Bioresource Technology 129 (2013) 156–163

Contents lists available at SciVerse ScienceDirect

Bioresource Technology

journal homepage: www.elsevier .com/locate /bior tech

A versatile and robust aerotolerant microbial community capable of cellulosicethanol production

Patrick Ronan a, C. William Yeung a, John Schellenberg b, Richard Sparling b, Gideon M. Wolfaardt a,c,Martina Hausner a,⇑a Department of Chemistry and Biology, Ryerson University, Toronto, Canadab Department of Microbiology, University of Manitoba, Winnipeg, Canadac Stellenbosch Institute for Advanced Study, Stellenbosch, South Africa

h i g h l i g h t s

" A cellulolytic aerotolerant microbial community was enriched from compost." Cellulolytic activity was observed in non-reduced as well as pre-reduced media." Ethanol and acetate were major fermentation products." Cellulolytic activity continued when sterile wastewater was provided as nutrient." The culture consisted of both facultative anaerobic and anaerobic members.

a r t i c l e i n f o

Article history:Received 21 August 2012Received in revised form 30 October 2012Accepted 31 October 2012Available online 16 November 2012

Keywords:Aerotolerant microbial consortiumBioethanolCellulose degradation

0960-8524/$ - see front matter � 2012 Elsevier Ltd. Ahttp://dx.doi.org/10.1016/j.biortech.2012.10.164

⇑ Corresponding author. Address: Department of ChUniversity, 350 Victoria Street, Toronto, ON, Canada5000x6553; fax: +1 416 979 5044.

E-mail address: martina.hausner@ryerson.ca (M. H

a b s t r a c t

The use of microbial communities in the conversion of cellulosic materials to bio-ethanol has the poten-tial to improve the economic competitiveness of this biofuel and subsequently lessen our dependency onfossil fuel-based energy sources. Interactions between functionally different microbial groups within acommunity can expand habitat range, including the creation of anaerobic microenvironments. Currently,research focussing on exploring the nature of the interactions occurring during cellulose degradation andethanol production within mixed microbial communities has been limited. The aim of this study was toenrich and characterize a cellulolytic bacterial community, and determine if ethanol is a major solubleend-product. Cellulolytic activity by the community was observed in both non-reduced and pre-reducedmedia, with ethanol and acetate being major fermentation products. Similar results were obtained whensterile wastewater extract was provided as nutrient. Several community members showed high similarityto Clostridium species with overlapping metabolic capabilities, suggesting clostridial functionalredundancy.

� 2012 Elsevier Ltd. All rights reserved.

1. Introduction

Fossil fuels are unsustainable, finite resources and their produc-tion and consumption has caused widespread environmental im-pacts and led to rapid climate change. With such a dramaticincrease in the global energy demand over the last century, thecomplete depletion of fossil fuel reserves is predicted to occurwithin 50 years (Rodolfi et al., 2009). Consequently, there is anundeniable need for better renewable energy sources.

Bio-ethanol has garnered significant attention as a potentiallong-term replacement for fossil fuels. Currently, bio-ethanol is

ll rights reserved.

emistry and Biology, RyersonM5B 2K3. Tel.: +1 416 979

ausner).

produced mainly from sugars derived from food crops such as cornand sugarcane. Because the use of edible crops for fuel productionputs a burden on agricultural lands and contributes to rising foodprices (Inderwildi and King, 2009), cellulose is an attractive alter-native feedstock for bio-ethanol production, due to its abundanceand renewability.

To date, much attention has been paid to utilizing pure culturesof anaerobic cellulose-degrading bacteria such as Clostridium ther-mocellum to overcome the challenges related to cellulose recalci-trance (Lynd et al., 2002; Xu et al., 2010). This organism iscapable of simultaneously hydrolyzing cellulose and fermentingthe resulting sugars to produce ethanol in a process known asconsolidated bioprocessing (CBP) (Lynd et al., 2005). Pure culturesystems, however, often persist within a narrow range of growthconditions (pH, temperature, oxygen content) and their activity

P. Ronan et al. / Bioresource Technology 129 (2013) 156–163 157

may be altered by contamination from other microorganisms(Brenner et al., 2008).

Naturally-occurring microbial communities conversely, areadapted to exploit a wide variety of nutrient and energy sources,often as specialized consortia that utilize specific resources(Brenner et al., 2008). Cellulose utilization, for example, iscommonly accomplished by microbial communities consisting ofphysiologically cooperative species. Consortia enriched from thesenatural communities can exhibit cellulolytic activity greater thanthe sum of their parts, and have been shown to perform morecomplex activities and persist within a wider range of environmen-tal parameters than the separated members (Brenner et al., 2008;Kato et al., 2004).

Previous studies have aimed to characterize cellulose-degradingcommunities enriched from natural sources such as manure, papermill waste, and various types of compost (Izquierdo et al., 2010;Okeke and Lu, 2011; O’Sullivan et al., 2005; Sizova et al., 2011;Zyabreva et al., 2001). Most often however, the communitiesdescribed in such studies are cultivated under strict anaerobic con-ditions using pre-reduced media.

In order to improve the economic competitiveness of biofuels,efforts must focus on simplifying and streamlining their produc-tion, and reducing process-related costs. Anaerobic cellulosedegraders such as C. thermocellum require strict anaerobic condi-tions, which necessitates the inclusion of reducing agents in theculture medium. On an industrial scale, this may represent asignificant extraneous cost (Maddipati et al., 2011). The use ofcooperative consortia enriched from natural cellulose-degradingcommunities that are aerotolerant and do not require pre-reducedmedia may help to overcome this limitation, as the diversity withinthem allows for efficient cellulolysis coupled with increased toler-ance to environmental fluctuations (Kato et al., 2008; Okeke andLu, 2011). Research focused on such aerotolerant communitieshas been somewhat limited with a relatively small number of stud-ies describing cellulolytic consortia cultivated in non-reducedmedium (Kato et al., 2004; Wang et al., 2011; Wongwilaiwalinet al., 2010).

