Appl Microbiol Biotechnol

DOI 10.1007/s00253-014-5835-z

ENVIRONMENTAL BIOTECHNOLOGY

Cultivation of nitrite-dependent anaerobic methane-oxidizing

bacteria: impact of reactor configuration
Baolan Hu & Zhanfei He & Sha Geng & Chen Cai &
Liping Lou & Ping Zheng & Xinhua Xu

Received: 25 March 2014 / Revised: 14 May 2014 / Accepted: 15 May 2014

# Springer-Verlag Berlin Heidelberg 2014

Abstract Nitrite-dependent anaerobic methane oxidation (ndamo) is mediated by bacteria that anaerobically oxidize
methane coupled with nitrite reduction and is a potential
bioprocess for wastewater treatment. In this work, the effect
of reactor configuration on n-damo bacterial cultivation was
investigated. A magnetically stirred gas lift reactor (MSGLR),
a sequencing batch reactor (SBR), and a continuously stirred
tank reactor (CSTR) were selected to cultivate the bacteria.
Microbial community was monitored by using quantitative
PCR, 16S rRNA gene sequencing, pmoA gene sequencing,
and fluorescence in situ hybridization (FISH). The effects of
substrate inhibition, methane mass transfer, and biomass
washout in the three reactors were focused on. The results
indicated that the MSGLR had the best performance among
the three reactor systems, with the highest total and specific ndamo activities. Its maximum volumetric nitrogen removal
rate was up to 76.9 mg N L1 day1, which was higher than
previously reported values (5.137.8 mg N L1 d1).

anaerobic oxidation of methane (AOM) coupled to nitrite

reduction (Eq. 1) independently (Ettwig et al. 2008). Under
the rules of the international code of nomenclature of bacteria,
the newfound n-damo bacteria were tentatively named
Candidatus Methylomirabilis oxyfera (M. oxyfera) in affiliation with the uncultured NC10 phylum (Ettwig et al.
2009). Compared with conventional heterotrophic denitrifier,
n-damo bacteria use methane, a major end product of anaerobic digestion, as electron donor for denitrification (Shen et al.
2012). N-damo process costs less to treat wastewater containing nitrogenous pollutions, because methane is inexpensive
and plentiful in wastewater treatment plants, while conventional heterotrophic denitrification requires more expensive
electron donor, like methanol. Moreover, N2O, a greenhouse
gas, is an important intermediate of conventional heterotrophic denitrification, while there is no N2O emission in n-damo
process (Shi et al. 2013). Therefore, it is deemed that n-damo
process is an excellent substitute of conventional heterotrophic denitrification process (Shen et al. 2012).

Keywords AOM . N-damo . Cultivation . Reactor

configuration

3CH4 8NO2 8H 3CO2 4 N2 10H2 O

Introduction
Nitrite-dependent anaerobic methane-oxidizing (n-damo) bacteria were first discovered and confirmed in a laboratory
enrichment culture (Raghoebarsing et al. 2006) and were
subsequently found in natural habitats (Deutzmann and
Schink 2011; Kojima et al. 2012; Shen et al. 2013; Hu et al.
2014). N-damo bacteria can mediate the bioprocess of
B. Hu : Z. He : S. Geng : C. Cai : L. Lou : P. Zheng : X. Xu (*)
Department of Environmental Engineering, Zhejiang University,
Hangzhou 310058, China
e-mail: xuxinhua@zju.edu.cn

However, M. oxyfera grow slowly, and their doubling time

is as long as weeks (Raghoebarsing et al. 2006; Ettwig et al.
2008; He et al. 2013), which makes their cultivation very
difficult. There are many factors that influence bacterial cultivation, like temperature, oxygen, substrate, growth factor, and
inhibitor, and reactor configuration is one of the important
impact factors, particularly for slowly growing bacteria (Jin
et al. 2008; Tao et al. 2012). Sequencing batch reactor (SBR)
and sequencing fed-batch reactor (SFBR) are powerful tools
for cultivating slowly growing bacteria (Strous et al. 1998)
and are commonly used to enrich n-damo bacteria
(Raghoebarsing et al. 2006; Ettwig et al. 2009; He et al.
2013). In addition, Ettwig et al. (2008) also obtained an

