The dynamics of the bacterial community associated with the denitrification process in a fixed bed column reactor (FBCR) were investigated using 454-pyrosequencing methodology. A FBCR filled with elemental sulfur and limestone was operated for about 94 days under autotrophic and mixotrophic (autotrophic + heterotrophic) conditions at 30 °C. Efficient simultaneous bromate and nitrate removal was achieved at feed concentrations of 500 µg/L bromate and 45 mg/L nitrate under autotrophic and mixotrophic conditions. Operational taxonomic units-based analysis (97% similarity cut-off) of bioreactor samples (three periods) revealed that the microbial diversity changed regardless of operational conditions. Sulfurimonas spp. was dominant in the reactor at the adaptation stage. Thiobacillus denitrificans is a chemolithoautotrophic bacterium that is capable of the oxidation of inorganic sulfur compounds. After the adaptation period, the microbial profile changed such that Spirochaetacea spp. and Denitratisoma spp. were major species in the column reactor. After 60 d of operation, Hyphomicrobium vulgare became dominant due to the mixotrophic denitrification conditions.

INTRODUCTION

Bromate (BrO3) can be formed during ozonation processes of natural water containing bromide (Br) (Gunten & Pinkernell 2000). Bromate is a drinking water pollutant that represents a risk for public health because of its carcinogenic potential. Therefore, the maximum allowable bromate concentration in drinking water has been set as 10 μg/L by the US Environmental Protection Agency (USEPA) and European Community Drinking Water Directive (Clark & Bouting 2001). In drinking water resources, nitrate and bromate can be present together, as nitrate is a common co-contaminant in surface and ground waters (Westerhoff & James 2001). The USEPA limits the maximum contaminant level of 10 mg/L nitrate for drinking water (Ucar et al. 2016). Nitrate can be found in drinking water resources also contaminated with bromate; however, bromate is usually present at low concentrations (<100 μg/L). It is a challenge to develop a remediation technology able to economically remove these pollutants from contaminated drinking water resources. Biological denitrification is a cost-effective process for removing nitrogenous compounds from water. Biological denitrification, either autotrophic or heterotrophic, has been proposed separately for nitrate and/or bromate removal. Our previous study (Demirel et al. 2013) showed that sulfur-based autotrophic and mixotrophic denitrification can be considered as an alternative process for nitrate and bromate removal from drinking water. In the sulfur-based autotrophic denitrification process, sulfur and nitrate act as an electron donor and an acceptor, respectively, without requirement of organic supplementation (Reaction (1)). Comparing to the electron donors, elemental sulfur (S0) as an attractive alternative electron donor possesses the characteristics of being non-toxic, water insoluble, stable under normal conditions, and readily available (Sahinkaya et al. 2013). 
formula
1

The main disadvantages of the autotrophic denitrification process are sulfate and acid generation. Sahinkaya et al. (2011) showed that elimination of sulfate generation by sulfur-based autotrophic denitrification processes is possible by stimulating a simultaneous autotrophic and heterotrophic (mixotrophic) denitrification process by methanol supplementation. This process minimizes individually the disadvantages of heterotrophic and autotrophic denitrification: the former generates alkalinity which the latter consumes; the result is an alkali-neutral reaction (Demirel et al. 2013).

In this research, an autotrophic and mixotrophic fixed bed column reactor (FBCR) was investigated to remove bromate and nitrate simultaneously. The objectives of this research were to (1) determine which autotrophic and heterotrophic denitrifying bacteria in the FBCR could reduce bromate to below the emission standard of 10 μg/L treatment objectives and (2) compare the microbial community structure under different operational conditions by using pyrosequencing technology.

METHODS

Design of column reactor and analytical methods

For a complete removal of nitrogenous compounds (NO3-N and NO2-N), denitrification conditions were provided in a laboratory-scale glass column reactor with an empty volume of 500 ml filled with elemental sulfur (0.5–1 mm) and limestone (0.5–1 mm) in ratios of 2/3 and 1/3, respectively. The inoculum was obtained from the first anoxic tank of a five-stage Bardenpho process. The reactor was supplemented with methanol during mixotrophic denitrification conditions. Hydraulic retention time (HRT) was calculated considering the empty bed volume (Table 1).