Yeast extract (YE) is often a stimulatory constituent of mediaused to cultivate cellulose-degrading bacteria (Haruta et al.,2002; Kato et al., 2004; Miyazaki et al., 2008; Wongwilaiwalinet al., 2010), and is not a cost-effective option at industrial scale(Maddipati et al., 2011). The ability to replace YE with a low-costand easily attainable waste-derived complex nutrient source, with-out compromising the cellulolytic activity of the culture, wouldhelp to circumvent such financial hurdle. It is also possible thatcross-feeding of essential nutrients between community membersmay eliminate the requirement for such supplementation.

In addition to the applied relevance related to the search for im-proved cellulose hydrolysis, such research provides an opportunityto study microbial interactions when challenged with a recalcitrantsubstrate. The objectives of this study were therefore to: (i) enrichan aerotolerant cellulolytic consortium, (ii) obtain a time-resolvedprofile of selected soluble end-products in non-reduced (initiallyaerobic) and pre-reduced (anaerobic) media, as well as YE-freenon-reduced media supplemented with either compost tea orwastewater from a municipal wastewater treatment plant, and(iii) obtain a time-resolved analysis of bacterial diversity withinthe consortium under these conditions.

2. Methods

2.1. Inoculum and enrichment conditions

Approximately 15 g of material from a household composterwas added to 60 mL of non-reduced RM medium (Ozkan et al.,

2001) containing urea (2 g/L), KH2PO4 (2 g/L), K2HPO4 (3 g/L), yeastextract (2 g/L), and the oxygen indicator resazurin (0.002 g/L).Whatman Grade 1 Qualitative Filter Paper (Whatman, Piscataway,NJ) was used as the cellulosic substrate and was added prior toautoclave sterilization. A filter-sterilized trace minerals solutioncontaining MgCL2�6H2O (20 g/L), CaCl2�2H2O (5 g/L), and FeSO4-

�7H2O (0.25 g/L) was added after autoclave sterilization at a ratioof 1:100 (v/v). Incubation was done at 60 �C under static, initiallyaerobic conditions. Flasks were tightly capped to eliminate diffu-sion of oxygen into the culture. Over time, oxygen present in theheadspace and medium at the beginning of the incubation periodwas depleted as a result of microbial activity, and was notreplenished.

The resulting culture was sequentially transferred to fresh med-ium at a 1:5 ratio (v/v). Consistent and predictable cellulose-degrading activity (disappearance of filter paper) was observedthrough six enrichments. The sixth enrichment culture was usedin subsequent bacterial diversity and end-product analyses.

2.2. Variation in growth medium composition and incubationconditions

Time-resolved analysis of selected end-products and microbialcommunity structure was carried out in four different media. Inaddition to the non-reduced RM medium, experiments were alsoconducted with pre-reduced RM medium by the addition ofL-cysteine hydrochloride monohydrate (Sigma–Aldrich, Oakville,ON) at a final concentration of 1 g/L, YE-free non-reduced composttea RM medium, and YE-free non-reduced wastewater RMmedium. Sealed 50 mL serum vials (Sigma–Aldrich, Oakville, ON)were used for anaerobic culturing, with the headspace in eachevacuated by vacuum and flushed with grade 4.8 nitrogen (BOCGasses, Mississauga, ON) for 30 s intervals until the resazurin hadchanged from pink to colourless.

YE-free non-reduced compost tea RM medium was used to as-sess the feasibility of replacing yeast extract with inexpensivesources of nutrients and growth factors. Compost tea was firstprepared by combining compost and water at �1:5 (w/v) and mix-ing for 30 min using a magnetic stirring-rod and stirrer at roomtemperature. After a 2.5 h settling period, the supernatant wasgently pumped into a clean bottle using a Masterflex� ConsoleDrive pump (Cole-Parmer, Montreal, QC). When preparing non-re-duced compost tea RM medium, this compost tea was used as thesolvent into which the various media components (excluding YE)were dissolved, prior to autoclave sterilization.

YE-free non-reduced wastewater RM medium was prepared in asimilar fashion. Wastewater collected from the aeration tanks ofthe secondary treatment process were obtained from the HumberWastewater Treatment Plant in Toronto, Canada. After mixing for30 min at room temperature using a magnetic stirring-rod andstirrer, flocs and other solid particles were allowed to settle andthe supernatant was gently pumped to a separate container andstored at 4 �C in the dark. When preparing non-reduced wastewa-ter RM medium, the various RM medium components (excludingYE) were dissolved into this wastewater supernatant and auto-clave-sterilized.

2.3. Experimental setup

Three millilitres of the enriched cellulolytic consortium grownin non-reduced RM medium was transferred to 27 mL aliquots ofthe respective media in 50 mL sealed serum vials and cultivatedat 60 �C. Each 30 mL culture contained a small 80 mg (±3 mg) pieceof Whatman Grade 1 Qualitative Filter Paper (Whatman, Piscata-way, NJ), generating a cellulose concentration of 2.66 g/L. Theconsortium was also inoculated into cellulose-free controls of each

158 P. Ronan et al. / Bioresource Technology 129 (2013) 156–163

medium which contained no added filter paper. All test and controlcultures were setup in triplicate.