Appl Microbiol Biotechnol

enrichment by running a continuously stirred tank reactor

(CSTR) with an external settler. However, the activity of ndamo bacteria in these reactors was far lower than that of
conventional heterotrophic denitrifier (Ettwig et al. 2009; He
et al. 2013; Kampman et al. 2012). The efforts should be
exerted to boost the n-damo activity and the bacterial growth
rate. Same factors have been known that limit the n-damo
process including substrate inhibition of nitrite (He et al. 2013;
Hu et al. 2009) and washout of biomass (Kampman et al.
2012), and an ideal reactor could push these limits and enhance the n-damo activity.
To identify the effect of reactor configuration on n-damo
bacterial cultivation and to obtain a high n-damo activity, three
different representative reactor systemsthe SBR, the CSTR,
and the magnetically stirred gas lift reactor (MSGLR)were
selected and operated. In the MSGLR, a magnetic stirring
system was placed in the bottom center of a normal gas lift
reactor, an O shape aeration pipe was installed around the
bottom, and a gas-liquid-solid separator was installed in the
top of the reactor (see Fig. 1). The methane was recirculated
by a gas-recycling pump from the headspace to the bottom of
the MSGLR, which provided sufficient methane to the reaction liquid. Compared with the SBR and the CSTR, the
MSGLR enhances the mass transfer of gas-liquid phases and
the mixing of liquid-solid phases. Biomass loss, substance
inhibition, and methane mass transfer were monitored, because these factors exert important effects on the n-damo
cultivation process. Quantitative PCR (qPCR), 16S ribosomal
RNA (rRNA) and pmoA gene sequencing, and fluorescence in
situ hybridization (FISH) were used to monitor changes in the
microbial community in the reactors. This work investigated
the effect of reactor configuration on n-damo bacterial cultivation and provided an efficient reactor to obtain more biomass and a higher n-damo activity, which would promote the
research on n-damo bacteria and their application.

MgSO4 7H2O, 0.1970.690 g NaNO2, 0.5 mL of an acidic

trace element solution, and 0.2 mL of an alkaline trace element
solution. The acidic (100 mmol L1 HCl) trace element solution
contained (per liter) 2.085 g FeSO4 7H2O, 0.068 g ZnSO4
7H2O, 0.12 g CoCl2 6H2O, 0.5 g MnCl2 4H2O, 0.32 g CuSO4,
0.095 g NiCl2 6H2O, and 0.014 g H3BO3. The alkaline
(10 mmol L1 NaOH) trace element solution contained (per
liter) 0.067 g SeO2, 0.050 g Na2WO4 2H2O, and 0.242 g
Na2MoO4 (Ettwig et al. 2009). During the operating period,
the nitrite concentration in the influent changed from 40 to
140 mg N L1, but stable pH was at 7.07.2.
Setup and operation

Materials and methods

Schematic diagrams of the three reactors with different configurations (SBR, CSTR, and MSGLR) are shown in Fig. 1.
Each reactor has a total volume of approximately 1.3 L, with a
working volume of 1.0 L. The CSTR and MSGLR had an
additional 0.1 L for the gas-liquid-solid separator, but their
headspaces were slightly smaller than those of the SBR.
The SBR was operated in successive 24-h cycles consisting
of 0.1 h of medium feeding, 22 h of aeration and stirring, 1.8 h
of settling, and 0.1 h of effluent discharge. The volume exchange ratio for the SBR was 0.5. The CSTR and MSGLR
were both continuous flow reactor systems with hydraulic
retention times (HRTs) of 48 h. Methane (99.99 %) was
supplied at 10 mL min1 into the three reactors, and gas was
recirculated in the MSGLR at 1,000 mL min1. All three
reactors were magnetically stirred at 150 rpm, and the temperature was controlled at 300.5 C. The feeding tank and
the effluent collection tanks were flushed with pure Ar
(99.999 %) for 15 min and subsequently sealed with argon
gas bags. Each day, 2 mL of influent and effluent were
sampled and centrifuged (5 min, 7,440g) to measure nitrite
and nitrate concentrations, and 2 mL of sludge was sampled
for molecular analyses on day 75. Moreover, suspensions in
the reactors and the effluent were collected to determine the
VSS level on days 0 and 75 and day 75, respectively.

Inoculum and medium

Measurement of the n-damo activity

The inoculum for the three reactors was taken from a previous
SBR enrichment cultivated after 650 days. The inoculum was
completely mixed and divided into three portions before the
reactors were seeded. The volatile suspended solid (VSS)
concentration of the seeding sludge was 5.30.3 g VSS L1,
and the inoculum volume was 30 % of the working volume for
each reactor.
The medium was modified from previous protocols (Ettwig
et al. 2009). The KHCO3 concentration was reduced from 1.0
to 0.25 g L1 to avoid the generation of CaCO3 and MgCO3
deposits. Here, the modified medium contained (per liter)
0.25 g KHCO3, 0.05 g KH2PO4, 0.3 g CaCl2 2H2O, 0.2 g

On days 0 and 75, stable isotope activity tests were conducted

to determine n-damo activity reliably. The tests were performed in a series of 150-mL serum bottles containing
10 mL of culture (sediment from the reactors after 1.8 h of
settling), 70 mL of medium, and 70 mL of headspace. The
culture was rinsed three times with nitrite-free medium and
was quickly added to the serum bottles along with 70 mL of
medium (nitrite added, 5.6 mg N L1). The serum bottles were
flushed with Ar (99.999 %) for 15 min and sealed.
Subsequently, 2 mL of 13CH4 (99 %) was directly injected
into the headspace. These serum bottles were incubated on a
shaking table at 300.5 C and 150 rpm. After 1.0 h of

