Nonelectroactive clostridium obtains extracellular electron transfer-capability after forming chimera with Geobacter

Abstract Extracellular electron transfer (EET) of microorganisms is a major driver of the microbial growth and metabolism, including reactions involved in the cycling of C, N, and Fe in anaerobic environments such as soils and sediments. Understanding the mechanisms of EET, as well as knowing which organisms are EET-capable (or can become so) is fundamental to electromicrobiology and geomicrobiology. In general, Gram-positive bacteria very seldomly perform EET due to their thick non-conductive cell wall. Here, we report that a Gram-positive Clostridium intestinale (C.i) attained EET-capability for ethanol metabolism only after forming chimera with electroactive Geobacter sulfurreducens (G.s). Mechanism analyses demonstrated that the EET was possible after the cell fusion of the two species was achieved. Under these conditions, the ethanol metabolism pathway of C.i was integrated by the EET pathway of G.s, by which achieved the oxidation of ethanol for the subsequent reduction of extracellular electron acceptors in the coculture. Our study displays a new approach to perform EET for Gram-positive bacteria via recruiting the EET pathway of an electroactive bacterium, which suggests a previously unanticipated prevalence of EET in the microbial world. These findings also provide new perspectives to understand the energetic coupling between bacterial species and the ecology of interspecies mutualisms.


Introduction
Extracellular electron transfer (EET) is the process whereby some microorganisms exchange electrons with the outside environment and is the prominent characteristic that defines an electroactive (EET-capable) microorganism [1].EET-capable microorganisms were first noted and reported by their ability to perform anaerobic respiration using insoluble metal oxides as electron acceptors [2,3].They have since been shown to utilize anodic electrodes [4], and even other bacteria [5,6] as electron acceptors.Other studies have revealed the ability to carry out chemoautotrophic growth by taking up electrons from low valence state iron as well as from other bacterial electron donors [7].Thus, EET can facilitate the electron transfer between different species, contributing to such key ecological and environmental processes as anaerobic methane generation and oxidation, anaerobic ammonia oxidation, anaerobic photosynthesis, and anaerobic dark CO 2 fixation [8][9][10][11][12].Moreover, it also lays the foundation for the development of bioelectrochemical systems that exploit the electron transfer between microorganisms and electrodes for current generation or value-added chemical production [13,14].Clearly, the study of EET has greatly expanded our understanding and appreciation of the interaction(s) between microorganisms and their outside world, and has garnered considerable attention in the fields of geomicrobiology and electromicrobiology, especially in anoxic ecosystems such as those found in sediments, the deep subsurface, and animal gut microbiomes [7,14].
Our understanding of the mechanisms that enable EET rests mainly on studies of Gram-negative bacteria [15,16].In general, two mechanisms of direct electron transfer (DET) and mediated electron transfer (MET) have been shown.Brief ly, to achieve DET, bacteria either express transmembrane porin-cytochrome protein complex to transfer electrons to the directly physically contacted electron acceptors, or express conductive nanowires to perform long-range DET to reduce remote electron acceptors [17][18][19].In contrast, for MET, redox active mediators either secreted by bacterial cells or preexisting in the environment have been shown to act at high concentrations as electron shuttles to mediate electron transfer between bacterial cells and extracellular electron acceptors [20][21][22][23][24].In contrast, Gram-positive bacteria contain thick non-conductive cell walls (20-80 nm) that are composed of peptidoglycan, teichoic acids, and sometimes are covered with a protective glycoprotein S layer [25,26].Therefore, Gram-positive bacteria have generally been thought to be nonelectroactive with only a few species having been reported to perform EET.For example, it was reported that Lactococcus lactis excreted the redox mediator quinone to enable anode reduction [27], and Listeria monocytogenes was shown to express f lavoproteins that cooperated with f lavin shuttles to transfer electrons extracellularly [28].Lysinibacillus varians utilized a conductive nanowire to achieve EET [29], while Thermincola potens and Carboxydothermus ferrireducens contained multiple c-type cytochrome encoding genes and were proposed to express extracellular cytochromes to aid in the reduction of extracellular electron acceptors [30,31].
Clostridium is a widely distributed Gram-positive genus in the phylum firmicutes.It is a strict fermentative anaerobe that typically produces hydrogen, organic solvents, and/or organic acids as fermentation by-products [32,33].Clostridia have been found as predominant genera in mixed species electroactive biofilms performing anode reduction [34,35].However, only a few Clostridia species have been reported to be electroactive with the EET mechanism(s) remaining unknown [36][37][38], and most were believed to only ferment the complex organics for the production of electron donors benefitting the cohabitating electroactive bacteria [39,40].
We report here that non-electroactive Clostridium intestinale (C.i) acquired EET ability for ethanol catabolism when cocultured with electroactive Geobacter sulfurreducens (G.s) (Fig. 1).C.i was not able to catabolize ethanol since the produced electrons or energy were accumulated as NADH in the cell which results from the lack of a mechanism to discharge from the cell for NAD + replenishment.When grown with G.s, these two species formed an intimate connection via cell fusion by which the cytochrome-based EET pathway of G.s was integrated with the energy metabolic pathway of C.i, providing an EET pathway to C.i for NADH oxidation and energy generation.Meanwhile, G.s transferred the electrons to extracellular electron acceptors for energy production supporting the growth of itself.In this manner, the ethanol oxidation by C.i was driven by and synchronized with the extracellular electron acceptor reduction of G.s.That is, C.i and G.s established energetic coupling (Fig. 1).

