possible as long as the free energy is minimal. However, the total magnetization of each magnetotactic bacterium containing chained magnetosomes is higher than that containing unchained magnetosomes. Therefore, it is advantageous for MTB to possess chained magnetosomes in terms of efficient migrations along geomagnetic field lines, which may explain why most of MTB form chained magnetosomes. Detailed characteristics that distinguish strain FCR-1 from Mf. australis IT-1, “ Ca . Mc. massalia” MO-1 and Mc. marinus MC-1 are summarized in Table 1 . In conclusion, strain FCR-1 is a novel freshwater magnetotactic coccus belonging to the family “ Ca . Magnetaquicoccaceae” that biomineralizes unchained truncated hexagonal prismatic Fe 3 O 4 NPs. We tentatively name strain FCR-1 “ Candidatus Magnetaquiglobus chichijimensis”. Description of novel candidate genus and species Candidatus Magnetaquiglobus Magnetaquiglobus (Mag.net.a.qui.glo′bus. L. n. magnês - etis a magnet; N.L. pref. magneto -, pertaining to a magnet; L. fem. n. aqua water; N.L. masc. n. globus , a sphere; N.L. masc. n. Magnetaquiglobus the magnetic sphere from water, referring to its sphere morphology and magnetotactic behaviour). Candidatus Magnetaquiglobus chichijimensis Magnetaquiglobus chichijimensis (chi.chi.jim.en′sis L. gen. n. chichijimensis , pertaining to Chichijima Island, referring to the source of the sediment sample from which the strain was isolated)."
} | 5,060 |
29900423 | PMC5995479 | pmc | 1,546 | {
"abstract": "Society is on the cusp of harnessing recent advances in synthetic biology to discover new bio-based products and routes to their affordable and sustainable manufacture. This is no more evident than in the discovery and manufacture of Synthetic Biological Materials , where synthetic biology has the capacity to usher in a new Materials from Biology era that will revolutionise the discovery and manufacture of innovative synthetic biological materials. These will encompass novel, smart, functionalised and hybrid materials for diverse applications whose discovery and routes to bio-production will be stimulated by the fusion of new technologies positioned across physical, digital and biological spheres. This article, which developed from an international workshop held in Manchester, United Kingdom, in 2017 [1], sets out to identify opportunities in the new materials from biology era. It considers requirements, early understanding and foresight of the challenges faced in delivering a Discovery to Manufacturing Pipeline for synthetic biological materials using synthetic biology approaches. This challenge spans the complete production cycle from intelligent and predictive design, fabrication, evaluation and production of synthetic biological materials to new ways of bringing these products to market. Pathway opportunities are identified that will help foster expertise sharing and infrastructure development to accelerate the delivery of a new generation of synthetic biological materials and the leveraging of existing investments in synthetic biology and advanced materials research to achieve this goal.",
"introduction": "1 Introduction: the dawn of a new era for advanced materials Strong synergies exist between materials and chemicals sciences and their allied technologies, which have driven understanding at the atomic and molecular levels of the complex relationships between the chemical compositions, structures and macroscopic properties of materials. This has been the predominant agenda behind the development of new Advanced Materials to modify properties and enhance performance. The market pull is defined in the main by societal grand challenges. These include The Digital Economy , Energy , Living with Environmental Change and Life-long Health and Well-being, each placing demands on overall performance to suit new applications. Advanced materials are at the core of Systems Engineering that relates to the design and management of complex systems over their complete life cycles, and its nexus with industrial engineering, manufacturing engineering, other branches of engineering and human-centered disciplines (e.g. project and risk management). Embedded in this is the need for Sustainable Materials Management and, increasingly, Sustainable Materials Manufacture , to reuse and sustain materials more productively, and affordably. All this places great demand on the need to engineer