Shewanella oneidensis in a lactate-fed pure-culture and a glucose-fed co-culture with Lactococcus lactis with an electrode as electron acceptor

Article (PDF Available)inBioresource Technology 102(3):2623-8 · October 2010with55 Reads
DOI: 10.1016/j.biortech.2010.10.033 · Source: PubMed
Abstract
Bioelectrochemical systems (BESs) employing mixed microbial communities as biocatalysts are gaining importance as potential renewable energy, bioremediation, or biosensing devices. While we are beginning to understand how individual microbial species interact with an electrode as electron donor, little is known about the interactions between different microbial species in a community: sugar fermenting bacteria can interact with current producing microbes in a fashion that is either neutral, positively enhancing, or even negatively affecting. Here, we compare the bioelectrochemical performance of Shewanella oneidensis in a pure-culture and in a co-culture with the homolactic acid fermenter Lactococcus lactis at conditions that are pertinent to conventional BES operation. While S. oneidensis alone can only use lactate as electron donor for current production, the co-culture is able to convert glucose into current with a comparable coulombic efficiency of ∼17%. With (electro)-chemical analysis and transcription profiling, we found that the BES performance and S. oneidensis physiology were not significantly different whether grown as a pure- or co-culture. Thus, the microbes worked together in a purely substrate based (neutral) relationship. These co-culture experiments represent an important step in understanding microbial interactions in BES communities with the goal to design complex microbial communities, which specifically convert target substrates into electricity.
Shewanella oneidensis in a lactate-fed pure-culture and a glucose-fed co-culture
with Lactococcus lactis with an electrode as electron acceptor
Miriam A. Rosenbaum
a
, Haim Y. Bar
b
, Qasim K. Beg
c
, Daniel Segrè
c,d,e
, James Booth
b
, Michael A. Cotta
f
,
Largus T. Angenent
a,
a
Department of Biological and Environmental Engineering, Cornell University, Ithaca, NY, USA
b
Department of Biological Statistics and Computational Biology, Cornell University, Ithaca, NY, USA
c
Department of Biomedical Engineering, Boston University, Boston, MA, USA
d
Department of Biology, Boston University, Boston, MA, USA
e
Bioinformatics Program, Boston University, Boston, MA, USA
f
Bioenergy Research Unit, United States Department of Agriculture, Agricultural Research Service (ARS), National Center for Agricultural Utilization Research (NCAUR), Peoria, IL, USA
article info
Article history:
Received 29 July 2010
Received in revised form 5 October 2010
Accepted 6 October 2010
Available online 12 October 2010
Keywords:
Bioelectrochemical system
Microbial fuel cell
Shewanella oneidensis
Microarray
Lactococcus lactis
abstract
Bioelectrochemical systems (BESs) employing mixed microbial communities as biocatalysts are gaining
importance as potential renewable energy, bioremediation, or biosensing devices. While we are begin-
ning to understand how individual microbial species interact with an electrode as electron donor, little
is known about the interactions between different microbial species in a community: sugar fermenting
bacteria can interact with current producing microbes in a fashion that is either neutral, positively
enhancing, or even negatively affecting. Here, we compare the bioelectrochemical performance of Shewa-
nella oneidensis in a pure-culture and in a co-culture with the homolactic acid fermenter Lactococcus lactis
at conditions that are pertinent to conventional BES operation. While S. oneidensis alone can only use lac-
tate as electron donor for current production, the co-culture is able to convert glucose into current with a
comparable coulombic efficiency of 17%. With (electro)-chemical analysis and transcription profiling,
we found that the BES performance and S. oneidensis physiology were not significantly different whether
grown as a pure- or co-culture. Thus, the microbes worked together in a purely substrate based (neutral)
relationship. These co-culture experiments represent an important step in understanding microbial inter-
actions in BES communities with the goal to design complex microbial communities, which specifically
convert target substrates into electricity.
Ó 2010 Elsevier Ltd. All rights reserved.
1. Introduction
The ability to respire with solid terminal electron acceptors,
such as mineral oxides or electrodes, has positioned Shewanella
oneidensis MR-1 as a model organism for microbial fuel cell
(MFC) or more generally bioelectrochemical system (BES) re-
search (Bretschger et al., 2007; Gorby et al., 2006; Kim et al.,
1999, 2002; Marsili et al., 2008; Nealson et al., 2002). In the anode
compartment of BESs, microorganisms transfer electrons, which
result from the breakdown of organic substrates, to the anode
(i.e., anaerobic respiration with a solid electron acceptor). Such
transfer of electrons is highly promising for novel applications in
wastewater treatment, bioremediation, and biosensing. All known
mechanisms for microbial extracellular respiration are proposed to
be employed by S. oneidensis: (i) direct electron transfer with outer
membrane redox-proteins (Kim et al., 1999; Myers and Myers,
2001), possibly with the help of conductive appendages (Gorby
et al., 2006); and (ii) mediated electron transfer with microbially
produced soluble redox-compounds (Marsili et al., 2008; von Can-
stein et al., 2008). Besides S. oneidensis, the direct electron-transfer
bacterium Geobacter sulfurreducens and the mediator (phenazine)
producer Pseudomonas aeruginosa represent important model
organisms for the functional investigation of extracellular electron
transfer in bioelectrochemical systems (often referred to as anode
respiring bacteria [ARB]).
Shewanella’s typical electron donor for anaerobic respiration
with solid electron acceptors is lactate, which is oxidized to ace-
tate, carbon dioxide, and four electrons. Recent studies have shown
that oxygen promotes the utilization of a wider spectrum of carbon
sources by S. oneidensis for electric current generation (e.g., ace-
tate) (Biffinger et al., 2008; Ringeisen et al., 2007; Rosenbaum
et al., 2010). However, this increase in versatility occurs at lower
electron transfer efficiencies from the substrate to the electrode
(i.e., coulombic efficiency). Since S. oneidensis was found in the
0960-8524/$ - see front matter Ó 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.biortech.2010.10.033
Corresponding author. Address: 214 Riley-Robb Hall, Ithaca, NY 14853, USA.