2.4. End-product analysis and calculations

Samples (1.5 mL) were collected from all cultures on days 1, 2,3, 4, 6, 9, and 14 using sterilized syringes and needles, placed in1.5 mL cryovials (Cole-Parmer, Montreal, QC), and immediatelyacidified to �pH 2 using 1.5 mM HCl. They were stored at 4 �C untilchemical analysis. The cultures inoculated into the wastewater-supplemented media had one additional sample taken on day 5.Additional sub-samples were used to measure pH using a bench-top meter (Cole-Parmer, Montreal, QC).

Concentrations of ethanol, acetate, lactate, formate, butyrate,cellobiose, and glucose in the samples were analyzed by HPLC(Model #1515 pump, #2707 autosampler and #2414 refractiveindex detector, Waters, Milford, MA) using a column for organicacid quantitation (Amicon # HPX-87H, Bio-Rad Laboratories, Miss-issauga, ON). Additional 1.5 mL samples were collected from onereplicate of each culture on days 4 and 14 and stored at �20 �Cfor subsequent bacterial community analysis and comparison withthe initial compost inoculum and an intermediate enrichment.

The ethanol:acetate concentration ratio in each of the four med-ia treatments, taken at the ethanol peak, is reported. The combinedethanol and acetate yield (symbolized by [ethanol + acetate]) atthis point is also reported and provided as a percentage of thetheoretical maximum, which was 32 mM. The theoretical maxi-mum was calculated by dividing the initial cellulose concentrationin each test culture (2.66 g/L) by the molar mass of glucose equiv-alents in cellulose (162 g/mol), which results in a total of 16 mMglucose equivalents. Given that one mole of glucose equivalentscan give rise to two moles of ethanol or acetate, a maximum [eth-anol + acetate] of 32 mM can be produced from 2.66 g/L cellulose.

2.5. DNA extraction and PCR amplification

Samples were washed three times by centrifuging at 6000g for3 min, with the pellets resuspended in 1 mL PBS at pH 7.4. DNAwas extracted using the ZR Soil Microbe DNA MiniPrep™ Kit(Cedarlane, Burlington, ON) following the manufacturer’s instruc-tions. Purified DNA was stored at �20 �C.

Bacteria-specific primers were used to amplify a 418 bp frag-ment of the 16S rRNA gene. The forward primer was U341F-GC(Muyzer et al., 1993) with a GC clamp affixed to the 50end(Sheffield et al., 1989). The reverse primer was U758R (Lee et al.,1993). PCR was carried out with minor modifications (supplemen-tary information) as described previously (Yeung et al., 2010). SixPCR amplifications for each sample were quantified on agarosegel and cleaned and combined using illustra GFX™ PCR DNA andGel Band Purification Kit (GE Healthcare, Piscataway, NJ) as de-scribed in Yeung et al. (2011).

2.6. Denaturing gradient gel electrophoresis (DGGE)

Approximately 300–500 ng of DNA from each sample was runon an 8% polyacrylamide (BioRad Laboratories, Mississauga, ON)DGGE gel with a 30–70% denaturant gradient, as described previ-ously (Yeung et al., 2010, 2011). Bands were visualized after stain-ing with 0.01% SYBR Gold (Invitrogen, Burlington, ON) in 1x TAEbuffer. Images were captured using a Gel Logic 1500 Imaging Sys-tem (Kodak, Rochester, NY). Bands of interest were excised andeluted in 25 ll sterile ddH2O at 4 �C for 5 days. One microlitre ofthe eluted DNA was reamplified using the same primers withoutthe added GC clamp as described previously (Yeung et al., 2010,2011). DGGE fingerprints were analyzed and dendrogram profiles(based on bands possessing peak intensity of at least 5% of the

most intense band in the gel) depicting the similarity indexes(SAB) between the various lanes were constructed as described pre-viously (Yeung et al., 2010).

2.7. DNA sequencing and phylogenetic analysis

Sequencing of reamplified DNA was performed by the SickKidsCentre for Sequencing (Toronto, Canada) using the Applied Biosys-tems SOLiD 3.0 System. Unreadable portions of the DNA segmentswere removed from the 30 and 50 ends using Chromas V2.01 soft-ware (Technelysium, Tewantin, Australia). A consensus sequencebetween the forward and reverse sequences was constructed usingBioEdit™ Biological Sequence Alignment Editor (BioEdit, Carlsbad,CA). The NCBI BLAST algorithm (http://blast.ncbi.nlm.nih.gov/Blas-t.cgi) was used to check the consensus sequence against a databaseof known 16S rRNA sequences. Sequence alignment was carriedout using ClustalW (http://www.ebi.ac.uk/clustalw/). Phylogeneticand molecular evolutionary analyses were conducted usingMEGAV5.05 (Tamura et al., 2011). The bootstrap analyses for thephylogenetic trees were calculated by running 1000 replicates ofthe neighbor-joining data. The 16S rRNA gene sequences have beendeposited in the GenBank database under accession Nos.JQ686213–JQ686220.

3. Results and discussion

3.1. Culture enrichment

A cellulose-degrading consortium was enriched from compostthrough a series of sequential enrichments in non-reduced RMmedium. The bacterial diversity in the original solid compost inoc-ulum, the fifth and sixth enrichments, as well a 2-week old sixthenrichment, was analyzed using PCR-DGGE (Fig. 1). Our DGGEresults showed that there was a high bacterial diversity in the ori-ginal solid compost material (Fig. 1). Through enrichment, thisdiversity was markedly reduced (SAB = 37.7) with the remainingDGGE bands representing species involved either directly or indi-rectly in cellulose hydrolysis. The fifth and sixth enrichments arecloser in similarity (SAB = 73.7) with the major bands (i.e. C, D, E,H) present in both. As the consortium aged and nutrients becamedepleted, there was a decrease in diversity (SAB = 48.3), as demon-strated by a reduction in the number of bands in the DGGE finger-print (SAB = 48.3). At this point, some members had likely entered adormant phase.