Bioelectrochemical system construction and operation
A three-electrode H-shaped anaerobic bioelectrochemical system was constructed (Fig. S1).It consisted of an anode chamber and a cathode chamber with a work volume of 22 ml for each.The two chambers were separated by a proton exchange membrane (Nafion N-117, Thermo Scientific Alfa Aesar, Waltham, MA).The anodic electrode was an indium tin oxide-coated glass plate (Sunyo, Suzhou, China), and the cathodic electrode was made of polished graphite carbon plate (HP Graphite Ltd, Handan, China).The salt bridge filled with saturated KCl was mounted with a saturated calomel electrode to act as the reference electrode [44].The electrolyte was the anaerobic fresh water medium (FWNN) except that a final concentration of 20 mM ethanol or 20 mM acetate was supplied in the anolyte as electron donor [42].Both chambers were kept under strictly anaerobic conditions (80:20 N 2 :CO 2 ).To inoculate, cells were washed with FWNN beforehand.When running electrochemical tests, a constant voltage of 0.3 V (versus Hg/HgCl sat.KCl) was applied to the working electrode and the current was recorded simultaneously using a potentiostat (CH Instrument, Inc., Shanghai, China).Cyclic voltammetry was performed in situ under non-turnover conditions using CHI660E (CH Instrument) by

Coulombic efficiency calculation
Coulombic efficiency (C E ; %) was determined by integrating the current over time: where M S is the molecular weight of ethanol, t b is the total running of time, I is the current, F is the Faraday's constant, b is the number of moles electrons generated per mole of ethanol, V An is the anode liquid volume, Δc is the change of ethanol concentration [45].

Sample characterization
The concentrations of Fe(II) were measured by the ferrozine assay as previously reported [43].The production of acetate was monitored by ultra-high-performance liquid chromatography (Thermo U-3000; Thermo Fisher Scientific, Waltham, MA) equipped with an Aminex NPX-87H column (Bio-Rad, Hercules, CA) and a UV-Vis detector.The eluent was 8 mM H 2 SO 4 running at a rate of 0.6 ml min −1 and acetate was detected at 210 nm according to a previous study [46].The consumption of ethanol was determined by a gas chromatograph (Agilent 7890 A; Agilent Technologies, Santa Clara, CA), which was equipped with a headspace automatic sampler, a HPINNOWAX column (Agilent), and a FID detector as previously reported [46].The NADH/NAD + ratio was measured using NAD + /NADH assay kit (Beyotime, China) following instructions.

Electron microscopic analyses
Cell cultures (10 μL) from the bioelectrochemical system were directly dropped onto a 400-mesh carbon-coated copper grid and were let to standstill for 5 min at room temperature.The grid was then wiped with filter paper and negatively stained with 2% uranyl acetate for 60 s and wiped again [42].The ready-made sample was imaged under a transmission electron microscope (TEM) operating at 80 kV (Tecnai T12, ThermoFisher Scientific).
For microtome sectioning, co-culture cells were embedded in LR White embedding resin.In particular, to stain cytochromes in the coculture, the cytochrome reactive histochemical stain 3,3 ′diaminobenzidine was applied to the co-culture before embedding as previously described [47].The obtained resin blocks were sectioned at 70 nm with a diamond knife (DiATOME Ultra 35 • , Diatome Ltd, Nidau, Switzerland), and f loating sections were applied to grids.Thin sections were then viewed under TEM operating at 80 kV.