Tel.: +1 607 255 2480; fax: +1 607 255 4449.
E-mail address: la249@cornell.edu (L.T. Angenent).
Bioresource Technology 102 (2011) 2623–2628
Contents lists available at ScienceDirect
Bioresource Technology
journal homepage: www.elsevier.com/locate/biortech
anaerobic community of undefined mixed-culture MFCs (Kim
et al., 2006), while lactate is not a common primary MFC substrate,
we believe that S. oneidensis may play an important role as a termi-
nal link between fermentation processes and electrode respiration
in the anaerobic food-web of a BES anode compartment. To im-
prove the community’s conversion efficiency for a specific organic
substrate into electric power, we need to understand how
microbes interact in the mixed culture anodic food web. Only
few recent studies have investigated the interactions of microor-
ganisms in defined co-cultures in bioelectrochemical systems.
Ren et al. (2007) studied the cellulose fermenter Clostridium cellul-
olyticum with the direct electron-transfer bacterium G. sulfurredu-
cens to generate an electric current with cellulose as substrate,
whereby the co-culture showed similar electrochemical perfor-
mances as a pure-culture of G. sulfurreducens when fed with ace-
tate. Read et al. (2010) studied biofilm formation and
electrochemical performance of the Gram
ARBs S. oneidensis,
G. sulfurreducens, and P. aeruginosa in combination with two Gram
+
fermenters: Clostridium acetobutylicum and Enterococcus faecium.
They found an increased current production for all the tested
binary co-cultures with E. faecium, while the co-cultures with
C. acetobutylicum produced lower currents compared to the Gram
pure-cultures alone. It seems therefore, that the interactions be-
tween fermenters and ARBs can be various: neutral with only a
food-chain relationship (G. sulfurreducens and C. cellulolyticum);
positive with enhanced current production of the co-culture
(all tested ARBs and E. faecium); and negative with reduced cur-
rent production of the co-culture (all tested ARB and C. acetobutyl-
icum) compared to the ARB pure-culture. At this time point, a
prediction of the type of interaction between a fermenter and an
ARB is not possible.
Here, we studied how the physiology of S. oneidensis is influ-
enced when it is grown in a co-culture with an unmetabolizable
substrate, such as glucose. We combined S. oneidensis MR-1 with
the homolactic fermenter Lactococcus lactis in a continuous-flow
BES. L. lactis ferments C-6 sugars into
L
-lactate, and therefore a
combination of both organisms should allow S. oneidensis to pro-
duce electric current with glucose as the primary fuel. Besides
investigating the electrochemical behavior of the pure- versus
the co-culture, we performed chemical, biochemical, and micros-
copy analyses, and at the end of the electrochemical experiment
we recovered mRNA from the pure- and co-culture biofilms. The
mRNA was used for a S. oneidensis transcriptional analysis with
Affymetrix GeneChips. This analysis enabled us to ascertain if
S. oneidensis MR-1 undergoes physiological changes in a synergistic
co-culture fed with glucose.
2. Methods
2.1. Strains and media for bioelectrochemical systems
S. oneidensis MR-1 (a gift from Tim Gardner, Boston University,
Boston, MA, USA) was grown in LB medium for strain maintenance.
L. lactis LM 0230 (Efstathiou and McKay, 1977) was obtained from
the Bioenergy Research Unit of the ARS, USDA, Peoria, IL. L. lactis
was grown in M17 lactic acid bacteria medium (DSMZ medium
449) and 5 g/L sterile-filtered glucose was added after autoclaving.
The defined anode medium for bioelectrochemical experiments
was prepared according to Myers and Nealson (1988) and was
modified by adding 1.27 mM K
2
HPO
4
, 0.73 mM KH
2
PO
4
, 125 mM
NaCl, 5 mM HEPES, 0.5 g/L yeast extract, 0.5 g/L tryptone, and
5 g/L sodium b-glycerophosphate a medium requirement of
L. lactis (no addition of amino acids). After autoclaving, sterile fil-
tered 1 g/L glucose, 20 mM sodium
L
-lactate, and 20 mM K
2
HPO
4
were added. Analytical chemicals were ACS grade.
2.2. Reactor set-up
Two identical H-type electrochemical reactors were made of
glass with an anode and cathode liquid chamber volume of
220 mL each (Supplementary Fig. S1). The anode and cathode
chambers were separated by an anion exchange membrane
(19.6 cm
2
, AMI-7001 Membranes International, Glen Rock, NJ,
USA). Each anode chamber was temperature controlled at 30 °C
with a water jacket, stirred, continuously fed with defined medium
at a hydraulic retention time (HRT) of 5–10 h, and was equipped
with a carbon paper electrode (50 cm
2
geometric surface area,
AvCarb P50, The Fuel Cell Store, San Diego, CA, USA), which served
as working electrode. It was bound to a graphite rod (Poco Graphite
Inc., Decatur, TX, USA) with carbon cement (CCC Carbon Adhesive,
EMS, Hatfield, PA, USA). We used an Ag/AgCl (saturated KCl) refer-
ence electrode to control the working electrode (anode) potential.
The cathode chambers (=counter electrode chamber) were oper-
ated in batch with a graphite block electrode (3 9 1 cm,
PocoGraphite, Decatur, TX, USA). The entire assembled setup,
including two 10-L feeding tanks, was autoclaved before the exper-
iment. The tanks were used consecutively, whereby one tank fed
both anode chambers simultaneously during the S. oneidensis
pure-culture stages. At all time, the medium tanks and the reactors
were kept anaerobic by keeping it under a positively pressured 20%
CO
2
/80% N
2
atmosphere.
