microbial community structure in a biofilm anode fed with a fermentable substrate: the significance...

ARTICLE

Microbial Community Structure in a Biofilm AnodeFed With a Fermentable Substrate: TheSignificance of Hydrogen Scavengers

Prathap Parameswaran, Husen Zhang, Cesar I. Torres, Bruce E. Rittmann,Rosa Krajmalnik-Brown

Center for Environmental Biotechnology, The Biodesign Institute at Arizona State University;

telephone: 1-480-727-7574; fax: 1-480-727-0889; e-mail: [email protected]

Received 12 June 2009; revision received 9 August 2009; accepted 12 August 2009

Published online 17 August 2009 in Wiley InterScience (www.interscience.wiley.com)
. DOI 10.1002/bit.22508
ABSTRACT: We compared the microbial community struc-tures that developed in the biofilm anode of two microbialelectrolysis cells fed with ethanol, a fermentable substrate—one where methanogenesis was allowed and another inwhich it was completely inhibited with 2-bromoethanesulfonate. We observed a three-way syntrophy among etha-nol fermenters, acetate-oxidizing anode-respiring bacteria(ARB), and a H2 scavenger. When methanogenesis wasallowed, H2-oxidizing methanogens were the H2 scavengers,but when methanogenesis was inhibited, homo-acetogensbecame a channel for electron flow from H2 to currentthrough acetate. We established the presence of homo-acetogens by two independent molecular techniques: 16SrRNA gene based pyrosequencing and a clone library from ahighly conserved region in the functional gene encodingformyltetrahydrofolate synthetase in homo-acetogens. Bothmethods documented the presence of the homo-acetogenicgenus, Acetobacterium, only with methanogenic inhibition.Pyrosequencing also showed a predominance of ethanol-fermenting bacteria, primarily represented by the genusPelobacter. The next most abundant group was a diversecommunity of ARB, and they were followed by H2-scaven-ging syntrophic partners that were either H2-oxidizingmethanogens or homo-acetogens when methanogenesiswas suppressed. Thus, the community structure in thebiofilm anode and suspension reflected the electron-flowdistribution and H2-scavenging mechanism.

Biotechnol. Bioeng. 2010;105: 69–78.

� 2009 Wiley Periodicals, Inc.

KEYWORDS: methanogenesis; homo-acetogenesis; anoderespiring bacteria; syntrophy; microbial electrolysis cells

Introduction

The microbial ecology at the anode of a microbial fuel cell(MFC) or a microbial electrolysis cell (MEC) is becoming a

Correspondence to: R. Krajmalnik-Brown

� 2009 Wiley Periodicals, Inc.

well-studied topic (Choo et al., 2006; Ishii et al., 2008; Jonget al., 2006; Kim et al., 2007; Lee et al., 2003). Many of thesestudies are based on molecular microbial analyses fromanodes fed with a variety of substrates (e.g., acetate, ethanol,glucose, glutamate, and cellulose). Fermentation appears tobe an important process for the utilization of complexorganic substrates, alcohols, and simple acids other thanacetate (Freguia et al., 2007; Ishii et al., 2008; Jung andRegan, 2007; Kim et al., 2007; Liu et al., 2005; Ren et al.,2007; Torres et al., 2007). However, coulombic efficienciesdropped significantly when fermentable substrates wereused at the anode (Lee et al., 2008; Min and Logan, 2004),and methane was a major electron sink accounting for thedecrease in coulombic efficiency. Fermenters can out-compete anode-respiring bacteria (ARB) that can directlyutilize glucose such as Rhodoferax ferrireducens in sedimentMFCs (Lovley, 2006b). The impact of fermenters towardanode respiration in the biofilm anode, either positivelyor negatively, is unknown as well. The above observationsunderscore that it is essential to understand the anode’smicrobial ecology and its relationship to communitystructure and MFC performance.

Ethanol is a common fermentation product at neutral pH(Combet-Blanc et al., 1995; Lee and Rittmann, 2008;Temudo et al., 2007), and it is usually fermented to acetateand H2 before producing electricity in an MFC (Kim et al.,2007; Parameswaran et al., 2009; Richter et al., 2007).Acetate is a well-known electron donor for ARB, butevidence regarding the ability of ARB to oxidize H2 isunclear. Electric current was produced with 100% H2 gasonly using a pure culture of Geobacter sulfurreducens (Bondand Lovley, 2003), and a mixed culture developing at ananode that was fed with acetate first and then switched to H2

(Lee et al., 2009; Rozendal et al., 2008). However, otherresearchers reported that their ARB consumed H2 slowly ornot at all (Lee et al., 2003; Torres et al., 2007).

In our earlier study (Parameswaran et al., 2009), wedemonstrated that ethanol was converted to acetate and H2

Biotechnology and Bioengineering, Vol. 105, No. 1, January 1, 2010 69

in our MEC system before producing electric current fromthe anode. When methanogenesis was allowed, H2 wasrouted to CH4 by H2-oxidizing Methanobacteriales. Thisresulted in a coulombic efficiency of 60%, while 26% ofelectrons from ethanol ended up in CH4, with biomassand soluble microbial products (SMP) comprising 14%.When we added 50mM 2-bromoethane sulfonate (2-BES),which inhibits methyl coenzyme reductase A activity in allmethanogens, the electron flow to current increased to 84%,no CH4 was observed, and 16% of the electrons ended up asbiomass and SMP.