Appl Microbiol Biotechnol

Fig. 1 Schematic diagram of the
three different reactor systems,
displaying the 1 SBR, 2 CSTR, 3
MSGLR, 4 feeding tank, 5
effluent collection tank, 6
peristaltic pump, 7 recycling
pump, 8 exhaust gas bag, 9
methane gas bag, 10 argon gas
bag, 11 gas-liquid-solid separator,
12 PTFE magnetic stirring bar, 13
draft tube, 14 aeration pipe, and
15 liquid sealing

8
15
9

15

15
7

11

11

3
13

12

12

14

12

10

preincubation, 1 mL of the medium was sampled and centrifuged (5 min, 7,440g) every 1.5 h to measure the nitrite and
nitrate concentrations, and 30 L of gas was extracted to
measure the 13CH4 and 13CO2 concentrations in triplicate.
To precisely determine the n-damo activity, two control
groups were established. Control group (I), without methane
in the headspace, was designed to eliminate the disturbances
of denitrification by heterotrophic denitrifier, and control
group (II), without nitrite in the medium, could present the
methane consumption of other methane oxidation processes.
Both of the two control groups were studied simultaneously
with the experimental group.
Analytical methods
The concentrations of VSS, nitrite, and nitrate were measured
according to the APHA standard methods (APHA 2005). The
13
CH4 and 13CO2 concentrations in the headspace were measured by a mass spectrometer (Agilent 7890/5975C inert
MSD; Agilent, USA) as previously described (Ettwig et al.
2009). The pH was measured with a Mettler-Toledo FE20 pH
meter (Mettler-Toledo Instruments Co., Ltd., Shanghai).
Molecular analyses
The inoculum and cultures of the three reactors on day 75
were collected (2 mL) for molecular analysis. M. oxyfera-like

bacteria were identified by 16S rRNA and pmoA gene sequencing and FISH microscopic analysis. The quantities of
M. oxyfera-like bacteria in the biomass were represented by
the abundance of the 16S rRNA genes of NC10 bacteria
(copies g1 VSS) determined by qPCR. The densities of
M. oxyfera-like bacteria in the reactors were characterized by
the copy numbers of the 16S rRNA genes of NC10 bacteria
(copies L1) that were calculated by multiplying the 16S
rRNA gene abundance (copies g1 VSS) times VSS concentrations (g VSS L1).
The total DNA was isolated with a PowerSoil DNA
Isolation Kit (Mo Bio Laboratories, Carlsbad, CA, USA)
according to the manufacturers instructions. Nested PCR
was employed to amplify the 16S rRNA and pmoA genes
from the DNA samples. The primers and the amplification
conditions were the same as previously reported (Luesken
et al. 2011a, b). The PCR products were examined by agarose
gel (1.0 %) electrophoresis. Cloning of the PCR products was
performed using the pMD19-T vector (TaKaRa Bio Inc.,
Shiga, Japan) according to the manufacturers instructions.
The vectors were transformed into competent cells. They were
incubated in SOC medium for 1 h and in LB medium containing ampicillin, X-gal, and IPTG for 16 h at 37 C. Using
blue-white screen technique, the white colonies were picked
out. About twenty positive clones were randomly selected
from each of the 16S rRNA and pmoA gene clone libraries
and were sequenced by BGI-Hangzhou, China.

Appl Microbiol Biotechnol

qPCR was performed to assess the abundance of the 16S

rRNA genes of NC10 bacteria in the biomass, using an
iCycler iQ5 thermocycler and a real-time detection system
(Bio-Rad, CA, USA), with qP1F/qP1R (Ettwig et al. 2009)
as the primer pair. The PCR amplification and the construction
of standard curves were conducted as previously described
(Hu et al. 2010). The copy numbers of the samples were
calculated based on the threshold cycle values of the standard
curve.
Phylogenetic analysis of the 16S rRNA and pmoA genes
was conducted with MEGA4.0 software (Tamura et al. 2007)
using the ClustalW algorithm. Phylogenetic trees were constructed in MEGA4.0 using the neighbor-joining method with
p distance correction and a 1,000-replicate bootstrap value
(Hu et al. 2012).
FISH was conducted as previously described (Ettwig et al.
2008). The following oligonucleotide probes were used: S-*DBACT-1027-a-A-18 (red) specific for NC10 phylum bacteria (Raghoebarsing et al. 2006) and the EUB I-III (blue)
mixture for most bacteria (Daims et al. 1999). FISH images
were acquired using a two-photon laser confocal microscope
(Zeiss, LSM710 NLO, Germany).
Nucleotide sequence accession numbers
The sequences reported in this work have been deposited in
the GenBank database, and the accession numbers are
KJ668605-KJ668624 for M. oxyfera 16S rRNA genes and
KJ668625-KJ668649 for M. oxyfera pmoA genes.