Throughout enrichment, the majority of the cellulolytic activityoccurred after the resazurin turned colourless, indicating reducedconditions. This suggests the co-existence of aerobic or facultativeanaerobic, as well as anaerobic community members. Similarly, re-sults of Kato et al. (2004) and Wongwilaiwalin et al. (2010) sug-gested that efficient cellulose utilization by compost-derivedmixed communities in non-reduced media was accomplishedthrough interactions between non-cellulolytic facultative anaer-obes that generate the reduced environment needed by cellulolyticanaerobes. A similar interaction was likely responsible for thecellulose-degrading activity of our consortium.

3.2. Analysis of microbial community structure dynamics

Fig. 2 summarizes community structure dynamics in the differ-ent experimental set-ups.

3.2.1. Non-reduced RM mediumWithin 4 days of inoculation into non-reduced RM medium, the

bacterial community structure remained relatively stable(SAB = 72.7) (Fig. 2a). All of the major bands (B, C, D, E, and H) were

Fig. 1. Cluster analysis of DGGE banding pattern in the original solid compostinoculum, the fifth enrichment culture, the stable consortium (enrichment 6), andthe aged consortium.

Fig. 2. Cluster analysis of DGGE banding pattern in the consortium at inoculation,as well as after 4 and 14 days of incubation in non-reduced RM medium (a), pre-reduced RM medium (b), YE-free non-reduced compost tea RM medium (c), and inYE-free non-reduced wastewater RM medium (d).

P. Ronan et al. / Bioresource Technology 129 (2013) 156–163 159

present. Band G, however, appeared for the first time in the day 4sample. After 14 days, the community structure shifted, resultingin a slightly lower SAB value of 60.2. The most notable differenceswere the disappearance of band D, as well as the appearance oftwo new (unsequenced) bands between bands B and C.

3.2.2. Pre-reduced RM mediumSince fermentation is an anaerobic process, and consortial cellu-

lolytic activity was most apparent after the medium becamereduced, the performance of the community was also assessed inpre-reduced media. The bacterial community structure changedconsiderably after inoculation into pre-reduced RM medium(SAB = 27.2) (Fig. 2b). At day 4, band D was no longer present. Thecommunity structure continued to shift between days 4 and 14(SAB = 57.1). Band A appeared for the first time in the day 4 laneand remained until day 14. Band B conversely, persisted through-out the entire incubation period.

3.2.3. YE-free non-reduced compost tea RM mediumThe consortium was cultured in non-reduced RM medium in

which yeast extract had been replaced by sterile compost tea. Inthis treatment, the bacterial community structure remained verystable by day 4 (SAB = 91.7) (Fig. 2c). All of the major bands (B, C,D, E, and H) in the consortium at inoculation were still presenton day 4, with band G appearing on day 4 and persisting throughday 14. Band D disappeared after day 4 resulting in a slight shiftin community structure at day 14 (SAB = 59.1).

3.2.4. YE-free non-reduced wastewater RM mediumChanges in the community structure were analyzed after 4 and

14 days of incubation in non-reduced wastewater RM medium

(Fig. 2d). The community at day 4 revealed a very high degree ofsimilarity to the stable consortium, resulting in an SAB value of78.3. All of the major sequenced bands (B, C, D, E, and H) persistedthroughout the incubation period. The community structureremained unchanged (SAB = 100) between days 4 and 14, which im-plies a rich nutrient composition in the medium. This is in contrastto the changes in community structure that occurred by day 14 inthe non-reduced RM (Fig. 2a, SAB = 60.2) and non-reduced composttea RM medium treatments (Fig. 2c, SAB = 59.1).

3.3. Community diversity

Eight DGGE bands were excised and sequenced; all showingP98% similarity to the 16S rRNA gene of known isolates from Gen-Bank. The phylogenetic relatedness of the sequenced consortiummembers is depicted in Fig. 3. Seven of the excised band sequencesclustered within the Firmicutes phylum, with five showing high

160 P. Ronan et al. / Bioresource Technology 129 (2013) 156–163

similarity to the anaerobic, spore forming members of the Clostrid-ium genus.

CDC-B showed 99% sequence similarity to a Clostridiaceae bacte-rium initially isolated from farmyard compost soil (Kim et al.,2008). CDC-C showed 99% similarity to a strain of C. thermocellum,one of the most widely studied cellulolytic bacteria. The fact that C.thermocellum is a cellulose hydrolysis specialist, exhibiting one ofthe fastest known rates of cellulose utilization (Lu et al., 2006),suggests that the member designated CDC-C is likely one of thekey cellulose-degraders within the consortium. Interestingly, thismember was able to persist through the initial non-reduced (aero-bic) stage, despite the fact it is most likely an obligate anaerobegiven that its closest relative is C. thermocellum.

CDC-D and CDC-E both exhibited a high similarity to Clostridiumcaenicola, initially isolated from the anaerobic sludge of a cellulose-degrading methanogenic bioreactor (Shiratori et al., 2009).Although non-cellulolytic, this microorganism can use the mainproducts of cellulose hydrolysis (glucose and cellobiose) toproduce hydrogen, carbon dioxide, acetate, lactate, and ethanol(Shiratori et al., 2009). CDC-H and CDC-F both exhibited a highsimilarity to Clostridium cellulosi, initially isolated by Yanlinget al. (1991) from cow manure compost. Aside from cellulose, C.cellulosi can utilize a wide range of carbon sources, including cello-biose, glucose, and xylose (Yanling et al., 1991). CDC-A showed ahigh level of similarity to a Flavobacterium species (phylum Bacter-oidetes). It was most prevalent in the pre-reduced treatment(Fig. 2b) and was the only member closely related to a Gram neg-ative species.