Florescence in situ hybridization
Florescence in situ hybridization (FISH) was conducted as previously described with some modifications [48].Cell suspensions were collected and fixed in 4% paraformaldehyde at 4

C.i obtained EET ability in the presence of G.s
C.i was predicted to be able to catabolize ethanol since there is an intact ethanol catabolism for an acetate generation pathway encoded on its genome.Here, ethanol will first be catalyzed by the ethanol dehydrogenases (ADHs) to generate acetaldehyde and then to acetate by acetaldehyde dehydrogenases with the concomitant reduction of NAD + to NADH (Fig. 1).However, C.i was not able to ferment ethanol for hydrogen production due to the thermodynamically unfavorable reaction of NADH oxidation for hydrogen generation catalyzed by [FeFe]-hydrogenase.In addition, C.i. alone was not able to oxidize ethanol via reducing extracellular electron acceptors, neither soluble Fe(III) citrate, solid Fe(III) oxide, nor ITO anode (Fig. 2, Fig. S2).The reason might be that C.i was not able to express EET pathways or be electroactive (Fig. S3) to use those extracellular electron acceptors to discharge the cell for NAD + regeneration then achieving intracellular redox balance, and thereafter the metabolism of ethanol was blocked.We speculated that if an EET pathway was provided, C.i would oxidize ethanol.To demonstrate this, electron shuttles were supplied into the pure C.i culture.Two electron shuttles were selected: (1) resazurin, a lipophilic mediator that is able to pass the cell wall and accepts electrons from periplasm or inner membrane [49], and (2) anthraquinone-2,6-disulfonate (AQDS), a hydrophilic mediator that can only accept electrons from the cell surface [50].As indicated in Fig. S4, the supplement of AQDS was unable to recover the ethanol oxidation.In contrast, the addition of resazurin achieved the oxidation of ethanol (ca.0.8 mM day −1 ) (Fig. S4) and the anode reduction in C.i with a maximum current generation of ca.4.7 μA (Fig. 2D).In particular, the addition of resazurin decreased the proportion of NADH/NAD + in contrast to what occurred in the presence of AQDS showing a ratio comparable to C.i without any treatment (Fig. S5A), as well as, the addition of resazurin also increased the intracellular ATP (Fig. S5B), suggesting a process of oxidative phosphorylation.It is not surprise considering there were NADH-quinone oxidoreductases and Ftype ATPase coding genes (Fig. S6) in the genome of C.i.All in all, the results demonstrated that only the exogenous lipophilic mediator of resazurin was able to facilitate the EET of C.i to achieve intracellular redox balance for energy generation and contributed to the sustained ethanol oxidation, and suggested that C.i also lacked the electron transfer pathway across the cell wall/membrane.G.s is an electroactive bacterium [13] that oxidizes acetate, using EET to reduce extracellular electron acceptors directly for energy generation.It is not capable of growth on ethanol [5,41].However, when G.s was incubated with C.i, the two species formed a coculture capable of ethanol oxidation with the concomitant reduction of extracellular electron acceptors (Fig. 2).Namely, it appeared that C.i obtained EET ability.Since G.s could not directly contribute to the oxidation of ethanol, we speculated that G.s probably facilitated or was involved in the EET of C.i in the coculture.Such assumption was further strengthened using a bioelectrochemical system inoculated with C.i.As expected, C.i was unable to oxidize ethanol to reduce the anode during the preliminary stage (Fig. 2C), but the addition of G.s enabled ethanol oxidization and anode reduction.Notably, neither the metabolites nor any cell components of G.s contribute to the EET of C.i, since neither the addition of the growth culture medium of G.s nor the supplementation with cell-free lysate of G.s was able to recover the EET of C.i (Fig. S7).This raises the issue of the mechanism involved with the participation of G.s in the apparent EET of C.i.

