Impact of temperature, CO2 fixation and nitrate reduction on selenium reduction, by a paddy soil Clostridium strain

Authors

  • P. Bao,

    1. State Key Lab of Urban and Regional Ecology, Research Center for Eco-Environmental Sciences, Chinese Academy of Sciences, Beijing, China
    Search for more papers by this author
  • H. Huang,

    1. State Key Lab of Urban and Regional Ecology, Research Center for Eco-Environmental Sciences, Chinese Academy of Sciences, Beijing, China
    Search for more papers by this author
  • Z.-Y. Hu,

    1. College of Resources and Environment, Graduate University of Chinese Academy of Sciences, Beijing, China
    Search for more papers by this author
  • M.M. Häggblom,

    1. Rutgers University, Department of Biochemistry and Microbiology, School of Environmental and Biological Sciences, New Brunswick, NJ, USA
    Search for more papers by this author
  • Y.-G. Zhu

    Corresponding author
    1. Key Lab of Urban Environment and Health, Institute of Urban Environment, Chinese Academy of Sciences, Xiamen, China
    • State Key Lab of Urban and Regional Ecology, Research Center for Eco-Environmental Sciences, Chinese Academy of Sciences, Beijing, China
    Search for more papers by this author

Correspondence

Y.-G. Zhu, Key Lab of Urban Environment and Health, Institute of Urban Environment, Chinese Academy of Sciences, Xiamen 361021, China. E-mail: ygzhu@rcees.ac.cn

Abstract

Aims

To elucidate the impact of CO2 fixation, nitrate reduction and temperature on selenium reduction by a newly identified acetogenic bacterium, Clostridium sp. BXM.

Methods and Results

A series of culture experiments were designed to evaluate the impact of temperature, CO2 fixation and nitrate reduction on the rate and extent of selenium reduction by strain BXM. The products of selenium reduction, CO2 fixation and nitrate reduction were determined. Molecular analysis was performed to identify the functional genes involved in the selenium reduction process. CO2 may have enhanced the activity of hydrogenase I and/or the level of cytochrome b, thus increasing selenium reduction. Nitrate may inhibit selenium reduction due to its higher reduction potential and/or by decreasing selenite/selenate reductase activity. The suitable temperature was 37 and 30°C for selenite reduction under anaerobic and aerobic conditions, respectively. The optimum temperature was 30°C for selenate reduction under both anaerobic and aerobic conditions. CO2 fixation and nitrate reduction by Clostridium sp. BXM stimulated each other.

Conclusions

Clostridium sp. BXM was capable of reducing up to 36–94% of 1 mmol l−1 selenate and selenite under anaerobic or aerobic conditions over 15 days. The strain might be used for the precipitation of Se from highly selenium-contaminated water or sediments.

Significance and Impact of the Study

The findings contribute to the current understanding about the role that micro-organisms play in the detoxification of toxic selenium compounds in paddy soils. Micro-organisms in paddy soils can influence selenium accumulation in rice grain and hence human selenium intake.

Introduction

Selenium ranks 66th in abundance among the elements in the Earth's crust (avg. concentration, 0·05 mg kg−1) (Greenwood and Earnshow 1997) and occurs in the environment with the oxidation states of +6, +4, 0 and −2. Selenium plays fundamental roles in both thyroid hormone and antioxidant production, as a vital component of glutathione peroxidase and thioredoxin reductase, in addition to a number of other structural and metabolic functions (Rayman 2000; Combs 2001). Selenium, however, is also a very toxic element to cells at concentrations above the critical level (Rayman 2000). The Se oxyanions, math formula and math formula, are soluble and highly toxic and have been shown to be assimilated and to bioaccumulate in cereals (Cao et al. 2001; Williams et al. 2009; Zhu et al. 2009; Sun et al. 2010; Winkel et al. 2012), thus affecting the nutritional quality of the grains. Approximately 50% of the human population lives on a diet of rice, not only for energy intake but also for a significant proportion of protein and micronutrients, including selenium. An insufficient dietary intake of Se has been estimated for up to 1 billion people worldwide (Williams et al. 2009). Selenium levels in rice vary greatly, depending on geochemical conditions, agricultural management regimes and nonoptimal selenium intakes (Welch and Graham 2005; Rayman et al. 2008; White and Broadley 2009; Williams et al. 2009). Thus, optimizing the levels of selenium in rice grain can have significant impact on the human selenium status.

