Content uploaded by Romana Marinšek Logar
Author content
All content in this area was uploaded by Romana Marinšek Logar on Dec 01, 2017
Content may be subject to copyright.
Biogas production from brewery spent grain enhanced
by bioaugmentation with hydrolytic anaerobic bacteria
Maša C
ˇater
a
, Lijana Fanedl
a
, Špela Malovrh
b
, Romana Marinšek Logar
a,
⇑
a
Division of Microbiology and Microbial Biotechnology, Department of Animal Science, Biotechnical Faculty, University of Ljubljana, Groblje 3, 1230 Domz
ˇale, Slovenia
b
Division of Animal Breeding Sciences, Department of Animal Science, Biotechnical Faculty, University of Ljubljana, Groblje 3, 1230 Domz
ˇale, Slovenia
highlights
The impact of bioaugmentation on methane production from spent grain was examined.
Methane production elevated for 17.8% with P. xylanivorans Mz5
T
bioaugmentation.
Bioaugmentation enhances the hydrolysis of brewery spent grain.
Changes in bacterial and archaeal communities were detected during biogas process.
Bioaugmentation enables successful biogas production from brewery spent grain.
article info
Article history:
Received 10 January 2015
Received in revised form 4 March 2015
Accepted 5 March 2015
Available online 20 March 2015
Keywords:
Brewery spent grain
Lignocellulose
Biogas production
Bioaugmentation
Hydrolytic bacteria
abstract
Lignocellulosic substrates are widely available but not easily applied in biogas production due to their
poor anaerobic degradation. The effect of bioaugmentation by anaerobic hydrolytic bacteria on biogas
production was determined by the biochemical methane potential assay. Microbial biomass from full
scale upflow anaerobic sludge blanket reactor treating brewery wastewater was a source of active
microorganisms and brewery spent grain a model lignocellulosic substrate. Ruminococcus flavefaciens
007C, Pseudobutyrivibrio xylanivorans Mz5
T
,Fibrobacter succinogenes S85 and Clostridium cellulovorans
as pure and mixed cultures were used to enhance the lignocellulose degradation and elevate the biogas
production. P. xylanivorans Mz5
T
was the most successful in elevating methane production (+17.8%),
followed by the coculture of P. xylanivorans Mz5
T
and F. succinogenes S85 (+6.9%) and the coculture of
C. cellulovorans and F. succinogenes S85 (+4.9%). Changes in microbial community structure were detected
by fingerprinting techniques.
Ó2015 Elsevier Ltd. All rights reserved.
1. Introduction
To reduce organic wastes, greenhouse gas emissions and to pro-
duce methane as an alternative energy source, anaerobic digestion
of continuously generated organic waste is an excellent choice and
an environmentally friendly technology. Energy crops, harvesting
residues, manure, household waste, wood industry waste and
brewery waste are examples of potentially useful lignocellulosic
substrates (LCS). Lignocellulose is the most abundant renewable
resource in the world, widely available and cheap, but its
hydrolysis is usually incomplete due to its complex structure.
Cellulose and hemicellulose solubilization is the rate-limiting
step in the anaerobic digestion of LCS used for biogas production.
Lignin degradation is poor, toxic digestion intermediates are
produced and crystalline cellulose is very structurally stable.
Therefore the biogas production is frequently not efficient enough
without additional substrate treatment before or during the biogas
production process. Physico-chemical pretreatments, which
proved to be effective in particle size reduction, crystalline
structure disruption and enhancing anaerobic digestion of LCS
(Hendriks and Zeeman, 2009), demand high energy inputs and
application of aggressive chemicals and are environmentally
unfriendly and economically unsuitable. There have been many
attempts to enhance lignocellulose biodegradability by biological
means where anaerobic bacteria, wood degrading fungi, actinomy-
cetes or enzymes that selectively degrade lignin, cellulose and
hemicellulose are added (C
ˇater et al., 2014).
In cellulolytic rumen bacteria, highly active cellulolytic and
hemicellulolytic enzymes are combined in extracellular multien-
zyme complexes, cellulosomes (Fontes and Gilbert, 2010). Rumen
bacteria are often used as an inoculum for biogas reactors for
http://dx.doi.org/10.1016/j.biortech.2015.03.029
0960-8524/Ó2015 Elsevier Ltd. All rights reserved.
⇑
Corresponding author. Tel./fax: +386 13203 849.
E-mail address: romana.marinsek@bf.uni-lj.si (R. Marinšek Logar).
Bioresource Technology 186 (2015) 261–269
Contents lists available at ScienceDirect
Bioresource Technology
journal homepage: www.elsevier.com/locate/biortech
increased LCS hydrolysis and fatty acids production (Yue et al.,
2013). In this study rumen bacteria Pseudobutyrivibrio xylanivorans
Mz5
T
,Fibrobacter succinogenes S85, Clostridium cellulovorans and
Ruminococcus flavefaciens 007C in addition to biogas digester
original microbiota were used for LCS hydrolysis enhancement in
biogas production from brewery spent grain (BSG). Although BSG
contains a relatively high content of proteins which supports the
growth of anaerobic microbiota, it is similarly resistant to anaero-
bic degradation as other woody biomass due to the high content of
lignin, cellulose and hemicelluloses. BSG was chosen as a model
lignocellulosic substrate in this study due to its wide availability
as brewery waste and its typical LCS characteristics.
2. Methods
2.1. Lignocellulosic substrate
BSG from Brewery Laško (Slovenia) was used as a substrate for
biogas production. Mussatto and Roberto (2005) and Russ et al.
(2005) reported total solids (TTS) of BSG contain approximately
17–25% cellulose, 28% arabinoxylan, 8–28% lignin, 15–23%
proteins, 4% ash and lipids in traces. According to the data of
Brewery Laško, used BSG contains approximately 26.2% of proteins,
5.8% of starch and 29.8% of lignin in TTS (data for cellulose and
hemicelluloses not available).