Band G was most pronounced at day 4 in non-reduced RM(Fig. 2a) and non-reduced compost tea RM media (Fig. 2c). Thismember was found to belong to the Geobacillus genus, closelyrelated to the facultative anaerobes G. thermoglucosidasius and G.stearothermophilus, and may be one of the key oxygen scavengers,since it respires in the presence of oxygen. This would account forits presence in the earlier stages of culturing in non-reduced med-ia. Although G. thermoglucosidasius and G. stearothermophilus arenon-cellulolytic, they can utilize the main products of cellulosehydrolysis (cellobiose and glucose) to produce lactate, formate,acetate, and ethanol (Cripps et al., 2009). G. thermoglucosidasiusspecifically has garnered attention as a good candidate for use inbio-ethanol production due to its high ethanol tolerance, which

Fig. 3. Phylogenetic relationship of the eight sequence

is approximately 10% (Taylor et al., 2009). Cripps et al. (2009) re-cently reported three genetically modified strains of G. thermoglu-cosidasius that were optimized to achieve ethanol yields of 90%theoretical.

The fact that six of the eight sequenced members showed a highdegree of similarity to members of the Clostridiaceae family, specif-ically the Clostridium genus, implies a certain level of functionalredundancy within the consortium. Based on the sequencing data,there were at least three members (CDC-C, H, and F) closely relatedto species with cellulose-degrading properties (C. cellulosi and C.thermocellum), while at least six members (CDC-C, D, E, F, G, andH) showed high similarity to known ethanol producers (Crippset al., 2009; Shiratori et al., 2009; Yanling et al., 1991). This mostlikely contributed to the observed versatile and robust nature ofthe consortium.

3.4. Soluble end-product analysis

Fig. 4 summarizes profiles of major fermentation products inthe different experimental set-ups. The extent of cellulosic sub-strate degradation in each treatment was complete. Qualitatively,the point at which the filter paper was no longer visible coincidedroughly with the ethanol concentration peak.

3.4.1. Non-reduced RM mediumIn non-reduced RM medium, ethanol and acetate were the ma-

jor observed fermentation products and reached peak concentra-tions of 12.0 mM and 21.3 mM respectively (Fig. 4a). The ratio ofethanol:acetate, determined at the point of highest ethanol con-centration (day 6) was 1:1.6. C. thermocellum in batch culture hasbeen shown to produce a higher ethanol:acetate ratio of 1:0.4(Lynd et al., 1989), though it is important to note that this micro-organism requires strict anaerobic conditions, while this consor-tium was active in non-reduced media. Rydzak et al. (2011)indicated that the ethanol:acetate ratio of the consortium couldbe enhanced by adding acetate to the culture medium, shiftingthe metabolism of members away from acetate formation and to-ward ethanol production. The [ethanol + acetate] yield on day 6was 31.29 mM (12.04 mM + 19.25 mM), which represents 98% ofthe theoretical maximum of 32 mM calculated for the completedegradation of 2.66 g/L cellulose.

d DGGE gel bands based on 16S rRNA similarity.

Fig. 4. Ethanol and acetate production by the consortium in non-reduced RM medium (a), reduced RM medium (b), YE-free non-reduced compost tea RM medium (c), andYE-free non-reduced wastewater RM medium (d). Ethanol production in the test cultures is represented by . Ethanol production in the control culture is represented by

. Acetate production in the test culture is represented by . Acetate production in the control culture is represented by . Error bars represent the standarderror of each data point.

Table 1Peak concentrations of minor fermentation and hydrolysis products.

Medium Lactate (mM) Formate (mM) Butyrate (mM)

Non-reduced RM 1.6 2.0 0.6Pre-reduced RM ND ND 0.7Compost tea RM 3.0 1.3 0.2Wastewater RM 2.2 5.5 0.3a

a Peak values of all minor fermentation products occurred at day 4, exceptbutyrate in wastewater RM, which occurred at day 14. ND means ‘‘not detected’’.

P. Ronan et al. / Bioresource Technology 129 (2013) 156–163 161

After day 6, the ethanol concentration decreased, while acetatelevels remained relatively stable. Lactate, formate, and butyratewere also detected in the medium, reaching peak concentrationsof 1.6 mM, 2.0 mM, and 0.6 mM respectively by day 4 (Table 1).No cellobiose or glucose was detected in the culture supernatant,suggesting that the consortium was efficiently utilizing the cellu-lose hydrolysis products as they were being produced. Wang andChen (2009) described that a lack of hydrolysate accumulationbenefits the overall cellulolytic activity of the consortium becauseit negates cellulase inhibition. Kato et al. (2004) reported a similarresult after comparing the activity of a cellulolytic Clostridium inpure culture in reduced medium and co-culture with severalaerobic and facultative anaerobic strains in non-reduced medium.They found that after 10 days, the cellulosic oligosaccharideconcentration in the supernatant was approximately 40% lowerin the co-culture as compared to the pure culture. This impliesthat the non-cellulolytic members of these cellulose-degradingcommunities utilize cellulose-hydrolysis products following theirinitial oxygen scavenging.

