Skip to main content

Thank you for visiting nature.com. You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

Meta-omics-aided isolation of an elusive anaerobic arsenic-methylating soil bacterium

Abstract

Soil microbiomes harbour unparalleled functional and phylogenetic diversity. However, extracting isolates with a targeted function from complex microbiomes is not straightforward, particularly if the associated phenotype does not lend itself to high-throughput screening. Here, we tackle the methylation of arsenic (As) in anoxic soils. As methylation was proposed to be catalysed by sulfate-reducing bacteria. However, to date, there are no available anaerobic isolates capable of As methylation, whether sulfate-reducing or otherwise. The isolation of such a microorganism has been thwarted by the fact that the anaerobic bacteria harbouring a functional arsenite S-adenosylmethionine methyltransferase (ArsM) tested to date did not methylate As in pure culture. Additionally, fortuitous As methylation can result from the release of non-specific methyltransferases upon lysis. Thus, we combined metagenomics, metatranscriptomics, and metaproteomics to identify the microorganisms actively methylating As in anoxic soil-derived microbial cultures. Based on the metagenome-assembled genomes of microorganisms expressing ArsM, we isolated Paraclostridium sp. strain EML, which was confirmed to actively methylate As anaerobically. This work is an example of the application of meta-omics to the isolation of elusive microorganisms.

Introduction

Soil microbiomes represent a rich source of novel metabolisms and taxa [1,2,3,4]. However, isolating microorganisms from them to study specific functions can be challenging, and even more so in cases for which the phenotype is not identifiable with high-throughput methods [5, 6]. An example of challenging microorganisms to isolate are anaerobic As-methylating strains. Arsenic methylation, catalysed by arsenite (As(III)) S-adenosylmethionine methyltransferase (ArsM, in prokaryotes), entails the binding of one to three methyl group(s) to the As atom [7]. At present, there are no available microorganisms capable of anaerobic As methylation. This is because, adding to the constraints associated with maintaining an anoxic environment [8], there is no assay for ArsM activity that can be adapted for high-throughput assessment, despite recent endeavours [9]. Arsenic methylation occurs in anoxic flooded rice paddy soils, is mediated by soil microorganisms [10], and results in the accumulation of methylated As in rice grains [11]. The bioaccumulation of methylated As in rice grains is considerably more efficient than that of inorganic As [12, 13].

The gene encoding ArsM (arsM) has been identified in phylogenetically diverse soil microorganisms [14,15,16,17]. Anaerobic As methylation is expected to produce a toxic trivalent monomethylated As species (MMAs(III)). The function of this transformation is hypothesised to be microbial warfare, by which the As-methylating organism inhibits microbial competitors via the production of MMAs(III) [18, 19]. If that is confirmed, it is conceivable that As methylation may not occur in pure cultures but only in microbial communities, triggered by metabolites produced by the microbiota. Alternatively, arsM-harbouring microorganisms that express As(III) efflux pump(s), the major pathway of As resistance within bacteria [20], may not methylate As due to the efficient removal of As(III) from the cytoplasm, which is the location of ArsM [21, 22]. This effect could be direct, i.e., insufficient substrate concentration, or indirect, i.e., the intracellular As(III) concentration is too low to induce arsM expression. Either occurrence (microbial warfare or rapid As(III) efflux) would render the isolation of pure cultures of As-methylating anaerobes very challenging using standard approaches. The latter hypothesis is supported by recent work showing the lack of As methylation by anaerobic pure cultures harbouring functional ArsM enzymes [22].

An additional complexity is evidence for the fortuitous methylation of As upon cell lysis and the release of methyltransferases. This fortuitous activity was suggested for the methanogen Methanosarcina mazei, for which As methylation was initiated only when cell viability decreased [22], and by the in vitro methylation of various metals, including As, by MtaA, a methyltransferase involved in methanogenesis [23]. Thus, As methylation activity in cultures incubated beyond the exponential phase may simply be an experimental artefact [22]. Finally, the detection of methylated As requires relatively complex analytical tools (high pressure liquid chromatography coupled to inductively-coupled plasma mass spectrometry, HPLC-ICP-MS) that do not lend themselves readily to high-throughput screening of a large number of colonies [9]. As a result of these challenges, there are no anaerobic microorganisms available known to actively methylate As despite many efforts to identify them. In one instance, researchers had identified a Gram-positive sulfate-reducing bacterium (SRB) [24] that was reported to methylate As but this isolate is no longer available, precluding further investigation.

Thus, this study aimed to conclusively identify an active anaerobic As methylator in soil-derived microbial cultures using a multi-omics approach. The experimental strategy was to build Metagenome-Assembled Genomes (MAGs) from metagenomic data and to identify the subset of MAGs harbouring the gene arsM that also expressed the arsM RNA transcript (metatranscriptomics) and/or the enzyme ArsM (metaproteomics). Based on the genetic information from the target MAG, an isolation strategy was devised that allowed the recovery of a pure culture, later confirmed to be a novel anoxic As-methylating strain.