H2 produced from ethanol fermentation must beconsumed efficiently either by ARB or other H2-oxidizingmembers of the community for fermentation to proceed.The requirement for H2 consumption gives rise tosyntrophic interactions based on interspecies H2 transferin mixed cultures (Nath and Das, 2004; Schink, 1997).Figure 1 shows the possible three-way syntrophies in anethanol-fed MEC.When we did not inhibit methanogenesis,H2-oxidizing methanogens (Parameswaran et al., 2009)formed a three-way syntrophy with the fermenters andacetate-oxidizing ARB. Acetoclastic methanogens, with ahalf-maximum rate concentration (Ks) of 177–427mgCOD/L were outcompeted for acetate by the acetate-oxidizing ARB, with a half-maximum rate concentration of0.64mg COD/L (Esteve-Nunez et al., 2005; Lawrence andMcCarty, 1969), and hence the acetoclastic methanogens arenot shown in Figure 1. However, when methanogens wereinhibited, current was produced as a result of H2 releasedin ethanol fermentation. One possibility is that homo-acetogens consumed the H2 and produced acetate consumedby acetate-oxidizing ARB. Since the metabolism of homo-acetogens is similar to that of H2-oxidizing methanogens(Zehnder, 1988), a three-way syntrophy can be establishedwith fermenters, H2-oxidizing homo-acetogens, and acet-ate-oxidizing ARB. Finally, H2-oxidizing ARB may haveconsumed H2 directly and formed another syntrophy, whichalso involves the fermenters and ARB that consume acetate.

Figure 1. Schematic showing the three-way syntrophic interactions that

develop at the biofilm anode utilizing a fermentable substrate, ethanol. Hydrogen

scavengers (hydrogenotrophic methanogens, homo-acetogenic bacteria, and

H2-consuming ARB) form the vital link for this three-way syntrophy, resulting in three

different scenarios for the syntrophic relationship, as shown.

70 Biotechnology and Bioengineering, Vol. 105, No. 1, January 1, 2010

Using typical biomass yields, we estimated the expecteddistributions of biomass among the various groups ofmicroorganisms when ethanol is consumed at the biofilmanode. Table I summarizes the distribution and providesthe sources for the yields. We based our values for ARB onG. sulfurreducens, a representative of the Geobacteraceaefamily, which is predominant in laboratory and sedimentfuel cells (Bond et al., 2002; Holmes et al., 2004; Jungand Regan, 2007; Lee et al., 2008; Phung et al., 2004).G. sulfurreducens consumes H2 and acetate (Bond andLovley, 2003; Esteve-Nunez et al., 2005; Richter et al., 2007),and this allows us to analyze the biomass electrondistribution for H2-oxidizing ARB in a simple manner.The biomass distribution in Table I is based on typical valuesfor f os (the fraction of electrons from the donor substratethat is utilized toward synthesis of biomass, which is aninherent property of each group of microorganism), fromthe literature. Even if we placed a range on the f os values from0.08 to 0.15 as observed for different fermenters (Chapter 13,Rittmann and McCarty, 2001), the amount of biomass fromfermenters (in meq) does not change by more than 20%from the original value of 10meq, which is the valuereported in Table I. The possible changes are similarly smallfor all other groups of bacteria and archaea in Table I as well.From the biomass distribution in Table I, fermenters formthe largest group (10meq of the 100meq from ethanol),followed by the acetate-oxidizing ARB (3meq); both arepart of a distinct syntrophic group. Among the possiblethird members of the three-way syntrophy (i.e., the H2

consumers), the homo-acetogens have the highest biomasssynthesis (3meq), followed by methanogens (2.4meq),and H2-oxidizing ARB (1.5meq).

Parameswaran et al. (2009) provided evidence throughelectron balances pointing to homo-acetogens as the H2

scavengers when methanogens were suppressed. However,proof of this hypothesis demands direct microbiologicalevidence for homo-acetogens. A challenge for gaining directproof is the high phylogenetic diversity of homo-acetogens.To date, homo-acetogens are found in 20 different genera,most of them affiliated with the phylum of Gram-positivebacteria with low GþC content, including 8 of the 19Clostridium clusters (Drake et al., 2002; Stackebrandt et al.,1999). On the other hand, the gene for formyltetrahydro-folate synthetase (FTHFS), catalyzing the ATP-dependentactivation of formate, is highly conserved in homo-acetogenic bacteria (Drake et al., 2002). Thus, a diversityassessment targeting the functional gene may complementthe phylogenetic limitations for the identification of homo-acetogenic bacteria.

The goals of this study are to evaluate if (1) homo-acetogens were significant members of the microbialcommunity when methanogenesis was completely inhibitedin an MEC fed with ethanol and (2) the microbial structurein an MEC fed with ethanol followed the general electron-distribution trends in Table I. To accomplish both goals, weused ethanol as the substrate in batch MECs and firstquantified the end-product electron distribution with and

Table I. Predicted distribution of biomass among the three groups of microorganisms when H2 is scavenged by three different mechanisms—

methanogenesis, homo-acetogenesis, and H2 consumption by ARB.

Syntrophic group

Substrate for

the group

Electrons available

(meq) to the group f osa

Electrons to

biomass (meq) References

Ethanol fermenters Ethanol 100 0.1 10 Thauer et al. (1968)

ARB-consuming acetateb Acetate 60.7 (88.06)d 0.05 3.04 (4.40)d Esteve-Nunez et al. (2005)

H2 scavengersc

H2-oxidizing methanogens H2 30.4 0.08 2.4 Rittmann and McCarty (2001)

Homo-acetogens H2 30.4 0.1 3.04 Bainotti and Nishio (2000)

ARB-consuming H2e H2 30.4 0.05 1.52 Bond and Lovley (2003)

af os represents the fraction of electrons from the donor substrate that is utilized toward synthesis of biomass, which is an inherent property of each group ofmicroorganism.

bRepresentative ARB considered here is Geobacter sulfurreducens.cThe processes occur independently and represent three different scenarios.dThe values in parentheses indicate the electrons available to acetate-oxidizing ARB, considering the possibility of having homo-acetogens convert H2 to

acetate.eGeobacter sulfurreducens is the representative ARB here as well.

without methanogenesis (Parameswaran et al., 2009). Weanalyzed the anode microbial community from these batchMECs with pyrosequencing to explore the full diversity ofthe biofilm anode. We used two independent methods—one targeting the 16S rRNA gene to capture most membersin the bacterial community and a second targeting theFTHFS gene—to establish the presence and diversity ofhomo-acetogens.