2.2. Microbial biomass
Anaerobic microbial biomass from up flow anaerobic sludge
blanket (UASB) bioreactor treating brewery waste water from
Brewery Laško (Slovenia) was used as a source of active
microorganisms. UASB bioreactor operated at pH 6.5, 35 °C,
average hydraulic retention time was 15.8 h, average flow was
1150 m
3
/day and residence time was 40–85 h.
2.3. Bacterial strains used for bioaugmentation
P. xylanivorans Mz5
T
(DSM 14809) was taken from our lab-
oratory collection as a successful xylan degrader (Zorec et al.,
2001). This Gram-negative anaerobic rumen bacterium synthesizes
many hydrolytic enzymes, possesing cellulolytic, xylanolytic, amy-
lolytic and pectinolytic activity. Its xylanolytic activity is 1.7 times
higher than of any other rumen xylanolytic bacterium (C
ˇepeljnik
et al., 2003; Kopec
ˇny et al., 2003). R. flavefaciens 007C and F. suc-
cinogenes S85 were both obtained from the Rowett Research
Institute in Scotland (Stewart and Flint, 1989; Stewart et al.,
1990). R. flavefaciens 007C is an anaerobic Gram-positive rumen
bacterium and possesses cellulosomes with cellulolytic and
hemicellulolytic activity (Krause et al., 2003; Fontes and Gilbert,
2010). F. succinogenes S85 is one of the most efficient cellulolytic
anaerobic Gram-negative bacteria found in the rumen
(Montgomery et al., 1988). Unlike other cellulolytic bacteria, it
does not use cellulosome for degrading cellulose (Suen et al.,
2011). C. cellulovorans 743B, which was purchased from DSMZ
collection (DSM 3052), is a Gram-negative anaerobic bacterium
and was first isolated from a batch methanogenic fermentation
sludge treating finely chopped hybrid poplar wood (Sleat et al.,
1984). Its cellulosomes are responsible for high cellulolytic and
xylanolytic activity (Doi and Tamaru, 2001).
2.4. Media and culture conditions
The bacteria were cultured in inducible M2 medium (Hobson,
1969) and incubated at 37 °C for 19–24 h to induce hydrolytic
enzyme activities. Inducible M2 medium contained xylan from
birchwood (0.3%) and carboxymethylcellulose (CMC) (0.3%)
instead of regular sugars, except for C. cellulovorans, which did
not grow successfully on CMC-cellulose and demanded xylan from
birchwood (0.25%), Avicel cellulose (0.25%) and the addition of
cellobiose (0.1%). When the bacteria reached optical density
(654 nm) 0.5 ± 0.05, they were transferred anaerobically into 1 L
batch reactors to bioaugment the anaerobic digestion of BSG.
2.5. Enzyme activity analysis
Xylanase and CMC/Avicel-cellulase activities of the hydrolytic
bacteria with induced enzymatic activity were measured with
the reducing sugars method (Lever, 1977); concentrations of
microbial proteins were determined by Lowry protein assay
(Lowry et al., 1951). Enzymatic activities were expressed as
nanomoles of reducing sugars per mg of proteins per minute.
2.6. Biochemical methane potential (BMP) assay
BMP assay was used to examine the effect of bioaugmentation
on the biogas production. Granular microbial biomass was
disrupted with Polytron System PT 1200E (Kinematica
AG, Switzerland) with the highest power for 10 min to suspend
the microorganisms into the medium. Distilled water was added
to BSG (2:1) for easier homogenization and disrupted for 5 min
to reduce the BSG particles size. Hydrolytic bacteria were added
into the bioreactors at the start of the BMP assay and their volume
was 5% of the total volume of experimental BMP mixtures. The
appropriate loading of the bioreactors was determined by measur-
ing TTS and total volatile solids (TVS) of microbial biomass and
chemical oxygen demand (COD) of BSG (American Public Health
Association et al., 2006). The microbial biomass concentration
was 2 g TVS/L and the BSG loading was 0.413 g TVS/L
(0.484 g COD/L).
Phosphate buffer and anoxic tap water were prepared for all
mixtures. All experiments were conducted at least twice in lab-
oratory batch bioreactors (1 L) at 37 °C and 120 rpm in dark for
30 days with 2 or more replicate experimental mixtures. While
mixing ingredients, anaerobic conditions were maintained by
sparging gaseous nitrogen. Sole microbial biomass served as nega-
tive control for residual methanogenic activity. Mixtures with non-
inoculated inducible M2 medium were tested to measure the med-
ium’s nutrients background effect. Bioaugmentation of selected
hydrolytic bacteria was tested with individual bacteria or coculture
additions into the bioreactors at the start of the BMP assays (Fig. 1).
Bioaugmentation of autoclaved hydrolytic bacteria was also tested
to measure the effect of the dead cells COD on biogas production. R.
flavefaciens 007C was excluded from bioaugmentation tests further
on as it did not elevate methane production from BSG.
Concentrations of methane and carbon dioxide and biogas vol-
ume were monitored 8 times during the whole investigation per-
iod of 30 days and short-chain fatty acids (SCFAs) concentrations
were monitored 4 times, while pH and COD were measured on
day 0 and 30. The volume of the newly produced biogas was
measured manually with syringe connected to a U-shaped water
column and released each time. The percentage contents of
nitrogen, methane and carbon dioxide were monitored by gas
chromatograph (GC) equipped with thermal conductivity detector
(TCD) (Shimadzu GC14A, Japan) and PORAPAK-Q column (Agilent)
(helium used as a carrier gas). The quantification was performed by
absolute calibration method. GC analysis conditions were: injector
T=50°C, column T=25°C, detector T=80°C, current = 60 mA. The
resulting methane yields were normalized to standard conditions
as described by Hansen et al. (2004). SCFAs were extracted by dou-
ble diether extraction (Holdeman et al., 1977) and determined by
GC equipped with flame ionization detector (FID) (Shimadzu
262 M. C
ˇater et al. / Bioresource Technology 186 (2015) 261–269
GC14A, Japan) and DB-WAXetr column (Agilent). GC analysis con-
ditions were: injector T= 160 °C, detector T= 210 °C, initial
T=90°C, initial time = 4 min, final T= 160 °C, final time = 2 min, T
program = 15 °C/min. Helium was used as a carrier gas and quan-
tification was performed by an internal standard method (crotonic
acid, 100 mg/ml).
