Microbes have great potential for arsenic (As) and antimony (Sb) bioremediation in heavily contaminated soil because they have the ability to biotransform As and Sb to species that have less toxicity or are more easily removed. In this study, we integrated a metagenomic method with physicochemical characterization to elucidate the composition of microbial community and functional genes (related to As and Sb) in a high As (range from 34.11 to 821.23 mg kg−1) and Sb (range from 226.67 to 3923.07 mg kg−1) contaminated mine field. Metagenomic analysis revealed that microbes from 18 phyla were present in the 5 samples of soil contaminated with high As and Sb. Moreover, redundancy analysis (RDA) of the relationship between the 18 phyla and the concentration of As and Sb demonstrated that 5 phyla of microbes, i.e. Actinobacteria, Firmicutes, Nitrospirae, Tenericutes and Gemmatimonadetes were positively correlated with As and Sb concentration. The distribution, diversity and abundance of functional genes (including arsC, arrA, aioA, arsB and ACR3) were much higher for the samples containing higher As and Sb concentrations. Based on correlation analysis, the results showed a positive relationship between arsC-like (R2 = 0.871) and aioA-like (R2 = 0.675) gene abundance and As concentration, and indicated that intracellular As(V) reduction and As(III) oxidation could be the dominant As detoxification mechanism enabling the microbes to survive in the environment. This study provides a direct and reliable reference on the diversity of microbial community and functional genes in an extremely high concentration As- and Sb-contaminated environment.
Citation: Luo J, Bai Y, Liang J, Qu J (2014) Metagenomic Approach Reveals Variation of Microbes with Arsenic and Antimony Metabolism Genes from Highly Contaminated Soil. PLoS ONE 9(10): e108185. https://doi.org/10.1371/journal.pone.0108185
Editor: John Parkinson, Hospital for Sick Children, Canada
Received: May 19, 2014; Accepted: August 18, 2014; Published: October 9, 2014
Copyright: © 2014 Luo et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: The authors confirm that all data underlying the findings are fully available without restriction. All relevant data are within the paper and its Supporting Information files. All the five original metagenomic datasets files were archived at the NCBI Sequence Read Archive (SRA) (accession number PRJNA239941).
Funding: This work is financially supported by the Major Program of the National Natural Science Foundation of China (No. 51290282) and National High Technology Research and Development Program of China (No. 2012AA062604). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Arsenic (As) and antimony (Sb) are naturally occurring toxic elements, which are released into the natural environment either by natural activities or anthropogenic sources from various industries . Both As and Sb can be strongly retained in soils , and their toxicity, behavior and bioavailability in the environment intensively rely on speciation and environmental conditions. Because of their similar chemistry, As and Sb are commonly associated with each other in the environment . Despite their toxicity, functional microbes have still been reported to have survived in As- and Sb-contaminated environments , , which has implications for As and Sb bioremediation.
As and Sb microbial metabolisms include arsenate (As(V)) reduction , arsenite (As(III)) oxidation , As methylation , Sb(V) reduction , Sb(III) oxidation , and Sb methylation . The biogeochemical cycling of As and Sb has been shown to be promoted based on redox reactions of microbes . As we know, 99% of microorganisms in the natural environment have not been cultured. Traditional fingerprint methods, such as T-RFLP , DGGE and TGGE , can identify designated DNA fragments and analyze the composition of a microbial community efficiently, whereas these techniques still have some limitations and deficiencies. T-RFLP can only roughly analyze the microbial composition, and has difficulty identifying the species present. As for DGGE/TGGE, it can only analyze DNA fragments that are below 500 base pairs. Assessment of the composition of the whole community gene pool of microbial communities is limited using these methods.
Currently the metagenomic method is regarded as the most efficient, reliable, rapid and accurate way to reveal the entire microbial composition and genetic content of a community under complex environment conditions. It can identify the functional genes related to the metabolism of As and Sb, in order to provide reliable data on the phylogenetic composition and microbial metabolism synchronously . A deep and direct insight into the soil biodiversity and microbial community and its functions can be investigated by using this novel molecular technique. Until now a few reports have been put forward studying water , sediment  and activated sludge  contaminated environments with relatively low concentrations of As by the metagenomic method. However, there are large numbers of As- and Sb-contaminated mine areas with high concentrations of As and Sb worldwide. And As and Sb are always co-exists in the natural environment. However, to the best of our knowledge, there are no reports analyzing As- and Sb-contaminated environments using the metagenomic method. Therefore, it is essential to gain a deep perspective and understanding of the diversity of microbes and functional genes under high concentrations of As and Sb, as well as co-existence of As and Sb through the metagenomic method.