C.i and G.s form energetic coupling
Previous studies of other Clostridium species showed that the formation of either hydrogen or formate provided a mechanism for recycling the NADH via discharge of the NADH [51][52][53], and hydrogen and formate could be further oxidized by G.s for extracellular electron acceptors reduction [54,55].To identify the possibility of hydrogen or formate facilitating the EET of C.i in coculture, both a formate dehydrogenase and uptake hydrogenase-deficient G.s strain, G.s-ΔfdnGΔhybL, which is not able to oxidize formate and hydrogen, was cocultured with C.i in the anode chamber of a bioelectrochemical system.As demonstrated in Fig. 3, the deficiency of formate dehydrogenase and uptake hydrogenase did not affect the ethanol oxidation and the anode reduction of the coculture, and a comparable maximum current of ca.0.4 mA compared to the wildtype coculture was generated.In addition, neither hydrogen nor formate could be detected in either C.i single culture or the coculture.So, neither hydrogen nor formate was involved in the EET of C.i and a direct interspecies electron transfer (DIET) between these two species was hypothesized.
C.i oxidizes ethanol to generate electrons and acetate (Fig. 3B).If DIET could be established in the coculture, those electrons could be transferred to G.s, to ultimately arrive at the anode.Meanwhile, the acetate could be further oxidized by G.s to contribute to the anode reduction for current generation.So, theoretically, two electron sources of C.i and acetate are projected to contribute to the current generation of G.s if the DIET occurred.To verify this process, a citrate synthase-deficient G.s strain G.s-ΔgltA, which was unable to oxidize acetate but could assimilate acetate, was cocultured with C.i.In theory, in this coculture, only the electrons released by C.i would be able to contribute to the current generation and a lower current would be generated.As expected, acetate was accumulated in the anolyte and a much lower current with a maximum of 0.04 mA was generated (Fig. 3A), and a much lower coulombic efficiency of 1.55% was calculated compared to the 31.80% of the wild-type coculture.The decline of current generation in the mutant coculture also suggested that acetate oxidation by G.s primarily contributed to the current generation of the wild-type coculture.Notably, ethanol oxidation was also impaired in this mutant coculture, with only 5 mM ethanol consumed in 5 days compared to 12 mM ethanol consumed in 5 days of the wildtype coculture (Fig. 3B).The reason may due to the accumulation of acetate, which restrained the ethanol oxidation to acetate generation thermodynamically.
Meanwhile, both strain G.s and strain G.s-ΔgltA grew alongside C.i and formed biofilms on the anode (Fig. S8).So, it shows a mutualistic symbiosis that G.s conferred the EET to C.i achieving energy metabolism via oxidative phosphorylation and thereafter contributed to the growth of C.i, and in return, the electrons released by C.i were also able to be used by G.s for energy generation supporting the growth of G.s.More precisely, those two species of C.i and G.s formed an intimate energetic coupling to  achieve electrosyntrophic growth via DIET.Previous studies indicated that extracellular cytochromes in electron donor bacterial species usually facilitated the DIET and were necessary for the formation of electrosyntrophy with the other electron acceptor bacterial species [ 41,56].However, C.i lacks cytochrome-encoding genes on its genome.The mechanism of electron transfer between those two species remains an enigma.