2.7. Analyses of microbial community shifts
Terminal restriction fragment length polymorphism (T-RFLP)
was applied to study the changes in the microbial composition
on day 0 and day 30. T-RFLP is one of the most powerful
fingerprinting techniques, because it allows high sample
throughput, precise fragment length determination and provides
more comprehensive examination of the microbial community
composition. Additional analysis of denaturing gradient gel
electrophoresis (DGGE) profiles was used only to sharpen up the
understanding of bioaugmentation. Dynamics of the bacterial
community in the bioaugmented bioreactors with P. xylanivorans
Mz5
T
or C. cellulovorans individually were tracked by DGGE on
day 3, 6 and 10 to gain more knowledge about their persistence
in the bioreactors during the first days of BMP assays.
Samples were first centrifuged at 3200 rpm for 5 min; the
precipitated biomass was dried at 45 °C over night and frozen for
further analyses. Total genomic DNA was extracted from 150 mg
of dried samples using PowerSoil DNA Isolation Kit (MO-BIO
Laboratories, Inc., Carlsbad, USA) according to the manufacturer’s
instructions. DNA from pure cultures of bacteria for
bioaugmentation was extracted separately by a common method
using proteinase K, CTAB/NaCl and phenol/chloroform.
2.8. T-RFLP
Primers used in PCR amplification of bacterial 16S rRNA gene
were fD1 (Weisburg et al., 1991) and 926r (Muyzer et al., 1993);
and 109f (Großkopf et al., 1998) and 915r (Peng et al., 2008) for
archaeal 16S rRNA gene. Both forward primers were labeled at
the 5
0
end with the dye 6-carboxyfluorescein. The PCR protocol
was as follows: initial denaturation at 95 °C for 5 min, 30 cycles
of denaturation at 95 °C for 30 s, annealing at 55 °C for 30 s
(bacteria) or at 52 °C for 45 s (archaea) and elongation at 72 °C
for 1 min, followed by a final extension at 72 °C for 7 min.
Amplicons were purified using High Pure PCR Product
Purification Kit (Roche, Switzerland) and quantified using a
NanoVue-Plus spectrophotometer (GE Healthcare, UK). Screening
for appropriate restriction endonuclease was made using HhaI
(Thermo Fisher Scientific Inc., USA), BsuRI (HaeIII) (Fermentas,
Canada) and MspI (Thermo Fisher Scientific Inc., USA). BsuRI was
chosen for further analysis, producing the highest numbers and
best size distribution of the terminal restriction fragments
(T-RFs). Amounts of 300 ng of purified PCR products were digested
at 37 °C for 16 h. The restriction enzyme was then inactivated by
incubation at 85 °C for 20 min. DNA fragments were purified with
the same kit as mentioned before and 5
l
l was mixed with 8
l
l
HiDi formamide and 0,5
l
l GeneScan ROX-500 internal size
standard (Applied Biosystems, USA), and denaturated for 5 min at
95 °C. T-RFs were separated by capillary gel electrophoresis on
an ABI PRISM 3130xl Genetic Analyser (Applied Biosystems, USA)
and electropherograms were analyzed by BioNumerics software
5.1 (Applied Maths, Belgium). Pearson’s correlation coefficient
with UPGMA clustering method (Unweighted Pair-Group Method
with Arithmetic Mean) was used to derive general similarity of
community profiles. The significance of clusters was determined
using the cluster cut-off tool in BioNumerics.
2.9. DGGE
For PCR amplification of bacterial DNA the primers used were
HDA1 with GC clamps and HDA2 (Walter et al., 2000). The PCR pro-
tocol was as follows: initial denaturation at 95 °C for 2 min, 30
Fig. 1. Experimental set-up for BMP assays. NC – negative control, Stand. – standard bioreactor, BSG – brewery spent grain, medium 1 – medium containing xylan and CMC
cellulose, medium 2 – medium containing xylan, Avicel cellulose and cellobiose, M – P. xylanivorans Mz5
T
,C–C. cellulovorans,F–F. succinogenes S85, R – R. flavefaciens 007C.
M. C
ˇater et al. / Bioresource Technology 186 (2015) 261–269 263
cycles of denaturation at 95 °C for 30 s, annealing at 58 °C for 30 s
and elongation at 72 °C for 30 s, followed by extension at 72 °C. All
PCR products were checked in a 1% agarose gel electrophoresis and
then separated on acrylamide gel with 30–65% denaturing gradient
(100% corresponded to 7 M urea and 40% formamide (vol/vol)).
DGGE was performed on a DCode universal mutation detection
system (Bio-Rad, USA) in 1TAE buffer at 60 °C for 16 h 30 min
at 75 V, 70 mA and 5 W. The gels were stained in a SYBR Gold
Nucleic Acid Gel Stain (Thermo Fisher Scientific Inc., USA) solution
for 45 min. Under UV light fluorescing bands were excised by
Chemi Genius (Syngene, UK) and visualized with computer
program Gene Snap (Syngene, UK). Gels were normalized to the
outside and middle lane ladder, which was generated in our lab-
oratory with equal-volume mixtures of PCR products from 16S
rRNA gene of 10 pure cultures, and analyzed by Bio Numerics soft-
ware 5.1.
2.10. Statistical analysis
To determine the significance of differences in methane
production per g COD
initial
of two noninoculated media with BSG
and different combinations of added hydrolytic bacteria, the non-
linear mixed models methodology was used (Ratkowsky, 1990).
The procedure NLMIXED with the Newton–Raphson ridge
optimization method in statistical package SAS/STAT (SAS Inst.
Inc., 2008) was applied. Due to preliminary analyses and practi-
cally no lag phase in methane production, the negative exponential
curve was chosen to describe cumulative methane production per
g COD
initial
over time. The following statistical model was used:
y
ijk
¼K
i
ð1e
r
i
t
ijk
Þþu
ij
þ
e
ijk
, where y
ijk
is observation at time
t
ijk
,K
i
and r
i
are curve parameters – final production and produc-
tion rate, respectively, e
()
is exponential function, u
ij
is random
sample effect which covers individual variation of four repetitions,
while
e
ijk
is residual term. Normal distribution was assumed for
random sample effect and residuals. Maximum likelihood
estimates of the parameters for final production and production
rate were pairwise compared between experimental mixtures.