In this study, we adopt a multidisciplinary approach to identify the inorganic components of the soil samples, and use a metagenomic-based molecular technique to decipher the distribution, diversity, and abundance of microbial communities in As- and Sb-contaminated environmental soil samples. The objectives of the current study are to (i) elucidate the whole microbial communities' composition and structure in soil contaminated by high As and Sb concentrations, (ii) identify the bacterium that plays a dominant role under different As and Sb contamination levels, (iii) investigate the abundance, distribution and diversity of functional genes related to As and Sb species transformations, and (iv) use statistical analysis methods to identify the relationship among microbes, redox genes' abundance and soil physicochemical parameters.
Materials and Methods
No specific permits were required for the described field studies in Lengshuijiang City, Hunan Province, China. The location is not privately-owned or protected in any way and the field studies did not involve endangered or protected species.
Soil Sample Collection and Preparation
Five soil samples (named LSJ-1, LSJ-2, LSJ-3, LSJ-4 and LSJ-5) were collected from an As- and Sb-contaminated mine field in Lengshuijiang City, Hunan Province, China (Table S1). Soil samples were collected from 2–5 cm below the surface. The ambient temperature was about 15°C during sample collection. A portable freezer was used to keep the temperature at minus fifteen centigrade during sample collection. Soil samples were dried, pulverized, and sieved to 100 meshes before analysis of physicochemical properties.
The sieved soil samples were freeze-dried by a freeze drier (LGJ-10B, China) before analysis. pH values were measured by pH meter (Thermo Orion 3 STAR, USA) in a soil sample to water ratio of 1∶5 (w/w) after shaking about 24 h. The basic elements C, H and N were analyzed with an elemental analyzer (Elementar Analysen system GmbH, Germany). Before analyzing the total Sb, As, Pb, Cu and Mg, the soil samples were dissolved by a Microwave Reaction System (Multwave 3000) and then measured with an Atomic Absorption Spectrophotometer (ContrAA 700, Germany). Fourier transform infrared (FT-IR) spectroscopy was used to identify the organic compounds present in soil samples. X-ray photoelectron spectroscopy (XPS) was used to determine the distribution of elements.
Soil DNA Extraction
Soil DNA was extracted from the contaminated field soil using the PowerSoil DNA Isolation Kit according to the manufacturer's instructions (Mobio, USA). A 0.25 g soil sample was added to each tube with three replicates, and the DNA was dissolved in 100 µL sterilized deionized water and stored at −20°C before use. The concentration of extracted nucleic acids was determined photometrically using an Eppendorf BioPhotometer plus (Hamburg, Germany).
Real-time polymerase chain reaction (qPCR) was used to determine the quantitative distribution of As- and Sb-related genes and 16S rRNA bacteria. As(V)-reducing functional genes arsC (representing As-reducing bacteria under aerobic conditions) , arrA (representing As-reducing bacteria under anaerobic conditions) , As methylation functional gene arsM , As(III) oxidation gene aroA , as well as As(III) and Sb transfer gene arsB  were quantified using the SYBR Green I method . The information about primer pairs and the PCR thermal programs is described in Table S2.
The SYBR Green I PCR reaction mixture includes 1 µL template (0.4 µL for each primers), 0.4 µL ROX Reference Dye II (50×), and 10 µL 2×SYBR Premix Ex Taq (Tli- RnaseH Plus) within 6.8 µL ddH2O in a 20 µL reaction. Reactions were performed on an ABI 7300 fast Real-time PCR system. Melting curves were performed for each reaction to confirm the purity of amplified products. Questionable products were verified on a 1.2% agarose gel.
Calibration curves were created for each sample by tenfold serial dilution of plasmids whose gene concentrations were quantified in triplicate. A standard curve of eight dilutions run in duplicate was generated for each qPCR plate. All DNA samples concentrations were measured with an Eppendorf BioPhotometer plus (Hamburg, Germany) immediately before diluting tenfold for standard curves. The concentrations of gene copies per gram of the soil were calculated from the gene copies detected in each microliter of DNA, multiplied by the dilution ratio of the DNA template as well as the volume of DNA extracted from one gram of original soil.
Metagenomic analysis was carried out to identify the composition and diversity of the microbial community in the soil.