Materials and methods

Rice paddy soil microbiomes

The soil-derived cultures consisted of two anaerobic microbial enrichments derived from a Vietnamese rice paddy soil and described in Reid et al. [25]. The microbiota from the first soil-derived microbiome was grown in ¼ strength tryptic soy broth (TSB) medium (7.5 g l−1 TSB), used previously to enrich As-methylating microbes from a lake sediment [26], and henceforth referred to as the TSB culture. The medium for the second soil-derived microbiome, in addition to ¼ strength TSB, included electron acceptors and two additional carbon sources to simultaneously allow the growth of nitrate-, iron-, and sulfate-reducers, as well as fermenters and methanogens (EA medium: 5 mM NaNO3, 5 mM Na2SO4, 5 mM ferric citrate, 0.2 g l−1 yeast extract (Oxoid, Hampshire, UK) and 1 g l−1 cellobiose, pH 7). This enrichment will be referred to as the EA culture. Both media were boiled, cooled down under 100% N2 gas and 50 ml of medium were dispensed into 100-ml serum bottles. The bottle headspace was flushed with 100% N2 gas prior to autoclaving. All culture manipulations were carried out using N2-flushed syringes and needles. Cultures were grown at 30 °C. Growth was quantified using optical density at 600 nm (OD600).

Arsenic methylation assays

Pre-cultures from each enrichment were started from −80 °C glycerol stocks. The EA culture started from the glycerol stock was transferred only after a dark precipitate, presumably iron sulfide resulting from sulfate reduction, was formed. The first experimental set-up consisted of bottles containing medium amended with As(III) as NaAsO2 (+As condition) pre-inoculation or unamended (no-As control). For this set-up, cell pellets were sampled for DNA sequencing and proteome characterisation during the stationary phase, and for RNA sequencing at the mid-exponential growth phase (see Supplementary Figs. S1, S2 and S3 in Supplementary Information (SI)). In a second experimental set-up, cultures were grown in unamended (no As(III) added) medium and As(III) was added at the mid-exponential growth phase. For this set-up, cell pellets were sampled before (no-As control) and 30 min after As amendment (+As condition) and were used solely for a second transcriptomic analysis. Triplicate biological experiments were performed for each condition (no-As, +As) and per soil-derived enrichment and were used for DNA and RNA sequencing and metaproteome characterisation. Sampling for soluble As species, determination of As speciation, and total As concentration are described in SI.

DNA sequencing and metagenomic analysis

DNA was extracted from the pellet (10 min, 4500 × g) of 4 ml of culture using the DNeasy Power Soil Kit (Qiagen, Hilden, Germany) homogenising with a Precellys 24 Tissue Homogeniser (Bertin Instruments, Montigny-le-Bretonneux, France) (6,500 rpm for 10 s, repeated 3× with 10 s pause intervals). Metagenomic sequencing was performed by the Genomics Platform of the University of Geneva, Switzerland (iGE3) on a HiSeq 4000 (Illumina, San Diego, CA, US). Libraries were multiplexed and prepared using 100-base reads with paired ends according to the Nextera DNA Flex Library Preparation Kit protocol (Illumina). The quality of sequence reads was assessed with FastQC [27] and duplicated reads eliminated by FastUniq [28]. Reads from all biological replicates within the same experimental condition were assembled into contigs using MegaHit [29]. The contig abundance was determined by aligning the sequencing reads from each biological replicate back to the assembled contigs using Kallisto [30, 31]. The abundance for each gene was considered equivalent to the abundance of the contig in which it was encoded. Gene abundance is reported as ‘transcripts per million’ (TPM), referred to as TPM-DNA when used for gene abundance. TPM includes normalisation for gene length and read sequencing depth [32]. Prodigal was used for the prediction of protein-coding genes [33], generating protein sequence libraries for each culture (EA, TSB) and condition (no-As control, +As condition). The annotation server GhostKOALA [34] was used to assign a KEGG Orthology (KO) database number to each protein-coding gene to identify its encoded function and taxonomic category. The 16S small subunit (SSU) rRNA sequences were identified in the contigs and their taxonomy assigned by Metaxa2 [35]. The relative abundance of the 16S SSU rRNA sequences identified in each of the four metagenomes was quantified using the Kallisto-calculated contig abundance. Contigs with length >2000 bp were clustered into bins based on composition and coverage using CONCOCT [36], MetaBAT2 [37] and MaxBin 2.0 [38]. The final bin set was obtained by using the Bin_refinement module from MetaWRAP [39]. Completeness, contamination, strain heterogeneity and community (%) in contigs for each bin were calculated using CheckM [40]. Matching bins between the no−As and +As metagenomes, and between the EA +As and TSB +As metagenomes were identified by pairwise comparison of the predicted genomes using dRep [41]. Bins with an average nucleotide identity >95% were considered identical genomes.