3. Results and discussion
3.1. Cellulolytic and xylanolytic enzyme activities of applied bacteria
R. flavefaciens 007C and F. succinogenes S85 exhibited the
highest xylanolytic activity among the four tested hydrolytic bac-
teria (78.6 and 77.6 nmol/mg/min, respectively) (Table 1). The
highest CMC-cellulolytic activity was expressed by R. flavefaciens
007C (4.05 nmol/mg/min). C. cellulovorans expressed a low
xylanolytic and CMC-cellulolytic activity, but was more successful
in degrading Avicel cellulose (21.6 nmol/mg/min). Comparing
hydrolytic bacteria used, the protein concentrations were similar
what indicates that a comparable quantity of bacterial biomass
was added to BMP assay in all experimental sets.
3.2. Methane production enhancement
The microbial biomass contained approximately 63 g/L of TTS
and 55 g/L of TVS, while BSG contained approximately 78 g/L of
TTS, 76 g/L of TVS and 73 g
COD
/L (disrupted mixture of 50 g BSG
and 90 ml distilled water). The highest elevation of net methane
production per g COD
initial
has been detected in the experimental
mixtures bioaugmented with P. xylanivorans Mz5
T
(17.8%),fol-
lowed by the coculture of P. xylanivorans Mz5
T
and F. succinogenes
S85 (6.9%) and the coculture of C. cellulovorans and F. succinogenes
S85 (4.2%) (Table 2). Bioaugmentation with C. cellulovorans
individually resulted in 3.9% and with F. succinogenes S85 individu-
ally in 1.9% elevation of methane production per g COD
initial
.
Bacteria from genus Clostridium have been proven before to
increase the BMP of LCS as Peng et al. (2014) increased methane
production by bioaugmenting wheat straw with C. cellulolyticum
and Lü et al. (2013) enhanced methane production from
microalgae by 17–24% with Clostridium thermocellum. In contrary
to F. succinogenes S85 and R. flavefaciens 007C, C. cellulovorans
and P. xylanivorans Mz5
T
both produce hydrogen, apart from
carbon dioxide and other metabolites (Sleat et al., 1984; Stewart
and Flint, 1989; Kopec
ˇny et al., 2003; Fontes and Gilbert, 2010),
Table 1
Bacterial cellulolytic and xylanolytic enzyme activities at OD
654nm
= 0.5.
Bacteria/enzyme activity Xylanase activity
(nmol/mg/min)
CMC-cellulase activity
(nmol/mg/min)
Avicel-cellulase activity
(nmol/mg/min)
Protein concentration
(mg/ml)
R. flavefaciens 007C 78.6 4.05 – 6.4
P. xylanivorans Mz5
T
71.8 1.6 – 6.2
F. succinogenes S85 77.6 0.9 – 6.4
C. cellulovorans 1.0 1.7 21.6 6.6
Table 2
BMP assay results showing the average elevation values of net methane production per g COD
initial
for individually added bacteria and bacteria added in cocultures; and
parameter estimates for final net methane production per g COD
initial
and methane production rate with standard errors (SEE) for experimental mixtures. Legend: BSG – brewery
spent grain, medium 1 – medium containing xylan and CMC cellulose, medium 2 – medium containing xylan, Avicel cellulose and cellobiose, M – P. xylanivorans Mz5
T
,C–C.
cellulovorans,F–F. succinogenes S85, R – R. flavefaciens 007C.
Bioreactor Average net methane production
elevation per g COD
initial
(%)
Final net methane production
per g COD
initial
in ml (K)
SEE Methane production
rate in ml/day (r)
SEE
BSG + medium 1 – 221.84 7.839 0.2262 0.02143
BSG + medium 2 – 242.81 7.850 0.2184 0.01865
P. xylanivorans Mz5
T
17.8 261.31 7.875 0.2084 0.01717
F. succinogenes S85 1.9 226.20 7.862 0.2132 0.02010
C. cellulovorans 3.9 252.36 7.845 0.2221 0.01917
R. flavefaciens 007C 5.6 – – – –
M+R 4.2 – – – –
M + F 6.9 237.16 7.939 0.1901 0.01702
M+C 2.5 226.45 7.874 0.2081 0.01956
C + F 4.2 242.01 7.843 0.2230 0.01999
264 M. C
ˇater et al. / Bioresource Technology 186 (2015) 261–269
what was proven by gas chromatography in our study. This may be
the reason for accelerated methanogenesis and elevated methane
production. Bioaugmentation with R. flavefaciens 007C individually
or in coculture with P. xylanivorans Mz5
T
resulted in reduced
methane production (5.6% and 4.2%). Adding a coculture of
P. xylanivorans Mz5
T
and C. cellulovorans also decreased methane
production for 2.5% which may indicate to an unsuccessful cohab-
itation of these two bacterial strains. BMP assay with autoclaved
bioaugmented hydrolytic bacteria showed that COD from the
added bacterial cells does not accelerate methane production from
the BSG. These results prove that only active bioaugmented hydro-
lytic bacteria were responsible for methane production elevation
due to enhanced BSG hydrolysis.
The gas chromatography results of biogas composition revealed
that on day 30 of the BMP assay, the highest methane contents
were reached in mixtures with C. cellulovorans individually and
P. xylanivorans Mz5
T
individually, where the methane content in
biogas was approximately 9% higher in comparison to the mixtures
where only BSG was fermented and about 2% higher than in mix-
tures with BSG and non-inoculated medium added (Fig. 2).
pH values on day 0 ranged from 6.7 to 7.2 and on day 30 from
7.1 to 7.4. SCFAs concentration monitoring showed that SCFAs
have not accumulated and were continuously degraded further
during the experiment by acetogenic bacteria and methanogenic
archaea. In all experimental mixtures acetic acid was the most
abundant, other SCFAs were detected in minimal concentrations.