Materials and Methods

Batch MEC Experiments and Inoculum

The inoculum in the anode chamber of an MEC consistedof 2mL each of thickened anaerobic digested sludge (�1%solids) and return activated sludge from theMesa NorthwestWastewater Reclamation Plant (Mesa, AZ). After threeserial acclimations to ethanol as substrate in MECs operatedin batch mode, we carried out two batch experimentssimultaneously to establish electron mass balances. Eachsuccessive acclimation was meant to select for an efficientcommunity to consume a given substrate before beginningthe actual electron balance experiments, as observed byother researchers (Lee et al., 2008; Rabaey et al., 2003). Wecarried out experiments in H-type MECs with an anionexchange membrane (AMI 7001, Membranes International,Glen Rock, NJ) between the anode and cathode compart-ments. The two electrodes in the anode compartmentwere graphite rods having an outer diameter of 0.4 cm(www.graphitestore.com) and each electrode had a surfacearea of 4.8 cm2. The anode electrodes were poised at apotential of �200mV versus Ag/AgCl to ensure that theARB were not limited by anode potential (Torres et al.,2007). Total volumes of the anode compartment were 320 or325mL for ethanol-only and ethanol-plus-BES reactors,respectively. We fed one MEC with ethanol only and theother with ethanol and 50mM 2-bromoethane sulfonicacid to inhibit all activities of methanogens. We operated

the reactors until the current production stopped inboth reactors (�33 days in both reactors) and we analyzedall electron sinks at regular intervals along the batchoperation.

DNA Extraction

At the end of batch operation, we extracted DNA frombiofilms and from the suspension for comparison purposes.We centrifuged the entire reactor volume to obtain a pelletfor DNA extraction from the suspension fraction. Forobtaining biofilm samples, we cut a defined cross-section ofthe anode electrode with the biofilm, removed the biofilmwith a pipette tip, and suspended 0.25 g of biomass in thebead tube provided by MOBIO1 Powersoil DNA extractionkit, quantified the DNA with a nanodrop spectrophot-ometer, and documented its yield and purity (characterizedby 260/280 nm absorbance ratio). We normalized the DNAto the same concentration for further analyses, namelyquantitative real time PCR (QPCR), FTHFS gene clonelibrary, and 16S rRNA gene based pyrosequencing.

Amplification of Formyltetrahydrofolate Synthetase(FTHFS) Genes

We used degenerate primers that targeted the highlyconserved FTHFS gene in homo-acetogens (Leaphart andLowell, 2001). We carried out the PCR amplificationreaction in 20mL reaction volumes. Each reaction tubecontained 80mM of each dNTP, TAQ polymerase, 2.5mMof MgCl2, TAQ DNA polymerase, PCR buffer, 100 nM eachof forward and reverse primer, and 0.02mg per reactionof template DNA. As a positive control, we used genomicDNA from Eubacterium limosum (ATCC no: 8486D), arepresentative homo-acetogen. We separated the resultingPCR products using agarose gel electrophoresis in a 1 g/Lagarose gel in Tris-acetate–EDTA buffer and visualized themusing the GELDOC (BIORAD Systems, Hercules, CA, 1997)unit.

Parameswaran et al.: Community Structure in a Biofilm Anode 71

Biotechnology and Bioengineering

Clone Library and Sequence Analysis

In order to sequence the FTHFS PCR products, we used theamplicons from the FTHFS-gene PCR reaction to establish aclone library.We purified the PCR products with a QIAGENPCR purification kit (Qiagen, Valencia, CA) per manufac-turer’s recommendations. Purified PCR products werecloned in TOP10 Escherichia coli cells using a TOPO TAcloning kit for sequencing (Invitrogen, Carlsbad, CA). Atotal of 49 clones were obtained (Genbank accessionnumbers: FJ848919–FJ848965). We amplified the insertwithin the clone using vector-targeted primers M13F andM13R. The resulting PCR products were cleaned upfor sequencing using EXOSAP-IT (USB Corporation,Cleveland, OH). The cleaned products were sequenced atthe DNA laboratory at Arizona State University, with anApplied Biosystems 3730 capillary sequencer.

Pyrosequencing

We used bacterial primers 967f and 1046r to amplify the V6region of the 16S rRNA gene as described by Sogin et al.(2006) and Zhang et al. (2009). We performed PCR asfollows: 948C for 2min, 25 cycles of denaturation at 948C for30 s each, 578C annealing for 45 s, 728C for 1min extension,and a final extension at 728C for 2min. Excess primer dimersand dNTPs were removed with QiaQuick spin columns(Qiagen). Amplicon pyrosequencing was performed withstandard 454/Roche GS-FLX protocols.

After a sequencing run and base-calling, we sorted thesequences by unique tags using the 454 script (sfffile) toseparate and group all data and then trimmed the sequencesusing the 454 script (sffinfo) for downstream analysis. Toassign 454 pyrosequencing tags to their closest relative in areference database with 44,011 non-identical V6 sequencesextracted from 119,480 bacterial rRNA genes (Sogin et al.,2006), we carried out four data-processing steps. First,454 reads were preprocessed to remove ambiguous and shortsequences, all sequences having mismatches with the PCRprimers, and all sequences having less than 50 nucleotidesafter the proximal primer (unless they reach the distalprimer). These filtering steps eliminated all the sequenceshaving more than one ambiguity (N). We used theRibosomal Database Project Classifier 2.0 (Wang et al.,2007) to assign taxonomy to pyrosequencing tags.