2.3. Electrochemical measurements
Steady-state S. oneidensis biofilms were grown with lactate as
electron donor at a working electrode under potentiostatic condi-
tions at 0.4 V (all potentials refer to standard hydrogen electrode,
SHE) (VSP potentiostat, BioLogic, Knoxville, TN, USA, in a three
electrode setup) in two parallel reactor setups. Before inoculation
with an overnight culture of S. oneidensis, blank electrochemical
measurements were conducted in the growth medium. In some
runs, complex media components that were required for the
growth of L. lactis (yeast extract, tryptone, b-glycerophosphate)
were added to a growing S. oneidensis culture one at a time to eval-
uate the bioelectrochemical effects of those components (after we
verified that those media components had no effect, they were
added to the medium from the start). Once stable bioelectrochem-
ical performance with S. oneidensis was achieved, one reactor was
switched to a medium tank without
L
-lactate (from now on called
reactor SL, as opposed to reactor S which still was fed with 20 mM
lactate), while 1 g/L glucose was fed to both reactors. When resid-
ual lactate was depleted, 3 mL overnight grown L. lactis was inoc-
ulated into reactor SL. Cyclic voltammetry tests (0.3 to 0.7 V,
v = 1 mV/s) were performed regularly (every 48 h) during the en-
tire experimental run. This experimental series was repeated sev-
eral times with similar performance but with different length of
operating periods. From two of these trials, biofilm RNA was col-
lected for transcriptional analysis. The purity of the culture was
confirmed by plating of the reactor liquid at the end of each exper-
iment (S. oneidensis forms characteristic pinkish colonies).
2.4. Chemical analysis
Anode effluent samples were taken every other day to deter-
mine HRT, effluent pH,
L
-lactate (Accutrend Lactate Analyzer,
Roche Diagnostics), and other soluble metabolites. Filtered sam-
ples (0.2-
l
m nitrocellulose filter, Millipore, Billerica, MA, USA)
were analyzed for sugars, organic acids, and ethanol using a Spec-
traSYSTEM liquid chromatography system equipped with a refrac-
tive index detector (Thermo-Fisher Scientific Inc.) and with an
organic acids column (Aminex HPX-87H Column, Bio-Rad
2624 M.A. Rosenbaum et al. / Bioresource Technology 102 (2011) 2623–2628
Laboratories Inc., Hercules, CA, USA). Samples were run at 65 °C
and eluted at 0.6 mL/min with 5 mM sulfuric acid.
2.5. SEM imaging
At the end of the experiment a part of the carbon paper elec-
trode was sampled for SEM imaging: the electrode samples were
fixed with 2.5% glutaraldehyde and 1% osmium tetroxide, followed
by a serial ethanolic dehydration, critical point drying, and thin
gold coating. Images were taken with a Hitachi S-450 scanning
electron microscope at 20 kV accelerating voltage.
2.6. Microarray analysis
2.6.1. Chemicals and reagents used for microarrays
All general chemicals for molecular biological work were certi-
fied RNase/DNase free. Qiagen Inc. (Valencia, CA) supplied the RNA
protect reagent and the DNA purification kit. Other specific re-
agents and chemicals used during isolation/purification of RNA
and during various steps of Shewanella chips hybridization
(Affymetrix Inc., Santa Clara, CA) were purchased form several dif-
ferent vendors: Superscript II reverse transcriptase, DTT, random
hexamers, and BSA from Invitrogen Inc. (Carlsbad, CA); GeneChip
labeling reagent, One-phor-all buffer, and B2 oligo from Affymetrix
Inc.; DNAse from Pierce Biochemicals Inc. (Rockford, IL); MES stock,
lysozyme, Goat IgG, and 200 proof ethanol from Sigma–Aldrich
(St. Louis, MO); Terminal transferase, Herring sperm DNA, and
dNTPs from Promega Inc. (Madison, WI); Biotinylated Anti-Strepta-
vidin antibody from Vector laboratories (Burlingame, CA); SSPE,
Streptavidin, SAPE, 10% Tween 20, NaOH, and HCl from Thermo-
Fisher Scientific (Waltham, MA); and RNAse-free DNAse I enzyme
for RNA purification, TE Buffer (pH 8.0), Superase 1n, 5 M NaCl,
and nuclease-free water were from Ambion (Austin, TX).
2.6.2. RNA sampling and isolation
RNA for microarray analysis was sampled from two S and two
SL bioelectrochemical reactors. The carbon paper electrodes were
removed from the reactor, bathed in Qiagen RNA-protect for 30 s
and immediately frozen at 80 °C. The biofilm samples were loos-
ened, but not removed, from the electrode backbone by scraping
with a sterile razor blade. The electrode backbone was then
washed with 2 mL RNA protect and the biofilm-carbon sludge
was transferred to a 15 mL tube. Seven millilitres ice-cold phos-
phate buffer saline (PBS) was added, the mix was vortexed on high-
est speed and centrifuged for 10 min at 5500g. The supernatant
was decanted and replaced with 7 mL ice-cold PBS. Then, the mix
was sonicated at 7 W for 30 s on ice. Vortexing, centrifuging, and
sonicating was repeated twice. Then, the pellets were resuspended
in 0.75 mL NAES buffer (50 mM sodium acetate buffer, 10 mM
EDTA, and 1% SDS at pH 5). RNA was isolated with a phenol:chlo-
roform extraction protocol similar to Cury and Koo (2007). The iso-
lated RNA was purified from genomic DNA contaminations with
Ambion DNase I treatment following the manufacturer’s instruc-
tions. RNA yields were quantified with a NanoDrop spectrometer
(Thermo Scientific, Wilmington, DE) and UV 260/280 ratios were
calculated to check purity of each RNA sample. RNA quality was
verified in a 1.5% agarose electrophoreses gel with ethidium bro-
mide staining.
2.6.3. Microarray hybridization
A previously described protocol (Driscoll, 2008; Faith et al.,
2007) was used for microarrays on S. oneidensis chips from Affymetrix
Inc. In brief, approximately 10
l
g of each RNA sample was used for
cDNA synthesis through reverse transcription, cDNA purification, and
cDNA-fragmentation. This was followed by labeling of cDNA and 16 h
of hybridization at 45 °ConS. oneidensis arrays. The labeled arrays were
subjected to several cycles of washing and staining using Affymetrix
Wash buffers A and B, Goat IgG, Streptavidin, Anti-streptavidin, and
SAPE according to the Affymetrix protocol for prokaryotic arrays. This
was followed by scanning of the stained arrays with Affymetrix Gene-
Chip Scanner Model 3000. The S. oneidensis MR-1 microarray data have
been submitted to Gene Expression Omnibus (accession number
GSE20343).