Micro-organisms play a key role in selenium transformations that involve the conversion of the highly soluble and biologically available math formula and math formula to insoluble and relatively nonavailable Se0. The bacterial reduction of math formula to math formula is known to occur through dissimilatory pathways (Macy et al. 1989; Switzer et al. 1998; Stolz and Oremland 1999; Zehrt and Oremland 1987; Narasingarao and Häggblom 2006, 2007a; Rauschenbach et al. 2011). The reduction of math formula has been shown to be mediated by a soluble periplasmic selenate reductase (SerABC) (Schröder et al. 1997), nitrate reductase (Steinberg et al. 1992; Gates et al. 2011a) or via a dissimilatory sulfate-reducing pathway (Stolz and Oremland 1999; Hockin and Gadd 2006). The reduction of math formula to Se0, a common feature of diverse micro-organisms, is not very well understood. It has been reported that math formula reduction may be catalysed by a periplasmic nitrite reductase (DeMoll-Decker and Macy 1993), hydrogenase I (Yanke et al. 1995) or through nonenzymatic reactions (Tomei et al. 1992).

Members of the genus Clostridium are chemoorganotrophic ubiquitous soil micro-organisms (Rui et al. 2009; Wang et al. 2009; Li et al. 2011). Most of them are acetogens and utilize the acetyl-CoA ‘Wood–Ljungdahl’ pathway for the reductive synthesis of acetate from CO2 as an energy-conserving, terminal electron-accepting process (Wood and Ljungdahl 1991; Drake 1994). Clostridium species are generally obligatory anaerobes (Holdeman et al. 1977; Sneath 1986), capable of utilizing inorganic sulfur compounds or nitrate as terminal electron acceptors in addition to diverse fermentations of organic matter or acetogenic reduction of CO2. We recently isolated a Clostridium, a paddy soil facultative anaerobic strain (Bao et al. 2012) that was found to reduce selenium oxyanions. The Clostridium sp. BXM is a good model to study the combined effect of carbon, nitrate and selenium in paddy soil for it can reduce selenium, nitrate and at the same time CO2 fixation. The aim of this study was to investigate the reduction of selenium (math formula and math formula) by Clostridium sp. BXM and determine the combined effects of selenium reduction, CO2 fixation and nitrate reduction, as well as the impact of common environmental conditions (temperature, pH and O2) on selenium reduction.

Materials and methods

Organism, physiological characterization, culture medium and H2–CO2 chemolithoautotrophic growth test

The bacterial strain used throughout this work was originally isolated from a paddy soil and identified as a Clostridium sp. (Bao et al. 2012). Cultures were maintained throughout in a basal medium containing (g l−1) K2HPO4, 0·2; MgCl2·6H2O, 0·4; CaCl2·2H2O, 0·1; NH4Cl, 0·25; and KCl, 0·5. The additional solutions were (ml l−1) as follows: Selenite–tungstate solution (0·02 mmol l−1), 0·1; vitamin mixture (10·0 mmol l−1), 1·0; thiamine solution (10·0 mmol l−1), 1·0; vitamin B12 solution (10·0 mmol l−1), 1·0; and yeast extract (10·0 g in a final volume of 100 ml), 5·0. The basal medium was autoclaved and cooled to room temperature under an atmosphere of N2 (purity = 99·999%) (Widdel and Bak 1992). The additional solutions were filter sterilized/autoclaved individually. The pH of the medium was adjusted to 7·0 with HCl solution after injection of additional solutions. Selenium (1·0 mmol l−1), nitrate (1·0 mmol l−1) and NaHCO3 (30 mmol l−1) solutions were added as delineated for the different experiments. Then, the medium was transferred into 100-ml serum bottles under an atmosphere of N2 and sealed with butyl rubber stoppers and aluminium crimp caps.

