Chemolithoautotrophic arsenite oxidation by a thermophilic Anoxybacillus flavithermus strain TCC9-4 from a hot spring in Tengchong of Yunnan, China

A new facultative chemolithoautotrophic arsenite (AsIII)-oxidizing bacterium TCC9-4 was isolated from a hot spring microbial mat in Tengchong of Yunnan, China. This strain could grow with AsIII as an energy source, CO2–HCO3- as a carbon source and oxygen as the electron acceptor in a minimal salts medium. Under chemolithoautotrophic conditions, more than 90% of 100 mg/L AsIII could be oxidized by the strain TCC9-4 in 36 h. Temperature was an important environmental factor that strongly influenced the AsIII oxidation rate and AsIII oxidase (Aio) activity; the highest Aio activity was found at the temperature of 40∘C. Addition of 0.01% yeast extract enhanced the growth significantly, but delayed the AsIII oxidation. On the basis of 16S rRNA phylogenetic sequence analysis, strain TCC9-4 was identified as Anoxybacillus flavithermus. To our best knowledge, this is the first report of arsenic (As) oxidation by A. flavithermus. The Aio gene in TCC9-4 might be quite novel relative to currently known gene sequences. The results of this study expand our current understanding of microbially mediated As oxidation in hot springs.


Introduction
Geothermal waters often contain high levels of arsenic (As) and may contribute to surface and ground water contaminations (Langner et al., 2001;Zhang et al., 2008). As the waters equilibrate with the oxygen in outflow channels, arsenite (As III ) is oxidized via microbial processes (Wilkie and Hering, 1998;Langner et al., 2001). To date, numerous As III oxidizing mircrobes have been isolated from geothermal, sewage, soil, and mining environments.
Tengchong is a typical volcanic geothermal area located in southwestern China (Guo, 2012). It has more than 50 volcanoes and 140 geothermal fields (Du et al., 2005), with the largest two hydrothermal fields being located at Rehai and Ruidian (Guo and Wang, 2012). High concentrations of As have been found in hot spring waters (up to 687 μg/L), sediments and sinters in these geothermal areas (Zhang et al., 2008). Previous studies showed that diverse microbial communities inhabited these thermal environments (Hou et al., 2013). Our recent study of aioA diversity implied that novel lineages of microbial As oxidizers might inhabit this area (Jiang et al., 2014); however, microbially mediated As oxidation in this geothermal area has yet to be fully understood. The current study describes a new facultative chemolithoautotrophic As III -oxidizing bacterium isolated from a hot spring mat from the Rehai hot spring features, expanding our understanding of microbially mediated As oxidation in hot spring environments.

Site Description and Sample Collection
The Tengchong geothermal area is located in Yunnan province in southwestern China. It lies at the eastern end of the Yunnan-Tibet geothermal belt, a major part of the Mediterranean-Himalayas geothermal belt (Guo and Wang, 2012). Quaternary volcanoes and faults are the most conspicuous geologic features in this area (Du et al., 2005;Shangguan et al., 2005). Rehai is one of the largest geothermal fields within Tengchong, with large numbers of hot springs (Guo and Wang, 2012). A hot spring referred to as SRBZ (N24.9500 • , E98.4375 • ) is located in the Rehai and is documented to have significant As loads; containing a total As concentration of 226.93 μg/L and arsenate concentration of 217.07 μg/L (Jiang et al., 2014). Microbial biomass in water was collected via filtration (0.22 μm × 142 mm diameter, Millipore Corporation, Billerica, MA, USA) and mat sample was collected with sterile spoons. All samples were placed into 50 mL sterile polypropylene tubes for transport to the lab.
About 1.0 g mat samples were incubated in 10 ml MSM containing 1.33 mM (100 mg/L) NaAsO 2 (MSM-As) for 1 week at 40 • C. Following four weekly 10-fold dilution transfers, the microbial suspension was diluted and spread onto solid plates with MSM-As agar (1.5%) plates amended with 0.1% yeast extract. After 2 days incubation, a number of different colonies were sub-cultured to purify and screen for As III oxidization activity. Concentrations of As III and arsenate (As V ) were determined by using liquid chromatography-hydride generationatomic fluorescence spectrometry (LC-HG-AFS, Haiguang AFS-9780, Beijing; Keller et al., 2014).