Results
Volumetric nitrogen removal rates
Figure 2 presents the volumetric nitrogen removal rate (NRR)
of the three reactors. The 75-day operating period can be
divided into three phases (marked by dashed lines in Fig. 2)
according to the nitrite loading rate (NLR). The NLR was
approximately 30 mg N L1 day1 in phase I,
50 mg N L1 day1 in phase II, and 70 mg N L1 day1 in
phase III (the NLRs of the SBR and CSTR decreased to
20 mg N L1 day1 after day 50 due to substrate inhibition).
In phase I (114 days), the NRRs in all three reactors
increased over time. The performances of the SBR and CSTR
were similar, and the MSGLR performed better than the other
two. The NRR in the MSGLR rapidly increased to the maximum (approximately 30 mg N L1 day1) for this phase during
the first 5 days and was then stable near the maximum for the
remaining time. In phase II (1545 days), the NRRs in the
CSTR and MSGLR continuously increased, but the SBR exhibited a stagnation period. The nitrite concentrations in the

reactors, the SBR, the CSTR, and the MSGLR, were 4.15
5.87, 2.753.31, and 0.021.42 mmol L1, respectively. In
phase III (4675 days), the NRR in the MSGLR continuously
increased after a short stagnation period, but those in the SBR
and CSTR were rapidly falling when the NLR increased to
70 mg N L1 day1 for 5 days, and the nitrite concentrations in
the SBR and the CSTR sharply increased to 9.17 and
8.71 mmol L1 on the fifth day after NLR increasing.
Moreover, the nitrite concentration in the MSGLR increased
to 3.80 mmol L1 on the fifth day after NLR increasing but
decreased gradually in the remaining time. After the NLR was
reduced, the NRR in the SBR and CSTR partially recovered,
but it was difficult to resume to the previous best level (in phase
II). Finally, the MSGLR achieved the highest NRR,
76.9 mg N L1 day1, which is almost twice as high as the
previously reported highest values of 37.8 mg N L1 day1
(Kampman et al. 2012) (see Table 1).
Biomass washout
The copy numbers of the 16S rRNA genes of the NC10
bacteria in all three reactors (copies L1) increased over the
75-day operation period, and the copy number for the
MSGLR was the highest (see Table 2), which agrees with
the NRR results. A comparison of the SBR and CSTR showed
that the SBR maintained more biomass (higher biomass concentration) but the abundance of the 16S rRNA genes of
NC10 bacteria in the SBR was slightly lower than that in the
CSTR. The MSGLR owned both the highest biomass concentration and 16S rRNA gene abundance among the three reactors, which indicated that the MSGLR could be the optimal
choice to cultivate n-damo bacteria.
To further investigate the sludge retention performances of
the three reactors, the concentrations of biomass in the effluent
of the SBR, CSTR, and MSGLR were determined on day 75,
which were 0.039, 0.038, and 0.050 g VSS L1, respectively.
The percentages of biomass washed out with the effluent of
the SBR, CSTR, and MSGLR were estimated to be 1.63, 2.00,
and 1.85 %, respectively, as calculated from the concentration
of biomass in the effluent divided by that in the reactor. The
ratios of washed-out biomass to produced biomass in the
SBR, CSTR, and MSGLR were estimated to be 57.2, 77.0,
and 55.3 %, respectively, by assuming that the percentages of
washed-out biomass were constant during the period of
operation.
Batch activities by stable isotope tests
The consumption of nitrite and 13CH4 occurred simultaneously, and 13CO2 was detected in the headspace after 3.0 h. The
generation of 13CO2 occurred later than 13CH4 consumption
(Fig. 3), which may be explained by the fact that 13CO2 was
dissolved into the liquid. The specific activities of n-damo

Appl Microbiol Biotechnol

80

Fig. 2 Volumetric nitrogen

removal rates in the three reactors:
MSGLR (black square), CSTR
(black circle), and SBR (black
triangle)

Phase II

Phase I

Phase III

volumetric nitrogen removal rate

-1 -1
(mg N.L .d )

70
60
50

MSGLR
CSTR
SBR

40
30
20
10
0
0

10

20

30

40

50

60

70

80

time (d)

bacteria were used to compare the performances of the different cultures and are represented by the generation rates of
13
CO2 and the conversion rates of nitrite and 13CH4 (Fig. 4).
The specific activities of the cultures from the SBR, CSTR,
and MSGLR after 75 days increased at varied levels. The
specific activity was the highest in the MSGLR, followed by
the CSTR and lowest in the SBR. Moreover, almost all of the
specific activities increased during the 75 days of operation
(except for the 13CO2 generation rate in the SBR). The results
of the specific activities support the qPCR findings (the abundance of the 16S rRNA genes of the NC10 bacteria in
Table 2).
According to the stable isotope batch tests, the molar ratios
of the 13CO2, 13CH4, and nitrite generation/conversion rates
for the inoculum and the cultures from the SBR, CSTR, and
MSGLR were 2.6:3:7.1, 2.3:3:7.5, 2.8:3:6.6, and 2.8:3:8.5

(Fig. 4), respectively. These values are close to the theoretical

stoichiometric ratio of 3:3:8 (Eq. 1). Moreover, the specific
per cell activities for n-damo from the inoculum, SBR culture,
CSTR culture, and MSGLR culture were calculated to be 0.28
0.06, 0.280.04, 0.310.09, and 0.340.10 fmol CH4
day1 cell1 (for one copy).
Phylogenetic analysis and FISH
The phylogenetic analyses showed that the 16S rRNA gene
sequences obtained from the three cultures belonged to group
A members (Fig. 5). The sequence identity among the three
cultures was 96.599.8 %. The sequence identities between
the M. oxyfera (FP565575) and the M. oxyfera-like bacteria
from the SBR, CSTR, and MSGLR were 97.497.8, 96.7
97.8, and 96.597.6 %, respectively. The phylogenetic