DNA samples of the five soils were sent to BGI Company for sequencing. An Illumina TruSeq DNA Sample Preparation Kit (Illumina, Inc., San Diego, USA) was used to dispose metagenomic samples and construct two paired-end libraries (350 bp insert size). The prepared templates were measured on the Illumina TruSeq 2000 sequencing system for analysis (90 bp pair-end reads). Sequence reads were filtered, in order to remove adapters, low quality reads and those belonging to the host. We used a velvet de novo assembler  to assemble the filtered reads (∼2 Gbp for both datasets) into long contigs, and they were calculated on the cloud computing platform of the Beijing Computing Center (Beijing, China). MG-RAST on-line software  (http://metagenomics.anl.gov) was used to analyze the assembled contigs in order to investigate the taxonomic classification and function annotation.
Since no As or Sb metabolism genes or protein databases have been reported, we used the same subdatabases of arrA, arsM and aioA which have been put forward by Cai et al.  to analyze arrA, arsM and aioA in our samples. For arsC, ACR3 and arsB analysis, the subdatabases of arsC, ACR3 and arsB were built (Table S3). In order to analyze the distribution, diversity, and abundance of functional gene groups, we compared the original datasets with these subdatabases, and displayed it in a heatmap.
Five original metagenomic datasets were archived at the NCBI Sequence Read Archive (SRA) under the accession number of PRJNA239941.
To compare the community composition and function among the five samples, we used the statistical analysis of metagenomic profiles bioinformation software . The two-sided Fisher's exact test was used to calculate the statistical significance, and the Newcombe-Wilson method with 95% confidence interval was used to analyze the differences between the proportions. The relationships between soil microbes and physicochemical parameters were examined by means of RDA applied to the correlation matrix of these variables. The analysis was performed in Canoco 4.5 for Windows . SPSS 18.0 software was used to analyze the significant differences (P≤0.05) and the least significant difference (LSD) test after one-way analysis of variance (ANOVA) among five soil samples in Real-time PCR analysis. The other statistical analyses were carried out through OriginPro 8.5 software.
Results and Discussion
The Distribution of Inorganic Components in the Soil
The main physicochemical characteristics of the soil samples are shown in Table 1, which included the pH values, the contents of main elements C, N and H and the concentrations of Sb, As, and other heavy metals (including Pb, Cu, Cd, Cr, Fe, and Mg). In this study, the concentrations of As and Sb were much higher than those reported in the literature (Table S4). The five soil samples possessed similar pH values (pH 7.19 to 7.75), which could be due to the high buffering capacity of the clay soils. Even though the concentrations of As and Sb were extremely high in these environments, they fluctuated among the five samples. For LSJ-4 and LSJ-5, despite the close location of the two sites, the concentrations of As and Sb in LSJ-4 were much higher than in LSJ-5. Being aware that the soil sample of LSJ-5 was close to the root of plants while LSJ-4 was far from the root of plants, the discrepancy possibly arose because the hyperaccumulating plants  could adsorb As and Sb from the soil, and contributed to the lower As and Sb concentrations in LSJ-5. FT-IR and XPS data were analyzed in order to gain a further understanding of the physicochemical properties of the soil samples that could contribute to the composition and function of the microbial community. As shown in Figure. 1a, the FT-IR spectra reveal that the dominant organic functional groups in the soil samples were -CO-, C = O (NH2- and -NH-), CH2-, and OH- as recognized by peaks at around 1050 cm−1, 1700 cm−1, 2970 cm−1, 3420 cm−1, respectively . XPS analysis identified that there was Fe, O, C, Si and Al in the soil samples (Figure. 1b). In summary, the results revealed that the concentrations of As and Sb varied among the five samples but the organic and inorganic compositions in the samples were identical.
The Diversity of Microbial Communities Based on the Metagenomic Datasets
After velvet assembly, five metagenomic datasets were uploaded to MG-RAST for taxonomic classification and function annotation. From 16S rRNA gene identification, bacteria play a dominant role among three domains, and account for 68.3%, 74.1%, 71.0%, 46.4%, and 62.9%, among the samples of LSJ-1, LSJ-2, LSJ-3, LSJ-4 and LSJ-5, respectively (Figure. 2a). Archaea was not observed in LSJ-1 and LSJ-3. As shown in Figure. 2b, Proteobacteria had a higher abundance than the others on the phylum level, including six classes (Betaproteobacteria, Gammaproteobacteria, Zetaproteobacteria, Alphaproteobacteria, Epsilonproteobacteria, and Deltaproteobacteria) and unclassified (derived from Proteobacteria). Gammaproteobacteria was the dominant species among the five soil samples.
