Bacillus sp. SXB and Pantoea sp. IMH, aerobic As(V)-reducing bacteria isolated from arsenic-contaminated soil

Authors

  • Q. Wu,

    1. State Key Laboratory of Environmental Chemistry and Ecotoxicology, Research Center for Eco-Environmental Sciences, Chinese Academy of Sciences, Beijing, China
    Search for more papers by this author
  • J. Du,

    1. State Key Laboratory of Environmental Chemistry and Ecotoxicology, Research Center for Eco-Environmental Sciences, Chinese Academy of Sciences, Beijing, China
    Search for more papers by this author
  • G. Zhuang,

    1. State Key Laboratory of Environmental Chemistry and Ecotoxicology, Research Center for Eco-Environmental Sciences, Chinese Academy of Sciences, Beijing, China
    Search for more papers by this author
  • C. Jing

    Corresponding author
    • State Key Laboratory of Environmental Chemistry and Ecotoxicology, Research Center for Eco-Environmental Sciences, Chinese Academy of Sciences, Beijing, China
    Search for more papers by this author

Correspondence

Chuanyong Jing, State Key Laboratory of Environmental Chemistry and Ecotoxicology, Research Center for Eco-Environmental Sciences, Chinese Academy of Sciences, Beijing 100085, China. E-mail: cyjing@rcees.ac.cn

Abstract

Aims

To isolate highly effective aerobic As(V)-reducing bacteria from arsenic(As)-contaminated soils in Northwest China and to identify their dynamic As(V) reduction processes and genomic detoxification mechanisms.

Methods and Results

Enrichment cultures were performed aerobically in tryptone, yeast extract and glucose medium in the presence of As(V). Strain SXB isolated from soil in Shanxi Province, belonging to Bacillus genus, reduced As(V) more effectively under aerobic conditions than under anaerobic conditions. Strain IMH, a strictly aerobic isolate obtained from soil in Inner Mongolia, identified as Pantoea, is reported for the first time to reduce As(V). Both isolates could reduce over 90% As(V) in 36 h under aerobic conditions. Putative gene fragments for the ArsB efflux pump gene were obtained from both strains. The putative arsenate reductase gene was only amplified from strain SXB. A putative arsH gene was amplified from strain IMH.

Conclusions

Strains SXB and IMH isolated from the As-contaminated soils reduce As(V) effectively under aerobic conditions via a detoxification mechanism regulated by ars operons.

Significance and Impact of the Study

Pantoea genus is reported to reduce As(V) for the first time. This study provides a full understanding of the highly effective As(V)-reducing bacteria SXB and IMH, which could influence the As biogeochemical cycle in soils.

Introduction

Arsenic (As) catastrophes are occurring worldwide and resulting in serious health problems today in many countries such as Bangladesh, India, Mexico, the US and China (Smedley and Kinniburgh 2002). Arsenic is released into the environment either by natural phenomena or from anthropogenic origins through industrial or agricultural activities (Lièvremont et al. 2009). The bioavailability and toxicity of As depend on the species. Two common biologically relevant forms of inorganic As are arsenite [As(III)] and arsenate [As(V)]. Usually, As(III) is more toxic and mobile than As(V) (Salnikow and Zhitkovich 2008; Jomova et al. 2011).

Micro-organisms play an important role in the As biogeochemical cycle in various environments, especially in As-contaminated sites (Jackson et al. 2005; Pepi et al. 2007a; Lièvremont et al. 2009; Tsai et al. 2009). As(V)-reducing bacteria are one of the dominant families involved in the global cycle of As (Lièvremont et al. 2009). To date, a series of As(V)-reducing bacteria have been isolated under anaerobic conditions in previous studies (Santini 2004; Handley et al. 2009; Chang et al. 2012). However, reports of aerobic As(V)-reducing bacteria are limited. The effects of aerobic As(V)-reducing bacteria on As behaviours in As-contaminated environment need to be clarified.