Table 1

Operation conditions used in FBCR

StagesPeriodDaysBromate feed (μg/L)NO3-N feed (mg/L)HRT (h)Methanol feed (mg/L)a
Start-up 0–10 – 45 10.1 – 
Autotropic 10–60 500 45 10.1 – 
Mixotrophic 60–94 500 45 10.1 75 (29.5) 
StagesPeriodDaysBromate feed (μg/L)NO3-N feed (mg/L)HRT (h)Methanol feed (mg/L)a
Start-up 0–10 – 45 10.1 – 
Autotropic 10–60 500 45 10.1 – 
Mixotrophic 60–94 500 45 10.1 75 (29.5) 

aValue in parentheses represents the methanol concentration as dissolved organic carbon (DOC).

Samples for analysis were collected from the reactor effluent at least three times a week for the measurement of NO3, NO2, bromate, bromide and pH. The feed solution was sampled once a week for the determination of NO3, NO2, bromate, bromide and pH. All samples were filtered over a 0.45 μm-pore-size sterile filter and stored at 4 °C until analysis. NO3, NO2 bromate and bromide concentrations were determined by ion chromatography. For identifying the changes in the microbial communities in the reactor, three different biomass samples (during the adaptation, autotrophic and mixotrophic periods) were taken from the sampling port located in the lower part of the reactor (Figure 1).
Figure 1

Scheme showing the application of 454-pyrosequencing analysis to the study of bacterial communities associated with denitrification.

Figure 1

Scheme showing the application of 454-pyrosequencing analysis to the study of bacterial communities associated with denitrification.

PCR amplification and high-throughput 16S rRNA gene pyrosequencing

DNA was extracted using a PowerSoil DNA isolation kit (MoBio) according to the manufacturer's instructions. Extracted DNA samples were stored at −20 °C until downstream applications. After linear amplification of DNA using the REPLI-G Mini Kit (Qiagen), amplified DNA samples were used for amplification of the bacterial V4 region of the 16S rRNA gene by using forward primer bac515f (GTGCCAGCMGCCGCGGTAA) and reverse primer bac806R (GGACTACVSGGGTATCTAAT). The polymerase chain reaction (PCR) mix contained 250 ng template DNA, 25 mM dNTP, 10× buffer with MgSO4, 2.5 U Pfu polymerase, and 0.2 μM of each primer in a 100 μL reaction mixture. The temperature conditions for PCR amplification was performed under the following conditions: initial denaturation at 94 °C for 3 min followed by 28 cycles of denaturation at 94 °C for 30 sec, primer annealing at 53 °C for 40 sec and primer extension at 72 °C for 1 min. The reaction was terminated after 5 min extension at 72 °C. Pyrosequencing of amplicons was performed on a Roche 454 Genome Sequencer FLX Titanium according to the manufacturer's protocol. The obtained DNA sequence was compared to search the closest 16S rRNA sequences of the reference microorganisms available in GenBank by BLAST.

RESULTS AND DISCUSSION

Nitrate and bromate removal performance in long-term operation

The column reactor was first operated under start-up conditions for 10 days (period 1, Table 1) at 45 mgNO3-N/L. After the adaptation period, the column reactor was operated under autotrophic conditions up to 60 days at 45 mg NO3-N/L and 500 μg BrO3/L. The reduction of bromate and nitrate was monitored during these periods (Figure 2). In the first 10 days of the operation of the autotrophic period, both nitrate and bromate were detectable in the effluent, although each decreased linearly to undetectable levels. Under steady-state conditions, the bromate concentration was less than 3 μg/L. The measured bromide concentration in the effluent demonstrated that all the bromate was stoichiometrically reduced without any by-product formation under steady-state conditions. Moreover, bromate and nitrate were removed simultaneously from the water, without detectable nitrite in the effluent.
Figure 2

Influent and effluent (a) NO3-N or NO2-N, and (b) BrO3 variations.

Figure 2

Influent and effluent (a) NO3-N or NO2-N, and (b) BrO3 variations.

After day 60, the FBCR was fed with methanol (75 mg/L) in addition to 45 mg NO3-N/L and 500 μg/L bromate (Table 1). Similarly to the autotrophic operating conditions, bromate and nitrate were removed under the mixotrophic conditions. During the operation period, a hydraulic residence time of 10.1 h was sufficient for complete nitrate and bromate transformation to dinitrogen gas and bromide, respectively (Figure 2).