All the information was obtained from metagenomic datasets revealed by Illumina.
The domain analysis among the five samples from the metagenomic datasets indicates that archaea exist universally in the surrounding environment, while archaea is not detected in extreme environments (with extremely high concentrations of As and Sb), such as the area where the LSJ-1 and LSJ-3 samples were collected. Proteobacteria is bacteria that can tolerate extremely severe environments and widely exist in the environment. In addition, it can be seen from Figure. 2b that Actinobacteria is another large group present under the rigorous circumstances, and have a high tolerance to extremely high concentrations of As and Sb as well. Nitrospirae and Gemmatimonadetes were only found in sample LSJ-3, which had an extremely high concentration of Sb (CSb = 3923.07 mg/kg). Compared to Nitrospirae and Gemmatimonadetes, Chlorobi and Bacteroidetes were only found in the sample LSJ-2 with extremely high concentration of As (CAs = 821.23 mg/kg).
Distribution, Diversity and Abundance of As and Sb Functional Genes
The species transformations of As and Sb is greatly influenced by microorganisms. Thus, it is essential to quantify the abundance of the functional genes related to As and Sb metabolism using Real-time PCR (qPCR). In order to obtain the relative expression of each functional gene, we compared each functional gene's abundance with 16S rRNA bacteria abundance. Rosen et al.  and Mukhopadhyay et al. , indicate that arsB is responsible for As (III) and Sb(III) transfer. While other four genes, such as arsC, arrA, arsM and aroA are responsible for As redox and methylation reaction.
As shown in Figure. S1, the relative expression of functional genes in different samples showed significant differences. The arsC gene had the highest mean relative expression with 1.32×10−2 (Figure. S1. c),while arsM was the lowest with 3.50×10−4 (Figure. S1. e). Since arsC is reported mostly under aerobic conditions, while arrA (1.73×10−3) is found under anaerobic conditions (Figure. S1. b) , the higher expression of arsC (compared to arrA) could indicate that the microbes that reduce As(V) under aerobic conditions may have higher abundance than the anaerobic microbes. The arsC and arrA gene mediated As(V) reduction process seems to be the most indispensable part of the As biogeochemical cycle. This process can reduce As(V) to As(III), and promote As mobility and bioavailability in the environment . It also contributes to the release of adsorbed As(V) from solid surfaces, such as ferrihydrite and a common soil mineral , to the aqueous phase. Relatively high expression was obtained for the aroA gene (5.39×10−3 (Figure. S1. a)), which mediates the As(III) oxidation process. As (III) oxidation is thought to be the microbial detoxification metabolism in various environments that oxidizes toxic As(III) to less toxic As (V). A bioreactor experiment has been carried out to remove As(III) in synthetic groundwater using immobilized As(III)-oxidation bacteria, which were isolated from activated sludge . Meanwhile, As(V) is immobilized and more easily removed than As(III) in aqueous environments , , especially using FeOOH . The expression of arsM was the lowest detected in the five samples. The arsM gene can regulate As(III) methylation, which is another detoxification mechanism in microorganisms. Qin et al.  showed that Rhodopseudomonas palustris can create volatile trimethylarsine (TMAs) as the final product of the As(III) methylation reaction. Since the methylation process of As can volatilize As as an organic form to the atmosphere , it is an efficient means to bioremediate As-contaminated soil. Based on the high abundance of redox genes and methylation genes, there is indication that high levels of microbial activities related to As and Sb metabolism exist in the area. The research on Sb metabolism genes and proteins is far more limited than that for As. Until now, only one Sb(V) reduction protein has been identified from Lershamania, named LmACR2 . The functional gene arsB is related to both Sb(III) and As(III) transport . The relative expression of arsB in LSJ-2 was significant higher than in the other four samples, which could be attributed to the high As and Sb concentration (Figure. S1. d).
It is quite essential to conduct an evaluation of As and Sb functional genes based on the distribution, diversity and abundance of gene levels among the five samples using heatmap analysis. In LSJ-3, arsC-like and arrA-like genes (Figure. 3a) show relatively high distribution, diversity and abundance. Based on analysis of arsC-like and arrA-like genes, even though the gene distribution was variable, a cluster analysis indicates that LSJ-1 and LSJ-2 share the closest distance and further cluster with LSJ-3, and finally cluster with LSJ-4 and LSJ-5. However, relatively high diversity, distribution and abundance of the aioA-like gene is observed in LSJ-2, with the highest concentration of As and relatively high concentration of Sb. Furthermore the cluster analysis of the aioA-like gene is the same as arsC-like and arrA-like genes (Figure. 3b). For arsM-like genes, it is evident that the LSJ-1 sample shows relatively high diversity when compared with the other four samples (Figure. 3c). Based on the analysis above, the diversity, distribution and abundance of all functional genes are relatively high even under extremely high concentrations of As and Sb.