Microbial As(V) reduction occurs via two common mechanisms, respiration and detoxification (Silver and Phung 2005). Respiration occurs only under anaerobic conditions. Detoxification is an efficient As(V)-reducing mechanism that occurs under aerobic conditions. The detoxification processes found in As(V)-reducing bacteria are mostly regulated via ars operons, chromosomally or plasmid-encoded. The five (arsRDABC) or three (arsRBC) genes are the two most common types of ars operons (Silver and Phung 2005). The arsR gene encodes a transcriptional regulator or repressor, and the arsB gene encodes a membrane carrier that extrudes As(III) but not As(V). The arsC gene encodes a cytoplasmic As(V) reductase (Rosen 2002). The arsA and arsD genes are usually adjacent to each other. An arsenite-stimulated ATPase, the transcription product of the arsA gene, co-expresses with arsB and forms a complex ArsAB efflux pump to remove As(III) (Silver and Phung 2005). The arsD gene was recently demonstrated to encode an arsenic chaperone that transfers the trivalent metalloids As(III) and Sb(III) to the ArsA subunit of the pump (Lin et al. 2007; Yang et al. 2011). The ArsH protein from the transcription of arsH gene is annotated as a NADPH-dependent FMN reductase with unknown biological function (Hervas et al. 2012).

Details about sedimentary As(V)-reducing bacteria associated with high arsenic groundwater in Shanxi Province and Inner Mongolia are still unknown. The aims of the present work were (i) to isolate and identify the culturable and effective aerobic As(V)-reducing bacteria from As-contaminated soils in Shanxi Province and Inner Mongolia; (ii) to evaluate the dynamic process of As(V) reduction for each strain; and (iii) to clarify the genomic As(V) reduction mechanism of the isolated strains.

Materials and methods

Soil collection and analysis

Soil SX and IM were collected in Shanxi Province (39°25′24″N; 112°52′50″E) and Inner Mongolia (40°58′1″N; 107°0′27″E), respectively. An auger was used to drill a hole in the soil, and the depth was 1 m from the upper layer. Soil samples were stored in sterilized sealed plastic bags and transported at 4°C in a cooler to minimize microbial activity before use. Analysis of the soil samples was performed using microwave-assisted acid digestion of sediments, sludges, soils and oils (Epa 2007). The concentrations of As and major metal ions (Ca, Mg, Na, Al, Fe, K, etc.) in the digestion solutions were analysed by ICP-OES (PerkinElmer, MA, USA).

Isolation of aerobic As(V)-reducing bacteria

One-gram soil was added into 20 ml tryptone, yeast extract and glucose (TYEG) medium (Anderson and Cook 2004; Ruta et al. 2011) at pH 7·0 with 5 mmol l−1 or 10 mmol l−1 sodium arsenate added. The culture was incubated aerobically for 48 h with shaking (150 rpm) at 28°C. Then, 1 ml of the culture was collected and injected into another 20 ml fresh medium at 24-h intervals. After repeating this incubation process two times, 1 ml culture was removed, resuspended in TYEG medium by serial dilutions to 10−6 and plated onto TYEG medium plates with the same content of sodium arsenate (Fan et al. 2008). The plates were incubated for another 24 h at 28°C. Isolates were purified by the streak plate method and kept at −80°C in the presence of 30% sterile glycerol. Isolates with higher As(V)-reducing capacity (over 90% As(V)) were chosen to be further investigated. As(V)-reducing capacity was determined by the ratio As(III)/total As in aqueous solution after a 36-h incubation.

Characterizations of different isolates and dynamic As(V)-reduction assay

Isolates cultured in the TYEG medium containing 1 mmol l−1 As(V) were characterized by transmission electron microscopy (TEM, JEM-1400 from JEOL Ltd, Tokyo, Japan). To examine the pH range for growth, MES (pH 4·0–6·0), PIPES (pH 7·0), Tricine (pH 8·0), CAPSO (pH 9·0) and CAPS (10·0–11·0) were added to the TYEG medium with a concentration of 25 mmol l−1, respectively. The pH was slightly adjusted using HCl or NaOH. All samples were incubated at 25°C for 48 h before the culture turbidity (OD600) variation was monitored. The temperature range for growth was determined from 4°C to 53°C.