Acetic acid concentrations did not overcome 0.13 g/L and propionic
acid concentrations were below 0.06 g/L. Although absolutely low,
the highest concentrations of acetic acid through the whole experi-
ment were detected in the bioaugmented mixtures containing
P. xylanivorans Mz5
T
individually which accelerated methane
production most successfully. The low concentrations of SCFAs
correlate with optimal pH during BMP assays and to a well-run
anaerobic digestion under given conditions.
Our results clearly show that the addition of hydrolytic bacteria
improved BSG hydrolysis, which resulted in more efficient acidoge-
nesis, acetogenesis and finally, accelerated methanogenesis.
3.3. Negative exponential curve fits net methane production per g
COD
initial
Final methane production and production rates with standard
errors for experimental mixtures were calculated (Table 2). The
highest final net methane production per g COD
initial
was found
in experimental mixtures containing P. xylanivorans Mz5
T
(261 ml), followed by C. cellulovorans (252 ml), while there were
no outstanding results found at the methane production rate.
Statistically significant differences for final net methane pro-
duction per g COD
initial
(p-value < 0.05) were found between
experimental mixtures containing P. xylanivorans Mz5
T
and experi-
mental mixture containing BSG and noninoculated medium
(Table 3). According to this statistical analysis it is clear that
bioaugmenting the biogas reactor with P. xylanivorans Mz5
T
pro-
duced the highest net methane volumes per g COD
initial
. Effects of
P. xylanivorans Mz5
T
on methane production are also significantly
different from those of F. succinogenes S85, the coculture of
P. xylanivorans Mz5
T
and F. succinogenes S85 and the coculture of
P. xylanivorans Mz5
T
and C. cellulovorans (Fig. 3). The physiology
of P. xylanivorans Mz5
T
may be the reason for its successfulness
in accelerating the methane production from BSG. It has been pre-
viously found that short-time exposure to air does not destroy the
ability of P. xylanivorans Mz5
T
to recover under anaerobic condi-
tions which gives it a great advantage to survive in comparison
to the other strictly anaerobic hydrolytic bacteria used in the
experiment. Additionally, P. xylanivorans Mz5
T
has been proven
to exert high xylanolytic activity (Kopec
ˇny et al., 2003) which
obviously helps in BSG hydrolysis and results in accelerated
methane production. P. xylanivorans Mz5
T
is an active bacteriocins
producer (C
ˇepeljnik et al., 2003) and this ability might help it to
persist more actively in the mixed methanogenic community and
express its hydrolytic activity following the addition into the
BMP bioreactors.
3.4. Changes in bacterial and archaeal community detected by T-RFLP
Adding microorganisms into the biogas reactor can lead to
bacterial and archaeal microbial community shifts which can be
55
56
57
58
59
60
61
62
Average methane content
(%) on day 30 of BMP assay
Fig. 2. Average methane content (%) in experimental mixtures on day 30 of BMP
assay. BSG – brewery spent grain, medium 1 – medium containing xylan and CMC
cellulose, medium 2 – medium containing xylan, Avicel cellulose and cellobiose, M
–P. xylanivorans Mz5
T
,C–C. cellulovorans,F–F. succinogenes S85.
Table 3
Significant differences with standard errors (SEE) between experimental mixtures for
final net methane production per g COD
initial
(K). Legend: BSG – brewery spent grain,
medium 1 – medium containing xylan and CMC cellulose, M – P. xylanivorans Mz5
T
,C
–C. cellulovorans,F–F. succinogenes S85.
Difference Difference in final net
methane production per
g COD
initial
in ml (K)
SEE p-Value
M vs. BSG + medium 1 39.47 11.11 0.0004
M vs. F 35.11 11.13 0.0018
M vs. M + F 24.15 11.18 0.0316
M vs. M + C 34.86 11.14 0.0019
C vs. F 26.16 11.11 0.0192
C vs. M + C 25.90 11.12 0.0205
Fig. 3. Estimated curves for net methane production per g COD
initial
over time.
Estimated curves are shown only for experimental mixtures which P. xylanivorans
Mz5
T
significantly differs from. BSG – brewery spent grain, medium 1 – medium
containing xylan and CMC cellulose, M – P. xylanivorans Mz5
T
,C–C. cellulovorans,F
–F. succinogenes S85.
M. C
ˇater et al. / Bioresource Technology 186 (2015) 261–269 265
beneficial for biogas production or they can have no impact or
negative impact. Bacterial community structure analysis showed
samples from day 0 and day 30 separate into two clusters which
indicates microbial community shift during BMP assay (Fig. 4). At
the end of the experiment the divergence of original microbial bio-
mass with no additions was approximately 30%, of biomass with
BSG and non-inoculated medium approximately 36% and in experi-
mental mixtures bioaugmented with F. succinogenes S85 25% and
with C. cellulovorans 29%. More prominent changes in bacterial
community were detected in experimental mixtures
bioaugmented with P. xylanivorans Mz5
T
(39%), with coculture of
P. xylanivorans Mz5
T
and C. cellulovorans (37%) and coculture of P.
xylanivorans Mz5
T
and F. succinogenes S85 (51%). The coculture of
C. cellulovorans and F. succinogenes S85 resulted in 30% of diver-
gence in bacterial community structure upon the initial state of
community at the start of BMP assay. P. xylanivorans Mz5
T
has
the strongest effect on methane production enhancement due to
the enhanced hydrolysis of BSG which provides more monosaccha-
rides to acidogenic bacteria. This provides as well more substrates
for acetogenic bacteria and further on for methanogenic archaea,
what might change the whole bacterial community structure.