Table 1 Overview of the maximum nitrogen removal rates for the n-damo process
Configuration

Temperature
(C)

HRT
(day)

Maximum NRR
(mg N L1 day1)

Reference

SBRa
SBRa

25
30

15.3b
29.3b

Raghoebarsing et al. (2006)

Ettwig et al. (2009)

SBRa
SBR
SBR
CSTR with an external settler
CSTR

2023
30
30
30
30

5.5b
1330.3b (initial)
4.3b (after 6 months)
1550b
2.0
2.0
16.2565b
2.0

5.1b
17.9b
11.4
9.6b
26.4

Luesken et al. (2011b)

He et al. (2013)
this work
Ettwig et al. (2008)
this work

SFBR with gas recirculation

MSGLR

30
30

4.855.0b
2.0

37.8
76.9

Kampman et al. (2012)

this work

These reactors are SFBRs actually, although they were called as SBRs in the references

Calculated value

Appl Microbiol Biotechnol

Table 2 Concentrations of biomass and 16S rRNA genes of NC10 bacteria on day 75
Inoculum/Reactor

Inoculum

SBR

CSTR

MSGLR

Concentration of biomass(g VSS L1)

Abundance of 16S rRNA genes (copies g1 VSS)
Copy numbers of 16S rRNA genes (copies L1)

1.60.1
9.32.01011
1.50.31012

2.40.1
9.61.31011
2.30.31012

1.90.2
1.30.41012
2.50.81012

2.70.1
1.30.51012
3.51.41012

Discussion
Substrate inhibition and methane mass transfer

0.05

0.10

0.04

0.09

0.03

0.08

0.02

0.07

0.01

13

0.11

13

Fig. 3 The conversion of 13CH4

and the generation of 13CO2 in the
headspace of the serum bottles.
The tested biomass was taken
from the SBR, CSTR, and
MSGLR on day 75. 13CH4 data
from the MSGLR (black square),
CSTR (black circle), and SBR
(black triangle) are plotted with
the ordinate on the left. 13CO2
data from the MSGLR (white
square), CSTR (white circle), and
SBR (white triangle) are plotted
with the ordinate on the right

CH4 in gas phase (mmol)

High concentrations of nitrite may inhibit the n-damo activities in the reactors, as is well known that nitrite is a substrate
inhibitor. The stagnation period of the SBR in phase II may be
resulted from the substrate inhibition of nitrite for its nitrite
concentration was slightly higher than the reported nitrite
inhibition constant for n-damo bacteria (K I NO2 , 4.1

0.5 mmol L1) (He et al. 2013). The sharp decreases of

NRRs of the SBR and the CSTR in the first 5 days of phase
III should also be caused by the substrate inhibition of nitrite,
because only nitrite concentration in the influent was changed
from phase II to phase III (NLR increased from 50 to
70 mg N L1 day1) and the nitrite concentrations in the
SBR and the CSTR were significantly higher than the KINO2
value. The NRRs in the SBR and CSTR partially recovered
after the NLR decreased to 20 mg N L1 day1, which further
validated the assumption of substrate inhibition of nitrite.
Furthermore, the NRR in the MSGLR may be also influenced
by the substrate inhibition of nitrite during the first 5 days of
phase III for the nitrite concentration was near the nitrite
inhibition constant. The results suggested that the nitrite concentration in the reactors should be strictly controlled at a low
value, such as around 2.0 mmol L1, an optimum concentration of nitrite obtained from previous batch tests (He et al.
2013).
A comparison of the CSTR and MSGLR performances
indicates that the mass transfer of methane might be an essential impact factor, although it was reported that methane was
not a limiting factor when solid-liquid was mixed well (He
et al. 2013). Foam and float sludge were observed at the gasliquid interface in the CSTR, which would hinder the gasliquid mass transfer. The stirring bar is in the bottom of the

CO2 in gas phase (mmol)

analysis showed that the pmoA gene sequences had a low

similarity (<92 %) with the other reported pmoA gene sequences (Fig. 6). The most similar sequence was JN609384
from a coculture of anammox and NC10 bacteria (Zhu et al.
2011). The sequence similarity between JN609384 and the
sequences of this study (from the SBR, CSTR, and MSGLR)
were 86.691.0, 86.690.0, and 86.691.5 %, respectively.
The FISH results (Fig. 7) indicate that the M. oxyfera-like
bacteria (purple) were the dominant bacteria in all three cultures. Moreover, the M. oxyfera-like bacteria were found in
microbial aggregates, particularly in the SBR and CSTR, in
which the shear stress was weaker than that in the MSGLR.