Composition of the microbial community

In recent years, scientists have been understanding the structure and function of target communities and clarifying the link between them by new-generation microbial ecology techniques. New-generation molecular approaches have enabled deeper and more accurate analysis of complex ecosystems, leading to significant progress in identifying active species in the environment (Baytshtok et al. 2009).

Pyrosequencing analysis yielded a total of 43,105 sequence reads from three genomic DNA samples after removal of short and low-quality reads. The 454-pyrosequencing results indicated that microbial community structure changed through the periods depending on operational conditions (electron donor, energy source because of changing conditions such as autotrophic or mixotrophic etc.) in the FBCR. Proteobacteria was the main phylum in all operational stages standing out in the presence of denitrifiers (Figure 3). This bacterial composition was consistent with the previous study which investigated the microbial diversity of different operational conditions on the effect of denitrification (Park et al. 2011). The Proteobacteria accounted for 71.05%, 58.15% and 65.95% of the sequences which were related to the known bacteria during each of the operational stages, respectively. In addition Bacteroidetes and Chloroflexi were phyla common to all periods. Despite the limited knowledge about the role of Chloroflexi in denitrification, it might be related to reducing nitrate to nitrite (Kohno et al. 2002). β-Proteobacteria represented the most abundant class in all operational periods (21.52%, 24.81% and 31.51% for adaptation, autotrophic and mixotrophic stages, respectively). A functional role for members of β-proteobacteria in denitrifying biofilm has been suggested by Wrighton et al. (2010).
Figure 3

Relative abundance of the main phyla identified in the sulfur-based FBCR. Only phyla with a relative abundance greater than 1% are shown.

Figure 3

Relative abundance of the main phyla identified in the sulfur-based FBCR. Only phyla with a relative abundance greater than 1% are shown.

As seen in Figure 4, Sulfurimonas sp. was the most abundant species (22.11%) in the reactor at the adaptation stage. Sulfurimonas was the most commonly reported autotrophic denitrifier, which reduced nitrate to N2 while oxidizing elemental sulfur or reduced sulfur compounds to sulfate (Koenig et al. 2005; Shao et al. 2010). These bacteria in the column reactor should be responsible for sulfur-based autotrophic denitrification at the adaptation stage. After the adaptation period, the column reactor was operated at 500 μg/L of bromate concentration under autotrophic conditions. At this stage, the microbial profile changed so that Spirochaetaceae sp. (15.72%) and Denitratisoma sp. (8.33%) were major species in the column reactor. Denitratisoma have been playing an important role in the denitrification of nitrate as in the references reported by Mao et al. (2008, 2010). The distribution and intensity of the samples taken at different stages seemed to be identical, which can be explained by ion distribution data, which was consistent with patterns of nitrate and nitrite reduction in the column reactor. A similar distribution of predominant bacterial species in denitrification columns has also been reported (Koenig et al. 2005).
Figure 4

Relative abundance of the main species identified in the sulfur-based FBCR. Only species with a relative abundance greater than 1% are shown.

Figure 4

Relative abundance of the main species identified in the sulfur-based FBCR. Only species with a relative abundance greater than 1% are shown.

After day 60 (period 3), the FBCR was fed with methanol (Table 1). The facultative methylotrophs, e.g., Hyphomicrobium, and Methylophaga, which prominently belong to β-proteobacteria, were dominant species when methanol was added during the mixotrophic stage. These populations can be classified as obligate (growing on C1 compounds as sole source of carbon and energy only) and facultative methylotrophs (growing on C1 and multi-carbon compounds) (Lu et al. 2014). In this stage, Hyphomicrobium vulgare was the most abundant species (12.68%) due to the heterotrophic denitrification conditions (Figure 4). H. vulgare has been shown to grow anaerobically using nitrate as an alternative terminal electron acceptor and methanol as a sole carbon and energy source (Nikolausz et al. 2005).