Peak profiles of pure cultures of the added hydrolytic bacteria were
65
P.xyl.Mz5 + F.succ.S85 1 d0
P.xyl.Mz5 + F.succ.S85 2 d0
P.xyl.Mz5 + F.succ.S851 d30
P.xyl.Mz5 + F.succ.S852 d30
100
989694929088868482807876
74
72706866
92.4
95.3
88.6
92.9
96.3
90.3
87.5
97.6
96
92.1
82
96
95.8
90.9
87
83.6
71.9
90
93.9
85.9
93.1
78.6
96.4
95.5
95.6
93.7
91.4
96.3
89
94
90.6
89.6
83.8
72.7
C.cell. + F.succ.S85 1 d0
C.cell. + F.succ.S85 2 d0
F.succ.S85 1 d0
F.succ.S85 2 d0
C.cell. 1 d0
C.cell 2 d0
NC 1 d0
NC 2 d0
BSG + medium1 2 d0
BSG + medium2 2 d0
BSG + medium2 1 d0
BSG + medium1 1 d0
P.xyl.Mz5 1 d0
P.xyl.Mz5 + C.cell. 2 d0
P.xyl.Mz5 2 d0
P.xyl.Mz5 + C.cell. 1 d0
BSG + medium1 1 d30
BSG + medium1 2 d30
BSG + medium2 1 d30
BSG + medium2 2 d30
NC 1 d30
NC 2 d30
F.succ.S85 2 d30
P.xyl.Mz5 1d30
F.succ.S85 1 d30
P.xyl.Mz5 2 d30
P.xyl.Mz5 + C.cell. 1 d30
C.cell 1 d30
C.cell 2 d30
P.xyl.Mz5 + C.cell. 2 d30
C.cell. + F.succ.S85 2 d30
C.cell. + F.succ.S85 1d30
Fig. 4. Pearson correlation dendrogram of bacterial T-RFLP fingerprints. NC – negative control; BSG – brewery spent grain; medium 1 – xylan and CMC cellulose; medium 2 –
xylan, Avicel cellulose and cellobiose; P. xyl. Mz5
T
–P. xylanivorans Mz5
T
;C. cell.–C. cellulovorans;F. succ. S85 – F. succinogenes S85; d0 – samples from day 0; d30 – samples
from day 30.
266 M. C
ˇater et al. / Bioresource Technology 186 (2015) 261–269
located using PCR products of extracted DNA, digested by BsuRI. As
the T-RFLP profiles of 16S rRNA amplicons of pure cultures resulted
in few peaks, we analyzed the samples also with DGGE technique
to monitor the presence of the added bacteria during BMP assay.
Archaeal community structure analysis results also divided into
two clusters for day 0 and day 30 (Fig. 5). The divergence was
approximately 20% in most of the samples, except for mixtures
where C. cellulovorans was added. This correlates to the fact that
archaeal community in methanogenic sludge is much smaller
and less diverse than bacterial (Zakrzewski et al., 2012). Archaeal
community does not change that much as bacteria during biogas
production process as archaea conduct only the last step of biogas
process, methanogenesis, and grow slowly due to the less energy
flux being available (Cord-Ruwisch et al., 1988). Bacterial commu-
nity is much richer and flexible as it is responsible for hydrolysis of
the substrate, acidogenesis and acetogenesis; and community
structure modifications are possible due to the substrate or process
conditions changes. Interestingly, there was approximately 50%
divergence in archaeal community structure in the mixtures with
added C. cellulovorans. The reason for this may lie in the higher
concentrations of hydrogen, carbon dioxide and acetic acid pro-
duced by C. cellulovorans (Sleat et al., 1984).The homoacetogenic
pathway for acetic acid production by acetogenic bacteria may
be accelerated (Nie et al., 2007) which can affect archaeal com-
munity, especially the acetoclastic methanogenic archaea.
3.5. The persistence of P. xylanivorans Mz5
T
and F. succinogenes S85 in
microbial community during the BMP assay as detected by DGGE
Bands of pure cultures of the added hydrolytic bacteria were
located using PCR products of extracted DNA. They were used to
monitor their presence in the experimental mixtures on day 0
and 30. The DGGE biases made following of C. cellulovorans
impossible as the position of the band of the pure culture was
55.7
F. succ. S85 2 d30
F. succ. S85 1 d30
BSG+medium1 1 d0
BSG+medium1 2 d0
F. succ. S85 1 d0
BSG+medium2 1 d0
BSG+medium2 2 d0
F. succ. S85 2 d0
100
95908580
75
7065
60
99.3
98.4
99.2
98.6
98.6
96.7
96.1
97.6
98.1
96.6
97.9
96.2
93.3
98.7
99
98.4
98.1
97.7
99.2
97.4
97.8
96.1
99.7
99.3
98.5
98.2
97.8
99
98.5
98.3
96.9
94.9
79.6
90.3
P.xyl. Mz5 1 d30
P.xyl. Mz5 2 d30
NC 1 d30
NC 2 d30
NC 1 d0
NC 2 d0
C. cell. 1 d0
P.xyl. Mz5 2 d0
P.xyl. Mz5 1 d0
C. cell. 2 d0
C. cell. 1 d30
C. cell.+F. succ. S85 1 d30
C. cell.+F. succ. S85 2 d30
P.xyl. Mz5+C. cell. 2 d30
P.xyl. Mz5+C. cell. 1 d30
P.xyl. Mz5+F. succ. S85 1 d30
P.xyl. Mz5+F. succ. S85 2 d30
BSG+medium2 2 d30
BSG+medium2 1 d30
C. cell.+F. succ S85 1 d0
C. cell.+F. succ S85 2 d0
P.xyl. Mz5+C. cell 2 d0
BSG+medium1 1 d30
BSG+medium1 2 d30
P.xyl. Mz5+F. succ. S85 1 d0
P.xyl. Mz5+F. succ. S85 2 d0
P.xyl. Mz5+C. cell 1 d0
C. cell. 2 d30
Fig. 5. Pearson correlation dendrogram of archaeal T-RFLP fingerprints. NC – negative control; BSG- brewery spent grain; medium 1 – xylan and CMC cellulose; medium 2 –
xylan, Avicel cellulose and cellobiose; P. xyl. Mz5
T
–P. xylanivorans Mz5
T
;C. cell.–C. cellulovorans;F. succ. S85 – F. succinogenes S85; d0 – samples from day 0; d30 – samples
from day 30.