Experiments were conducted in duplicate to test the pH and temperature range for growth of strain BXM in anoxic freshwater medium. The strain was subcultured at least once under the same experimental conditions. For pH studies, the medium was adjusted with anaerobic stock solutions of either HCl or NaOH to give the desired pH. Growth temperature was determined from 5 to 50°C. The DNA base composition of strains BXM was determined using a Tm technique using a Lambda 35 UV/VIS Spectrometer (Perkin/Elmer, Waltham MA, USA). E. coli K12 (CGMCC 1·365) was used as a reference sample. The DNA G + C content (mol%) of strain BXM was calculated using the formula: G + C mol% = G + C mol%AS1·365+2·08 (TmBXM-TmAS1·365). Cellular fatty acids of strain BXM were analysed from cells grown on anoxic freshwater medium at 30°C for 2 weeks. The bacterial culture material was used for lipid extraction according to the procedure of White et al. (1979) and identified by Sherlock Microbial Identification System (MIDI). DNA for DNA–DNA hybridizations analysis was prepared according to the procedure of Ausubel et al. (1995). DNA–DNA hybridization was performed according to the method of Ezaki et al. (1989). The DNA–DNA hybridization values are the means of two independent experiments. All hybridization reactions were carried out in quadruplicate and calculations were based on the mean fluorescence values. Growth experiments were performed in duplicate to test for carbon source and electron acceptor utilization by strain BXM. Elemental sulfur, nitrate, selenate and selenite as electron acceptors were added to the medium at initial concentrations of 1·0 mmol l−1. Yeast extract (1·0 g l−1 final concentration), lactate and glucose (1·0 mmol l−1 final concentration) were tested as carbon sources. Growing cultures were transferred two times into fresh media with the same carbon source/electron acceptor combination to ensure that growth was not due to residual carbon source/electron acceptor from the original stock. Growth was determined after 8-days cultivation and compared to controls without carbon source/electron acceptor.

The H2–CO2 chemolithoautotrophic growth of strain BXM was measured in medium with math formula (10 mmol l−1) as the only carbon source. The cultivation gas phase was 100% H2 that served as an energy source (Fröstl et al. 1996). Samples were taken from the culture after 2, 6, 9, 12, 15 days and analysed for cell growth (OD; A600, absorbance at 600 nm) and acetate production. Cell growth was measured by a spectrophotometer (Pgeneral T6, Beijing Purkinje General Instrument Co., Ltd, Haidian, Beijing, China). Acetate was measured by ion chromatography as described in experimental protocols and analytical methods section.