2.6.4. Statistical analysis
Microarray data were analyzed with the lemma (Laplace
approximated EM Microarray Analysis) package, available from
http://www.cran.r-project.org/ by Bar and Schifano. The software
is based on the statistical model proposed by Bar et al. (in press).
To account for the large number (3949) of hypotheses (i.e., genes
tested), we used the Benjamini–Hochberg adjustment to the p-val-
ues, which allows to control the false discovery rate (fdr) at any de-
sired level (Benjamini and Hochberg, 1995). Visualization of the
results as heatmaps was performed in JMP 8.0 (SAS Institute Inc.).
3. Results and discussion
3.1. BES performance of a S. oneidensis pure- and co-culture
S. oneidensis was grown on carbon paper electrodes in two,
simultaneously-operated, continuously-fed BES reactors under
potentiostatic control at 0.4 V (Fig. S1). One of the two reactors
(SL) was transitioned to co-culture operation once stable perfor-
mance of S. oneidensis was observed. Within a few hours after
switching from a lactate-fed S. oneidensis pure-culture to a
glucose-fed co-culture with L. lactis, the current generation recov-
ered to the original value of the pure-culture. We performed four
different trials and report final steady-state current densities and
the steady-state coulombic efficiencies (Table 1). The cyclic vol-
tammogram in Fig. 1 shows the catalytic current wave of the
S. oneidensis pure-culture in trial 2 at different time points of the
experiment with an onset potential of +0.1 V. The current pro-
duction and coulombic efficiencies within each trial were very sim-
ilar for the two reactors, but varied between the individual trials
(we commonly observed broad variations in S. oneidensis perfor-
mance between trials although the experimental conditions were
unaltered). Therefore, we performed paired, two-tailed t-tests to
evaluate the experimental differences between S. oneidensis in
pure-culture and co-culture, and found no statistical difference be-
tween the current production (p-value = 0.709) and coulombic effi-
ciency (p-value = 0.804) of reactor S versus reactor SL. Thus, from
an electrochemical standpoint, S. oneidensis showed the same per-
formance in pure-culture with lactate feeding and in co-culture
with glucose feeding.
We operated each trial for weeks with a continuous flow of
nutrients (i.e., in a chemostat) and added sufficient phosphate buf-
fer (up to 100 mM) to our medium to prevent a pH value lower
than 5.8 (especially during the L. lactis acid formation phase it is
important to maintain a high enough pH for sufficient S. oneidensis
activity). The pH levels and metabolic regime during the transition
from a pure-culture to a co-culture in reactor SL during experimen-
tal trial 2 is given (Fig. 2). The feeding concentration of glucose was
adjusted from initially 1 g/L at the starting time of glucose feeding
to achieve excess lactate concentrations in the effluent (to prevent
substrate limitation effects). Acetate was mainly formed as a by-
product during lactate oxidation by S. oneidensis, but L. lactis can
also produce small amounts of acetate as a by-product of glucose
fermentation. Even though L. lactis is considered a homolactic fer-
menting organism, the growth and substrate conditions define
how complete the homo-fermentative reaction is. Other products
typical of a mixed acid fermentation (e.g., pyruvate, acetate,
M.A. Rosenbaum et al. / Bioresource Technology 102 (2011) 2623–2628
2625
ethanol) may compose up to 10% of the fermentation products
(Cocaign-Bousquet et al., 1996). Many microbial fermentation
reactions can be enhanced by specific product removal to over-
come product toxicity effects or feed-back inhibition (Ezeji et al.,
2007). In our study, where S. oneidensis removed lactate formed
during glucose fermentation by L. lactis, we did not see such shift
in fermentation patterns because L. lactis converts glucose almost
stoichiometrically into lactate. In addition, at our low glucose con-
centrations (1–4 g/L) no toxicity of lactate was observed.
3.2. L. lactis alone showed no significant electrochemical activity
Despite recent reports on electric current generation by L. lactis
(Freguia et al., 2009; Masuda et al., 2010), our results showed no
significant current generation by a L. lactis pure-culture. We mea-
sured inconsequently low electrochemical activity for a pure-cul-
ture of L. lactis in defined (minimal) medium (0.4
l
A/cm
2
; note
that we report up to 20
l
A/cm
2
for a pure-culture of S. oneidensis).
Freguia et al. (2009), who also worked with a minimal medium for
pure-culture growth of L. lactis, did not report current densities/
surface area, but we estimated that 2.5
l
A/cm
2
was the upper limit
for their study (based on the given geometric surface area of the
utilized three-dimensional graphite felt). These authors reported
that a total of 1% of the electrons from glucose were redirected
from lactate production to pyruvate, acetate, and electric current,
with a very low coulombic efficiency of 0.275% for the current gen-
eration. With the knowledge that fermentation side products are
well known for L. lactis (Cocaign-Bousquet et al., 1996) and com-
pared to the much higher current production of S. oneidensis,we
concluded that L. lactis cannot generate significant electric current
in absence of externally-supplied redox shuttles. In rich medium
(#449, DSMZ, Germany, containing 15 g peptones and 2.5 g yeast
extract), we measured 7
l
A/cm
2
for a L. lactis pure-culture with a
much higher abiotic background current (20
l
A/cm
2
compared to
0.2
l
A/cm
2
background current in the defined medium with
0.5 g/L each yeast extract and tryptone). This indicates that media
redox components can serve as electron shuttles for L. lactis and,
indeed, Masuda et al. (2010) reported current production by
L. lactis in the presence of flavin-type redox-mediators contained
in yeast extract (the electrode surface related current density could
not be derived from the publication). But since the influence of
L. lactis on the S. oneidensis bioelectrochemical physiology was
the target of this study, all experiments were performed in defined
medium with low amounts of complex medium components.