M. C
ˇater et al. / Bioresource Technology 186 (2015) 261–269 267
the same as a band appearing in all other samples, where C. cel-
lulovorans was not added. We assume that we detected also other,
closely related clostridiales, which are abundant in biogas reactors
and have similar GC contents (Krause et al., 2008). As P. xylanivo-
rans Mz5
T
and F. succinogenes S85 possess similar GC content
(42% and 48%, respectively (Kopec
ˇny et al., 2003; Suen et al.,
2011), the DGGE technique could not differ between them (Kisand
and Wikner, 2003) and as a result only one band appears when they
are both present in the sample (Fig. 6). DGGE profiles revealed that
the added P. xylanivorans Mz5
T
and F. succinogenes S85 are clearly
present in the samples from day 0 but could not be detected
undoubtedly on day 30. The reason for the last result may be in
DGGE technique, which was probably not sensitive enough.
Additional analysis with DGGE with samples taken on day 3, 6,
and 10 revealed that P. xylanivorans Mz5
T
remains in the bioreac-
tors for at least 6 days, while C. cellulovorans was detected only
on day 0, anyway further presence cannot be proven because a
slight bend of clostridiales was seen in all samples.
P. xylanivorans Mz5
T
proved to be successful in methane pro-
duction enhancement (17.8%), although it was not detected in
the microbial community at the end of the experiment.
According to the results of methane production and SCFAs concen-
tration we speculate that the majority of hydrolytic work was per-
formed during the first 14 days of batch anaerobic digestion and
later on the added bacteria were outcompeted by the original
microbial community under the given experimental conditions.
We can speculate further that bioaugmentation of biogas produc-
tion from BSG on full-scale level would be more successful in
two-phase biogas digesters, where P. xylanivorans Mz5
T
would be
applied in the first phase (hydrolytic, acidogenic). P. xylanivorans
Mz5
T
would probably find more suitable conditions for its persis-
tence in acidogenic microbial community. Similar approaches were
suggested by Yue et al. (2013) for the application of rumen
microorganisms in anaerobic bioconversion of lignocellulosic
biomass.
4. Conclusions
Improved lignocellulose hydrolysis was achieved by bioaug-
menting biogas batch reactors treating BSG with P. xylanivorans
Mz5
T
,C. cellulovorans and F. succinogenes S85, which promoted
biogas production and significantly elevated methane production.
The persistence of the most successful P. xylanivorans Mz5
T
in
bioreactor is not stable but appears to be long enough to improve
the anaerobic digestion efficiency. Bioaugmentations caused
greater changes in bacterial than in archaeal community structure,
although the addition of C. cellulovorans affected archaea
significantly, too. Bioaugmentation is a promising method for
increasing methane production from BSG and it may broaden the
usage of lignocellulose in biogas industry.
Acknowledgement
This research was supported by the World Federation of
Scientists and Slovenian Science Foundation.
Appendix A. Supplementary data
Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.biortech.2015.03.
029.
References
American Public Health Association, American Water Works Association, Water
Environment Federation, 2006. Standard Methods for the Examination of Water
and Wastewater, 22nd ed. American Public Health Association, American
Water Works Association, Water Environment Federation, Washington, D.C..
C
ˇater, M., Zorec, M., Marinšek Logar, R., 2014. Methods for improving anaerobic
lignocellulosic substrates degradation for enhanced biogas production. Springer
Sci. Rev. 2, 51–61.
C
ˇepeljnik, T., Zorec, M., Kostanjšek, R., Nekrep, F.V., Marinšek-Logar, R., 2003. Is
Pseudobutyrivibrio xylanivorans strain Mz5
T
suitable as a probiotic? An in vitro
study. Folia Microbiol. 48, 339–345.
Cord-Ruwisch, R., Seitz, H.J., Conrad, R., 1988. The capacity of hydrogenotrophic
anaerobic bacteria to compete for traces of hydrogen depends on the redox
potential of the terminal electron acceptor. Arch. Microbiol. 149, 350–357.
Doi, R.H., Tamaru, Y., 2001. The Clostridium cellulovorans cellulosome: an enzyme
complex with plant cell wall degrading activity. Chem. Rec. 1, 24–32.
Fontes, C.M.G.A., Gilbert, H.J., 2010. Cellulosomes: highly efficient nanomachines
designed to deconstruct plant cell wall complex carbohydrates. Annu. Rev.
Biochem. 79, 655–681.
Großkopf, R., Janssen, P.H., Liesack, W., 1998. Diversity and structure of the
methanogenic community in anoxic rice paddy soil microcosms as examined by
cultivation and direct 16S rRNA gene sequence retrieval. Appl. Environ.
Microbiol. 64, 960–969.
Hansen, T.L., Schmidt, J.E., Angelidaki, I., Marca, E., Jansen, J.C., Moskæk, H.,
Christensen, T.H., 2004. Method for determination of methane potentials of
solid organic waste. Waste Manage. 24, 393–400.
Hendriks, A.T.W.M., Zeeman, G., 2009. Pretreatments to enhance the digestibility of
lignocellulosic biomass. Bioresour. Technol. 100, 10–18.
Hobson, P.N., 1969. Rumen bacteria. In: Noriss, J.R., Ribbons, D.W. (Eds.), Methods in
Microbiology. Academic Press, New York, pp. 133–149.
Holdeman, L.V., Cato, E.P., Moore, W.E.C., 1977. Anaerobe Laboratory Manual,
fourth ed. Virginia Polytechnic Institute and State University, Blacksburg,
Virginia.
Kisand, V., Wikner, J., 2003. Limited resolution of 16S rDNA DGGE caused by melting
properties and closely related DNA sequences. J. Microbiol. Methods 54, 183–191.
Kopec
ˇny, J., Zorec, M., Mrazek, J., Kobayashi, Y., Marinšek-Logar, R., 2003. Butyrivibrio
hungatei sp. nov. and Pseudobutyrivibrio xylanivorans sp. nov., butyrate-producing
bacteria from the rumen. Int. J. Syst. Evol. Microbiol. 53, 201–209.
Krause, D.O., Denman, S.E., Mackie, R.I., Morrison, M., Rae, A.L., Attwood, G.T.,
McSweeney, C.S., 2003. Opportunities to improve fiber degradation in the rumen:
microbiology, ecology, and genomics. FEMS Microbiol. Rev. 27, 663–693.