Results and Discussion

Bacterial and Archaeal Populations in the Ethanol-OnlyControl MEC

Figure 2a shows the overall distribution of bacteria andarchaea in the suspension fraction of the ethanol-only MEC.The distributions of total bacteria versus archaea in thesamples are based on results from QPCR analysis of thebacterial and archaeal communities in the reactors, asexplained in Parameswaran et al. (2009). The fraction of


Archaeal 16S rRNA genes compared to total bacterial16S rRNA genes in the suspension was 4.2%, and it wascomprised entirely of the hydrogenotrophic methanogenicorderMethanobacteriales. Figure 2b shows a smaller fractionof archaeal 16S rRNA genes compared to total bacterial 16SrRNA genes in the corresponding biofilm sample (0.7%),and this indicates that the methanogens were predominantlyin the suspension fraction when the MEC was operatedin batch mode. This is further supported by Q-PCR datashown in Parameswaran et al. (2009), where Methanobac-teriales was observed at 7.68� 105 and 1.55� 105 genecopies, in the suspension and biofilm fractions of the EtOH-only control reactor, respectively.

Based on pyrosequencing analysis of the V6 region in thebacterial 16S rRNA gene amplicons in the MEC samples,about 92% of the total bacterial 16S rRNA genes in thesuspension fraction belonged to the phylum Proteobacteria,of which 86% were Deltaproteobacteria. The predominantgenus among Deltaproteobacteria was Pelobacter (99% oftotal Deltaproteobacteria), followed by Geobacter (0.5%).Epsilonproteobacteria formed the remaining 6% of theProteobacteria phylum. We observed other phyla inrelatively small numbers: Firmicutes (2% of total pyrose-quencing tags) and Bacteroidetes (2% of total bacterialpyrosequencing tags).

The genus Pelobacter is associated with the syntrophicoxidation of ethanol via ‘‘interspecies hydrogen transfer’’in the presence of hydrogen-scavenging homo-acetogenicand methanogenic partners (Bryant et al., 1967; Seitz et al.,1988). Experiments by Richter et al. (2007) using Pelobactercarbinolicus in an ethanol-fed MFC confirmed that it cannotperform anode respiration. However, when P. carbinolicuswas mixed with G. sulfurreducens, an acetate/hydrogen-consuming ARB that cannot ferment ethanol, they observedcurrent generation, with about a 1:1 ratio of eachP. carbinolicus and G. sulfurreducens at the anode surface.Most other species of Pelobacter also cannot reduce solidFe(III), and their ability to utilize the anode as electronacceptor has not been proven (Schink, 1984; Seitz et al.,1988). As outlined in the Introduction Section, we expectedto observe a predominance of ethanol-fermenting bacteriain our system, and this observation of Pelobacter supportsour hypothesis. The very low fraction of Geobacter in thesuspension agrees with the fact that they are ARB and use theanode for their metabolism. It is important to note here thatmetabolic contribution from the microbial communityin the suspension was miniscule during the continuousoperation due to the washout of the slow growing anaerobicmicroorganisms.

The bacterial population in the biofilm of the EtOH-onlyMEC revealed that the Proteobacteria (72% of totalpyrosequencing tags) phylum was more diverse in thebiofilm than in the suspension, being distributed amongDeltaproteobacteria (38% of Proteobacteria), Epsilonproteo-bacteria (31%), and Betaproteobacteria (7%). Deltaproteo-bacteria was mostly represented by the genus Pelobacter(92% of total Deltaproteobacteria), probably performing

Figure 2. Distribution of bacterial and archaeal populations in the EtOH-only MEC. The ratio of archaea to bacteria is based on QPCR results (Parameswaran et al., 2009),

while the bacterial phyla 16S rRNA gene distributions are from pyrosequencing analysis. a: EtOh-only MEC—suspension fraction. b: EtOh-only MEC—biofilm fraction. [Color figure

can be seen in the online version of this article, available at www.interscience.wiley.com.]

ethanol fermentation. Epsilonproteobacteria comprised ofthe genus Wolinella (100% of total Epsilonproteobacteria),which are physiologically, morphologically, and structurallysimilar to the genus Campylobacter, except for their GCcontent. Both genera contain anaerobic pathogenic bacteriathat can utilize acetate, H2 gas, or formate as an electrondonor (Cord-Ruwisch et al., 1998; Logan, 1998; Wallaceet al., 1998). Additionally, the inability of Wolinella toferment carbohydrates and simple alcohols (Penner, 1988)and their ability to perform dissimilatory Fe(III) reduction(Lovley, 2006a) could possibly suggest that they are ARB inour system. Known ARB, such asGeobacter, were very low inthe biofilm samples as well.

Betaproteobacteriawere predominantly represented by thegenus Rhodocyclus (70% of total Betaproteobacteria), whichare phototrophic non-sulfur purple bacteria that are similarin function to the genus Rhodospirillum.

The phylum Firmicutes accounted for 13% of the totalbacteria in the biofilm, compared to the 2% in thesuspension fraction (Fig. 3), and Bacteroidetes accounted

for 8% of the total bacteria. Both the phyla were very diversein distribution, and their roles with reference to ethanolfermentation and anode respiration are not clear from ouranalysis.

Bacterial and Archaeal Populations in theEthanol-Plus-BES MEC

We did not detect any significant methanogenic activityin the ethanol-plus-BES MEC, and QPCR analysis (Para-meswaran et al., 2009) indicated that the ratio of archaeal16S rRNA genes to bacterial 16S rRNA genes was below 0.3%in the suspension and biofilm fractions.