3.3. Continuous co-cultures were dominated by S. oneidensis-direct
electron transfer processes
Although S. oneidensis can produce flavin-type mediators to
enhance electron transfer to the electrode (Marsili et al., 2008;
von Canstein et al., 2008), our cyclic voltammetry analyses in
S. oneidensis pure-culture experiments did not indicate the involve-
ment of flavin-type mediators (Fig. 1), which, if they had been
present, could have enabled L. lactis to participate in current
generation. In medium of similar composition to ours, at a slightly
Table 1
Current density and coulombic efficiency (CE) of a S. oneidensis pure-culture (S) and a co-culture with L. lactis (SL).
Trial Duration (days)
a
Current density (
l
A/cm
2
)CE
b
(%)
(S) (SL) (S) (SL)
1 13 22.8 19.8 20.3 13.7
2 30 22.5 21.5 17.9 16.7
3 32 6.8 6.4 1.3 1.4
4 32 11.34 19.9 4.8 9.9
Average 26.75 15.9 16.9 11.1 10.4
t-Test (paired, two tailed) p-value: 0.709 0.804
a
The significantly shorter experimental duration of trial 1 is due to starting this trial with the final media composition containing all complex components, while in the
other trials complex media components (yeast extract, tryptone) were added one by one during the experiment to exclude influences to the electrochemical results.
b
The coulombic efficiency CE represents the ratio of the current (as C/s) and the charge resulting from lactate oxidation as follows: CE = I/[z F fr c
Lactate
] with
I—current in C/s, z—number of electrons transferred per molecule, F—Faraday’s constant, fr—reactor flow rate in L/s, and c
Lactate
—molar concentration of consumed lactate for
S, and for SL ([glucose in] [glucose out]) 2 [lactate out].
Fig. 1. Cyclic voltammetry plots of a S. oneidensis pure-culture from different time
points (days 1, 7, 14, and 28) during the continuous BES reactor operation of trial 2.
Bold black lines represent reactor S with a S. oneidensis pure-culture, dashed gray
lines represent reactor SL that was operated with a S. oneidensis/L. lactis co-culture
and glucose substrate from day 23 on (before that time point it contained only
S. oneidensis and lactate substrate similar to reactor S). Potentials were scanned at
1 mV/s.
Fig. 2. Chemical analyses of metabolites (left axis) and pH (right axis) during a
continuous-flow co-culture BES experiment with S. oneidensis MR-1. L. lactis was
inoculated on day 22 after the medium was switched from lactate to glucose as
carbon source.
2626 M.A. Rosenbaum et al. / Bioresource Technology 102 (2011) 2623–2628
more basic pH (pH 7 instead of 6.2) Marsili et al. (2008) found the
(ribo)flavin mid-peak potential (and corresponding onset in the
catalytic current curve) to be 0.2 V. Malinauskas (2008) deter-
mined the mid-peak potential of riboflavin on graphite electrodes
to be 0.22 V. Our voltammograms (Fig. 1) show the steady in-
crease of a catalytic current with increasing activity of the culture
at an onset potential of 0 to +0.1 V, which indicates a different
catalytic component than the flavins, such as an outer membrane
c-type cytochrome, that exhibits direct electron transfer to the
electrode. Meitl et al. (2009) determined the standard potentials
for S. oneidensis outer membrane cytochromes (MtrC and OmcA)
to be 0.0–0.1 V vs. SHE, which is very close to the onset potential
found in our study. In another study, Cho and Ellington (2007) used
anaerobically-grown, washed cells of S. oneidensis in a phosphate
buffer electrolyte to measure non-turnover cyclic voltammograms
(i.e., no catalytic wave occurs, since no substrate is oxidized during
the test) and found a redox system with a mid-peak potential of
0.0 V vs. SHE. Since those cells were washed before the cyclic vol-
tammetry tests, the observed redox system likely does not result
from soluble mediators, but instead from membrane bound redox
components, such as c-type cytochromes. Electron microscopy
analysis of the electrodes from reactors S and SL showed an open
monolayer of S. oneidensis cells, instead of a complex biofilm,
regardless of the presence of L. lactis (Fig. S2). The latter microbe
grew almost exclusively in a planktonic state at high densities with
only some cells (cocci) attached on the outside of the monolayer of
S. oneidensis for reactors SL (Fig. S2). The combination of a thin
and open biofilm structure of primarily S. oneidensis for the
co-culture experiments with continuous-flow reactor operating
conditions at short hydraulic retention times of 5–7 h, assured an
electron-transfer mechanism that was dominated by direct elec-
tron transfer. Our data shows that a mediated electron-transfer
mechanism was not important, likely because the soluble redox
shuttles were washed out and never reached high enough concen-
trations. This not only explains why we did not measure mediator
activity, but also eliminated redox mediators from S. oneidensis
that possibly could have been used by L. lactis to respire with the
anode. Therefore, our experimental design is pertinent for BESs
that are dominated by the direct-electron-transfer mechanism of
exoelectron transfer to the anode.
3.4. Gene expression of S. oneidensis in pure- and in co-culture is
indifferent
To evaluate not only the experimental changes, but also molec-
ular physiological changes, we performed gene expression analysis
on 3949 S. oneidensis genes with Affymetrix gene chips (after data
normalization and exclusion of chip miss-reads from a total of
4077 S. oneidensis gene probes on the chip) for the S. oneidensis
pure-culture (S; n = 2) and co-culture electrode biofilms (SL;
n = 2) from the BES electrodes (for trials 1 and 2). The statistical
analysis was performed using the R package lemma (Laplace
approximated EM Microarray Analysis), and no statistically signif-
icant differences in expression between pure- or co-culture were
detected. This result is in accordance with our electrochemical
and metabolic data: the presence of L. lactis as a metabolic partner
organism had no significant effect on the physiological state or
activity of S. oneidensis. Supporting our statistical conclusions, a
heatmap visualizes only slight differences in the color pattern be-
tween reactor S and SL (Fig. 3). The similarity becomes even more
clear when the expression levels of S and SL are averaged (column
S + SL), or when the difference is shown (column S SL). The latter
heatmap should cancel out the expression levels of individual
genes if they are similarly expressed in both samples. The fact that
only colors in the immediate range around the general mean gene
expression level of the genome (mean{log
2
[signal inten-
sity]} = 8.56) are shown in this column, confirms visually that S.
oneidensis gene expression is statistically not different for both
samples.