Krause, L., Diaz, N.N., Edwards, R.A., Gartemann, K.H., 2008. Taxonomic composition
and gene content of a methane-producing microbial community isolated from a
biogas reactor. J. Biotechnol. 136, 91–101.
Lever, M., 1977. Carbohydrate determination with 4-hydroxybenzoic acid
hydrazide (PAHBAH): effect of bismuth on the reaction. Anal. Biochem. 81,
21–27.
Lowry, O.H., Rosebrough, N.J., Farr, A.L., Randall, R.J., 1951. Protein measurement
with the Folin phenol reagent. J. Biol. Chem. 193, 265–275.
Lü, F., Ji, J., Shao, L., He, P., 2013. Bacterial bioaugmentation for improving methane
and hydrogen production from microalgae. Biotechnol. Biofuels 6, 92.
Montgomery, L., Flesher, B., Stahl, D., 1988. Transfer of Bacteroides succinogenes
(Hungate) to Fibrobacter gen. nov. as Fibrobacter succinogenes comb. nov. and
description of Fibrobacter intestinalis sp. nov. Int. J. Syst. Evol. Microbiol. 38,
430–435.
Mussatto, S.I., Roberto, I.C., 2005. Acid hydrolysis and fermentation of brewer’s
spent grain to produce xylitol. J. Sci. Food Agric. 85, 2453–2460.
Muyzer, G., de Waal, E.C., Uitterlinden, A.G., 1993. Profiling of complex microbial
populations by denaturing gradient gel electrophoresis analysis of polymerase
Fig. 6. DGGE profiles showing the presence of added bacteria in samples from BMP
assay. S – standard, M1 and M2 – samples with added P. xylanivorans Mz5
T
, F1 and
F2 – samples with added F. succinogenes S85, MF1 and MF2 – samples with added
coculture of P. xylanivorans Mz5
T
and F. succinogenes S85, d0 – samples from day 0,
d30 – samples from day 30. The white box indicates the position of pure culture
bands; an arrow indicates the line where the band of P. xylanivorans Mz5
T
and F.
succinogenes S85 appears if they are present in the sample.
268 M. C
ˇater et al. / Bioresource Technology 186 (2015) 261–269
chain reaction-amplified genes coding for 16S rRNA. Appl. Environ. Microbiol.
59, 695–700.
Nie, Y., Liu, H., Du, G., Chen, J., 2007. Enhancement of acetate production by a novel
coupled syntrophic acetogenesis with homoacetogenesis process. Process
Biochem. 42, 599–605.
Peng, J., Lü, Z., Rui, J., Lu, Y., 2008. Dynamics of the methanogenic archaeal
community during plant residue decomposition in an anoxic rice field soil.
Appl. Environ. Microbiol. 74, 2894–2901.
Peng, X., Aragão Börner, R., Achu Nges, I., Liu, J., 2014. Impact of bioaugmentation on
biochemical methane potential for wheat straw with addition of Clostridium
cellulolyticum. Bioresour. Technol. 152, 567–571.
Ratkowsky, D.A., 1990. Handbook of Nonlinear Regression Models. Marcel Dekker,
New York.
Russ, W., Mortel, H., Meyer-Pittroff, R., 2005. Application of spent grains to increase
porosity in bricks. Constr. Build. Mater. 19, 117–126.
Sleat, R., Mah, R.A., Robinson, R., 1984. Isolation and characterization of an
anaerobic, cellulolytic bacterium, Clostridium cellulovorans sp. nov. Appl.
Environ. Microbiol. 48, 88–93.
Stewart, C.S., Flint, H.J., 1989. Bacteroides (Fibrobacter)succinogenes, a cellulolytic
anaerobic bacterium from the gastrointestinal tract. Appl. Microbiol.
Biotechnol. 30, 433–439.
Stewart, C.S., Duncan, S.H., McPherson, C.A., Richardson, A.J., Flint, H.J., 1990. The
implications of the loss and regain of cotton-degrading activity for the
degradation of straw by Ruminococcus flavefaciens strain 007. J. Appl.
Bacteriol. 68, 349–356.
Suen, G., Weimer, P.J., Stevenson, D.M., Aylward, F.O., Boyum, J., Deneke, J.,
Drinkwater, C., Ivanova, N.N., Mikhailova, N., Chertkov, O., Goodwin, L.A., Currie,
C.R., Mead, D., Brumm, P.J., 2011. The complete genome sequence of Fibrobacter
succinogenes S85 reveals a cellulolytic and metabolic specialist. PLoS ONE 6,
e18814.
Walter, J., Tannock, G.W., TiLCSala-Timisjarvi, A., Rodtong, S., Loach, D.M., Munro, K.,
Alatossava, T., 2000. Detection and identification of gastrointestinal
Lactobacillus species by using denaturing gradient gel electrophoresis and
species-specific PCR primers. Appl. Environ. Microbiol. 66, 297–303.
Weisburg, W.G., Barns, S.M., Pelletier, D.A., Lane, D.J., 1991. 16S ribosomal DNA
amplification for phylogenetic study. J. Bacteriol. 173, 697–703.
Yue, Z.-B., Li, W.-W., Yu, H.-Q., 2013. Application of rumen microorganisms for
anaerobic bioconversion of lignocellulosic biomass. Bioresour. Technol. 128,
738–744.
Zakrzewski, M., Goesmann, A., Jaenicke, S., Jünemann, F., Eikmeyer, F.,
Szczepanowski, R., Al-Soud, W.A., Sørensen, S., Pühler, A., Schlüter, A., 2012.
Profiling of the metabolically active community from a production-scale biogas
plant by means of high-throughput metatranscriptome sequencing. J.
Biotechnol. 158, 248–258.
Zorec, M., C
ˇepeljnik, T., Nekrep, F.V., Marinšek-Logar, R., 2001. Multiple
endoxylanases of Butyrivibrio sp. Folia Microbiol. 46, 94–96.
M. C
ˇater et al. / Bioresource Technology 186 (2015) 261–269 269