3.4.2. Pre-reduced RM mediumAs described above, the consortium could degrade cellulose and

produce ethanol in non-reduced media, with the majority of cellul-olysis occurring after the medium had been reduced to the pointwhere resazurin turned colourless. Therefore, the consortium wasinoculated into pre-reduced RM medium to determine if pre-reduced conditions improved the overall performance of the con-sortium. Ethanol and acetate were once again the major observed

162 P. Ronan et al. / Bioresource Technology 129 (2013) 156–163

fermentation products, although their production was diminishedin comparison to non-reduced RM medium. The peak ethanoland acetate concentrations (day 14) were 8.7 mM and 19.1 mMrespectively, resulting in an ethanol:acetate ratio of 1:2.2(Fig. 4b). The [ethanol + acetate] yield at this point was 27.8 mM,representing 87% of the theoretical maximum. The fact that thisyield and ethanol:acetate ratio is lower than in non-reducedmedia (98% and 1:1.6 respectively), suggests that a pre-reducedenvironment did not improve the performance of the consortiumwith respect to end-product formation. Conditioning in non-reduced media might have led to increased oxygen toleranceof the consortium, possibly increasing cellulose-degrading andethanol-producing activities. This observation suggests that thecontribution of aerobic or facultative anaerobic members likelygoes beyond simply providing an anaerobic environment, by alsoproviding growth factors to the cellulolytic consortium members,consuming inhibitory metabolites, or neutralizing pH, as suggestedby others (Kato et al., 2004; Wongwilaiwalin et al., 2010).

Unlike non-reduced RM medium, only butyrate was detected asa minor fermentation product and reached a concentration of0.7 mM at day 4. No cellobiose or glucose was detected in theculture supernatant (Table 1). An appreciable amount of end-products were detected in the cellulose-free control cultures in thispre-reduced medium, compared to the non-reduced medium. Itappears that in the non-reduced medium, carbon sources withinYE are required by aerobes and facultative anaerobes in order tobe able to respire and generate the anaerobic conditions neededfor cellulolysis and fermentation. In pre-reduced medium, how-ever, YE is not needed for this aerobic metabolism, as conditionsare already reduced. The carbon within it therefore is subject tofermentation by anaerobic community members.

3.4.3. YE-free non-reduced compost tea RM mediumYeast extract is a rich source of carbon, nitrogen, amino acids,

and vitamins, and is often used in the cultivation of cellulolyticbacteria in non-reduced media (Haruta et al., 2002; Kato et al.,2004; Miyazaki et al., 2008; Wongwilaiwalin et al., 2010). Sinceone of the aims of consolidated bioprocessing is to reduce or elim-inate the need for added YE in fermentations (Lynd et al., 2005), theconsortium was cultured in media lacking yeast extract, supple-mented with compost tea, a more abundant and less costlywaste-derived complex nutrient source. In this treatment, ethanoland acetate were again major fermentation products, reachingpeak concentrations of 9.9 mM by day 4, and 26.8 mM by day 14respectively (Fig. 4c). The [ethanol + acetate] yield on day 4 was18.393 mM (9.876 mM + 8.517 mM), representing 57% of the theo-retical maximum. The ethanol:acetate ratio on day 4 was 1:0.9.This ratio is higher than the one achieved in non-reduced RM med-ium (1:1.6) and pre-reduced RM medium (1:2.2), and is closer tothe ratio reportedly achieved by C. thermocellum in batch culture,which was 1:0.4 (Lynd et al., 1989).

Trace levels of lactate, formate, and butyrate were also detected,achieving peak concentrations of 3.0 mM, 1.3 mM, and 0.2 mMrespectively, all by day 4 (Table 1). No cellobiose or glucose wasdetected in the culture supernatant.

3.4.4. YE-free non-reduced wastewater RM mediumMunicipal wastewater is a plentiful resource and its use in the

production of value-added products such as bio-ethanol may beparticularly beneficial. In this medium, the ethanol peak concen-tration was 14.3 mM (day 6), while the highest acetate concentra-tion was 28.8 mM (day 4, Fig. 4d). The ethanol:acetate ratio at day6 was approximately 1:1.9. This ratio was slightly lower than innon-reduced RM medium, although still higher than the oneachieved in pre-reduced RM medium (1:2.2). Small amounts oflactate, formate, and butyrate were also detected, achieving peak

concentrations of 2.2 mM (day 4), 5.5 mM (day 4), and 0.3 mM(day 14) respectively (Table 1). No cellobiose or glucose wasdetected in the culture supernatant.

The [ethanol + acetate] yield at day 6 was 40.76 mM(14.27 mM + 26.49 mM). Since only a maximum of 32 mM can beproduced from 2.66 g/L cellulose, it is apparent that some non-cellulosic carbon present in the medium must have contributedto the observed end-product accumulation. This was confirmedby the cellulose-free control culture in this treatment, in which asubstantial amount of acetate was produced. Wastewater is a richsource of nutrients and can contain a variety of carbon sources.Therefore, it is likely there was far more carbon present in themedium than was initially required by aerobic members to gener-ate anaerobic conditions. Thus, once the medium was sufficientlyreduced, this remaining carbon was fermented by anaerobicmembers.

In the three non-reduced treatments (non-reduced RM, com-post tea RM, wastewater RM medium), ethanol levels peaked andeither reached a plateau or subsequently declined (Fig. 4a, c, andd). The lack of a sustained increase in ethanol concentration at latertime points in the non-reduced treatments may be attributed to acombination of substrate depletion and ethanol utilization as a car-bon source. The same trend was observed in a compost-derivedcellulolytic community described by Haruta et al. (2002). Nonethe-less, wastewater appears to be an effective replacement for YE andmay be a good candidate for use as a complex nutrient source inlarge scale CBP bio-ethanol production.

The consortium’s ethanol:acetate ratio may be improved infuture experiments through slight modifications to the culturemedium. Rydzak et al. (2011) noted that the addition of formateto the medium used to cultivate C. thermocellum causes ethanoland H2 production to increase by �10%, while simultaneouslydecreasing acetate production. Similarly, the addition of acetatecauses an increase in ethanol and decrease in formate production(Rydzak et al., 2011).