Physiological and Biochemical Characteristics
Physiological and biochemical tests were carried out according to Pikuta et al. (2000). The effects of oxygen, temperature, pH, and NaCl optima were determined in MSM containing 0.01% yeast extract (MSM-YE). When potential carbon sources and inorganic substrates were tested, the strain was cultured in MSM, and the concentrations of carbon sources and inorganic substrates added to the medium were 10 mM. The anaerobic growth was observed in MSM-YE which was incubated in a Hungate tube at 60 • C for 2 days.
Colony morphology was determined visually and microscopic characteristics were observed under the optical microscope (Leica DM5000 B). Cells of TCC9-4 were sent to Wuhan Institute of Virology, Chinese Academy of Sciences (CAS) after preculture with Luria-Bertani (LB) solid medium for negative-stain transmission electron microscopy (TEM, Hitachi H-7000FA) detection.

DNA Extraction and Amplification of 16S rRNA and As III Oxidase Genes
One isolate was obtained and was referred to as TCC9-4. DNA was extracted from over-night cultures or from the mat and the biomass-containing filter sampling from the SRBZ using FastDNA R SPIN Kit for Soil (Qbiogene, Inc. Carlsbad, CA, USA) according to the manufacturer's protocols. DNA yields were quantified with a NanoDrop 2000 spectrophotometer (Thermo Scientific, USA). All PCRs were undertaken in Master cycler ep Gradient S Thermal Cycler block (Eppendorf, Hamburg, Germany). 16S rRNA gene was amplified with the primers P1 (5 AGAGTTTGATCCTGGCTCAG 3 ) and P2 (5 GGTTACCTTGTTACGACTT 3 ; Weisburg et al., 1991). A typical 50 μL PCR mixture contained 0.4 pmol/L of primers (MWG), 1 μL DNA template (50 ng), 10x reaction buffer (Promega Corporation), 5 mM MgCl 2 , 1.5 U Biotaq DNA polymerase (Promega), 0.25 m dNTP (Promega).

Clone Library Construction and Phylogenetic Analysis
PCR products were purified using the E.Z.N.A. Gel Extraction Kit (Omega Bio-tek, Inc. Norcross, GA, USA) according to manufacturer's instructions. The purified PCR products were ligated into pMD-18T vectors (TaKaRa) and transformed into Escherichia coli DH5α competent cells. The transformed cells were plated on LB agar containing 100 μg/mL of ampicillin, 80 μg/mL of 5-bromo-4-chloro-3-indolyl-β-D-galactopyranoside (X-Gal) and 0.5 mmol isopropyl-β-D-thiogalactopyranoside (IPTG), and incubated overnight at 37 • C. Randomly selected clones were commercially sequenced. The 16S rRNA gene sequence was aligned with closely related sequences in the GenBank database using the BLAST program. aioA gene sequences were edited in MEGA 5.05 and binned into various operational taxonomic units (OTUs) using DOTUR 1.53. One representative sequence from each OTU (0.03 cutoff) was selected for phylogenetic analysis. Prior to phylogenetic analysis, representative aioA sequences were translated into amino acid sequences and searched for closely related aioA amino acid sequences in the GenBank database (BLASTX), which were then used to construct a phylogenetic tree with MEGA 5.05. The 16S rRNA fragment nucleotide sequence of TCC9-4 has been deposited in the GenBank database under accession numbers KP027949. The aioA gene fragment nucleotide sequences obtained have been deposited in the GenBank database under accession numbers KC777327-KC777338.

Chemolithoautotrophic and Heterotrophic Experiments
TCC9-4 was cultured in MSM-As for the chemolithoautotrophic growth and As III oxidization experiments or MSM-YE for experiments examining heterotrophic growth. Cells were washed twice with distilled water and then inoculated (1% inoculum), and samples were taken periodically for the detection of cell growth and As speciation. Cell growth was measured by plate cell counting or optical density at 600 nm (OD 600 ). Experiments were run in triplicate and duplicate, respectively.