Figure 3a shows that more than 98% of the bacterial 16SrRNA genes pyrosequencing tags in the suspension fractionwere comprised of the Proteobacteria phylum, all belongingto the Deltaproteobacteria and Pelobacter genus; the latterwas probably performing ethanol fermentation. Other phylawere diverse, but formed very small fractions of the totalbacteria. The bacterial 16S rRNA genes pyrosequencing tags



Figure 3. Distributions of bacterial 16S rRNA genes in the biofilm and suspen-

sion fractions of the EtOH-plus-BES MEC. The much larger diversity in the biofilm

sample is evident. a: EtOh-plus-BES—suspension. b: EtOh-plus-BES—biofilm. [Color

figure can be seen in the online version of this article, available at www.interscience.

wiley.com.]

in the biofilm fraction of the ethanol-plus-BES MEC, shownin Figure 3b, were predominantly Proteobacteria (84%of total bacteria), distributed among Deltaproteobacteria(73%), Betaproteobacteria (4%), Epsilonproteobacteria (3%),Alphaproteobacteria (2%), and Gammproteobacteria (1%).While Pelobacter was the predominant genus in theDeltaproteobacteria (43% of total Deltaproteobacteria), theremaining was composed of 9% from the genus Desulfovi-brio, and 21% from the Geobacter genus. Compared to theethanol-only biofilm (previous sub-section), the fraction ofGeobacter, which is a known ARB, was much higher. Thisexplains the true ecological advantage for ARB to be in thebiofilm anode, which is to carry out anode respiration.However, the role of other groups of bacteria in the biofilmanode toward biofilm initiation and electron transfer to thesolid anode remains unknown, apart from the syntrophicinteractions discussed in Parameswaran et al. (2009).Betaproteobacteria were predominantly comprised of thegenus Rhodocyclus (49% of total Betaproteobacteria) andDelftia (23%). The role of Delftia with respect to the biofilmanode is not clear from our analysis.

Alphaproteobacteria were comprised entirely of the genusRhodopseudomonas. Recently, Xing et al. (2008) reportedelectricity generation by Rhodopseudomonas palustris DX 1with a variety of substrates. Epsilonproteobacteria had adistribution similar to the ethanol-only biofilm, with a 100%


predominance of the genusWolinella. Gammaproteobacteriawere primarily made up of the genus Pseudomonas (76%of the total Gammaproteobacteria), a genus of known ARBthat performs extracellular electron transfer through solubleshuttles (Pham et al., 2008; Rabaey et al., 2004).

Other phyla present in the biofilm fraction includedActinobacteria (2%), Bacteroidetes (1%), and Firmicutes(13%). The fraction of Firmicutes in the anode biofilm wasthe same in the ethanol-plus-BES and the ethanol-onlybiofilms. However, the compositions of the Firmicutes weredifferent in each case and may have important implicationsfor community structure analysis, as we discuss in thesection below.

Distribution of Firmicutes in Ethanol-Onlyand Ethanol-Plus-BES MECs

Firmicutes 16S rRNA genes were mainly confined to thebiofilm in both reactors, accounting for 13% of the totalunique tags (Figs. 2b and 3b). Figure 4a shows that thephylum Firmicutes was more diverse in the biofilm ofethanol-only MEC, with no clear predominance of aparticular group of Firmicutes (Incertae Sedis XII accountingfor 34% of the total Firmicutes, formed the largest fraction).This was in contrast to the biofilm of EtOH-plus-BES MEC,where Firmicutes predominantly were comprised of thefamily Eubacteriaceae (62% of the total Firmicutes), and theAcetobacterium genus, a known homo-acetogen, accountedfor 100% of the Eubacteriaceae family. Correspondingly,acetate accumulated to greater than stoichiometric levels inthe EtOH-plus-BES experiments, but H2 did not accumulate(Parameswaran et al., 2009). All these observations suggestthat interspecies H2 transfer was probably happeningfrom ethanol-fermenting Pelobacter to acetate-producingAcetobacterium in the biofilm anode with methanogenesissuppressed.

Identification of Homo-Acetogenic Bacteria—FTHFSGene-Targeted Method

Using FTHFS-targeted primers, we PCR amplified frag-ments of the expected size (1,102 bp) in some of our samplesand in the positive control (data not shown). PCRamplicons were present in the suspension and biofilm ofthe EtOH-plus-BES MEC, but were not detected in thebiofilm and suspension fractions of the EtOH-only samples,sequence analysis confirmed that the amplicons were FTHFSgene sequences present in acetogens.

Table II shows a comparative analysis of translatedproteins from the DNA sequences obtained from a clonelibrary analysis of the EtOH-plus-BES biofilm fraction withavailable protein sequences at NCBI; 68% of the total clonesresembled the FTHFS gene isolated from Acetobacteriumcarbinolicum subsp. kysingense, with 96% identity to theprotein sequence. A. carbinolicum is known to act as a H2

scavenger during syntrophic oxidation of fermentable

Figure 4. Distribution of the phylum Firmicutes 16S rRNA genes in the biofilm fractions of the ethanol-only and ethanol-plus-BES MECs. Predominance of the homo-

acetogenic family, Eubacteriaceae, only with methanogenesis inhibited is evident from this figure. Various groups of Insertae sedis presented in the figure refer to the unclassified

fraction of Firmicutes. a: EtOh-only MEC—biofilm. b: EtOh-plus-BES MEC—biofilm. [Color figure can be seen in the online version of this article, available at www.interscience.

wiley.com.]

substrates, such as ethanol (Eichler and Schink, 1985;Schink, 1983), to acetate. Homo-acetogenic bacteria fromenvironmental samples are generally phylogeneticallydiverse (Drake et al., 2002), but our clone library resultsindicated a predominance of members from the cluster XVunder Clostridiales, to which Acetobacterium species belong.Some sequences that were not previously amplified(Leaphart and Lowell, 2001) by the same degenerate primersused in this study, such as Moorella thermoacetica, weredetected in this analysis.