new advanced materials. Whilst synthetic chemistry has, and continues to, advance core synthetic technologies to ‘build’ new materials through monomer provision, higher order polymerisation and functionalisation, synthetic biology is beginning to identify new ways of accessing chemical space [ [2] , [3] , [4] ]. This opens up possible new chemical connectivities not accessible to the synthetic chemist and the rapid exploration of diverse (bio)-molecular structures. The conflation of synthetic biology and (combinatorial) synthetic chemistry, and exploration of potential connections with contemporary manufacturing platforms such as Additive Manufacturing (3D printing), defines a new era in the exploration of new advanced materials extending from basal materials with new (desirable) properties to complex and well-defined 3D meso-structures (3D topologies). Supplement that with developments in Artificial Intelligence (e.g. Machine Learning ) to learn and predictively design new advanced materials in rapidly implemented (automated) iterative Design, Build, Evaluate, Learn cycles, and one has a powerful series of technology platforms with which to navigate the new advanced materials landscape. The unification of synthetic biology with other frontier sciences and technologies will usher in the new synthetic biological materials era. In the main, the definition of Biomaterials has been associated traditionally with healthcare applications, for example in the development of biocompatible scaffold materials (tissue regeneration), structural biocompatible materials (prosthetics) and new materials for drug delivery (biomedical devices) [ 5 ]. This can be classified as Materials for Biology . With synthetic biological materials the focus is more on Materials from Biology and the harnessing of new capability platforms (e.g. synthetic biology; additive manufacturing) in an integrated fashion with leading developments in more established fields (e.g. ‘Click’ chemistry; machine learning; automation; miniaturisation of materials evaluation platforms). By bringing deeper biological thinking to advanced materials societal grand challenges can be met. Biology will bring sustainable and affordable manufacture of complex new materials that will impact not only in Healthcare, but also in other sector challenge areas (e.g. Energy , Military , Advanced Manufacturing , Living with Environmental Change , Digital Economy etc.). This will give rise to a wide range of new advanced materials, especially – although not exclusively – in the realm of soft materials that can be functionalised, elaborated and assembled hierarchically, and validated rapidly, for target applications. The opportunity: By harnessing the power of synthetic biology, existing materials discovery platforms and fabrication technologies would be augmented to widen the materials development space and define a new materials paradigm – defined as Synthetic Biological Materials . This would enable delivery of next generation advanced materials with new and extended functional properties to address a wide range of unmet needs. Realising this opportunity would also provide access to affordable and sustainable routes to the production of synthetic biological materials. This would be an ambitious, high-risk and high-gain proposition, dependent on the emerging science of synthetic biology, but one that would deliver a new landscape for advanced materials and in a way that is fundamentally different to more traditional materials discovery and fabrication platforms. Through the engineering of biology, the synthetic biological materials paradigm would give access to i) rapid expansion of materials diversity, ii) hierarchical assembly of new multi-component, multi-functional materials (e.g. by uniting synthetic biology and materials fabrication platforms including additive manufacturing, spinning, coating technologies), iii) affordable and/or sustainable routes to production (e.g. fermentation on waste feedstocks) and iv) lessened environmental impact in next generation materials manufacture. This defines a large landscape for synthetic biological materials through the ‘writing of DNA’, and assembly of these biological ‘parts’ with other (e.g. non-biological) components. Synthetic biological materials will therefore span soft and hard materials, composites and other more complex multi-component and multi-functional structures."