The As resistance of an isolate was determined by the minimum inhibitory concentrations (MICs), namely the lowest concentrations of As(III) or As(V) that completely inhibited bacterial growth after 48-h incubation (Fan et al. 2008; Liao et al. 2011). Isolates were inoculated in liquid TYEG medium supplemented with increasing concentrations of either As(III) (0–80 mmol l−1, NaAsO2) or As(V) (0–300 mmol l−1, NaAsO4) and shaking (150 rpm) at 30°C for 48 h.

The capacities for As(V) reduction in the obtained bacteria were measured under aerobic conditions. Isolates were inoculated separately in 1 mmol l−1 As(V) liquid TYEG medium and incubated with shaking (150 rpm) at 30°C. Suspension samples were obtained from 0 to 60 h. The samples were passed through a 0·22-μm syringe filter for analysis of soluble As speciation.

A modified chemically defined medium (CDM) was used to verify the anaerobic growth with 10 mmol l−1 of As(V) under anoxic condition. The CDM medium contains all the reagent in CDM (Weeger et al. 1999) with 1 ml vitamin solution (0·25% pyridoxal HCl, 1% thiamine, 1% calcium pantothenate, 1% riboflavin, 1% niacin, 0·5% p-aminobenzoate, 2% pyridoxine HCl, 0·002% vitamin B12) and 0·5 ml trace elements solution (MnCl2·4H2O 10, CoCl2·6H2O 12, ZnCl2 7, H3BO3 6, NiCl2·6H2O 2·5, CuCl2·2H2O 1·5, Na2MoO4·2H2O 2·5, FeCl2·4H2O 150, all in mg per litre) added. The pH was adjusted to 7·2. The anaerobic growths for both bacteria were conducted in a glovebox (99·99% N2) using the CDM medium at 30°C with an incubation time of 72 h. To compare with the aerobic As(V)-reducing process, the same TYEG medium was used in the dynamic As(V)-reduction process investigation with the method described before from 0 to 102 h.

Abiotic controls were used both under aerobic and anaerobic conditions without bacteria added. The density of the cultures, that is, OD600, was detected using a UV spectrophotometer (HACH DR2800, US). The As speciation was determined by HPLC-AFS (Jitian, China) system following the method described in previous studies (Jesus et al. 2011; Hu et al. 2012).

Identification of 16S rDNA and detection of ars genes

Bacterial genomic DNA was extracted using a Bacterial DNA kit (Omega, GA, USA) and purified using an AxyPrep PCR Cleanup kit (Axygen, CA, USA). The amplifications of 16S rDNA were carried with universal primers, 27f and 1492r. The sequences of each primer for the ars gene polymerase chain reaction (PCR) are shown in Table 1 (Chang et al. 2008).

Table 1. The primer sequences used for the PCR of arsRBC genes (Chang et al. 2008)
PrimerSequence (from 5′ to 3′)Tm (°C)
SEL 0904f (universal)ATCATGGCTCAGATTGAACGC55
SEL 1226 r (universal)TACCTTGTTACGACTTCTACCT
arsR fATCCAGCTCTTCAAAACC53
arsR rGTTTTTCAGCTTCATAC
arsB fGTGGAATATCGTCTGGAATGCGAC57
arsB rGGTAATTTTCGGCCCCAAATCG
arsC fTGCGGCACTTCGTGAAACAC57
arsC rAAGTATATCCAGAACCACTT
arsH fATGGACCAGTTCCCAGAC55
arsH rCTGATTGGGGATGGTGAACA

PCR amplifications were performed with a Mastercycler (Eppendorf, Hamburg, Germany). The 50-μl reaction mixtures were composed of 0·2 mmol l−1 dNTP, 0·08 μmol l−1 of each primer, 5 μl of 10× PCR buffer, 1·5 U of Taq DNA polymerase (Fermentas Thermo Scientific, MA, USA) and 50 ng template DNA. The protocol for each primer set consisted of an initial denaturation step (94°C for 3 min) followed by 30 cycles of 94°C for 30 s, 55°C for 45 s (SEL 0904 ars), 51°C for 45 s (arsB), 54°C for 45 s (arsC), 53°C for 45 s (arsH) and 72°C for 1 min. A final extension for 10 min at 72°C was used. Negative controls included a deionized water reagent control.