Experimental protocols and analytical methods

To examine the impact of nitrate, CO2 on the rate and extent of selenate and selenite reduction by Clostridium sp. BXM, four batches of experiments were initiated (Table S1): (I) selenium (1·0 mmol l−1), (II) selenium (1·0 mmol l−1) + NaHCO3 (30·0 mmol l−1), (III) selenium (1·0 mmol l−1) + NaNO3 (1·0 mmol l−1) and (IV) selenium (1·0 mmol l−1) + NaHCO3 (30·0 mmol l−1) + NaNO3 (1·0 mmol l−1), and cultivated anaerobically in 100-ml serum bottles in the dark. Each batch of experiments were established in triplicate, inoculated with 1·0 ml (OD600 0·16) Clostridium sp. BXM exponential-phase culture and incubated at 25, 30 or 37°C, respectively. Samples were taken from the medium after 0, 6, 9, 12 and 15 days of incubation and analysed for selenite, selenate and acetate in set (I) and (II), and selenite, selenate, acetate, nitrate, nitrite and ammonium in set (III) and (IV). To determine selenite, selenate, acetate, nitrate and nitrite, 0·5 ml of the sample was filtered (0·22 μm) to remove particulates (e.g. selenium granules) that could interfere with ion chromatography. The ion chromatography system consisted of a GP50 gradient pump (Thermo Fisher Scientific Inc. Sunnyvale, CA, USA), a column oven LC25, an electrochemical detector ED50. The ion chromatography column system used was a Dionex Ionpac AS14 column (4·6 × 3100 mm). The operating condition was modified from Knight et al. (2002) with an eluent of 3·5 mmol l−1 Na2CO3 and 1·0 mmol l−1 NaHCO3 at a flow rate of 1·2 ml min−1. The concentration of ammonium was measured by the indophenol-blue colorimetric method and monitored by spectrophotometry (Pgeneral T6).

Two sets of experiments were designed to evaluate the impact of pH (6·5–7·5) on anaerobic selenium reduction at 30°C (Table S1). The initial selenium concentration was 1·0 mmol l−1. The concentration of selenate and selenite was measured at day 15 by ion chromatography. Another batch of experiment was adopted to evaluate the impact of O2 on the selenium reduction. The culture medium was prepared under aerobic conditions, and the dissolved O2 was measured with a Hach sensION 156 Meter (Hach Company, Loveland, CO, USA). Cultures were incubated at 25, 30 and 37°C. The other experimental conditions (concentrations of chemicals, sampling intervals, etc.) and analytical methods were as described above. All studied experimental conditions were summarized in Table S1.

Electron microscopy and measurement of precipitated selenium

Electron microscope–based energy-dispersive X-ray (EDX) spectrometry analysis was carried out using a scanning electron microscope (Hitachi S-2000N, Hitachi, Ltd, Tokyo, Japan) equipped with scanning and point EDX facilities. Cells from an exponential-phase culture were collected by centrifugation, fixed with 2·5% glutaraldehyde solution (pH 7·5), critical point dried through an ethanol series and mounted on aluminium stumps. The specimens were coated with gold in a sputter coater (Polaron SC7620, Quorum, UK) prior to microscopy. The material was examined immediately at an accelerating voltage of 15 kV.

Molecular analysis

Cell material for the sequencing of the selenium-reducing genes was taken from 4 ml of an exponential-phase culture. All general molecular techniques were performed according to described standard methods (Sambrook et al. 1989). Extraction of genomic DNA was performed with a FastDNA Spin kit (Takara) according to the manufacturer's protocol. The nasA (nitrate assimilation) gene of strain BXM was amplified using published primers and PCR conditions (Allen et al. 2001) (Table S2). The nirK (nitrite reduction gene) PCR amplification procedure was modified from a previously published procedure (Chénier et al. 2003) (Table S2). Five different degenerate PCR primer pairs for hydrogenases were designed, and PCR amplification was carried out according to Calusinska et al. (2011) (Table S2). Sequencing of PCR products was performed by GIGA Genomics Facility in Invitrogen Life Technologies Corporation, Shanghai, China. Phylogenetic analyses were conducted using MEGA4 (Tamura et al. 2007).