Phylogenetic analysis used full-length proteins derived from each subdatabase. Complete linkage clustering of five samples was calculated based on the composition and relative abundance of arsenite methylation, arsenite oxidation and arsenate reduction genes. Scales of completely black (−4), pale (−4 to −3), yellow (−3 to −2), orange (−2 to −1), and red(>−1) indicated the abundance of 0%, 0.01–0.1%, 0.1–1%, 1–10%, and >10%, respectively.
Two different membrane protein phylogenetic clades, called ArsB and ACR3, can pump out As(III) and Sb(III) from microbial cells , . These two proteins were selected to conduct further analysis on their distribution, diversity and abundance among the five soil samples. For ACR3-like genes and arsB-like genes, LSJ-2 shows relatively high diversity compared to the other four samples. Even though the gene distribution was variable, cluster analysis of the two samples indicates that LSJ-1 and LSJ-2 share the closest distance and further cluster with LSJ-3, and finally cluster with LSJ-4 and LSJ-5 (Figure. 4). The results for the two functional genes could be mainly attributed to the high concentration of As and Sb.
Phylogenetic analysis using full-length proteins derived from each subdatabase. Complete linkage clustering of five samples was calculated based on the composition and relative abundance of arsenite methylation, arsenite oxidation and arsenate reduction genes. Scales of completely black (−4), pale (−4 to −3), yellow (−3 to −2), orange (−2 to −1), and red(>−1) indicated the abundance of 0%, 0.01–0.1%, 0.1–1%, 1–10%, and >10%, respectively.
Comparing the results between qPCR and heatmap analysis, high distribution, diversity and abundance of arsB was obtained in both analyses. The results of other functional genes in the heatmap were positive with As and Sb concentration, while this tendency was not evidently observed in qPCR. One can mainly attribute the possible discrepancy to the differences between the two methods in analyzing the distribution, diversity and abundance of As- and Sb-related functional genes . Such inconsistent results between qPCR and metagenomic data have also appeared in Yang et al.  and Ye et al.'s  studies. Ye et al. also claim that metagenomic sequencing was demonstrated to be a better approach than qPCR.
Statistical Analysis of Microbes, Redox Gene Abundance and Soil Physicochemical Factors
Based on Detrended Correspondence Analysis (DCA) analysis, the length of gradient is less than 3, so a linear model was used in this study . In order to analyze the relationship between microbial species and soil physicochemical factors, RDA analysis was used. As shown in Figure. 5, there are multivariate relationships between physicochemical factors and microbial species in the soil samples. The environmental factors, such as pH value and Sb concentration, have great influence on species. Five species (including Actinobacteria, Firmicutes, Nitrospirae, Tenericutes and Gemmatimonadetes) are positively correlated with As and Sb, indicating that these five species can tolerate high concentrations of As and Sb, and are also dominant species in the environment. Clearly, Nitrospirae, Tenericutes and Gemmatimonadetes are more positively correlated with Sb. This is mainly due to the existence of a detoxification mechanism in these species. Jackson et al.  have identified and isolated Firmicutes as As(V)-resistant bacteria from As-contaminated woodland soil. Handley et al.  have identified Firmicutes as a dissimilatory As(V)-reducing prokaryote as well. And based on Joseph et al.  report, Gemmatimonadetes, as a previously uncultured soil bacterium, has been isolated under laboratory cultivation. Recently, Wang et al.  have focused on the study of the inhibition rate of culturable Actinobacteria populations under a compound As and Sb pollution system. Integrating these results with our RDA analysis results, we may deduce that Firmicutes was more related to As bioremediation, while Gemmatimonadetes and Actinobacteria, which have been isolated from the soil environment, can serve as a potential bioremediation in both As and Sb contaminated soil. In addition there are another four species highly positively correlated with soil pH, including Poribacteria, Fusobacteria, Bacteroidetes and Chlorobi.