The PCR products (5 μl) were separated on 1× TAE – 1% agarose gels by electrophoresis and visualized on a UV transilluminator. The positive DNA products were purified using a Gel Extraction kit (Omega). Standard procedures were used for manipulation. The DNA sequencing was performed by the Tsingke Company (Beijing, China). The BlastN (http://www.ncbi.nlm.nih.gov/blast) searching program was used to analyse similarities.

Phylogenetic tree construction

Reference sequences used in generating phylogenetic trees were obtained from the GenBank in the search tool (Blast; NCBI; http://www.ncbi.nlm.nih.gov). Sequence alignments were performed using the ClustalX2 program, European Bioinformatics Institute, Cambridge, UK. Phylogenetic trees of 16S rDNA, arsB genes and arsC genes were generated by neighbour-joining (NJ) method using mega ver. 5.0 program based on p-distance (Tamura et al. 2011).

The accession numbers are JX861127 (strain SXB) and JX861128 (strain IMH) for 16S rRNA genes. Accession numbers JX861129, JX861130, JX861131 and JX861132 are for ars genes.

Results

Isolations and characterizations of aerobic As(V)-reducing bacteria

The As content was 25·0 and 14·1 mg kg−1 in soil IM and in soil SX, respectively. Other metal contents in soil samples are shown in Table S1. Isolate SXB from soil SX and isolate IMH from soil IM were chosen according to the result of the As(V)-reducing capacity test (over 90% in 36 h).

The characteristics of isolate SXB and isolate IMH were examined. Isolate SXB is rod-shaped and mostly grew in pair (Fig. 1a). Isolate IMH is rod-shaped or ellipse-shaped with flagella around (Fig. 1b). Isolate SXB grows above pH 4 with an optimum pH of 7 and at temperatures of from 15°C to 45°C with an optimum temperature of 28°C (Table 2, Figs S1 and S2). Growth of isolate IMH occurs at pH 4–10 (optimum at 6) and at temperatures of 20–40°C (Table 2, Figs S1 and S2).

Table 2. Identification and characteristics of As-reducing bacteria isolated from As-contaminated sites
IsolateCloset GenBank match (similarity)Accession no.Optimum pHOptimum temp (°C)As resistance MIC (mmol l−1)
As(III)As(V)
SXBBacillus thuringiensis sp.IAM 12077 (99%) D16281 72860250
IMHPantoea agglomerans sp. DSM 3493T (98%) AJ233423 63010150
Figure 1.

Transmission electron microscopy micrographs of bacteria SXB (a) and bacteria IMH (b) at 80 kv.

The MICs were tested for each bacterium (Table 2 and Fig. S3). The tolerances of isolate SXB were 60 mmol l−1 As(III) and 250 mmol l−1 As(V). As a comparison, the tolerance level was 10 mmol l−1 As(III) and 150 mmol l−1 As(V) for isolate IMH. The tolerance capacity of isolate SXB is higher than that of isolate IMH.

Sequencing of 16S rDNA and phylogenetic analysis

According to the results of 16S rDNA sequences analysis using the BlastN searching program, strain SXB showed 99% similarity to Bacillus thuringiensis sp.IAM 12077 (D16281), and strain IMH had 98% similarity to Pantoea agglomerans sp.DSM 3493T (AJ233423) (Table 2).

The phylogenetic tree was inferred for the As(V)-reducing isolates (Fig. 2). The 16S rDNA sequence analysis indicated that strain SXB was phylogenetically dispersed in the Bacillus but in a separate line near a clade originated by Bacillus sp. Rice-C (FM163606) isolated from agricultural soil in Bangladesh (Bachate et al. 2009) and Bacillus sp. F1 (EF428913) isolated from noncontaminated soil in Italy (Pepi et al. 2007a). Strain IMH resulted in a separate line with the nearest strains included within a clade originated by Pseudomonas putida sp. RS-5 (DQ182328) and Pseudomonas putida sp. RS-5 (DQ182328) isolated from Korea (Chang et al. 2008).

Figure 2.