Results

Physiological and phylogenetic characterization of strain BXM

Strain BXM is a facultative anaerobic, nonflagellated, nonmotile, spore-forming, Gram-positive fusiform rod. Individual cells were 2–3 μm in length and 0·3–0·5 μm wide. Based on 16S rRNA gene sequence phylogeny (GenBank accession number: JN092128), strain BXM is identified as a Clostridium spp. (Bao et al. 2012) with Clostridium sporogenes ATCC 3584T (X68189) and Clostridium sulfidigenes SGB2T (EF199998) as its closest relatives. The genomic DNA G+C content was 26·3 mol%. The major cellular fatty acids of strain BXM were C14:0 (27·7%) and C16:0 (34·6%). DNA–DNA hybridizations indicated that strain BXM had 36% (range of duplicate samples 33·8–39·2%) similarity with Clostridium sulfidigenes SGB2T, which suggests that strain BXM may represent a novel species. Strain BXM had a growth temperature range of 21–45°C and pH range of 6·0–7·6, with an optimum at 7·0. It could use yeast extract, lactate and glucose as carbon substrates and reduce sulphur, nitrate, selenate and selenite. Strain BXM represented H2–CO2 chemolithoautotrophic growth, and acetate accumulated up to 2·7 mmol l−1 until the decline phase (Fig. S1).

Reduction of selenium oxyanions by Clostridium sp. BXM

Clostridium sp. BXM reduced math formula and math formula with production of a red precipitate (data not shown), suggesting the formation of amorphous Se0, as opposed to crystalline Se0, which is grey in colour. EDX spectroscopy elemental analysis of the precipitate from a stationary-phase culture showed that all cultures contained Se0 (Fig. 1). A characteristic energy peak for Se was found at 1·45 keV. The elemental selenium particles have nearly the same diameter about 0·5 μm. It should be noted that EDXs does not provide definitive proof that the Se is the only element in the deposit, but it confirms the presence of this element. Other selenium compounds were not determined and we did not quantify elemental selenium produced by strain BXM.

Figure 1.

Scanning electron microscopy (SEM) of cells grown in the presence of math formula (a) and math formula (c): small granules were Se0; Energy dispersive X-ray spectroscopy (EDX) spectrum of the electron-dense particles (X-ray emission line for selenium is at 1.45 keV) (b for math formula, d for math formula).

Impact of CO2 fixation, math formula on the reduction of selenium and interaction of math formula and math formula

Clostridium sp. BXM reduced math formula and math formula as shown in Figs 2 and 3. No loss of math formula and math formula was observed in the medium controls, indicating that the reduction was biotically driven. The addition of CO2 to the culture medium resulted in the increase in math formula (Fig. 2) and math formula (Fig. 3) reduction by strain BXM. Over 60% of math formula was reduced after 9-days incubation with math formula amendment compared to 14% without (Fig. 2). By 15 days, the extent of math formula reduction was 63% and 44%, respectively, for math formula amended and un-amended cultures. The extent of math formula reduction and accumulation of math formula in cultures amended with CO2 were 15% and 0·13 mmol l−1 after 15-days incubation at 30°C, which was higher than with math formula alone (Fig. 3).

Figure 2.

Effect of math formula and math formula on selenite reduction by Clostridium sp. BXM: (I) math formula (1.0 mmol l−1), (II) math formula (1.0 mmol l−1) + math formula (30 mmol l−1), (III) math formula (1.0 mmol l−1) + math formula (1.0 mmol l−1), (IV) math formula (1.0 mmol l−1) + math formula (30 mmol l−1) + math formula (1.0 mmol l−1). Time-course plots show the math formula concentration. Error bars represent ± SD (n = 3). (image_n/jam12084-gra-0001.png) Control; (image_n/jam12084-gra-0002.png) Se(IV); (image_n/jam12084-gra-0003.png) Se(IV) (+C); (image_n/jam12084-gra-0004.png) Se(IV) (+N); (image_n/jam12084-gra-0005.png) Se(IV) (+C+N).

Figure 3.

