Perchlorate reduction by an isolated Serratia marcescens strain under high salt and extreme pH

Authors

  • Anupama Vijaya Nadaraja,

    1. Environmental Technology, CSIR-National Institute for Interdisciplinary Science & Technology, Thiruvananthapuram, India
    Search for more papers by this author
  • Prajeesh Gangadharan Puthiya Veetil,

    1. Environmental Technology, CSIR-National Institute for Interdisciplinary Science & Technology, Thiruvananthapuram, India
    Search for more papers by this author
  • Krishnakumar Bhaskaran

    Corresponding author
    • Environmental Technology, CSIR-National Institute for Interdisciplinary Science & Technology, Thiruvananthapuram, India
    Search for more papers by this author

Correspondence: Krishnakumar Bhaskaran, Environmental Technology, CSIR-National Institute for Interdisciplinary Science & Technology, Thiruvananthapuram 695 019, India. Tel.: +91 471 2515262; fax: +91 471 2491712; e-mail: krishna@niist.res.in

Abstract

An isolated Serratia marcescens strain exhibited growth-coupled perchlorate (math formula) reduction under anoxic conditions. Perchlorate was reduced completely with stoichiometric chloride buildup and equimolar acetate consumption. Polymerase chain reaction confirmed the presence of pcrA and cld genes coding for key enzymes involved in the math formula degradation pathway. The isolate degraded math formula under high salt (up to 15% NaCl) and a wide range of pH (4.0–9.0), as well as simultaneously reduced nitrate and math formula.

Introduction

Perchlorate (math formula) is an emerging environmental contaminant reported from many parts of the world. As per World Health Organization guidelines, the provisional maximum tolerable daily uptake of math formula in drinking water is 0.1 μg kg−1 body weight per day. Meanwhile, math formula has been traced in different matrices including human consumption products in many countries (Hogue, 2003; Sanchez et al., 2005; Kannan et al., 2009). Bioremediation has been identified as an effective method for detoxifying math formula (Attaway & Smith, 1993; Bardiya & Bae, 2011). Dissimilatory perchlorate-reducing bacteria under anoxic conditions utilize math formula as an electron sink, sequentially reducing it to Cl. Bacterial math formula reduction involves two enzymes, perchlorate reductase and chlorite dismutase, coded by pcrABCD and cld genes, respectively (Bender et al., 2004, 2005). Studies on the biochemical mechanism and genetic regulation of math formula reduction and diversity of perchlorate-reducing bacteria including their kinetics of growth and math formula reduction have been well reported (Logan, 1998; Coates & Achenbach, 2004; Yu et al., 2006). Perchlorate-laden waste/discharges like ion exchange spent resin and regenerate solution can have salinity up to 15% (Chung et al., 2007) and pH either alkaline or acidic (Batista et al., 2002). Moreover, inhibition of math formula reduction by competitive and co-occurring electron acceptors like nitrate (math formula) is encountered in a number of studies (Chaudhuri et al., 2002). Therefore, high salt- and exteme pH-tolerant perchlorate-reducing bacteria that reduce math formula and math formula simultaneously are significant. In spite of this, only few studies have reported math formula reduction under high salt and extreme pH conditions. The maximum tolerance to salinity and pH by perchlorate-reducing bacteria reported so far was 7.5% and 6–9, respectively, by a Citrobacter sp. (Okeke et al., 2002). Perchlorate-reducing consortia tolerating 11% salinity (Logan et al., 2001) and pH 5–9 (Wang et al., 2008) were also reported previously. In this study, we report a Serratia marcescens strain, tolerating high salt and extreme pH and reducing math formula and NO3 simultaneously.