Energetic coupling is established after the formation of interspecies cell fusion
It was also presumed that the formation of direct physical connection is necessary for the DIET between electron donor and electron acceptor bacterial species, either by direct contact or by forming conductive cell aggregates (Fig. S9A) [5,41,57].However, no visible cell aggregates could be seen directly in the coculture ( Fig. S9B).Transmission electron microscopy was performed further to examine the possible physical connection between those two species.The length of a C.i is about 10 μm (Fig. 4A), while the G.s cells have an average length of about 1-2 μm (Fig. 4B), making it easy to distinguish the two species.Figure 4C shows that G.s attached to C.i at the cell pole (Fig. 4C and D) and fused with C.i but kept itself integral by way of cell fusion (probably via the fusion of cell walls) (Fig. 4D and E and Fig. S11).This is in contrast to the direct cell surface contact of the previous reported DIET cocultures [5,9,58], and it is reminiscent of previously reported interspecies cell fusion also involving Clostridium cells [59].In addition, some C.i cells had multiple G.s cells fused to them (Fig. S10).The direct contact between C.i and G.s was also verified by FISH after labeling each species with different species-specific f luorescent probes (Fig. 4F).
Cytochromes are necessary for DIET [41,60].The distribution of cytochromes in the coculture was examined after staining by the histochemical stain 3,3 ′ -diaminobenzidine, which could be catalyzed by hemes of cytochromes to form oxidized precipitate with a high electron opacity in transmission electron microscopy images in the presence of hydrogen peroxide [47].As indicated in Fig. S11, only G.s cells were stained dark, indicating that there were abundant cytochromes on G.s but not on C.i. OmcS was thought to be necessary for G.s to directly accept electrons from the electron-donating species in the DIET coculture [5,41].To identify the possibility that OmcS facilitated the DIET in the coculture, the omcS mutant G.s strain G.s-ΔomcS was cocultured with C.i in a bioelectrochemical system and the current generation was monitored.Considering OmcS was also thought to be involved in the anode reduction of G.s [42], the current generation of single G.s-ΔomcS was also compared.As shown in Fig. S12, the deletion of omcS only partially impaired the current generation of G.s as previously reported [42], with a maximum of ca.0.23 mA, which is in contrast to the maximum of ca.0.4 mA in wildtype G.s and the coculture.However, the deletion of omcS greatly impaired the current generation of the coculture, only generating a maximum current of ca.0.02 mA (Fig. 5A).This result indicated that OmcS participated in the interspecies electron transfer in the coculture.We further deleted three other of the most abundant extracellular c-type cytochromes, namely OmcB, OmcE, and OmcT, in strain G.s-ΔomcS, generating strain G.s-ΔomcBEST, and cocultured this strain with C.i in the bioelectrochemical system.As demonstrated in Fig. 5A, the current generation of this coculture was completely blocked, even though the single G.s-ΔomcBEST strain was still able to reduce the anode, generating a maximum current of ca.0.1 mA (Fig. S12).All of these results indicated that the DIET between C.i and G.s was facilitated by extracellular cytochromes of G.s or cytochromes from G.s conferred the EET on C.i.