Effect of math formula and math formula on selenite reduction by Clostridium sp. BXM: (I) math formula (1.0 mmol l−1), (II) math formula (1.0 mmol l−1) + math formula (30 mmol l−1), (III) math formula (1.0 mmol l−1) + math formula (1.0 mmol l−1), (IV) math formula (1.0 mmol l−1) + math formula (30 mmol l−1) + math formula (1.0 mmol l−1). Time-course plots showing culture math formula and math formula concentration at 30oC. Error bars represent ± SD (n = 3). (image_n/jam12084-gra-0001.png) Control; (image_n/jam12084-gra-0002.png) Se(IV); (image_n/jam12084-gra-0003.png) Se(IV); (image_n/jam12084-gra-0004.png) Se(IV) (+C); (image_n/jam12084-gra-0005.png) Se(IV) (+C); (image_n/jam12084-gra-0006.png) Se(VI) (+N); (image_n/jam12084-gra-0007.png) Se(IV) (+N); (image_n/jam12084-gra-0008.png) Se(VI) (+C+N); (image_n/jam12084-gra-0009.png) Se(IV) (+C+N).

Nitrate inhibited selenite and selenate reduction, with only approximately 5% math formula reduction (Fig. 2) and no significant math formula reduction detected after 15-days incubation with math formula amendment (Fig. 3). The extent of selenite and selenate reduction of CO2 + math formula -amended cultures was higher than with math formula only. Approximately 20% of math formula was reduced after 15-days incubation at 30°C in CO2 + math formula -amended cultures (Fig. 2), while the extent of math formula reduction was only 6% (Fig. 3). Although nitrate inhibited selenium reduction, this result was in accord with the positive impact of CO2 fixation on selenium reduction.

Figure 4 shows the production of acetate in bicarbonate-amended cultures after 15 days of incubation. The amounts of acetate in math formula  + CO2 treatments were 0·8 mmol l−1, 1·0 mmol l−1 and 0·6 mmol l−1, respectively, and 2·3 mmol l−1, 1·8 mmol l−1 and 2·0 mmol l−1 in math formula  + CO2 treatments at 25, 30 and 37°C. Interestingly, nitrate increased the production of acetate in this study.

Figure 4.

Acetate production in bicarbonate amended cultures (with selenium and/or nitrate) at day 15. Error bars represent ± SD (n = 3). Acetate production in cell free controls was not detected. (image_n/jam12084-gra-0010.png) 25°C; (image_n/jam12084-gra-0011.png) 30°C; (image_n/jam12084-gra-0012.png) 37°C.

Figure 5 shows the reduction of math formula after 15 days of incubation. Nearly all the math formula was reduced to nitrite and/or ammonium, except for the math formula  + CO2 + math formula treatment group at 25°C. Assimilation of nitrate may account for the imbalance of the total nitrogen in the cultures.

Figure 5.

Nitrate reduction and nitrite, ammonium production in nitrate-amended (initial concentration 1.0 mmol l−1) cultures of strain BXM (with math formula and math formula) at 15 days: (a) math formula + math formula, (b) math formula + math formula + math formula, (c) math formula + math formula, (d) math formula + math formula + math formula. The columns for each group (left to right) represent incubation at 25, 30 and 37°C, respectively. No loss of nitrate was observed in cultures without BXM (not shown). Error bars represent ± SD (n = 3). (image_n/jam12084-gra-0012.png) Nitrate; (image_n/jam12084-gra-0011.png) Nitrite; (image_n/jam12084-gra-0010.png) Ammonium.

Impact of temperature, pH and O2 on selenium reduction

The extent of math formula reduction by strain BXM increased with increasing temperature, with 25%, 44% and 94% of math formula loss after 15-days incubation at 25, 30 and 37°C, respectively (Table S3). The extent of math formula reduction was lower in comparison with math formula at the same temperature. The math formula reduction levels were 14%, 8% and 12% after 15-days incubation at 25, 30 and 37°C, respectively (Table S3). math formula is an intermediate during the reduction of math formula to Se0. A higher temperature had no significant stimulating effect on the reduction of math formula with CO2/nitrate amendment. The optimum pH for the math formula reduction was approximately 7·0. A lower or higher pH decreased the reduction of selenium (Table 1). The reduction of selenate and selenite also occurred under aerobic conditions with dissolved oxygen of 7·7 mg l−1 (Table 1).