As III Oxidase Assays
As III oxidase activity was assayed as previously described by Anderson et al. (1992). Briefly, TCC9-4 was inoculated in 13 liters MSM-As and incubated for 48 h. Cells were harvested by vacuum filtration, washed three times with 20 mM Tris-HC1, 0.1 mM PMSF, 0.6 mM EDTA, pH 8.4, containing 0.9% NaCl, suspended and incubated in 20 mM Tris-HC1, 0.6 mM PMSF, 0.6 mM EDTA, pH 8.4, containing lysozyme (0.25 mg/ml), and then sonicated. Unbroken cells were removed by centrifugation, and then the crude extract was diluted with 50 mM 2-[N-morpholino]ethanesulfonic acid (MES) buffer (pH 6.0). Aio activity was determined by measuring the reduction of 2,4dichlorophenolindophenol (DCIP) at OD 600 in 50 mM MES containing 200 μM sodium As III . The blank was 50 mM MES buffer containing 120 μM DCIP and 200 μM sodium As III . The enzyme activity was defined as μmol of DCIP reduced/min/mg of protein. Protein concentrations were determined by Modified Bradford Protein Assay Kit (Sangon) using BSA as a standard according to manufacturer's instructions.
the genus Anoxybacillus (99.66%). Phylogenetic analysis places TCC9-4 closest to two organisms referred to as A. flavithermus WL and A. tunisiense isolated in China and Tunisia, respectively (Figure 1). Phenotypic analysis found TCC9-4 to be a gram-positive, facultatively aerobic bacterium with a straight rod morphology and peritrichous flagella (Figure 2). The main physiological and biochemical characteristics of this isolate were investigated and compared with related species (Table 1). TCC9-4 has a pH range of 6.0-10.0 (optimum at 8.0), and is positive for methyl red test and nitrate reduction, but negative for gelatin hydrolysis, starch hydrolysis, tyrosine hydrolysis, and lysozyme tests. Growth was inhibited at NaCl concentrations higher than 3.0% (w/v). Utilizable carbon sources for heterotrophic growth include: maltose, sucrose, glucose, galactose, mannose, xylose, sodium citrate, tyrosine, mannitol, sorbitol, sodium acetate, sodium succinate, sodium formate, glycerin, and sodium lactate. Fructose, sodium oxalate, and inorganic substrates including Fe II , S 0 , and S 2− will not support growth. TCC9-4 is capable of growing in a chemolithotrophic mode using As III as an energy source and CO 2 -HCO 3 − as carbon source with oxygen as the electron acceptor (see below). The mol% G + C content of DNA is 56.63%. Based on 16S rRNA gene sequence and the above phenotype and biochemical characteristics, we conclude that TCC9-4 represents a new strain of A. flavithermus.

As III Oxidization
TCC9-4 temperature response varied as a function of mode of metabolism. When cultured chemolithotrophically with As III as an electron donor in MSM-As, the optimal temperature was around 40 • C ( Figure 3A). The generation times of TCC9-4 at 30 • C, 35 • C, 40 • C, 45 • C, 50 • C were 11.0 h, 9.1 h, 9.0 h, 14.0 h, 43.5 h, respectively. However, in MSM-YE, the temperature growth range was 30-68 • C, with an optimum at 60 • C ( Figure 3B). Subsequent experiments then examined As III oxidation in culture and crude cell lysates as a function of temperature (Figure 4). Optimum activity for both types of assays was 40 • C. In whole cell culture assays at this temperature, more than 90% of As III could be oxidized in 36 h and the highest As III oxidization rate was 2.78 mg As III oxidized/L/hour. The highest Aio activity was 0.037 U/mg at 40 • C, matching cell growth behavior ( Figure 3A) and suggesting the Aio enzyme temperature range is at least one critical limiting factor that constrains this organism in nature when growing in a chemolithotrophic mode. DCIP reduction was not detected in the blank without crude cell extract.
Additional experiments then examined if TCC9-4 could grow in both modes simultaneously. In MSM containing 0.01% yeast extract and 100 mg/L As III (MSM-YE-As), As III oxidation did not commence until the cells attained stationary phase (Figure 5), presumably due to the cells having already consumed all of the growth yielding substrates of yeast extract. We note, however, that As III oxidation at this stage failed to yield growth benefits as no further cell growth was observed. This may was resulted from key inorganic nutrients (e.g., P, K, Mg, etc) being consumed and thus an additional energy source was perhaps not useful at this point. Also apparent from these studies, TCC9-4 clearly preferred heterotrophic growth on yeast extract. This behavior deviates from that observed in previous studies with Rhizobium  sp. NT-26 and Sinorhizobium sp. KGO-5, where additions of yeast extract enhanced growth and accelerated As III oxidizing significantly (Santini et al., 2000;Dong et al., 2014). Although the cell growth was enhanced due to the relatively higher optical densities, no obvious As III oxidization has been observed during the logarithmic phase ( Figure 5).