Finding homo-acetogens when methanogenesis was sup-pressed is consistent with our earlier results (Parameswaranet al., 2009), which showed initial build up of acetate andno accumulation of H2 in the head space. Thus, homo-acetogenic bacteria channeled the electrons from H2 toacetate, which was then utilized by the ARB to produceelectricity. A previous study (Freguia et al., 2008) had ruled

Table II. Summary of key homo-acetogenic members identified by clone libr

Protein sequence with closest match Organism with closest

Formyltetrahydrofolate synthetase Acetobacterium carbinolicum su

(Genbank accession no: DQ

Formyltetrahydrofolate synthetase Alkaliphilus oremlandii OHILA

Formyltetrahydrofolate synthetase Moorella thermoacetica (ATCC

Formyltetrahydrofolate ligase (synthetase) Acetobacterium carbinolicum (A

Formyltetrahydrofolate synthetase Desulfitobacterium hafniense (A

Hypothetical protein CLOBAR 00933 Clostridium bartletii (DSM 167

Formyltetrahydrofolate synthetase Clostridium acidurici (Genbank

aValues obtained from BLASTX search in the NCBI database.bTotal number of clones analyzed¼ 49.

out a role of homo-acetogenic bacteria based solely onshort-term kinetic tests (90min) involving H2 spargingof the anode. However, 90min is an insufficient time forhomo-acetogens to accumulate detectable amounts of acetate,based on the slow growth kinetics of homo-acetogens(Leadbetter et al., 1999; Zeikus et al., 1985). On the otherhand, our new analysis of earlier microbial ecology studies atthe anode of an MFC fed with a fermentable substrate,glucose, revealed that homo-acetogenic bacteria, such asEubacteriaceae (Acetobacterium) and Spirochaetes, consti-tuted 22% of the total clones analyzed (Choo et al., 2006),which supports our finding of homo-acetogenic bacteria as ahydrogen scavenger. The finding of Treponema denticola, ahomo-acetogenic Spirochaete, and the hydrogenotrophicmethanogenic order, Methanobacteriales, by Ishii et al.(2008) in an MFC fed with cellulose adds new insight forpossible co-existence of two types of H2 scavengers in the

ary analysis targeting the FTHFS gene.

matching sequence

Percentage identity

with proteina (%)

Percentage of

total clonesb (%)

bsp. kysingense

152901)

96 68

s (ATCC no: NC 009922) 79 16

no: 39073) 80 6

TCC no: DQ152906) 95 4

TCC no: NC 007907) 77 2

95) 90 2

Accession no: P13419) 78 2



same system. However, the reasons for the simultaneous co-existence of the H2 scavengers were not discussed andremain to be understood. Both studies mentioned above didnot analyze the possible role of these bacterial communitieswith respect to the biofilm anode. Our study could put theirfindings in perspective and provide further opportunity forbeneficial ecological management at the biofilm anode.

Population Structure and Role of the Anode MicrobialCommunities

Table III shows the distribution of the major bacterial generain the various samples that were analyzed with pyrosequen-cing during this study. Ethanol fermenters, represented bythe genus Pelobacter, formed the predominant group of allthe bacteria. ARB formed the next major functional group,and they exhibited a diverse distribution. Among knowngenera of ARB, we detected Geobacter, Pseudomonas, andRhodopseudomonas; other ARB might have been present andwould have increased the fraction of ARBs in our reactors.The predominance of the genusWolinella only in the biofilmfraction of the EtOH-only MEC, their inability to consumesugars and simple alcohols, their ability to use acetate andH2

as electron donors, and their ability to respire on Fe(III)solid suggest that they could be possibly ARBs. The presenceof the H2-oxidizing methanogens Methanobacteriales inthe ethanol-only MEC agrees with the general trends of ourproposed biomass distribution among the fermenters,methanogens, and ARB (Table I). The presence of homo-acetogens, represented by Acetobacterium, in the ethanol-plus-BES MEC agrees with the general trends of ourproposed biomass distribution among the fermenters,homo-acetogens, and ARB when methanogenesis is sup-pressed (Table I). These findings indicate that the biofilmanode was structured according to the electron-flow

Table III. Distribution of the predominant bacterial genera in the suspension

Fraction of total un

EtOH-only MEC

Genus/order Suspensiona Biofilmb

Pelobacter 88 35

Geobacter 0.5 1.2

Wolinella 5.8 31

Rhodocyclus 0.2 5.1

Rhodopseudomonas ND ND

Pseudomonas ND 1.0

Methanobacteriales 4.2 0.7

Acetobacterium ND ND

Ruminococcaceae 0.8 2.4

Total 95.3e 77.1e

ND, Not detected.aTotal tags¼ 21,761.bTotal tags¼ 15,726.cTotal tags¼ 18,365.dTotal tags¼ 38,432.eSum of all fractions in this table.


distribution in Figure 1 and Table I. Encouraging homo-acetogenesis may be a novel and promising means tominimize the methane sink, select for more efficient ARB,and increase the H2 yield in MECs. Higher tolerance ofhomo-acetogens to oxygen exposure (Kusel et al., 2001) andtheir ability to outcompete hydrogenotrophic methanogensat lower temperatures and low pH (Phelps and Zeikus, 1984)present directions to find practical ways of promotinghomo-acetogenesis in the biofilm anode.

Conclusions

We proved the hypothesis that homo-acetogens weresignificant in the microbial community of the biofilmanode when methanogenesis was not active. We obtaineddirect evidence for the presence of homo-acetogens throughtwo independent molecular microbial ecology techniques:amplification and sequencing of the FTHFS gene conservedin homo-acetogens and pyrosequencing analysis whichtargets the V6 region of the bacterial 16S rRNA genes presentin the sample. Both results indicate the predominance of thehomo-acetogenic genus, Acetobacterium, and also supportthe three-way syntrophy involving H2-oxidizing homo-acetogens when methanogenesis was inhibited.

We also demonstrated that the distribution of thebacterial populations in the anode was in agreement withour proposed distribution of biomass in the electron flow(Table I). Ethanol-fermenting bacteria (largely Pelobacter)dominated the tags in the pyrosequencing analysis, followedby a diverse community of ARB, and then followed by eitherthe H2-oxidizing methanogens or homo-acetogens. Evalu-ating methods to enrich for homo-acetogens so that they canbe sustained at the biofilm anode has promise as an effectivemeans to compete with undesirable sinks, such as

and biofilm samples from the two reactors.

ique tags (%)

EtOH-plus-BES MEC

Suspensionc Biofilmd Probable role

96 44 Ethanol fermentation

1.2 21 Anode respiration

0.3 3 Unknown

ND 1.8 Unknown

ND 1.7 Anode respiration

ND 0.8 Anode respiration

0.2 0.01 Methanogenesis

0.4 8.1 Homo-acetogenesis

0.2 0.7 Fermentation

98.3e 82e

methanogenesis, and improve overall electron recoveriesfrom fermentable substrates.