A biotechnological approach toward improving the ethanol:ace-tate ratio conversely, could involve the targeted silencing of one orboth of the genes coding for acetyl-CoA phosphotransacetylase andacetate kinase, the enzymes responsible for converting acetyl-CoAto acetate. Since ethanol and acetate share acetyl-CoA as a directprecursor in clostridial metabolism (Rydzak et al., 2011), this couldserve to increase ethanol production while simultaneouslydecreasing acetate production.

3.5. pH profiles

The pH profiles in the test cultures in all four media werecharacterized by a rapid decrease within the first 4–6 days(Fig. 5). Despite differences in starting pH, the final pH in thenon-reduced RM, non-reduced compost tea RM, and non-reducedwastewater RM medium test cultures was approximately 6.4.The pH of the pre-reduced medium conversely decreased toaround 5.9 by day 14. In the cellulose-free control cultures, thepH remained fairly stable throughout the incubation period, withthe final and initial pH in each respective treatment being less than0.26 apart. Once again, the pH in the non-reduced RM, non-re-duced compost tea RM, and non-reduced wastewater RM controlcultures at the end of the experiment were very close at approxi-mately 7.4–7.5.

4. Conclusions

The consortium enriched and characterized in this study de-graded cellulose in the absence of reducing agent, and appearedto generate anaerobic conditions through oxygen-consumingaerobic respiration. Ethanol and acetate were major fermentation

Fig. 5. Profiles of pH throughout the 14 day culturing period. In non-reduced RMmedium, represents the test culture, while represents thecontrol. In pre-reduced RM medium, represents the test culture, while

represents the control. In non-reduced compost tea RM medium,represents the test culture, while represents the control. In aerobiccompost tea RM medium, represents the test culture, whilerepresents the control. The dark arrows indicate the close proximity of final pHachieved in the non-reduced RM, non-reduced compost tea RM, and non-reducedwastewater RM test and control cultures.

P. Ronan et al. / Bioresource Technology 129 (2013) 156–163 163

products and the consortium’s activity was stable in non-reducedand reduced media, as well as yeast extract-free non-reducedmedia supplemented with waste-based nutrient sources. Severalcommunity members showed high similarity to Clostridium spe-cies, suggesting the presence of some functional redundancy.

Reducing agent and yeast extract both represent significantcosts in the culturing of cellulolytic, ethanologenic microorgan-isms. The community described here exhibited this activity inthe absence of both.

Acknowledgements

This work was supported by The Natural Sciences and Engineer-ing Research Council of Canada (NSERC) Strategic Projects GrantSTPGP 365076 and Genome Canada.

References

Brenner, K., You, L., Arnold, F.H., 2008. Engineering microbial consortia: a newfrontier in synthetic biology. Trends Biotechnol. 26, 483–489.

Cripps, R.E., Eley, K., Leak, D.J., Rudd, B., Taylor, M., Todd, M., Boakes, S., Martin, S.,Atkinson, T., 2009. Metabolic engineering of Geobacillus thermoglucosidasius forhigh yield ethanol production. Metab. Eng. 11, 398–408.

Haruta, S., Cui, Z., Huang, Z., Li, M., Ishii, M., Igarashi, Y., 2002. Construction of astable microbial community with high cellulose-degradation ability. Appl.Microbiol. Biotechnol. 59, 529–534.

Inderwildi, O.R., King, D.A., 2009. Quo vadis biofuels? Energy Environ. Sci. 2, 343–346.

Izquierdo, J.A., Sizova, M.V., Lynd, L.R., 2010. Diversity of bacteria and glycosylhydrolase family 48 genes in cellulolytic consortia enriched from thermophilicbiocompost. Appl. Environ. Microbiol. 76, 3545–3553.

Kato, S., Haruta, S., Cui, Z.J., Ishii, M., Igarashi, Y., 2004. Effective cellulosedegradation by a mixed-culture system composed of a cellulolytic Clostridiumand aerobic non-cellulolytic bacteria. FEMS Microbiol. Ecol. 51, 133–142.

Kato, S., Haruta, S., Cui, Z.J., Ishii, M., Igarashi, Y., 2008. Network relationships ofbacteria in a stable mixed culture. Microbiol. Ecol. 56, 403–411.

Kim, J.J., Kim, H.N., Masui, R., Kuramitsu, S., Seo, J., Kim, K., Sung, M.H., 2008.Isolation of uncultivable anaerobic thermophiles of the family Clostridiaceaerequiring growth-supporting factors. J. Microbiol. Biotechnol. 18, 611–615.

Lee, S.H., Malone, C., Kemp, P.F., 1993. Use of 16S rRNA-targeted fluorescent probesto increase signal strength and measure cellular RNA from natural planktonicbacteria. Mar. Ecol. Prog. Ser. 101, 193–201.

Lu, Y., Zhang, Y.P., Lynd, L.R., 2006. Enzyme-microbe synergy during cellulosehydrolysis by Clostridium thermocellum. Proc. Natl. Acad. Sci. USA 103, 16165–16169.

Lynd, L.R., Grethlein, H.E., Wolkin, R.H., 1989. Fermentation of cellulosic substratesin batch and continuous culture by Clostridium thermocellum. Appl. Environ.Microbiol. 55, 3131–3139.

Lynd, L.R., Weimer, P.J., Van Zyl, W.H., Pretorius, I.S., 2002. Microbial celluloseutilization: fundamentals and biotechnology. Microbiol. Mol. Biol. R. 66, 506–577.

Lynd, L.R., Van Zyl, W.H., McBride, J.E., Laser, M., 2005. Consolidated bioprocessingof cellulosic biomass: an update. Curr. Opin. Biotechnol. 16, 577–583.