As III Oxidase Genes
We next sought to clone and characterize the Aio genes in this organism. To our knowledge, the apparent temperature sensitivity of this enzyme has not been reported previously and so it was of interest to attempt to understand potential structural features that might be associated with temperature sensitivity. All attempts to PCR amplification these genes failed. This included all known primers that span the known genes associated with As III oxidation: genes referred to as aro (Hamamura et al., 2009); aox (Quemeneur et al., 2008); arx (Zargar et al., 2012). Failure to obtain a PCR product was not due to template issues because the 16S rRNA gene readily amplified. Rather, it would appear that the gene encoding the Aio in this organism is sufficiently different so that even highly degenerate primers failed to recognize this gene. Moreover, all available genome sequences of A. flavithermus including strain 25, WK1, AK1, TNO-09.006, E13, and Kn10 with GenBank assembly accession nos. GCA_000753775, GCA_000019045, GCA_000353425, GCA_000327465, GCA_000753835, and GCA_000367505, respectively, from GenBank have been checked. No aioA gene sequence could be identified in these genomes. However, the primers designed to amplify aioA (previously referred to as aro, aox, and aso) robustly generated PCR products when used with DNA extracted from the hot spring sediments as template. An aioA gene clone library of the mat was constructed and a total of 46 aioA gene clone sequences were obtained (Figure 6). Those aioA gene sequences were mainly classified into two putative groups and were closely related (72-80% amino acid level) to the Aio protein sequences from unidentified pure culture or gene clones of the order Burkholderiales of Betaproteobacteria.
As III -oxidizing bacteria have been isolated from geothermal, sewage, soil, and mine environments, including both heterotrophic As III oxidizers and chemolithoautotrophic As III oxidizers. Investigations of the As III oxidation genes aioA indicated that As III oxidizers have a widespread distribution in 11 phyla (Heinrich-Salmeron et al., 2011). And Heinrich-Salmeron et al. (2011) also reported As III oxidizer-Bacillus sp. 21AAIII, expanding known As III oxidation to the phylum Firmicutes. In this study, As III oxidation in strain TCC9-4 further expands this trait in the Firmicutes exhibiting the capacity to grow chemolithotrophically with As III and CO 2 . The carbon/energy substrates that can support the heterotrophic growth and phenotype characteristics of TCC4-9 were consistent with those reported for A. flavithermus DSM 2641T (Heinen et al., 1982).
However, there are no reports of As oxidation in the genus Anoxybacillus. Moreover, no PCR product of aioA gene was obtained, nor could aioA be identified in any of the six available genome sequences of A. flavithermus in GenBank. Furthermore, no aioA gene has been found related to Firmicutes in the aioA gene clone library of the environmental sample from which strain TCC9-4 was isolated. This implies the Aio gene in TCC9-4 either has sufficiently low homology to currently known gene sequences so as to escape detection via PCR, or that this particular gene is quite novel relative to known enzymes catalyzing this reaction. Further studies that focus on the genes involved in As III oxidation with genome sequencing of TCC9-4 and metagenomes sequencing of the source spring sample are warranted.