References

Bainotti AE, Nishio N. 2000. Growth kinetics of Acetobacterium sp. on

methanol-formate in continuous culture. J Appl Microbiol 88:191–201.

Bond DR, Lovley DR. 2003. Electricity production by Geobacter sulfurre-

ducens attached to electrodes. Appl Environ Microbiol 69(3):1548–

1555.

Bond DR, Holmes DE, Tender LM, Lovley DR. 2002. Electrode-reducing

microorganisms that harvest energy from marine sediments. Science

295:483–485.

Bryant MP, Wolin EA, Wolin MJ, Wolfe RS. 1967. Methanobacillus

omelianskii, a symbiotic association of two species of bacteria. Arch

Microbiol 59:20–31.

Choo YF, Lee J, Chang IS, Kim BH. 2006. Bacterial communities in

microbial fuel cells enriched with high concentrations of glucose

and glutamate. J Microbiol Biotechnol 16(9):1481–1484.

Combet-Blanc Y, Kalamba KK, Kergoat PY. 1995. Effect of pH on Bacillus

thermoamylovorans growth and glucose fermentation. Appl Environ

Microbiol 61(2):656–659.

Cord-Ruwisch R, Lovley DR, Schink B. 1998. Growth of Geobacter sulfur-

reducens with acetate in syntrophic cooperation with hydrogen-oxidiz-

ing anaerobic partners. Appl Environ Microbiol 64(6):2232–2236.

Drake HL, Kusel K, Matthies C. 2002. Ecological consequences of the

phylogenetic and physiological diversities of acetogens. Antonie Von

Leeuwenhoek 81:203–213.

Eichler B, Schink B. 1985. Fermentation of primary alcohols and diols and

pure culture of syntrophically alcohol-oxidizing anaerobes. Arch

Microbiol 143(1):60–66.

Esteve-Nunez A, Rothermich M, Sharma M, Lovley DR. 2005. Growth of

Geobacter sulfurreducens under nutrient limiting conditions in con-

tinuous culture. Environ Microbiol 7(5):641–648.

Freguia S, Rabaey K, Yuan Z, Keller J. 2007. Electron and carbon balances in

microbial fuel cells reveal temporary bacterial storage behavior during

electricity generation. Environ Sci Technol 41:2915–2921.

Freguia S, Rabaey K, Yuan Z, Keller J. 2008. Syntrophic processes drive the

conversion of glucose in microbial fuel cells. Environ Sci Technol

42(21):7937–7943.

Holmes DE, Bond DR, Neil RA, Reimers CE, Tender LR, Lovley DR. 2004.

Microbial communities associated with electrodes harvesting electricity

from a variety of aquatic sediments. Microbial Ecol 48:178–190.

Ishii S, Shimoyama T, Hotta Y, Watanabe K. 2008. Characterization of a

filamentous biofilm community established in a cellulose-fed microbial

fuel cell. BMC Microbiol 8(6):1–12.

Jong BC, Kim BH, Chang IS, Liew PWY, Choo YF, Kang GS. 2006.

Enrichment, performance, and microbial diversity of a thermophilic

mediatorless microbial fuel cell. Environ Sci Technol 40(20):6449–

6454.

Jung S, Regan M. 2007. Comparison of anode microbial communities and

performance in microbial fuel cells with different electron donors. Appl

Microbiol Biotechnol 77:393–402.

Kim JR, Jung SH, Regan JM, Logan BE. 2007. Electricity generation and

microbial community analysis of alcohol powered microbial fuel cells.

Biores Technol 98:2568–2577.

Kusel K, Karnholz A, Trinkwalter T, devereux R, Acker G, Drake HL. 2001.

Physiological ecology of Clostridium glycolicum RD-1, an aerotolerant

acetogen isolated from sea grass roots. Appl Environ Microbiol

67:4734–4741.

Lawrence AW, McCarty PL. 1969. Kinetics of methane fermentation in

anaerobic treatment. J Water Pollut Control Fed 41(2P2):R1–R17.

Leadbetter JR, Schmidt TM, Graber JR, Breznak JA. 1999. Acetogenesis

from H2 plus CO2 by Spirochetes from termite guts. Science 283(5402):

686–689.

Leaphart AB, Lovell CR. 2001. Recovery and analysis of formyltetrahydro-

folate synthetase gene sequences from natural populations of aceto-

genic bacteria. Appl Environ Biotechnol 67(3):1392–1395.

Lee HS, Rittmann BE. 2008. Thermodynamic evaluation on H2 produc-

tion in glucose fermentation. Environ Sci Technol 42(7):2401–

2407.

Lee JY, Phung NT, Chang IS, Kim BH, Sung HC. 2003. Use of acetate for

enrichment of electrochemically active microorganisms and their 16S

rDNA analyses. FEMS Microbiol 223:185–191.

Lee HS, Parameswaran P, Torres CI, Marcus AK, Rittmann BE. 2008.

Evaluation of energy conversion efficiencies in microbial fuel cells

(MFC’s) utilizing fermentable and non-fermentable substrates. Water

Res 42(6–7):1501–1510.

Lee HS, Torres CI, Parameswaran P, Rittmann BE. 2009. Fate of H2 in a

single-chamber microbial electrolysis cell using a metal-catalyst-free

cathode. Environ Sci Technol 10.1021/es900204j.

Liu H, Cheng S, Logan BE. 2005. Production of electricity from acetate or

butyrate using a single chambered microbial fuel cell. Environ Sci

Technol 39(2):658–662.

Logan BE. 1998. A review of chlorate and perchlorate reducing micro-

organisms. Bioremediation 2(2):69–79.

Lovley DR. 2006a. Dissimilatory Fe-III and Mn-IV reducing prokaryotes.

Prokaryotes 2:635–658.