Axis 1 and axis 2 account for 51.2% and 22.7% of the variance, respectively. a-r indicate different microbes under phylum level: a (Acidobacteria); b (Actinobacteria); c (Aquificae); d(Bacteroidetes); e (Chlorobi); f (Cyanobacteria); g (Deferribacteres); h (Deinococcus-Thermus); i (Firmicutes); j (Fusobacteria); k (Gemmatimonadetes); l (Nitrospirae); m (Planctomycetes); n (Poribacteria); o (Proteobacteria); p(Spirochaetes); q (Tenericutes); r (Verrucomicrobia); TC (total C content); TN (total N content); TH (total H content)
As shown in Figure. 6, based on correlation analysis of the relationships between the functional gene abundance and As concentration, a significant linear relation of arsC-like genes and As concentration (R2 = 0.871) (Figure. 6a), as well as aioA-like genes and As concentration (R2 = 0.675) (Figure. 6b), were obtained. The abundance of arsC-like and aioA-like genes increases with increasing As concentration. arsC has been reported to be the dominant gene involved in the intracelluar microbial As detoxification process, which reduces As(III) and then pumps it out to the surrounding environment . aioA is also the key gene for As(III) oxidation, which is one of the As detoxification mechanisms as well . With the relatively high R2 values for both arsC and aioA, we strongly propose that the survive of these detected microbes under such a heavily As-contaminated mine field could be attributed to these As detoxification processes.
Points are arranged according to the local concentration of arsenic (from low concentration to high concentration, which from 34.1 mg L−1 to 821.2 mg L−1) corresponding with the abundance of functional proteins
In summary, using a molecular method (based on metagenomics) and statistical analysis with physicochemical data, this study demonstrates that Proteobacteria is the dominant species in high As- and Sb-contaminated environments. Our findings also reveal that As and Sb metabolism genes are diverse, abundant and universal in highly As- and Sb-contaminated environments, and the positive relation between the abundance of As detoxification genes (including arsC and aioA) and increasing As concentration. This present work promotes our understanding of microbial activities under highly contaminated environments, and opens up an avenue to investigate the relationships between microorganisms and the environment. It could also have implications for us to harness microbes in dealing with environmental issues.
The relative expression of aioA/16S bacteria, arrA/16S bacteria, arsC/16S bacteria, arsB/16S bacteria and arsM/16S bacteria.
The accurate locations for the sample collection.
The qPCR primer pairs and thermal programs of the functional genes, including arsB, arsC, arrA, arsM and aroA.
Accession numbers of arsenic metabolism protein sub-databases involved in arsC-like, arsB-like and ACR3-like genes.
This work is financially supported by the Major Program of the National Natural Science Foundation of China (No. 51290282) and National High Technology Research and Development Program of China (No. 2012AA062604). The authors give thanks to Nana Zhou (Xi'an University of Architecture and Technology) for valuable suggestions during the experiments. The authors also would like to acknowledge the assistance of the Hsikwangshan Twinkling Star Co., Ltd. in soil collection.
Conceived and designed the experiments: J. Luo YB J. Liang JQ. Performed the experiments: J. Luo YB J. Liang JQ. Analyzed the data: J. Luo YB J. Liang JQ. Contributed reagents/materials/analysis tools: J. Luo YB J. Liang JQ. Wrote the paper: J. Luo YB J. Liang JQ.
- 1. Marcovecchio JE, Botté SE, Freije RH (2007) Heavy metals, major metals, trace elements. Handbook of Water Analysis: 275–311.
- 2. Flynn HC, Meharg AA, Bowyer PK, Paton GI (2003) Antimony bioavailability in mine soils. Environ Pollut 124: 93–100.
- 3. Lehr CR, Kashyap DR, McDermott TR (2007) New insights into microbial oxidation of antimony and arsenic. Appl Environ Microbiol 73: 2386–2389.
- 4. Wang X, Rathinasabapathi B, de Oliveira LM, Guilherme LR, Ma LQ (2012) Bacteria-mediated arsenic oxidation and reduction in the growth media of arsenic hyperaccumulator Pteris vittata. Environ Sci Technol 46: 11259–11266.
- 5. Li J, Wang Q, Zhang S, Qin D, Wang G (2013) Phylogenetic and genome analyses of antimony-oxidizing bacteria isolated from antimony mined soil. Int Biodeterior Biodegrad 76: 76–80.
- 6. Oremland RS, Stolz JF (2003) The ecology of arsenic. Science 300: 939–944.
- 7. Bentley R, Chasteen TG (2002) Microbial methylation of metalloids: arsenic, antimony, and bismuth. Microbiol Mol Biol R 66: 250–271.