3.5. S. oneidensis and L. lactis establish a neutral food-web relationship
in BES
Here, we showed that the performance of S. oneidensis in a BES
under controlled experimental conditions (lactate feeding) and in a
specific microbial food-web (with no food competitors being pres-
ent) was comparable. This suggests the possibility of designing
microbial communities to exploit specific food webs for enhanced
bioelectrochemical conversion of a wide spectrum of organic waste
products. For instance, while S. oneidensis or L. lactis alone cannot
generate electric current from glucose, they can do so in a co-cul-
ture. We found that the S. oneidensis biofilm structure (i.e.,
electrode coverage) and physiological state (i.e., gene expression)
Fig. 3. Heatmap showing the clustered gene expression levels of S. oneidensis in
pure-culture (column S, n = 2), co-culture (column SL, n = 2), as the mean of pure-
and co-culture (column S + SL, n = 4), and as the normalized difference of the two
culture tests (column [S SL], normalized by the mean genomic expression, n = 4).
Individual genes in all columns are ordered by the mean expression level of S + SL.
The expression level represents the log
2
of the Affymetrix GeneChip fluorescence
signal intensity and is color-coded following the scheme on the left. The mean of
the gene specific mean-square-errors (m
g
) of the statistical expression analysis is
mean [m
g
] = 0.51. The smallest p-value was greater than 0.3, yielding no discoveries
at the selected 5% fdr level. The p-values are obtained by fitting a parametric model
to the data.
M.A. Rosenbaum et al. / Bioresource Technology 102 (2011) 2623–2628
2627
were the same for the pure-culture and the co-culture. The fermen-
tation reaction of L. lactis occurred in planktonic state and did not
influence the bioelectrochemical reaction of S. oneidensis, besides
the provision of lactate as the electron donor (i.e., a neutral rela-
tionship). Thus, the interaction between the two organisms oc-
curred only at the substrate level. Our work supports the
hypothesis that fermenting bacteria only provide intermediate
products to ARB, such as from the genera Shewanella and Geobacter,
in BESs that are dominated by direct electron transfer. Under such
conditions when external redox shuttles are absent, the very small
electric currents generated by fermenters are inconsequential. This
finding is in agreement with the results of Ren et al. (2007) for a co-
culture of G. sulfurreducens and C. cellulolyticum, but not with the
study of Read et al. (2010), who found a negative impact of the fer-
menter C. acetobutylicum and a positive, synergistic effect of the
fermenter E. faecium on the current production by S. oneidensis,
G. sulfurreducens, and P. aeruginosa. It is, thus, clear that the inter-
actions between ARB and fermenters cannot be simply predicted,
especially since some ARB only perform direct electron transfer,
some only interact with the electrode through redox mediators,
and some are capable of both. This opens many possible ways of
interaction between the ARB and fermenters. While substrate level
interactions between microorganisms in a mixed-culture BES are
important to uncover (such as in our study), the identification
and investigation of true synergistic fermenter–ARB relationships
will be much more essential for the improvement of BES
performance.
4. Conclusion
The understanding of metabolic network relationships in
mixed-culture BES, which can be of neutral, positively enhancing,
or negatively impeding kind, becomes an important step in the
advancement of BES development. Here, we studied the relation-
ship between the homolactic fermenter L. lactis, which converts
glucose to lactate, and the electricigen S. oneidensis, which converts
lactate into electric current at conditions that are pertinent to con-
ventional BES operation. With electrochemical, metabolic, and
gene expression analysis we determined that the two cultures
establish a pure food-based relationship, because the physiology
and electrochemical activity of S. oneidensis was similar regardless
of the presence of L. lactis.
Acknowledgements
Financial support for this work was provided through a specific
collaborative agreement between LTA and the Bioenergy Research
Unit, USDA, Agricultural Research Service, Peoria, IL and the Na-
tional Science Foundation through grant no. 0939882. D.S. and
Q.K.B. are partially supported by US Department of Energy grants
DE-FG02-07ER64388 and DE-FG02-07ER64483. We thank Pat
O’Bryan and Bruce Dien of the USDA-ARS, Peoria, IL, for their help
with the HPLC analysis, Mike Veith of Washington University in
St. Louis, St. Louis, MO, for his help with SEM imaging, and Gretta
Serres of the Marine Biological Laboratory, Woods Hole, MA for the
provision of the S. oneidensis gene annotation.
Appendix A. Supplementary data
Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.biortech.2010.10.033.
References
Bar, H.Y., Booth, J., Schifano, E.D., Wells, M.T., in press. Laplace approximated EM
microarray analysis: An empirical Bayes approach for comparative microarray
experiments. Stat. Sci.
Benjamini, Y., Hochberg, Y., 1995. Controlling the false discovery rate—a practical
and powerful approach to multiple testing. J. R. Stat. Soc. B 57, 499–517.
Biffinger, J.C., Byrd, J.N., Dudley, B.L., Ringeisen, B.R., 2008. Oxygen exposure
promotes fuel diversity for Shewanella oneidensis microbial fuel cells. Biosens.
Bioelectron. 23, 820–826.
Bretschger, O., Obraztsova, A., Sturm, C.A., Chang, I.S., Gorby, Y.A., Reed, S.B., Culley,
D.E., Reardon, C.L., Barua, S., Romine, M.F., Zhou, J., Beliaev, A.S., Bouhenni, R.,
Saffarini, D., Mansfeld, F., Kim, B.H., Fredrickson, J.K., Nealson, K.H., 2007.