Lovley DR. 2006b. Bug juice: Harvesting electricity with microorganisms.

Nat Rev Microbiol 4:497–508.

Min B, Logan BE. 2004. Continuous electricity generation from domestic

wastewater and organic substrates in a flat plate microbial fuel cell.

Environ Sci Technol 38(21):5809–5814.

Nath K, Das D. 2004. Improvement of fermentative hydrogen produc-

tion: Various approaches. Appl Microbiol Biotechnol 65(5):520–

529.

Parameswaran P, Torres CI, Lee HS, Krajmalnik-Brown R, Rittmann BE.

2009. Syntrophic interactions among anode respiring bacteria (ARB)

and non-ARB in a biofilm anode: Electron balances. Biotech Bioeng

103(3):513–523.

Penner JL. 1988. The genus Campylobacter: A decade of progress. Clin

Microbiol Rev 1(2):157–172.

Pham HT, Boon N, Aeltermann P, Clauwaert P, Schamphelaire LD,

Oostveldt PV, Verbeken K, Rabaey K, Verstraete W. 2008. High shear

enrichment improves the performance of the anodophilic microbial

consortium in a microbial fuel cell. Microbial Biotechnol 1(6):487–

496.

Phelps TJ, Zeikus JG. 1984. Influence of pH on terminal carbon metabolism

in anoxic sediments from a mildly acidic lake. Appl Environ Microbiol

48:1088–1095.

Phung NT, Lee J, Kang KH, Chang IS, Gadd GM, Kim BH. 2004. Analysis of

microbial diversity in oligotrophic microbial fuel cells using 16S rDNA

sequences. FEMS Microbiol 233:77–82.

Rabaey K, Lissens G, Siciliano SD, Verstraete W. 2003. A microbial fuel cell

capable of converting glucose to electricity at high rate and efficiency.

Biotechnol Lett 25(18):1531–1535.

Rabaey K, Boon N, Siciliano SD, Verhaege M, Verstraete W. 2004. Biofuel

cells select for microbial consortia that self-mediate electron transfer.

Appl Environ Microbiol 70(9):5373–5382.

Ren Z, Ward TE, Regan JM. 2007. Electricity production from cellulose in a

microbial fuel cell using a well defined binary culture. Environ Sci

Technol 41:4781–4786.

Richter H, Lanthier M, Nevin KP, Lovley DR. 2007. Lack of

electricity production by Pelobacter carbinolicus indicates that the

capacity for Fe(III) oxide reduction does not necessarily confer electron

transfer ability to fuel cell anodes. Appl Environ Microbiol 73:5347–

5353.

Rittmann BE, McCarty PL. 2001. Environmental biotechnology: Principles

and applications. NY: McGraw-Hill (Chapter 13).

Rozendal RA, Jeremiasse AW, Hamelers VMH, Buisman CJN. 2008.

Hydrogen production with a microbial biocathode. Environ Sci Tech-

nol 42(2):629–634.



Schink B. 1983. Fermentation of 2,3-butanediol by Pelobacter carbinolicus

sp. nov., and Pelobacter propionicus sp. nov., and evidence for propio-

nate formation from C2 compounds. Arch Microbiol 137:33–41.

Schink B. 1997. Energetic of syntrophic cooperation in methanogenic

degradation. Microbiol Mol Biol Rev 61(2):262–280.

Seitz HJ, Schink B, Conrad R. 1988. Thermodynamics of hydrogen meta-

bolism in methanogenic co-cultures degrading ethanol or lactate.

FEMS Microbiol Lett 55:119–124.

Sogin ML, Morrison HG, Huber JA, Welch DM, Huse SM, Neal PR,

Arrieta JM, Herndl JH. 2006. Microbial diversity in the deep sea

and the unexplored ‘‘rare biosphere’’. Proc Natl Acad Sci USA

103:12115–12120.

Stackebrandt E, Kramer I, Swiderski J, Hippe H. 1999. Phylogenetic basis

for a taxonomic dissection of the genus Clostridium. FEMS Immun

Med Microbiol 24:253–258.

Temudo MF, Kleerebezem R, Van Looschdrecht M. 2007. Influence of

the pH on (open) mixed culture fermentation of glucose: A chemostat

study. Biotechnol Bioeng 98(1):69–79.

Thauer RK, Jungerman K, Henninger H, Wenning J, Decker K. 1968. The

energy metabolism of Clostridium kluyveri. Eur J Biochem 4:173–180.


Torres CI, Marcus AK, Rittmann BE. 2007. Kinetics of consumption of

fermentation products by anode-respiring bacteria. Appl Microbiol

Biotechnol 77:689–697.

WallaceW, Beshear S,WilliamsD, Hospadar S, OwensM. 1998. Perchlorate

reduction by a mixed film culture in an up-flow fixed bed reactor. J Ind

Microbiol Biotechnol 20:126–131.

Wang Q, Garrity GM, Tiedje JM, Cole JR. 2007. Naıve Bayesian classifier for

rapid assignment of rRNA sequences into the new bacterial taxonomy.

Appl Environ Microbiol 73:5261–5267.

Xing D, Zuo Y, Cheng S, Regan JM, Logan BE. 2008. Electricity generation

by Rhodopseudomonas palustris DX-1. Environ Sci Technol 42(11):

4146–4151.

Zehnder A. 1988. Biology of anaerobic microorganisms. NY: John Wiley

and Sons.

Zeikus JG, Kerby R, Krzycki JA. 1985. Single-carbon chemistry of

acetogenic and methanogenic bacteria. Science 227(4691):1167–1173.

Zhang H, DiBaise JK, Zuccolo A, Kudrna D, Braidotti M, Yu Y,

Parameswaran P, Crowell MD, Wing R, Rittmann BE, Krajmalnik

Brown R. 2009. Human gut microbiota in obesity and after gastric-

bypass. Proc Natl Acad Sci USA 106:2365–2370.

microbial community structure in a biofilm anode fed with a fermentable substrate: the significance...

Documents