} | 1,843 |
35102192 | PMC8803863 | pmc | 1,547 | {
"abstract": "Developing a single-phase self-rectifying memristor with the continuously tunable feature is structurally desirable and functionally adaptive to dynamic environmental stimuli variations, which is the pursuit of further smart memristors and neuromorphic computing. Herein, we report a van der Waals ferroelectric CuInP 2 S 6 as a single memristor with superior continuous modulation of current and self-rectifying to different bias stimuli (sweeping speed, direction, amplitude, etc.) and external mechanical load. The synergetic contribution of controllable Cu + ions migration and interfacial Schottky barrier is proposed to dynamically control the current flow and device performance. These outstanding sensitive features make this material possible for being superior candidate for future smart memristors with bidirectional operation mode and strong recognition to input faults and variations.",
"introduction": "Introduction Recent advances in controlling the ionic migration processes in solid-state thin films have led to the rapid development of ionic functionalities 1 , 2 . One representative application is ionic memristors for neuromorphic computing 3 – 5 , whereas the main obstacle is the undesired current path flowing through neighboring memory cells in the crossbar array architecture, resulting in write/read inaccuracy and unnecessary energy consumption 6 . To overcome this issue, ionic memristors must possess a high current rectification or nonlinearity, which can be solved by artificially connecting a series of transistors 7 , diodes 8 , or nonlinear selectors 9 for constructing self-nonlinearity/self-rectifying. However, these complex methods inevitably bring sophisticated micro-nano fabrication/integration, low stacking density, and incompatible 3D vertical crossbar array under certain circumstances. Therefore, developing a single memristor system with self-rectifying characteristics is desirable due to its simplicity, but the progress towards this goal is hampered to a large extent by a lack of a suitable material. Hitherto, the research of single memristors mainly focuses on oxide-based multicomponent layers by tuning oxygen defects or a single insulating material by grafting the Ag filament. The former excessively depends on a multilayer interface and inevitably suffers from the current leakage from oxygen defects 10 , 11 . The latter needs assistance by doping fast diffusive silver ions or using active silver electrodes 12 , 13 . It’s also worth mentioning that the current developed ionic memristors usually show a unidirectional rectifying direction. For example, in Ag-based memristors, the relatively abrupt set process makes it difficult to continuously control the conducting states and current rectifying direction as a response to external stimuli 14 . However, these dynamic tunable features are critical for bidirectional operation mode and environment-adaptable learning through the neural network as a response to dynamic environmental stimuli. Therefore, a single-phase memristor simultaneously with highly mobile ions and continuously tunable feature is in urgent need of exploration. Electric field tuning of physical properties in Van der Waals materials is a very lively field 15 – 17 . In recent years, the memristor behavior of 2D van der Waals material has also aroused great attentions 18 – 20 . One veritable star compound from the thio/selono-phosphate family, CuInP 2 S 6 (CIPS), meets the requirements for a tunable single-phase memristor from the following three advantages. (1) An identifying feature of this material is the electrochemically active Cu + ions in lattice, which endows itself with an excellent insulating property at a low voltage, and exhibits highly sensitive response of current to electric stimuli above a threshold voltage 18 , 21 . (2) Due to the van der Waals interaction, CIPS can be easily exfoliated and transferred to desired substrates. (3) Current developed ionic memristors are mostly based on metallic oxide materials. These materials are usually rigid and non-piezoelectric, which are insensitive to external mechanical stimuli. On the contrary, CIPS material (Cc phase) is the only discovered 2D van der Waals ferroionic material so far 22 , and possesses good flexibility and piezoelectric characteristic 22 , 23 , which makes it sensitive to the external strain/strain gradient field. Taking above mentioned advantages, in this work, we study the ionic migration of Cu + and current rectifying behavior in single-phase CIPS materials by using conductive scanning tips to apply different bias and mechanical stimuli. A 120-nm-thick CIPS single material shows a high ON/OFF ratio of ~10 3 along with a self-rectifying ratio of ~10 3 . More importantly, this material possesses an excellently continuous modulation of current and rectifying direction by different bias (including the amplitude, speed, and direction) and mechanical strain, which is physically based on the synergetic contribution of controllable Cu + ions migration and interfacial Schottky barrier. These outstanding sensitive features make this material possible for being superior candidate for smart memristors with bidirectional operation mode and strong recognition to input faults and variations.",
"discussion": "Discussion In summary, we report a single-phase memristor based on van der Waals material, CuInP 2 S 6 (CIPS), which satisfies continuously tunable electric features. An identifying feature of this material is the electrochemically active Cu + ions, which endows CIPS with a superior continuous modulation of current and self-rectifying by the sweeping speed of voltage, voltage direction, voltage amplitude, etc. or mechanical strain. The synergetic contribution of controllable Cu + ions migration and interfacial Schottky barrier change is proposed to control the current flow in the bulk as well as the injection-limited electrode contact. The dynamically tunable feature is in urgent demand for further evolution of smart memristors with bidirectional operation mode and strong recognition to input faults and variations."
} | 1,522 |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.