Materials and methods

Isolation and characterization of the bacterium

This bacterium was isolated from a laboratory-scale fed-batch bioreactor (2.5 L working volume, hydraulic retention time = 9 days, solid retention time = 150 days) maintained in our laboratory for math formula reduction study since last 3 years. The reactor was fed with an inorganic mineral medium containing 225 mg L−1 K2HPO4, KH2PO4, and (NH4)2SO4, 200 mg L−1 yeast extract, 50 mg L−1 MgSO4·7H2O, 5 mg L−1 CaCO3, 0.5 mg L−1 FeSO4·7H2O, and 1.0 mL trace metal solution. math formula (Sigma) and sodium acetate equivalent to 1.5 g L−1 math formula and 3 g L−1 CH3COO were added to the medium. Redox potential of the reactor sludge was around −200 to −240 mV, pH around 7.0 ± 0.5, and temperature 30 ± 3 °C. Overall reduction of math formula in the reactor was 99% in 24 h during the study period. A portion of the reddish biofilm (Supporting Information, Fig. S1) from the bioreactor inner wall was scrapped off, rinsed with sterile water, and suspended in sterile saline. The suspension was subsequently spread plated on nutrient agar added with 25 mg L−1 math formula and incubated at 30 ± 3 °C for 24 h. Isolated, red-pigmented colonies formed were subcultured, and pure colonies obtained were transferred to nutrient agar slants and maintained under refrigerated conditions as stock culture.

Biochemical tests such as oxidase, urease, gelatinase, nitrate reduction, methyl red, Voges–Proskauer, and indole were performed as described earlier (Benson, 2002), and growth with various carbon sources such as glucose, lactose, starch, sucrose, and citrate were carried out to characterize the isolate. Genotypic characterization of the isolate was performed using 16S rDNA gene sequence analysis with universal primers (27f and 1492r). PCR amplification was performed with specific primers for pcrA (pcrA320F 5′-GCGCCCACCACTACATGT-AYGGNCC-3′ and pcrA598R 5′-GGTGGTCGCC-GTACCARTCRAA-3′) and cld (UCD-238F 5′-T(C/T)GA(A/C/G)AA(A/G)CA(C/T)AAGGA(A/T/C)AA(A/C/G)GT-3′ and UCD-646R 5′-GAGTGGTA(A/C/G)A(A/G)(C/T)TT(A/C/G)CG(C/T)TT-3′) genes (Bender et al., 2004; Nozawa-Inoue et al., 2008). The primers were procured from IDT-USA. Phylogenetic analysis was carried out using the mega software.

Growth and perchlorate reduction

Anoxic growth-coupled math formula reduction studies were conducted in 25-ml-capacity crimp cap vials. Medium supplemented with 10 mg L−1 math formula and 20 mg L−1 CH3COO was used in all the experiments. Dissolved oxygen in the medium was expelled by sparging air-free nitrogen from a cylinder filtered through 0.22-μm filters (Millipore) for 10 min. Cells from an anoxically grown culture at late log phase were centrifuged (8000 g), washed, and suspended in de-aerated medium in the vial (initial OD600 nm 0.01–0.02). Negative control vials with heat-killed inoculums were run parallel. The vials were incubated at 30 ± 3 °C, and samples were withdrawn using a syringe after every 12 h for estimating math formula, Cl, CH3COO, and OD600 nm. Specific growth rate (μ) of the bacterium was calculated from the slope of line from lnOD600 nm against time plot. Biomass yield (Y) was calculated by accounting biomass (dry wt.) produced and corresponding math formula reduced. Growth and math formula reduction were also tested with alternate electron donors like propionate, butyrate, succinate, and glucose in place of acetate under similar experimental conditions.

Perchlorate reduction under high salt and extreme pH

Effect of NaCl and pH on math formula reduction was studied under a similar experimental setup with modifications. A culture acclimatized to 3% NaCl was used for math formula reduction under high salt condition. The inoculum was added to medium containing 5, 10, and 15% NaCl in separate bottle sets. A control experiment was also run without NaCl. The effect of pH was studied within the range of 4.0–9.0. Culture acclimatized to pH 4.0 was used as inoculum for medium at pH 4.0.

Perchlorate reduction under denitrifying conditions

The effect of math formula on math formula reduction was studied by adding KNO3 (10 mg L−1 math formula) to medium under similar experimental conditions as above. Nitrate-free positive control and heat-killed (non biological) controls were run parallel. The effect of oxygen on math formula reduction was studied by conducting the experiment under aerobic conditions in an environmental shaker at 125 r.p.m. (dissolved oxygen = 7.6 mg L−1). The medium used and sampling interval were similar to that in anoxic experiments.