Current production and metal oxide reduction by Shewanella oneidensis MR-1
wild type and mutants. Appl. Environ. Microbiol. 73, 7003–7012.
Cho, E.J., Ellington, A.D., 2007. Optimization of the biological component of a
bioelectrochemical cell. Bioelectrochemistry 70, 165–172.
Cocaign-Bousquet, M., Garrigues, C., Loubiere, P., Lindley, N.D., 1996. Physiology of
pyruvate metabolism in Lactococcus lactis. Antonie Van Leeuwenhoek 70, 253–
267.
Cury, J.A., Koo, H., 2007. Extraction and purification of total RNA from Streptococcus
mutans biofilms. Anal. Biochem. 365, 208–214.
Driscoll, M.E., 2008. Inference of a Genome-wide Regulatory Network for the Metal-
reducing Microbe Shewanella Oneidensis Mr-1 via a Compendium of
Microarrays. Boston University, Boston, MA.
Efstathiou, J.D., McKay, L.L., 1977. Inorganic salts resistance associated with a
lactose-fermenting plasmid in Streptococcus lactis. J. Bacteriol. 130, 257–265.
Ezeji, T.C., Qureshi, N., Blaschek, H.P., 2007. Bioproduction of butanol from biomass:
from genes to bioreactors. Curr. Opin. Biotechnol. 18, 220–227.
Faith, J.J., Hayete, B., Thaden, J.T., Mogno, I., Wierzbowski, J., Cottarel, G., Kasif, S.,
Collins, J.J., Gardner, T.S., 2007. Large-scale mapping and validation of
Escherichia coli transcriptional regulation from a compendium of expression
profiles. PLoS Biol. 5, e8.
Freguia, S., Masuda, M., Tsujimura, S., Kano, K., 2009. Lactococcus lactis catalyses
electricity generation at microbial fuel cell anodes via excretion of a soluble
quinone. Bioelectrochemistry 76, 14–18.
Gorby, Y.A., Yanina, S., McLean, J.S., Rosso, K.M., Moyles, D., Dohnalkova, A.,
Beveridge, T.J., Chang, I.S., Kim, B.H., Kim, K.S., Culley, D.E., Reed, S.B., Romine,
M.F., Saffarini, D.A., Hill, E.A., Shi, L., Elias, D.A., Kennedy, D.W., Pinchuk, G.,
Watanabe, K., Ishii, S., Logan, B., Nealson, K.H., Fredrickson, J.K., 2006.
Electrically conductive bacterial nanowires produced by Shewanella oneidensis
strain MR-1 and other microorganisms. Proc. Natl. Acad. Sci. USA 103, 11358–
11363.
Kim, B.-H., Kim, H.-J., Hyun, M.-S., Park, D.-H., 1999. Direct electrode reaction of
Fe(III)-reducing bacterium, Shewanella putrefaciens. J. Microbiol. Biotechnol. 9,
127–131.
Kim, H.J., Park, H.S., Hyun, M.S., Chang, I.S., Kim, M., Kim, B.H., 2002. A mediator-less
microbial fuel cell using a metal reducing bacterium, Shewanella putrefaciens.
Enzyme Microb. Technol. 30, 145–152.
Kim, G.T., Webster, G., Wimpenny, J.W.T., Kim, B.H., Kim, H.J., Weightman, A.J., 2006.
Bacterial community structure, compartmentalization and activity in a
microbial fuel cell. J. Appl. Microbiol., 698–710.
Malinauskas, A., 2008. Electrochemical study of riboflavin adsorbed on a graphite
electrode. Chemija 19, 1–3.
Marsili, E., Baron, D.B., Shikhare, I.D., Coursolle, D., Gralnick, J.A., Bond, D.R., 2008.
Shewanella secretes flavins that mediate extracellular electron transfer. Proc.
Natl. Acad. Sci. USA 105, 3968–3973.
Masuda, M., Freguia, S., Wang, Y.F., Tsujimura, S., Kano, K., 2010. Flavins contained
in yeast extract are exploited for anodic electron transfer by Lactococcus lactis.
Bioelectrochemistry 78, 173–175.
Meitl, L.A., Eggleston, C.M., Colberg, P.J.S., Khare, N., Reardon, C.L., Shi, L., 2009.
Electrochemical interaction of Shewanella oneidensis MR-1 and its outer
membrane cytochromes OmcA and MtrC with hematite electrodes. Geochim.
Cosmochim. Acta 73, 5292–5307.
Myers, J.M., Myers, C.R., 2001. Role for outer membrane cytochromes OmcA and
OmcB of Shewanella putrefaciens MR-1 in reduction of manganese dioxide. Appl.
Environ. Microbiol. 67, 260–269.
Myers, C.R., Nealson, K.H., 1988. Bacterial manganese reduction and growth with
manganese oxide as the sole electron acceptor. Science 240, 1319–1321.
Nealson, K.H., Belz, A., McKee, B., 2002. Breathing metals as a way of life: geobiology
in action. Antonie Van Leeuwenhoek 81, 215–222.
Read, S.T., Dutta, P., Bond, P.L., Keller, J., Rabaey, K., 2010. Initial development and
structure of biofilms on microbial fuel cell anodes. BMC Microbiol. 10, 98.
Ren, Z., Ward, T.E., Regan, J.M., 2007. Electricity production from cellulose in a
microbial fuel cell using a defined binary culture. Environ. Sci. Technol. 41,
4781–4786.
Ringeisen, B.R., Ray, R., Little, B., 2007. A miniature microbial fuel cell operating with
an aerobic anode chamber. J. Power Sources 165, 591–597.
Rosenbaum, M., Cotta, M.A., Angenent, L.T., 2010. Aerated
Shewanella oneidensis in a
continuously-fed bioelectrochemical system for power and hydrogen
production. Biotechnol. Bioeng. 105, 880–888.
von Canstein, H., Ogawa, J., Shimizu, S., Lloyd, J.R., 2008. Secretion of flavins by
Shewanella species and their role in extracellular electron transfer. Appl.