Discussion
The results presented here are consistent with a new EET mechanism for Gram-positive bacteria via the recruitment of the EET pathway from an EET capable Gram-negative bacteria.C.i could not metabolize ethanol since it could not perform EET to discharge NADH, thus achieving intracellular redox balance.In contrast, G.s is able to express abundant extracellular cytochromes to facilitate EET.Our results showed that the pole end of G.s attached on C.i after cell fusion by which the G.s cytochrome dependent EET pathway was integrated with the oxidation of NADH in C.i, allowing the cycling regeneration of NAD + to drive ethanol metabolism and thereby discharging the C.i cell and generating the energy for C.i.We conclude that those electrons should not be transferred to extracellular electron acceptors directly, otherwise they would be transferred into G.s and used by G.s to contribute to the reduction of extracellular electron acceptors for energy generation.In the meantime, acetate was generated during ethanol metabolism.Although G.s could couple the oxidation of acetate with the reduction of extracellular electron acceptors, the direct oxidation of acetate by G.s contributed to the reduction of the majority of extracellular electron acceptors.Ultimately, those two species formed energetic coupling to achieve ethanol oxidation and extracellular electron acceptor, such as anode, reduction cooperatively ( Fig. 5B).
The energetic coupling represents a cooperating energy metabolism between C.i and G.s, which is in contrast to the previously reported coculture between Clostridium pasteurianum and G.s displaying a substance metabolism interaction [61].In the coculture, C. pasteurianum was able to grow independently via glucose fermentation with the production of acetate.G.s oxidized the acetate to release electrons which were further taken up by C. pasteurianum via its transmembrane f lavin-bound polyferredoxin and thereafter affected its intracellular redox state.Meanwhile, G.s secreted cobamide molecules that could modify the metabolic pathway of C. pasteurianum [37].So, in the coculture, C. pasteurianum showed a metabolic shift with a higher acetate and hydrogen yield.In our coculture, the interspecies electron transfer was coupled with the energy metabolism of both species and the secretion of metabolite did not contribute to the energetic coupling.
Our results suggest a heretofore unknown contribution of cell fusion to energetic coupling.However, the interspecies cell fusion in our coculture seems logical and inevitable, considering the nonconductivity of C.i cell wall.In previous studies, exchange of materials between cells after cell fusion has been shown to be necessary for the formation of interspecies energetic coupling [62,63].For example, in the Clostridium acetobutylicum and Desulfovibrio vulgaris coculture, the exchange of cytoplasmic molecules between those two species contributed to the survival of D. vulgaris in its nutrient-deficient environment and the enhancement of hydrogen fermentation of C. acetobutylicum [62].Similarly, in a Clostridium coculture of C. ljungdahlii and C. acetobutylicum, the direct cell fusion allowed the exchange of metabolites that enabled complete carbon utilization and expanded the metabolic product spectrum of individual species [59].In contrast, the energetic coupling via C.i and G.s cell fusion of this study was based upon direct electron exchange.We did not think the possibility of material exchange, such as the exchange of cytoplasmic reducing power of NAD(P)H or FdH 2 , between C.i and G.s cells, considering that both cells kept intact cell shapes after cell fusion and the extracellular EET pathway of G.s is necessary for the energetic coupling of the coculture.A previous study also suggested that nutritional stress could trigger cell fusion between bacterial species [62].So, we speculated that the nutritional stress from both C.i and G.s should also trigger the cell fusion of those two species in our study which is warranted to further study.
The formation of DIET between non-electroactive bacteria and electroactive bacteria is different from previously wellcharacterized classical electronic syntrophic coculture composed by both electroactive microorganisms [31,41,56].In those electrosyntrophic cocultures, cytochromes from the electron donor species are required for the DIET, and the expression of cytochromes in electron donor species acts as predictors to select electrosyntrophic cocultures [9].Here, we showed that the cytochrome-free Clostridium species could utilize cytochromes from the partner electron acceptor species to set up DIET.In addition, a new interspecies connection mode in DIET coculture via cell fusion was suggested.So, the discovery extends the knowledge on DIET.Furthermore, our study also provides a new perspective on the understanding of ectosymbiosis i.e. a symbiont dwelling on the external body of the host, which has never been reported between bacterial species.Notably, even though G.s invaded the cell wall of C.i, the relationship between those two species should also be mutualism, considering that the survival of any species is dependent on the metabolism of its partner species in the coculture.
Our study also provides a perspective that may lead to a better understanding of the ecology of microbial communities.In general, Clostridium performs fermentation to generate energy for cell growth [64].So, a low energy production mechanism of substrate level phosphorylation was involved in the ATP generation.Clostridium also has NADH quinone oxidoreductase and ATPase.By forming energetic coupling with electroactive Geobacter species, Clostridium activated the NADH quinone oxidoreductase (Fig. S13) and performed energy-efficient oxidative phosphorylation via EET to f lare off electrons on extracellular electron acceptors.Here, more energy could be harvested and the survival of Clostridium in microbial communities should be favorable.Actually, previous studies demonstrated that Clostridium species are widely present in mixed-culture BESs and there is a positive correlation between the dominance of Clostridium and current generation [35,65].Correspondingly, a "division-of-labor" model to explain the interspecific relationship in those mixed species electroactive microbial communities was suggested [66,67].In this model, fermentative bacteria such as Clostridium species ferment complex organics to produce small molecular organic acids, which further are able to act as electron donors to be oxidized by electroactive bacteria for the current generation.However, this is not contrary to our electrosyntrophic coculture model since the majority of electroactive Geobacter still lived on the catabolite of Clostridium, and our study illustrated another possible interaction of direct electronic connection between those two species.
It is striking that Clostridium received a new mode of energy metabolism after capturing Geobacter cells, which is reminiscent of the evolution of the mitochondrion [68].The mitochondrion is generally recognized as an exogenously evolved organelle functioning as a specialized energy generation machine for the host cell by which a symbiotic relationship via energetic coupling was established.Even though the origin of mitochondrion is controversial, the recent data indicated that mitochondrion originated from endosymbiosis of bacteria in archaeal host cell.However, geologic evidences and genetic analyses of archaea exclude an endocytic event to contribute to this intracellular symbiosis [69,70], and the formation of mitochondrion is mysterious.In contrast, our study showed that Geobacter cells contributed to a new energy metabolism of Clostridium cells after attached on and only partially fused with the host.This leads us to the speculation that interspecies cell fusion could be the mechanism by which archaea acquired new abilities and became pre-mitochondria; partial fusion between bacterial species may have played an important role in the genesis and evolution of mitochondria.
As a final note, we point out that the single non-electroactive Clostridium achieved electroactivity after coculture with electroactive Geobacter, a result that suggest that we may need a more expansive definition of electroactive bacteria.To this end, a new biotechnology to generate electroactive bacteria via synthetic microbial communities rather than the complex synthetic biology is not irrational.In addition, these findings also imply more functional diversity of bacteria in a community than the single cell, highlighting the importance of studying microorganisms at the community level.Finally, our study also provides an explanation for the persistence of unculturable bacteria and has profound implications for ecological or community engineering of microbial ecosystems.