Phylogenetic tree based on 16S rDNA sequences of isolate SXB, isolate IMH and aerobic As(V)-reducing bacteria. Sequences from this study are in bold type. The tree was constructed using the Neighbor-Joining method. The evolutionary distances were computed using the Kimura 2-parameter method. The scale bar represents 0·02 substitutions per site.

As(V)-reducing assays

The aerobic dynamic reduction processes were studied to verify the presence of detoxification mechanisms in the two isolates (Fig. 3). As(V) reduction occurred slowly in the initial 4 h for isolate SXB in aerobic conditions. Subsequently, As(V) was reduced much more rapidly when the logarithmic phase began. The highest OD600 value was reached at 16 h with an 88% As(V)-reducing ratio. After 22 h, isolate SXB reduced more than 95% of the As(V) in the culture. No more than 2% As(V) remained at 60 h.

Figure 3.

As(V) reduction to As(III) measured by time variation in the ratios of As(III)/Total As in aqueous solution and culture turbidity (OD600) in TYEG medium containing 1 mmol l−1 As(V). (×, culture turbidity, i.e. OD600; +, culture turbidity, i.e. OD600 of abiotic control; ▿, the ratios of As(III)/Total As in culture of abiotic control; ○, the ratios of As(III)/Total As in culture of strain SXB under aerobic condition; Δ, the ratios of As(III)/Total As in culture of strain IMH under aerobic condition; □, the ratios of As(III)/Total As in culture of strain SXB under anaerobic condition).

Compared with isolate SXB, isolate IMH reduced As(V) much more slowly in the initial 10 h under aerobic conditions. After 10 h, when isolate IMH growth entered the logarithmic phase, the As(V) concentration decreased rapidly. As(V) concentration in the culture decreased to less than 10% in 40 h. After 60 h, bacterium IMH reduced more than 97% of the As(V) in the culture.

The growths of both bacteria under anaerobic conditions were also investigated. Isolate SXB grew using As(V) as the electron acceptor in the CDM medium under anaerobic condition at 30°C. Isolate SXB grew quickly under the anaerobic condition and entered the steady phase no more than 5 h after incubation (Fig. 3). However, the anaerobic As(V) reduction was much slower than that under aerobic conditions. After 100 h incubation, only 22% As(V) was reduced in the culture. The As(V)-reducing efficiency under anaerobic circumstances is much lower than that under aerobic conditions. Isolate IMH cannot survive without oxygen.

In the control experiment, no culture turbidity variation and As(V) reduction were observed during aerobic and anaerobic growth.

The ars genotype of the aerobic As(V)-reducing bacteria

The ars genes were searched to identify the genomic mechanism of As resistance (Table 3 and Fig. S4). The arsB gene was observed in these two isolates. The arsC gene was found in strain SXB, while the arsH gene was detected in strain IMH. The length of putative arsB gene fragments was 936 bp for both strains. The arsB from strain SXB (JX861130) was phylogenetically dispersed close to the arsB from strain IMH (JX861131). The nearest amino acid strain within a clade was originated by the ArsB from Enterobacter cloacae (ACO54370) (Fig. 4). A 275-bp fragment of putative arsC gene was amplified from strain SXB. The ArsC product of arsC gene from strain SXB (JX861129) was phylogenetically dispersed close to the ArsC protein from Pseudoxanthomonas sp.KAs5-3 (AFP73420) (Fig. 5). A 964-bp fragment of putative arsH gene (JX861132) was obtained from strain IMH. No positive results were obtained using the designed arsR primer. Primers for arsD and arsA genes were also designed, while no positive fragments were obtained in this study (Table S2). Besides, primers for arrA gene were also designed to clarify the possible existing As(V)-respiratory mechanism. No positive fragments were obtained (Table S3).

Table 3. The ars genes detection of isolate SXB and IMH by PCR using the designed primers
ars genesIsolate
Strain SXBStrain IMH
arsR
arsB++
arsC+
arsH+
Figure 4.