- 8. Zhou Y, Messier N, Ouellette M, Rosen BP, Mukhopadhyay R (2004) Leishmania major LmACR2 is a pentavalent antimony reductase that confers sensitivity to the drug pentostam. J Biol Chem 279: 37445–37451.
- 9. Cutter GA, Cutter LS, Featherstone AM, Lohrenz SE (2001) Antimony and arsenic biogeochemistry in the western Atlantic Ocean. Deep-Sea Res Pt II: Topical Studies In Oceanography 48: 2895–2915.
- 10. Schütte UM, Abdo Z, Bent SJ, Shyu C, Williams CJ, et al. (2008) Advances in the use of terminal restriction fragment length polymorphism (T-RFLP) analysis of 16S rRNA genes to characterize microbial communities. Appl Microbiol Biotechnol 80: 365–380.
- 11. Muyzer G, Smalla K (1998) Application of denaturing gradient gel electrophoresis (DGGE) and temperature gradient gel electrophoresis (TGGE) in microbial ecology. Antonie Van Leeuwenhoek 73: 127–141.
- 12. Thomas T, Gilbert J, Meyer F (2012) Metagenomics-a guide from sampling to data analysis. Microb Inform Exp 2.
- 13. Bertin PN, Heinrich-Salmeron A, Pelletier E, Goulhen-Chollet F, Arsene-Ploetze F, et al. (2011) Metabolic diversity among main microorganisms inside an arsenic-rich ecosystem revealed by meta- and proteo-genomics. ISME J 5: 1735–1747.
- 14. Plewniak F, Koechler S, Navet B, Dugat-Bony É, Bouchez O, et al. (2013) Metagenomic insights into microbial metabolism affecting arsenic dispersion in Mediterranean marine sediments. Mol Ecol 22: 4870–4883.
- 15. Cai L, Yu K, Yang Y, Chen B-w, Li X-d, et al. (2013) Metagenomic exploration reveals high levels of microbial arsenic metabolism genes in activated sludge and coastal sediments. Appl Microbiol Biotechnol 97: 9579–9588.
- 16. Silver S, Phung LT (2005) Genes and enzymes involved in bacterial oxidation and reduction of inorganic arsenic. Appl Environ Microbiol 71: 599–608.
- 17. Malasarn D, Saltikov C, Campbell K, Santini J, Hering J, et al. (2004) arrA is a reliable marker for As (V) respiration. Science 306: 455–455.
- 18. Yin X-X, Chen J, Qin J, Sun G-X, Rosen BP, et al. (2011) Biotransformation and volatilization of arsenic by three photosynthetic cyanobacteria. Plant Physiol 156: 1631–1638.
- 19. Inskeep WP, Macur RE, Hamamura N, Warelow TP, Ward SA, et al. (2007) Detection, diversity and expression of aerobic bacterial arsenite oxidase genes. Environ Microbiol 9: 934–943.
- 20. Dey S, Rosen BP (1995) Dual mode of energy coupling by the oxyanion-translocating ArsB protein. J Bacteriol 177: 385–389.
- 21. Noble RT, Fuhrman JA (1998) Use of SYBR Green I for rapid epifluorescence counts of marine viruses and bacteria. Aquat Microb Ecol 14: 113–118.
- 22. Zerbino DR, Birney E (2008) Velvet: algorithms for de novo short read assembly using de Bruijn graphs. Genome Research 18: 821–829.
- 23. Meyer F, Paarmann D, D'Souza M, Olson R, Glass EM, et al. (2008) The metagenomics RAST server–a public resource for the automatic phylogenetic and functional analysis of metagenomes. BMC Bioinformatics 9: 386.
- 24. Parks DH, Beiko RG (2010) Identifying biologically relevant differences between metagenomic communities. Bioinformatics 26: 715–721.
- 25. Braak Ct, Smilauer P (2002) CANOCO reference manual and CanoDraw for Windows user's guide: software for canonical community ordination (version 4.5).
- 26. Rascio N, Navari-Izzo F (2011) Heavy metal hyperaccumulating plants: how and why do they do it? And what makes them so interesting? Plant Sci 180: 169–181.
- 27. China SEPAo (1989) Monitoring and analysis method of water and wastewater. 3rd edn. Beijing. China Environmental Science Press (In Chinese).
- 28. Rosen BP (2002) Biochemistry of arsenic detoxification. FEBS Lett 529: 86–92.