Figure 1 .
Figure 1.The proposed EET mechanism of C.i when cocultured with G.s; C.i catabolized ethanol for the generation of acetaldehyde by ethanol dehydrogenases (ADHs) and further oxidized to acetate by acetaldehyde dehydrogenases (ALDHs), and electrons were released and accumulated as NADH during this process simultaneously; in the coculture, G.s fused its cell with C.i by which the cytochrome-based EET pathway of G.s integrated with the energy metabolic pathway of C.i and thereafter discharged the C.i cell for NADH oxidation; such process contributed to the formation of proton motive force for the energy generation of C.i; at the same time, the electrons transferred into G.s and finally used for the reduction of extracellular electron acceptors with the energy generated in G.s concomitantly; here, the two species formed an energetic coupling that enabled both ethanol oxidation and extracellular electron acceptor reduction synchronously; Cyts is cytochromes.

Figure 2 .
Figure 2. The reduction of extracellular electron acceptors; (A) Fe(III) citrate; (B) ferrihydrite; (C) anode.C.i and G.s were tested; C.i and G.s were inoculated simultaneously (C.i and G.s) or sequentially (C.i + G.s).The arrow indicates the inoculation of G.s; (D) averaged current generation of C.i with and without the supply of 100 μM resazurin; the results shown were the means ± s.d. for quadruple cultures; the shaded area represents one standard deviation.

Figure 3 .
Figure 3. Ethanol oxidation and anode reduction of mixed species coculture; (A) averaged current generation of C.i and wild-type G.s coculture, C.i and G.s-ΔfdnGΔhybL (formate dehydrogenase and uptake hydrogenase double deletion strain of G.s) coculture, and C.i and G.s-ΔgltA (citrate synthase deletion strain of G.s) coculture; the shaded area represents one standard deviation; four independent tests were performed for each culture; (B) ethanol consumption and acetate accumulation in coculture of C.i and G.s, C.i and G.s-ΔfdnGΔhybL, C.i and G.s-ΔgltA growing in a bioelectrochemical system; the results shown were the means ± s.d. for quadruple cultures.

Figure 4 .
Figure 4. Microscopy of C.i and G.s in pure culture and coculture; transmission electron microscopy image of C.i (A) and G.s (B); (C) transmission electron microscopy image of the C.i and G.s coculture; (D) transmission electron microscopy image of the intersection between C.i and G.s; (E) Image of thin section of coculture cells ; arrows indicate the interspecies cell fusion; (F) confocal laser scanning microscope image of the coculture after f luorescence in situ hybridization of species-specific probes; the C.i cell was labeled with a green probe, and the G.s cell was labeled with a red probe; coculture cells were collected from the bioelectrochemical system; shown were representative images from at least 10 images.

Figure 5 .
Figure 5.The reduction of extracellular electron acceptors by coculture; (A) anode reduction by cocultures of C.i and G.s, C.i and G.s-ΔomcS (strain of G.s-deficient in the expression of extracellular cytochrome OmcS), and C.i and G.s-ΔomcBEST (strain of G.s-deficient in the expression of quadruple cytochromes of OmcB, OmcE, OmcS, and OmcT); the shaded area represents one standard deviation; four independent tests were performed for each culture; (B) the extracellular electron acceptor reduction model in C.i and G.s coculture; C.i oxidized ethanol and produced electrons and acetate; those electrons were directly transferred into G.s via the extracellular cytochromes of G.s and were further transferred to extracellular electron acceptors by G.s; meanwhile, the acetate was released into the culture medium which would be oxidized by G.s to reduce extracellular electron acceptors directly.