Maddipati, P., Atiyeh, H.K., Bellmer, D.D., Huhnke, R.L., 2011. Ethanol productionfrom syngas by Clostridium strain P11 using corn steep liquor as a nutrientreplacement to yeast extract. Bioresour. Technol. 102, 6494–6501.

Miyazaki, K., Irbis, C., Takada, J., Matsuura, A., 2008. An ability of isolated strains toefficiently cooperate in ethanolic fermentation of agricultural plant refuse underinitially aerobic thermophilic conditions: oxygen deletion process appended toconsolidated bioprocessing (CBP). Bioresour. Technol. 99, 1768–1775.

Muyzer, G., De Waal, E.C., Uitterlinden, A.G., 1993. Profiling of complex microbialpopulations by denaturing gradient gel electrophoresis analysis of polymerasechain reaction amplified gene coding for 16S rRNA. Appl. Environ. Microbiol. 59,695–700.

Okeke, B.C., Lu, J., 2011. Characterization of a defined cellulolytic and xylanolyticbacterial consortium for bioprocessing of cellulose and hemicelluloses. Appl.Biochem. Biotechnol. 163, 869–881.

O’Sullivan, C.A., Burrell, P.C., Clarke, W.P., Blackall, L.L., 2005. Structure of a cellulosedegrading bacterial community during anaerobic digestion. Biotechnol. Bioeng.92, 871–878.

Ozkan, M., Desai, S.G., Zhang, Y., Stevenson, D.M., Beane, J., White, E.A., Guerinot,M.L., Lynd, L.R., 2001. Characterization of 13 newly isolated strains of anaerobic,cellulolytic, thermophilic bacteria. J. Ind. Microbiol. Biotechnol. 27, 275–280.

Rodolfi, L., Zittelli, G.C., Bassi, N., Padovani, G., Biondi, N., Bonini, G., Tredici, M.R.,2009. Microalgae for oil: strain selection, induction of lipid synthesis andoutdoor mass cultivation in a low-cost photobioreactor. Biotechnol. Bioeng.102, 100–112.

Rydzak, T., Levin, D.B., Cicek, N., Sparling, R., 2011. End-product induced metabolicshifts in Clostridium thermocellum ATCC 27405. Appl. Microbiol. Biotechnol. 92,199–209.

Shiratori, H., Sasaya, K., Ohiwa, H., Ikeno, H., Ayame, S., Kataoka, N., Miya, A., Beppu,T., Ueda, K., 2009. Clostridium clariflavum sp. nov. and Clostridium caenicola sp.nov., moderately thermophilic, cellulose-/cellobiose-digesting bacteria isolatedfrom methanogenic sludge. Int. J. Syst. Evol. Microbiol. 59, 1764–1770.

Sizova, M.V., Izquierdo, J.A., Panikov, N.S., Lynd, L.R., 2011. Cellulose- and xylan-degrading thermophilic anaerobic bacteria from biocompost. Appl. Environ.Microbiol. 77, 2282–2291.

Sheffield, V.C., Cox, D.R., Lerman, L.S., Myers, R.M., 1989. Attachment of a 40-base-pair G + C-rich sequence (GC-clamp) to genomic DNA fragments by thepolymerase chain reaction results in improved detection of single-basechanges. Proc. Natl. Acad. Sci. USA 86, 232–236.

Tamura, K., Peterson, D., Peterson, N., Stecher, G., Nei, M., Kumar, S., 2011. MEGA5:molecular evolutionary genetics analysis using maximum likelihood,evolutionary distance, and maximum parsimony methods. Mol. Biol. Evol. 28,2731–2739.

Taylor, M.P., Eley, K.L., Martin, S., Tuffin, M.I., Burton, S.G., Cowan, D.A., 2009.Thermophilic ethanologenesis: future prospects for second-generationbioethanol production. Trends Biotechnol. 27, 398–405.

Wang, Z., Chen, S., 2009. Potential of biofilm-based biofuel production. Appl.Microbiol. Biotechnol. 83, 1–18.

Wang, W., Yan, L., Cui, Z., Gao, Y., Wang, Y., Jing, R.W., 2011. Characterization of amicrobial consortium capable of degrading lignocelluloses. Bioresour. Technol.102, 9321–9324.

Wongwilaiwalin, S., Rattanachomsri, U., Laothanachareon, T., Eurwilaichitr, L.,Igarashi, Y., Champreda, V., 2010. Analysis of a thermophilic lignocellulosedegrading microbial consortium and multi-species lignocellulolytic enzymesystem. Enzyme Microb. Technol. 47, 283–290.

Xu, C., Qin, Y., Li, Y., Ji, Y., Huang, J., Song, H., Xu, J., 2010. Factors influencingcellulosome activity in consolidated bioprocessing of cellulosic ethanol.Bioresour. Technol. 101, 9560–9569.

Yanling, H., Youfang, D., Yanquan, L., 1991. Two cellulolytic Clostridium species:Clostridium cellulosi sp. nov. and Clostridium cellulofermentans sp. nov. Int. J. Syst.Bacteriol. 41, 306–309.

Yeung, C.W., Lee, K., Whyte, L.G., Greer, C.W., 2010. Microbial communitycharacterization of the Gully: a marine protected area. Can. J. Microbiol. 56,421–431.

Yeung, C.W., Woo, M., Lee, K., Greer, C.W., 2011. Characterization of the bacterialcommunity structure of Sydney Tar Ponds sediment. Can. J. Microbiol. 57, 493–503.

Zyabreva, N.V., Isakova, E.P., Biryukov, V.V., 2001. Selection of a mixed culture ofcellulolytic thermophilic anaerobes from various natural sources. Appl.Biochem. Microbiol. 37, 363–367.