0.06

0.00
0

time (h)

generation/conversion rates of CO2 / CH4 / nitrite

-1 -1
(mmol.gVSS .h )

Appl Microbiol Biotechnol

0.05
13

CO2

13

Fig. 4 Generation rates of 13CO2

and conversion rates of 13CH4 and
nitrite. The biomass was taken
from the inoculum and the
cultivations from the SBR,
CSTR, and MSGLR on day 75.
The generation rates of 13CO2
were calculated from the data
points after 3.0 h in Fig. 3. The
conversion rates of 13CO2 are
plotted by the blue bar, 13CH4 by
the green bar, and nitrite by the
red bar

13

CH4

0.04

13

nitrite
0.03

0.02

0.01

0.00

Inoculum

CSTR

MSALR

the MSGLR makes the flocs small (Biggs and Lant 2000;
Jarvis et al. 2005), which reduces the solid-liquid mass transfer resistance and improves the apparent activity of the ndamo bacteria.

CSTR, and the gas-liquid-solid separator is equipped in the

middle-upper of the CSTR, which slowed down the flow
velocity of liquid near the gas-liquid interface and allowed
foam and float sludge to accumulateon the liquid surface. This
phenomenon showed that the solid-liquid was not mixed well
in the CSTR, and the previous batch tests also showed that ndamo activity was limited by methane mass transfer when
solid and liquid phases were separated at low agitation rates
(He et al. 2013). Hence, the NRR in the CSTR may also be
limited by methane mass transfer for the bad mixing of solidliquid phase.
The MSGLR finally achieved a high NRR,
76.9 mg N L1 day1, and the high performance must be
linked to the innovative reactor configuration of MSGLR.
The magnetic stirring bar can mix the reaction system horizontally, and the gas lift loop can mix the reaction system
vertically, which jointly intensify the mixing of gas-liquidsolid phases. Then, the intensive mixing improves the gasliquid mass transfer, reduces substrate inhibition, and enhances the n-damo process. Furthermore, high shear stress in
Fig. 5 Phylogenetic tree of the
NC10 phylum 16S rRNA gene
sequences recovered from
cultures in the SBR, CSTR, and
MSGLR (Acidobacteria is the
out-group). The tree was
constructed using the neighborjoining method with the TajimaNei correction. The bootstrap
values represent the percent
occurrence in 1,000 replicates.
The scale bar represents a 2 %
sequence divergence

SBR

Biomass retention property

The ability of the sludge to settle in the reactor is an important
impact factor in the cultivation of low-growth bacteria, and
poor sludge settlement could lead to sludge loss (Chen et al.
2010). Previous research detected denitrification and anaerobic methane oxidation occurring simultaneously in biomass
from an external settler equipped to collect biomass in the
effluent, and M. oxyfera-like bacteria were detected in the
biomass with a nested PCR approach for the 16S rRNA genes
of NC10 bacteria (Luesken et al. 2011), which indicated that a
certain amount of n-damo biomass was washed out with the
effluent. Kampman et al. (2012) investigated n-damo biomass
washout from two SBFRs, and they estimated that 4148 % of
the produced biomass washed out from the reactors in spite of
MSGLR; CSTR; SBR

28
40

WWTP Heerenveen, Netherlands (JF706194)

59

Paddy soil, China (JN704464)

Donana coastal aquifer, Spain (DQ837250)
25
100

49

Lake Biwa sediment, Japan (AB661586)

group A

DAMO enrichment culture, Netherlands (DQ369742)

12
25
83

DAMO enrichment culture, Netherlands (FJ621560)

Candidatus Methylomirabilis oxyfera
DAMO enrichment culture, Australia (FJ907182)

98

Lake Biwa sediment, Japan (AB661499)

Paddy soil, China (JN704421)
Acidobacteria (D26171)

0.02

group B

Appl Microbiol Biotechnol

Fig. 6 Phylogenetic tree of the
pmoA gene sequences recovered
from cultures in the SBR, CSTR,
and MSGLR (Methylacidiphilum
is the out-group). The tree was
constructed using the neighborjoining method with the TajimaNei correction. The bootstrap
values represent the percent
occurrence in 1,000 replicates.
The scale bar represents a 5 %
sequence divergence

MSGLR; CSTR; SBR

24

Co-culture of anmmox and DAMO bateria, Netherlands (JN609384)

DAMO enrichment culture, Netherlands (JF706203)

69

Co-culture of DAMO and anammox bacteria, Netherlands (JN006735)

Candidatus Methylomirabilis oxyfera (FP565575)

10

Paddy soil, China (JN704408)

35

Saline lake sediment, China (JQ429432)

74

West Lake sediment, China (JX531982)

Ditches sediment, Netherlands (HQ698931)

32

Lake Biwa sediment, Japan (AB661621)

31
95
36
60

Lake Constance sediment, Germany (HQ906566)

Peat soil, Netherlands (JX262153)
Methylacidiphilum (FJ462788)