Table 1. Impact of pH and O2 on selenium reduction. Cultures were spiked with 1 mmol l−1 math formula and math formula and loss measured after 15 days of incubation. Mean values are ± SD (n = 3)
SeleniumReduction (%) at different pH at 30°CReduction (%) under aerobic conditions at different temperatures
pH 6·5pH 7·0pH 7·525°C30°C37°C
math formula 50 ± 4·262 ± 2·246 ± 1·019 ± 3·241 ± 3·118 ± 2·5
math formula 10 ± 2·115 ± 0·29 ± 3·920 ± 0·936 ± 3·74 ± 2·0

Reductase genes identified in strain BXM

Strain BXM possesses a nitrite reductase gene nirK, and a 377-bp portion of this gene was amplified (GenBank accession no. JX075255) and compared with sequences from GenBank (Fig. 6a). The gene sequence of the nirK of BXM was highly similar to Paracoccus sp. R-24650 (AM230830.1) based on a Blast analysis. A portion of a [FeFe] hydrogenase gene hydB2 (GenBank accession no. JX075254) was also amplified (Fig. 6b). We hypothesized that nirK and hydB2 might function in the reduction of selenite. We were unable to amplify a soluble periplasmic selenate reductase (SerABC) or a dissimilatory nitrate reductase gene from strain BXM using several different primers sets (data not shown). The assimilatory nitrate reductase gene, nasA (GenBank accession no. JX075256), could be amplified from strain BXM (Fig. 6 C).

Figure 6.

Phylogenetic relationship of the (a) nirK, (b) hydB2 and (c) nasA gene sequences of strain BXM with other bacteria. The Neighbour-Joining method was used and bootstrap values derived from 100 replicates are shown. 377, 523 and 885 nucleotide positions used for analysis, respectively. Scale bar indicates changes per nucleotide position.

Discussion

Clostridium sp. strain BXM represents the classic characteristics of an acetogenic phenotype reducing CO2 to acetate (summarized in Eqn (1)). Strain BXM is thus a good model for the investigation of how selenium reduction may be affected by CO2 fixation and nitrate reduction.

display math(1)

The elemental selenium particles produced by strain BXM have nearly the same diameter (0·5 μm) and were approximately the same size of Se particles produced by Sulfurospirillum barnesii, Bacillus selenitireducens and Selenihalanaerobacter shriftii (Oremland et al. 2004). The elemental selenium particles occur outside the cell envelope from which they are released into the medium (Zehrt and Oremland 1987; Narasingarao and Häggblom 2007b). Strain BXM thus reduced math formula to math formula and then Se0. eqns (2) and (3) summarize the overall reactions:

display math(2)
display math(3)

The reduction of math formula to Se0, which is a common feature of diverse micro-organisms, is not very well understood. It has been reported that math formula reduction may be catalysed by a periplasmic nitrite reductase (DeMoll-Decker and Macy 1993), by hydrogenase I (Yanke et al. 1995) or by a nonenzymatic process (Tomei et al. 1992). DeMoll-Decker and Macy (1993) showed that nitrate induced a nonspecific reduction of math formula by nitrite reductase. The decrease in math formula reduction activity in nitrate-amended cultures suggests that hydrogenase might function as a math formula reductase and that nitrite reductase (nirK) is not involved. The reduction mechanism of math formula by strain BXM might be similar to that reported by Yanke et al. (1995) for another Clostridium species. Nitrite reductase (nirK) is likely involved in the reduction of nitrite to ammonia.

The increase in math formula reduction might be due to an enhancement of the growth of Clostridium by CO2 (Artin et al. 2010) and/or enhancement of the activity of hydrogenase (Arendsen et al. 1999), which has previously been shown to catalyse the reduction of math formula (Yanke et al. 1995). Alternatively, CO2 may have induced expression of a membrane-bound b-type cytochrome involved in electron transport in the periplasm (Arendsen et al. 1999), therefore increasing the reduction of math formula. CO2 might also increase math formula reduction by increasing the level of a b-type cytochrome that may mediate electron transport to math formula.