Analysis

math formula, Cl, and CH3COO were estimated through IC (Dionex-1100) with AS 11 column and anion self-regenerating suppressor (ASRS 300, 4 mm) with NaCO3/NaHCO3 (1 : 4.8 mM) buffer as mobile phase at flow rate of 1.5 mL min−1. The mean recovery of the estimated compounds with the present column and analytical condition was 100 ± 10%.

All batch experiments were repeated with duplicates, and the average values were expressed with standard deviation.

Result and discussion

Identification and characterization of the isolate

The biofilm suspension when spread on to math formula-supplemented nutrient agar plates produced red-pigmented colonies (2–3 mm diameter) in 24-h period under light exposure and ambient temperature (30 °C). Colony pigmentation was not observed in plates incubated in the dark. But nonpigmented colonies developed pigmentation on 2-hr exposure to light. Based on phenotypic, biochemical, and phylogenetic data (Table S1 and Fig. S2), the isolated bacterium was identified as Serratia sp. (γ-proteobacteria, Enterobacteriaceae). Analysis of 16S rDNA gene sequence (GenBank No. JQ807993) revealed 99% identity with S. marcescens, and hence, the present strain was designated as S. marcescens strain NIIST5. It has the unique property of reducing math formula owing to the presence of pcrA and cld genes (Fig. S3) coding for key respiratory enzymes, perchlorate reductase and chlorite dismutase, respectively, involved in perchlorate reduction pathway (Bender et al., 2004; Nozawa-Inoue et al., 2008). GenBank Nos. JX993983 and KC190515 were assigned to pcrA and cld gene sequences, respectively. Using acetate as electron donor, the isolate reduced 10 mg L−1 math formula to 1.3 mg L−1 with stoichiometric chloride buildup and equimolar acetate consumption in 72 h (Fig. 1a). When acetate was substituted with propionate, butyrate, succinate, and glucose, the isolate failed to grow and reduce math formula. Acetate is the preferred electron donor for most (not all) of the heterotrophic perchlorate-reducing bacteria reported. However, the molar consumption of acetate for math formula reduction has varied among pure as well as mixed cultures. Studies have reported acetate/math formula ratio up to 6.0 for complete reduction of math formula in pure cultures (Kim & Logan, 2001). Lower acetate requirement as with the case of present Serratia sp. might reduce operational cost of math formula remediation systems. The specific growth rate (μ) of the bacterium during the period was 0.08/h, and biomass yield was 0.24 mg cell dry weight per mg math formula. Growth rates and cell yield of dissimilatory perchlorate-reducing bacteria under heterotrophic condition ranged between 0.07–0.28 h−1 and 0.24–0.50 mg cells mg−1 math formula reduced, respectively (Bardiya & Bae, 2011).

Figure 1.

(a) Anoxic reduction of math formula (○) by the Serratia marcescens strain and corresponding Cl (▲) build up, CH3COO (●) consumption and growth (□: OD600 nm) increase in inorganic mineral medium (pH = 7.0); (b) anoxic reduction of math formula in the presence of alternative electron acceptors (□: math formula; ♢: O2). Controls with math formula (○) and math formula (∆) alone. Data points are mean (n = 4) with standard deviation.