Phylogenetic tree constructed from ArsB amino acid sequences of strain SXB, strain IMH and related As(V)-reducing bacteria. Sequences from this study are in bold type. The tree was constructed using the Neighbor-Joining method. The evolutionary distances were computed using the Poisson method. The scale bar represents 0·01 substitutions per site.

Figure 5.

Phylogenetic tree based on ArsC amino acid sequences of isolate SXB and related As(V)-reducing bacteria. Sequences from this study are in bold type. The tree was constructed using the Neighbor-Joining method. The evolutionary distances were computed using the Poisson method. The scale bar represents 0·2 substitutions per site.

Discussion

Strain SXB was classified as Bacillus thuringiensis sp. IAM 12077 (D16281) belonging to the Bacillus genus in the Gram-positive groups. The bacteria belonging to Bacillus have been isolated from sediments in Italy (Ruta et al. 2011) and agricultural soils in Bangladesh (Bachate et al. 2009). Meanwhile, the detoxification mechanisms of these Bacillus isolates had been clarified in previous studies. However, based on the published works, no As(V)-reducing associated characterizations for Bacillus thuringiensis sp. IAM 12077 were reported. This strain IAM 12077 was only identified to accumulate poly(-β-)hydroxybutyrate (PHB) (Adwitiya et al. 2009).

Furthermore, strain SXB can reduce As(V) not only under aerobic conditions but also under anaerobic conditions. Analogous facultative anaerobic As(V)-reducing bacteria belonging to Pseudomonas, Psychrobacter, Vibrio, Citrobacter, Enterobacter and Bacillus have been found in groundwater in Taiwan (Liao et al. 2011). According to the previous study, only the As(V)-reduction process of Citrobacter sp. AR-7 was further investigated. No details of the As(V)-reduction assay for facultative anaerobic Bacillus strains were present. Our new findings about strain SXB offer an overall insight about the Bacillus involved in As(V) reduction both under aerobic and anaerobic conditions at an As(V)-contaminated site.

Strain IMH showed 98% similarity to Pantoea agglomerans sp. DSM 3493T (AJ233423), belonging to the Gram-negative groups. To the best of our knowledge, this is the first time that a strain belonging to Pantoea has been reported to reduce As(V). Aerobic As(V)-reducing bacteria have been isolated from various habitats and classified as Pseudomonas, Bacillus, Exiguobacterium, Aeromonas, Arthrobacter, Thiobacillus and Staphylococcus, etc. (Butcher et al. 2000; Pepi et al. 2007b; Chang et al. 2008; Fan et al. 2008; Bachate et al. 2009; Ruta et al. 2011). Those As(V)-reducing isolates have been investigated a great deal. However, no reports about As(V)-reducing Pantoea can be found. This new finding will lead to the joining of Pantoea into the big family of aerobic As(V)-reducing bacteria. Nevertheless, how this Pantoea strain evolved to reduce As(V) and how it is involved in the As biogeochemical cycle remain to be clarified. Our results should inspire further study concerning the Pantoea genus.

Based on the results of phylogenetic analysis, the two effective As(V)-reducing bacteria isolated from Inner Mongolia and Shanxi Province are distantly related. As-reducing strains G5-G14 isolated from sediments in the same district were identified as Pseudomonas, Acinetobacter and Sphingomonos (Fan et al. 2008). These results, together with previous results (Bachate et al. 2009; Liao et al. 2011), indicate that As(V)-reducing bacteria are widely distributed in different environments and are phylogenetically diverse.

The arsenic detoxification mechanism was investigated by the amplification of ars genes in strains SXB and IMH. Although As(III) is more toxic than As(V), As(V) reduction is considered to be a detoxification method for micro-organisms (Oremland 2003). Aerobic As(V)-reduction is catalysed via ars operons (Silver and Phung 2005). The ArsB As(III) efflux pump and ArsC arsenate reductase are the core parts of the aerobic detoxification system (Rosen 2002). The existence of the arsB gene in strain SXB and IMH indicated that both strains could eliminate the reduced toxic As(III) outside of the cells. Strain SXB is positive for the arsC gene. Namely, As(V) reduction by strain SXB is catalysed by ArsC arsenate reductase. Then, As(III) is eliminated by the ArsB efflux pump. The absence of the arsC gene in strain IMH can be attributed to the diversity of ars operons. The amplification of ars genes can be influenced by many factors such as the type of designed primers, the cycling conditions, the buffers or agents in the PCR liquor and conformational variations in the extracted DNA (Chang et al. 2008; Liao et al. 2011). Several studies showed similar results of failed amplification of arsC genes in aerobic As(V)-reducing bacteria (Anderson and Cook 2004; Chang et al. 2008; Bachate et al. 2009).