0.05

Microbiological characteristics

a long settling time of 2 h was applied. He et al. (2013)

determined the rates of n-damo biomass growth, decay, and
washout in the SBR, and the ratio of washed-out biomass to
produced biomass was estimated to be 46.4 % according to the
rates above. Nevertheless, the ratios of washed-out to produced biomass in this work (55.377.0 %) were even higher
than the reported values above (4148 %), which may severely limit n-damo bacterial cultivation. Obviously, the cultivation of n-damo bacteria will come to failure, if the ratio of
washed-out to produced biomass exceeds 100 %; so, the ratio
is very essential and should be reduced in the present state.
The SBR is the optimal reactor for biomass retention
among the three reactors for the lowest percentage of
biomass washout, mainly because of its ideal settlement.
While the MSGLR has the highest NRR, biomass concentration, and 16S rRNA gene copy number among the
three reactors, its biomass retention ability was poorer
than that of SBR. To obtain a further higher performance,
the ability of the MSGLR to retain biomass should be still
improved, potentially by using a membrane module, enlarging the settling tank volume, increasing the HRT,
altering the biomass aggregation form, etc.

The specific per cell activities for n-damo (0.280.34 fmol

CH4 day1 cell1) estimated in this work were higher than the
reported values, 0.09 and 0.11 fmol CH4 day1 cell1 assessed
with the same primer pair qP1F/qP1R (Ettwig et al. 2009; Zhu
et al. 2011) and 0.20 fmol CH4 day1 cell1 assessed with
anther primer pair qP2F/qP2R (Ettwig et al. 2009). The specific per cell activities for n-damo bacteria were low but still
on the same order of magnitude as the values for other AOM
organisms, like 0.7 fmol CH4 day1 cell1 for sulfatedependent AOM (Raghoebarsing et al. 2006). Cautiously,
the existing primer pairs for qPCR, qP1F/qP1R and
qP2F/qP2R (Ettwig et al. 2009), are designed for NC10 phylum bacteria, not specific for n-damo bacteria (M. oxyfera
bacteria, belonging to NC10 phylum). In theory, the qPCR
findings may get a risk of overestimation, but the previous
studies (Kojima et al. 2012; Shen et al. 2013; Ettwig et al.
2009) and this work did not detect the drawback of these
primers.
As expected, no significant divergence in the bacterial
composition occurred among the cultures in the three reactors

Fig. 7 FISH of the biomass from the three reactors after 75 days. The
cells were hybridized with probe S-*-DBACT-1027-a-A-18 (Cy3, red)
specific for NC10 bacteria and a mixture of probes EUB I-III and EUB

V (dark blue), which detects nearly all eubacteria. The NC10 bacteria
appear purple due to double hybridization with the specific (red) and
general (blue) probes. Scale bar = 5 m

Appl Microbiol Biotechnol

after 75 days cultivation (Figs. 5 and 6) because of the same

inoculum and a relatively short operation period. Therefore,
reactor configuration has no significant effect on the structure
of microbial community in a short period but obviously influences the structure of microbial aggregates (Fig. 7).
In conclusion, this work proposes that the MSGLR is a
useful approach for supplying sufficient methane mass transfer and substrate mixing to efficiently cultivate n-damo bacteria. The maximum NRR for the n-damo process increased to
76.9 mg N L1 day1 in the MSGLR, which is higher than that
(5.137.8 mg N L1 day1, listed in Table 1) reported previously. M. oxyfera-like bacteria were the dominant bacteria,
and the n-damo process occurred in all three reactor systems,
as indicated by FISH and stable isotope tests. Substrate inhibition and biomass loss strongly limited the cultivation of ndamo bacteria, and future work should focus on overcoming
these limitations.

Acknowledgments This work was supported by two grants from the

Natural Science Foundation (Nos. 51108408 and 41276109).

References
Apha A (2005) Standard methods for the examination of water and
wastewater. American Public Health Association, Washington, DC
Biggs CA, Lant PA (2000) Activated sludge flocculation on-line determination of floc size and the effect of shear. Water Res 34(9):2542
2550
Chen J, Ji Q, Zheng P, Chen T, Wang C, Mahmood Q (2010) Floatation
and control of granular sludge in a high-rate anammox reactor.
Water Res 44(11):33213328
Daims H, Brhl A, Amann R, Schleifer K-H, Wagner M (1999) The
domain-specific probe EUB338 is insufficient for the detection of all
bacteria: development and evaluation of a more comprehensive
probe set. Syst Appl Microbiol 22(3):434444
Deutzmann JS, Schink B (2011) Anaerobic oxidation of methane in
sediments of Lake Constance, an oligotrophic freshwater lake.
Appl Environ Microbiol 77(13):44294436
Ettwig KF, Shima S, van de Pas-Schoonen KT, Kahnt J, Medema MH,
Op den Camp HJ, Jetten MS, Strous M (2008) Denitrifying bacteria
anaerobically oxidize methane in the absence of Archaea. Environ
Microbiol 10(11):31643173
Ettwig KF, van Alen T, van de Pas-Schoonen KT, Jetten MS, Strous M
(2009) Enrichment and molecular detection of denitrifying
methanotrophic bacteria of the NC10 phylum. Appl Environ
Microbiol 75(11):36563662
He Z, Cai C, Geng S, Lou L, Xu X, Zheng P, Hu B (2013) Modeling a
nitrite-dependent anaerobic methane oxidation process: parameters
identification and model evaluation. Bioresour Technol 147C:315
320
Hu BL, Zheng P, Tang CJ, Chen JW, van der Biezen E, Zhang L, Ni BJ,
Jetten MS, Yan J, Yu HQ, Kartal B (2010) Identification and
quantification of anammox bacteria in eight nitrogen removal reactors. Water Res 44(17):50145020
Hu BL, Shen LD, Zheng P, Hu AH, Chen TT, Cai C, Liu S, Lou LP
(2012) Distribution and diversity of anaerobic ammonium-oxidizing