- 29. Mukhopadhyay R, Rosen BP, Phung LT, Silver S (2002) Microbial arsenic: from geocycles to genes and enzymes. FEMS Microbiol Rev 26: 311–325.
- 30. Smedley P, Kinniburgh D (2002) A review of the source, behaviour and distribution of arsenic in natural waters. Appl geochem 17: 517–568.
- 31. Zobrist J, Dowdle PR, Davis JA, Oremland RS (2000) Mobilization of arsenite by dissimilatory reduction of adsorbed arsenate. Environ Sci Technol 34: 4747–4753.
- 32. Ito A, Miura J-i, Ishikawa N, Umita T (2012) Biological oxidation of arsenite in synthetic groundwater using immobilised bacteria. Water Res 46: 4825–4831.
- 33. Raven KP, Jain A, Loeppert RH (1998) Arsenite and arsenate adsorption on ferrihydrite: kinetics, equilibrium, and adsorption envelopes. Environ Sci Technol 32: 344–349.
- 34. Green H (1919) Isolation and Description of a Bacterium causing Oxidation of Arsenite to Arsenate in Cattle Dipping Baths. Union S Africa Dept Agrie 5th and 6th Repts Direc Vet Research, April 1918: 595–610.
- 35. Waychunas G, Rea B, Fuller C, Davis J (1993) Surface chemistry of ferrihydrite: Part 1. EXAFS studies of the geometry of coprecipitated and adsorbed arsenate. Geochim Cosmochim Ac 57: 2251–2269.
- 36. Qin J, Rosen BP, Zhang Y, Wang G, Franke S, et al. (2006) Arsenic detoxification and evolution of trimethylarsine gas by a microbial arsenite S-adenosylmethionine methyltransferase. Proc Natl Acad Sci USA 103: 2075–2080.
- 37. Chen J, Qin J, Zhu Y-G, de Lorenzo V, Rosen BP (2013) Engineering the soil bacterium Pseudomonas putida for arsenic methylation. Appl Environ Microbiol 79: 4493–4495.
- 38. Meng Y-L, Liu Z, Rosen BP (2004) As (III) and Sb (III) uptake by GlpF and efflux by ArsB in Escherichia coli. J Biol Chem 279: 18334–18341.
- 39. Filella M, Belzile N, Lett M-C (2007) Antimony in the environment: A review focused on natural waters. III. Microbiota relevant interactions. Earth-Sci Rev 80: 195–217.
- 40. Schmittgen TD, Livak KJ (2008) Analyzing real-time PCR data by the comparative CT method. Nature Protoc 3: 1101–1108.
- 41. Yang Y, Li B, Ju F, Zhang T (2013) Exploring variation of antibiotic resistance genes in activated sludge over a four-year period through a metagenomic approach. Environ Sci Technol 47: 10197–10205.
- 42. Ye L, Zhang T, Wang T, Fang Z (2012) Microbial structures, functions, and metabolic pathways in wastewater treatment bioreactors revealed using high-throughput sequencing. Environ Sci Technol 46: 13244–13252.
- 43. Lepš J, Šmilauer P (2003) Multivariate analysis of ecological data using CANOCO: Cambridge university press.
- 44. Jackson CR, Dugas SL, Harrison KG (2005) Enumeration and characterization of arsenate-resistant bacteria in arsenic free soils. Soil Biol Biochem 37: 2319–2322.
- 45. Handley KM, Héry M, Lloyd JR (2009) Redox cycling of arsenic by the hydrothermal marine bacterium Marinobacter santoriniensis. Environ Microbiol 11: 1601–1611.
- 46. Joseph SJ, Hugenholtz P, Sangwan P, Osborne CA, Janssen PH (2003) Laboratory cultivation of widespread and previously uncultured soil bacteria. Appl Environ Microbiol 69: 7210–7215.
- 47. Wang Q, He M, Wang Y (2011) Influence of combined pollution of antimony and arsenic on culturable soil microbial populations and enzyme activities. Ecotoxicology 20: 9–19.
- 48. Martin P, DeMel S, Shi J, Gladysheva T, Gatti DL, et al. (2001) Insights into the structure, solvation, and mechanism of ArsC arsenate reductase, a novel arsenic detoxification enzyme. Structure 9: 1071–1081.
- 49. Cavalca L, Corsini A, Zaccheo P, Andreoni V, Muyzer G (2013) Microbial transformations of arsenic: perspectives for biological removal of arsenic from water. Future Microbiol 8: 753–768.