The inhibition of math formula reduction by nitrate may be explained by the lower reduction potential (E′0) for the math formula –Se0 couple (−0·35 V) (Yanke et al. 1995) compared with the math formulamath formula couple (+0·42 V) (Steinberg et al. 1992). Furthermore, nitrate was shown to decrease hydrogenase specific activity in CO2-sparged cultures of a Clostridium thermoaceticum strain (Arendsen et al. 1999), which might therefore explain the decrease in math formula reduction.

Although an assimilatory nitrate reductase gene product (885 bp; JX075256) was identified as nasA, no investigation has confirmed that a nasA expression product might function as a selenate reductase. In a previous study, nasA was shown to contribute to nitrate/nitrite transport (Gates et al. 2011b). In this study, nasA might account for the assimilation of nitrate, which could explain the imbalance in the total nitrogen (math formula, math formula and math formula) in the culture medium (Fig. 5). Additional work is needed to determine the enzymology involved in math formula reduction. A plausible explanation for the repression of nitrate over math formula in Clostridium sp. BXM may be that the selenate reductase has a higher affinity for nitrate than math formula, and/or selenate reductase is inhibited by nitrate.

CO2 fixation and nitrate reduction stimulated each other, which is in contrast to previous experiments in which Arendsen et al. (1999) showed that nitrate (30 mmol l−1) repressed CO2 fixation in a CO2-sparged culture system. In our study, the concentration of amended nitrate (1 mmol l−1) was relatively low and showed no significant repression of CO2 fixation. The promotion of nitrate reduction by CO2 amendment might be due to an increasing level of a membrane-bound b-type cytochrome involved in electron transport in the periplasm. An intensive study of the mechanisms of interaction between acetyl-CoA ‘Wood–Ljungdahl’ pathway and assimilatory nitrate reduction is necessary.

An increase in temperature (37°C under anaerobic conditions and 30°C under aerobic conditions) stimulated the reduction of math formula to Se0 in this study. This has also been found in Shewanella sp. HN-41, as the reduction levels of math formula after 7 days of incubation at 30°C were significantly higher than at 4 and 15°C (Lee et al. 2007). Ikram and Faisal (2010) indicated that 37°C was an optimum temperature for math formula reduction by a Bacillus sp. The suitable temperature for math formula reduction by strain BXM was 37°C under anaerobic conditions and 30°C under aerobic conditions. Oxygen is a known inhibitor of hydrogenase activity (Adams et al. 1981) and therefore would inhibit math formula reduction. On the contrary, the math formula reduction rates at 25 and 30°C under aerobic conditions were higher than under anaerobic conditions. This was in accordance with the observations of Watts et al. (2003) who showed that selenate reductase activity in Enterobacter cloacae SLD1a-1 was enhanced under aerobic conditions.

Members of the genus Clostridium are chemoorganotrophic and ubiquitous paddy soil micro-organisms (Rui et al. 2009; Wang et al. 2009; Li et al. 2011). Specifically, Clostridium spp. were the predominant micro-organisms in paddy soil amended with different carbon substrates, such as formate, acetate, propionate, pyruvate, succinate and citrate. Clostridia in paddy soils have been considered to have the ability of respiratory as well as fermentative metal reduction (Guan et al. 2008). In this study, Clostridium sp. BXM was shown to reduce both math formula and math formula under either anaerobic or aerobic conditions. Clostridia may thus play an important role in the reduction and detoxification of selenium compounds in view of the periodic changes in the redox conditions of paddy soils. Our findings expand on the current understanding of the transformations of Se, C and N in paddy soils.

Acknowledgements

This research was financially supported by the National Natural Science Foundation of China (no. 41090280) and (no. 41090282).

Ancillary