bacteria in the sediments of the Qiantang River. Environ Microbiol

Rep 4(5):540547
Hu BL, Shen LD, Lian X, Zhu Q, Liu S, Huang Q, He ZF, Geng S, Cheng
DQ, Lou LPXXY, Zheng P, He YF (2014) Evidence for nitritedependent anaerobic methane oxidation as a previously overlooked
microbial methane sink in wetlands. Proc Natl Acad Sci U S A. doi:
10.1073/pnas.1318393111
Hu S, Zeng RJ, Burow LC, Lant P, Keller J, Yuan Z (2009) Enrichment of
denitrifying anaerobic methane oxidizing microorganisms. Environ
Microbiol Rep 1(5):377384
Jarvis P, Jefferson B, Gregory J, Parsons SA (2005) A review of floc
strength and breakage. Water Res 39(14):31213137
Jin RC, Hu BL, Zheng P, Qaisar M, Hu AH, Islam E (2008) Quantitative
comparison of stability of ANAMMOX process in different reactor
configurations. Bioresour Technol 99(6):16031609
Kampman C, Hendrickx TL, Luesken FA, van Alen TA, Op den Camp
HJ, Jetten MS, Zeeman G, Buisman CJ, Temmink H (2012)
Enrichment of denitrifying methanotrophic bacteria for application
after direct low-temperature anaerobic sewage treatment. J Hazard
Mater 227228:164171
Kojima H, Tsutsumi M, Ishikawa K, Iwata T, Mussmann M, Fukui M
(2012) Distribution of putative denitrifying methane oxidizing bacteria in sediment of a freshwater lake, Lake Biwa. Syst Appl
Microbiol 35(4):233238
Luesken FA, Sanchez J, van Alen TA, Sanabria J, Op den Camp HJ,
Jetten MS, Kartal B (2011a) Simultaneous nitrite-dependent anaerobic methane and ammonium oxidation processes. Appl Environ
Microbiol 77(19):68026807
Luesken FA, van Alen TA, van der Biezen E, Frijters C, Toonen G,
Kampman C, Hendrickx TL, Zeeman G, Temmink H, Strous M, Op
den Camp HJ, Jetten MS (2011b) Diversity and enrichment of
nitrite-dependent anaerobic methane oxidizing bacteria from wastewater sludge. Appl Microbiol Biotechnol 92(4):845854
Raghoebarsing AA, Pol A, van de Pas-Schoonen KT, Smolders AJ,
Ettwig KF, Rijpstra WI, Schouten S, Damste JS, Op den Camp
HJ, Jetten MS, Strous M (2006) A microbial consortium couples
anaerobic methane oxidation to denitrification. Nature 440(7086):
918921
Shen LD, He ZF, Zhu Q, Chen DQ, Lou LP, Xu XY, Zheng P, Hu BL
(2012) Microbiology, ecology, and application of the nitritedependent anaerobic methane oxidation process. Front Microbiol
3:269
Shen LD, Zhu Q, Liu S, Du P, Zeng JN, Cheng DQ, Xu XY, Zheng P, Hu
BL (2013) Molecular evidence for nitrite-dependent anaerobic
methane-oxidising bacteria in the Jiaojiang Estuary of the East Sea
(China). Appl Microbiol Biotechnol. doi:10.1007/s00253-0145556-3
Shi Y, Hu S, Lou J, Lu P, Keller J, Yuan Z (2013) Nitrogen removal from
wastewater by coupling anammox and methane-dependent denitrification in a membrane biofilm reactor. Environ Sci Technol 47(20):
1157711583
Strous M, Heijnen JJ, Kuenen JG, Jetten MSM (1998) The sequencing
batch reactor as a powerful tool for the study of slowly growing
anaerobic ammonium-oxidizing microorganisms. Appl Microbiol
Biotechnol 50:589596
Tamura K, Dudley J, Nei M, Kumar S (2007) MEGA4: molecular
evolutionary genetics analysis (MEGA) software version 4.0. Mol
Biol Evol 24(8):15961599
Tao Y, Gao DW, Fu Y, Wu WM, Ren NQ (2012) Impact of reactor
configuration on anammox process start-up: MBR versus SBR.
Bioresour Technol 104:7380
Zhu B, Sanchez J, van Alen TA, Sanabria J, Jetten MS, Ettwig KF, Kartal
B (2011) Combined anaerobic ammonium and methane oxidation
for nitrogen and methane removal. Biochem Soc Trans 39(6):1822
1825