Perchlorate reduction under high salt and extreme pH conditions

The isolated bacterium exhibited growth-coupled math formula reduction under high NaCl and extreme pH conditions (Fig. 2). Compared with normal medium (μ = 0.08 h−1), growth rate was reduced to 0.018 and 0.017 h−1 at 10% and 15% NaCl, respectively. Similarly, the growth rate was reduced to 0.065 and 0.076 h−1 at pH 4.0 and 9.0, respectively. Growth and math formula reduction were completely absent above 15% NaCl and pH below 4.0 and above 9.0. High salt and extreme pH tolerance of perchlorate-reducing bacteria are preferred especially for treating brine from ion exchange systems for math formula treatment. But, only a few studies have reported high salt- and/or extreme pH-tolerating perchlorate-reducing bacteria. Bacteria belonging to genera Halomonas and Marinobacter were identified as dominant population in ion exchange brine (5.6% NaCl) from math formula treatment unit (Zuo et al., 2009). Furthermore, cloning and sequencing have revealed Clostridium sp. and Rhodocyclaceae as dominant math formula-degrading population in a packed-bed bioreactor at 10% NaCl (Chung et al., 2009). The maximum salinity tolerance in perchlorate-reducing bacteria reported so far was 11% by an acclimatized mixed culture (Logan et al., 2001). Most of perchlorate-reducing bacteria reported so far have optimum pH around 7.0 for math formula reduction (Wolterink et al., 2005; Thrash et al., 2010). However, Wang et al. (2008) have reported math formula reduction in pH range 5.0–9.0 by a mixed culture, where a drastic decline in reduction beyond pH 6.0 and 8.0 was observed. A pure culture study with Citrobacter sp. has reported math formula reduction within pH 6.0–9.0 (Okeke et al., 2002). Compared with previous reports, the S. marcescens strain in this study has comparatively high salt and wide pH tolerance.

Figure 2.

Specific growth rate of Serratia marcescens strain under different pH (∆) and NaCl (○) during incubation in IMM containing 10 mg L−1 math formula and 20 mg L−1 acetate in batch culture.

Perchlorate reduction under denitrifying conditions

Simultaneous reduction of nitrate and math formula was another characteristic of the present S. marcescens strain. Although both compounds were reduced simultaneously, nitrate was reduced more rapidly compared with math formula (Fig. 1b). Oxygen completely inhibited math formula reduction by the isolate. Contrasting results have been reported about the effect of competitive electron acceptors like nitrate and oxygen on math formula reduction. A number of studies have reported oxygen and nitrate inhibition of math formula reduction in varying intensities, depending on the culture type, concentration, and duration (Coates et al., 1999; Chaudhuri et al., 2002; Song & Logan, 2004). Studies have also reported simultaneous nitrate and math formula reduction by pure cultures as well as in bioreactor with mixed cultures (Chaudhuri et al., 2002; Cang et al., 2004; Min et al., 2004). In simultaneous nitrate and math formula reduction by the pure culture Dechlorosoma suillum and in environmental samples, complete reduction of nitrate was observed before the commencement of math formula reduction (Chaudhuri et al., 2002; Nozawa-Inoue et al., 2011). Simultaneous reduction of math formula and math formula was reported in a known perchlorate-reducing bacterium, perlace, where the latter was reduced more rapidly (Herman & Frankenberger, 1999). A similar observation was recorded with the S. marcescens strain. Probably, math formula and math formula reduction pathways are different in the bacterium, but this needs to be verified experimentally. A previous study with Dechloromonas sp. KJ has revealed the presence of separate pathway for nitrate and math formula reduction, and both are induced separately by the respective compound (Xu et al., 2004).

Characteristics of the isolated S. marcescens strain in this study, complete reduction of math formula, preference to CH3COO as electron donor, and inhibition of math formula reduction by O2 are similar to most of perchlorate-reducing bacteria reported (Attaway & Smith, 1993; Song & Logan, 2004). Unlike previously reported perchlorate-reducing bacteria, the present S. marcescens strain degrades math formula under high salt and extreme pH conditions, as well as simultaneously reducing nitrate and math formula. This is the first report of a Serratia sp. exhibiting anoxic respiration of math formula. Only a few perchlorate-reducing bacteria belongs to γ-proteobacteria have been reported including Citrobacter sp. (Okeke et al., 2002) and Pseudomonas chloritidismutans (Wolterink et al., 2003), strains JB101 and 109 (Bardiya & Bae, 2004). Therefore, the new strain in this study will expand the math formula-reducing, γ-proteobacteria lineage.

Acknowledgements

This study was conducted with the financial assistance from Kerala State Biotechnology Commission (KSTEC), project 03/YIPB/KBC/2009/CSTE. The infrastructural support from CSIR-India is also acknowledged. Anupama Vijaya Nadaraja thanks CSIR-India for the research fellowship.

Ancillary