The microbial activity in the bioreactor depends mainly on three steps: (i) sulfur particles (electron donor) must be microbially oxidized, (ii) nitrate (electron acceptor) must be microbially reduced, and (iii) bromate (via the co-metabolic action of nitrate reductase) must be reduced in a specific reduction pathway. In the mixotrophic denitrification ecosystem, the sulfate-reducing bacteria utilized the sulfate produced by the autotrophic denitrifiers, and the alkalinity generated in the heterotrophic denitrification could effectively compensate for alkalinity consumption by the autotrophic denitrification. Although the composition of the microbial community might reflect the source of the bacteria, it is believed that the experimental conditions must have led to the development of a stable and specialized community of denitrifying bioreactor.

CONCLUSION

A laboratory-scale mixotrophic denitrification column, using elemental sulfur (in the autotrophic stage) and sulfur/methanol (in the mixotrophic stage) as an electron donor, was successfully operated for over 94 days. During the operation, the performance and microbial community structure of a FBCR with combined autotrophic and mixotrophic denitrification was investigated. The 454-pyrosequencing analysis showed that the dominant species were Sulfurimonas spp. in the reactor at the adaptation stage. When bromate was added to the reactor, the microbial profile changed so that Spirochaetacea spp. and Denitratisoma spp. were major species in the column reactor. The 454-pyrosequencing analysis revealed that the autotrophic and heterotrophic bacteria coexisted in the mixotrophic environment.

ACKNOWLEDGEMENTS

This study was funded by TUBITAK (Project No. 111Y165).

REFERENCES

REFERENCES
Baytshtok
V.
Lu
H.
Park
H.
Kim
S.
Yu
R.
Chandran
K.
2009
Impact of varying electron donors on the molecular microbial ecology and biokinetics of methylotrophic denitrifying bacteria
.
Biotechnology and Bioengineering
102
(
6
),
1527
1536
.
Clark
R.
Bouting
B.
2001
Controlling Disinfection By-Products and Microbial Contaminants in Drinking Water. US Environmental Protection Agency, Cincinnati, OH, USA
.
Gunten
U. V.
Pinkernell
U.
2000
Ozonation of bromide-containing drinking waters: a delicate balance between disinfection and bromate formation
.
Water Science and Technology
41
,
53
59
.
Kohno
T.
Sei
K.
Mori
K.
2002
Characterization of type 1851 organisms isolated from activated sludge samples
.
Water Science and Technology
46
,
111
114
.
Mao
Y.
Zhang
X.
Xia
X.
Zhong
H.
Zhao
L.
2010
Versatile aromatic compound-degrading capacity and microdiversity of Thauera strains isolated from a coking wastewater treatment bioreactor
.
Journal of Industrial Microbiology and Biotechnology
37
(
9
),
927
934
.
Nikolausz
M.
Kappelmeyer
U.
Nijenhuis
I.
Ziller
K.
Kastner
M.
2005
Molecular characterization of dichloromethane-degrading Hyphomicrobium strains using 16S rDNA and DCM dehalogenase gene sequences
.
Systematic and Applied Microbiology
28
,
582
587
.
Sahinkaya
E.
Kilic
A.
Calimlioglu
B.
Toker
Y.
2013
Simultaneous bioreduction of nitrate and chromate using sulfur-based mixotrophic denitrification process
.
Journal of Hazardous Materials
262
,
234
239
.
Shao
M. F.
Zhang
T.
Fang
H. H. P.
2010
Sulfur-driven autotrophic denitrification: diversity, biochemistry, and engineering applications
.
Applied Microbiology and Biotechnology
88
,
1027
1042
.
Ucar
D.
Cokgor
E. U.
Sahinkaya
E.
2016
Heterotrophic–autotrophic sequential system for reductive nitrate and perchlorate removal
.
Environmental Technology
37
(
2
),
183
191
.
Westerhoff
P.
James
J.
2001
Nitrate removal in zero-valent iron packed columns
.
Water Research
37
(
8
),
1818
1830
.
Wrighton
K.
Virdis
B.
Clauwaert
P.
Read
S. T.
Daly
R. A.
Boon
N.
Piceno
Y.
Andersen
G. L.
Coates
J. D.
Rabaey
K.
2010
Bacterial community structure corresponds to performance during cathodic nitrate reduction
.
International Society for Microbial Ecology Journal
4
,
1443
1455
.