Environ. Microbiol. 74, 615–623.
2628 M.A. Rosenbaum et al. / Bioresource Technology 102 (2011) 2623–2628
    • "For example, Clostridium celluloyticum (cellulose degrader) and Geobacter sulfurreducens (electrochemically active bacteria) were applied to generate electricity from cellulose [16]. In other studies, a co-culture MFC using an exoelectrogen (MR-1) and a glucose fermenter, Lactococcus lactis (glucose to lactate) and Escherichia coli (glucose to formate), were reported [18, 24]. Only a few studies have attempted to convert glycerol to electricity using a non-defined coculture system (i.e., anaerobic digester sludge) or pureculture system with low electrochemically active bacteria [4, 14, 15] . "
    [Show abstract] [Hide abstract] ABSTRACT: Glycerol is an attractive feedstock for bioenergy and bioconversion processes but its use in microbial fuel cells (MFCs) for electrical energy recovery has not been investigated extensively. This study compared the glycerol uptake and electricity generation of a co-culture of Shewanella oneidensis MR-1 and Klebsiella pneumonia J2B in a MFC with that of a single species inoculated counterpart. Glycerol was metabolized successfully in the co-culture MFC (MFC-J&M) with simultaneous electricity production but it was not utilized in the MR-1 only MFC (MFC-M). A current density of 10 mA/m2 was obtained while acidic byproducts (lactate and acetate) were consumed in the co-culture MFC, whereas they are accumulated in the J2B-only MFC (MFC-J). MR-1 was distributed mainly on the electrode in MFC-J&M, whereas most of the J2B was observed in the suspension in the MFC-J reactor, indicating that the co-culture of both strains provides an ecological driving force for glycerol utilization using the electrode as an electron acceptor. This suggests that a co-culture MFC can be applied to electrical energy recovery from glycerol, which was previously known as a refractory substrate in a bioelectrochemical system.
    Full-text · Article · Jul 2016
    • "There is a body of evidence (Busalmen et al., 2008; Rosenbaum et al., 2011; Feng et al., 2013 ) demonstrating that the applied potential of 0.2 V to the anode enables the generation of bioelectricity from organic substrate oxidation catalyzed by a variety of EAB. To determine the anodic current, the chronoamperometric current density normalized to the geometric anode surface area over time was recorded in the BERs with and without acetate addition. "
    [Show abstract] [Hide abstract] ABSTRACT: The application of a low-voltage electric field as an electron donor or acceptor to promote the bioremediation of chlorinated organic compounds represents a promising technology meeting the demand of developing an efficient and cost-effective strategy for in situ treatment of PCB-contaminated sediments. Here, we reported that bioanode stimulation with an anodic potential markedly enhanced dechlorination of 2,3,4,5-tetrachlorobiphenyl (PCB 61) contained in the sediment at an electronic waste recycling site of Qingyuan, Guangdong, China. The 110-day incubation of the bioanode with a potential poised at 0.2 V relative to saturated calomel electrode enabled 58% transformation of the total PCB 61 at the initial concentration of 100 μmol kg(-1), while only 23% was reduced in the open-circuit reference experiment. The introduction of acetate to the bioelectrochemical reactor (BER) further improved PCB 61 transformation to 82%. Analysis of the bacterial composition showed significant community shifts in response to variations in treatment. At phylum level, the bioanode stimulation resulted in substantially increased abundance of Actinobacteria, Bacteroidetes, and Chloroflexi either capable of PCB dechlorination, or detected in the PCB-contaminated environment. At genus level, the BER contained two types of microorganisms: electrochemically active bacteria (EAB) represented by Geobacter, Ignavibacterium, and Dysgonomonas, and dechlorinating bacteria including Hydrogenophaga, Alcanivorax, Sedimentibacter, Dehalogenimonas, Comamonas and Vibrio. These results suggest that the presence of EAB can promote the population of dechlorinating bacteria which are responsible for PCB 61 transformation.
    Full-text · Article · Apr 2016
    • "Minimal media meet better the environmental conditions of surface-waters than complete media, reason why they are often used in environmental microbiology to grow wild-type micro- organisms [53]. Minimal media have been generally employed in MFCs with pure cultures of Shewanella oneidensis, Lactococcus lactis, Enterobacter cloacae and Bacillus subtilis [54][55][56][57][58][59][60][61][62][63][64] . Nevertheless , mixed culture bioanodes have predominantly derived from microbial communities enriched from wastewaters, and reports on their direct adaptation to minimal media are scarce [65]. "
    [Show abstract] [Hide abstract] ABSTRACT: One of the greatest challenges in the development of Microbial Electrochemical Technologies (METs) is the achievement of efficient bioanodes, which not only can operate for long term but which can also be effectively and promptly regenerated for the sustained and successive production of electricity out of waste organics contained in aqueous streams. Simple strategies that facilitate the engineering of these systems are then pursued. Sustainable electricity generation was here achieved using electrochemically-active marine biofilms, which reached up to 6.8 A m-2 in the best case. These biofilms showed deteriorated current generation after successive transfers in fresh natural media. The electricity-generation functionality of these marine biofilms was recuperated after their relocation into synthetic minimal media (i.e. up to about 3.8 A m-2 after a decay down to about 1-2 A m-2). Upon this relocation, the overall electrochemical mechanisms were preserved. Fluctuating nutrient stress intensified the effect of minimal media on current generation. The change from natural to minimal media showed an important impact on the selection and adaptation of microbial communities, characterized by CE-SSCP profiles; yet, robust bioanodes in which some microbial species were preserved were obtained in synthetic minimal media, these being sufficient for a reproducible electrochemical functionality. The systematic cycling between natural media and minimal media, regarded as periodic stress conditioning, is therefore proposed as a convenient strategy to boost current generation in robust electrochemically-active marine bioanodes.
    Full-text · Article · Apr 2016
Show more