Another special gene, arsH, was detected in strain IMH. The function of ArsH is not yet clear. In Y. enterocolitica, the arsH gene is required for the resistance to As(III) and As(V) (Neyt et al. 1997). The ArsH protein from cyanobacterium Synechocystis has been reported to be an efficient NADPH-dependent quinine reductase (Hervas et al. 2012). More about the role of ArsH protein in As resistance remains to be identified.

Bacteria SXB and IMH both exhibit resistance to As(V) and As(III). The MICs for aerobic As(V)-resistant bacteria are normally in the range of 100–350 mmol l−1 for As(V) and 2–70 mmol l−1 for As(III) (Matlakowska et al. 2008; Bachate et al. 2009; Liao et al. 2011; Ruta et al. 2011). The As(V) concentration of As resistance even exceeds 1000 mmol l−1 in Chile (Escalante et al. 2009). An As(V)-reducing strain G5-14 from the same district in a previous study exhibited resistance to As(III) from 2 to 25 mmol l−1. Besides, no As(V) tolerance data of those strains were mentioned in that work (Fan et al. 2008). Moreover, Aerobic As-resistant bacteria, Arthrobacter sp. A-1 and GOLO 1, isolated from soils in the same site from Shanxi Province, exhibited As(III) resistance of 20 mmol l−1 and As(V) resistance of 300–400 mmol l−1 (Tian et al. 2012). The MICs of bacteria SXB and IMH are comparable with the MICs in the reported works. The two soils investigated in this study both exhibit high As concentrations over 10 mg kg−1. Perhaps owing to the detoxification mechanism existing in the As(V)-reducing bacterium, isolate SXB is capable of surviving with higher As(III) levels, which were three times greater than the reported As-resistant strains. Isolate SXB shows a slightly higher As(V) resistance than strain IMH, while the MIC of As(III) of isolate SXB is six times larger than that of isolate IMH.

The isolated bacteria SXB and IMH both have effective As(V)-reducing capacities under aerobic conditions. As(V) reduction occurred rapidly when the logarithmic phase began. Both bacteria reduced nearly 80% of the As(V) within 20 h. This is a high reducing ratio in 20 h compared with that of bacteria isolated from Bangladesh (23% for Bacillus sp. Rice-C, 46% for Alcaligenes sp. Dhal L and 0% for Arthrobacteragilis Dhal Rr) (Bachate et al. 2009), Italy (12% for strain MT3) (Ruta et al. 2011) and Korea (45% for strain RS-4 and 23% for strain RS-5) (Chang et al. 2008). By transforming the As(V) into As(III), these As(V)-reducing bacteria will change the fate of As in natural environments. Furthermore, the impacts made by strain SXB on the biogeochemical cycle of As are not only restricted to aerobic conditions. Although much lower As(V)-reducing capacity is exhibited by strain SXB under anaerobic conditions, the effect of the As(V) reduction process due to this strain on the biogeochemical cycle of As cannot be ignored.

The results obtained in this study exhibit the presence of aerobic As(V)-reducing bacteria in severely As-contaminated soils. Pantoea genus is reported to have As(V)-reducing capacity for the first time. The ars genotype illustrates the detoxification mechanism in As(V)-reducing bacteria. Although the roles of bacteria SXB and IMH in As transformation and cycling remain to be clarified, their highly effective As(V)-reducing capacities enable them to release more mobile and toxic As(III) into the environment. Thus, the role of these As(V)-reducing bacteria in the As biogeochemical cycle urgently deserves further investigation in the future.

Acknowledgements

We acknowledge the financial support of the National Natural Science Foundation of China (20977098, 20890112, and 20921063).

Ancillary