RNA sequencing and metatranscriptomic analysis

Each culture (5 ml) was harvested at mid-exponential phase for metatranscriptomic analysis. The cells were lysed and the RNA purified using the RNeasy Mini Kit following the manufacturer’s instructions (RNAprotect Bacteria, Qiagen). The purified RNA was DNase-I treated (Promega, Madison, WI, US) (1 h, 37 °C) and cleaned using the RNeasy Mini Kit a second time. Ribosomal RNA (rRNA) depletion (kit QIAseq FastSelect −5S/16S/23S, Qiagen), library preparation using single-end 100 bases reads (TrueSeq Stranded mRNA, Illumina) and RNA sequencing (on a HiSeq 4000) were performed by the iGE3 Platform. Reads were quality-assessed by FastQC, trimmed by Trimmomatic [42], post-sequencing rRNA-depleted by SortMeRNA [43] and aligned to their corresponding protein sequence library by Bowtie2 [44]. The program featureCounts [45] was employed to count the number of RNA reads aligned to the Prodigal-predicted protein-coding genes. The raw counts were used to calculate the TPM, referred as TPM-RNA when employed for transcript abundance. Finally, to assess RNA expression changes in the +As condition relative to the no−As condition, a differential abundance analysis was performed using the DESeq2 package [46] using the protein sequence libraries from the +As condition to align the RNA reads. A gene was considered to have a significant difference in transcription when the absolute log2 fold change was ≥1 (i.e., 0.5 ≥ fold change ≥2) and the adjusted q value ≤0.05.

Metaproteome characterisation and metaproteomic analysis

The metaproteome analysis was performed at Oak Ridge National Laboratory (Oak Ridge, TN, US). Biomass pellets from 100 ml of culture were washed with 100 mM NH4HCO3 buffer (ABC) (pH 8.0), re-suspended in lysis buffer (4% sodium dodecyl sulfate, 100 mM Tris-HCl, pH 8.0) and disrupted by bead-beating. Lysate proteins were reduced with 5 mM dithiothreitol (30 min, 37 °C), alkylated with 15 mM iodoacetamide (30 min in the dark, room temperature) and isolated by a chloroform-methanol extraction. Extracted proteins were solubilized in 4% sodium deoxycholate (SDC) in ABC and the concentration estimated with a Nanodrop (Thermo Fisher Scientific, Waltham, MA, US). Sequencing-grade trypsin (Promega) at a 1:75 enzyme:protein ratio (w/w) was used to digest the proteins and formic acid (1% final concentration) was used to precipitate the SDC and collect tryptic peptides. Aliquots of 12 μg of peptides were analysed by 2D LC-MS/MS consisting of a Vanquish UHPLC connected to a Q Exactive Plus MS (Thermo Fisher Scientific). Spectral data were collected using MudPIT (multidimensional protein identification technology) as described previously [47, 48]. Peptides were separated in three steps (35, 100, and 500 mM ammonium acetate eluent) with organic gradients after each step. Eluted peptides were measured and sequenced by data-dependent acquisition using previously described parameters [49].

Protein databases were created for the +As experimental condition (EA +As and TSB +As) from the corresponding protein sequence libraries generated by Prodigal. The MS/MS spectra raw files were processed in Proteome Discoverer version 2.4 (Thermo Fisher Scientific) with MS Amanda 2.0 [50] and Percolator [51]. Spectral data were searched against the protein database of the corresponding culture (i.e., EA or TSB). The following parameters were used in the search algorithm MS-Amanda 2.0 to derive tryptic peptides: MS1 tolerance = 5 ppm; MS2 tolerance = 0.02 Da; missed cleavages = 2; carbamidomethyl (C, +57.021 Da) as static modification; and oxidation (M, +15.995 Da) as dynamic modifications. The false discovery rate (FDR) threshold was set to 1% for strict FDR and 5% for relaxed FDR at the peptide-spectrum matched (PSM), peptide, and protein levels. FDR-controlled peptides were then quantified according to the chromatographic area-under-the-curve and mapped to their respective proteins. Areas were summed to estimate protein-level abundance.

For differential abundance analysis of proteins, the spectral data from the no−As control, EA no As and TSB no As, were searched against the EA +As and TSB +As protein databases, respectively. All the above-described parameters were maintained. The proteins with at least one peptide detected were exported from Proteome Discoverer. Protein data matrix from EA no As and EA +As were merged and TSB no As and TSB +As were merged. Protein abundance values were log2 transformed, LOESS-normalised among biological replicates and mean-centred across all conditions using the software InfernoRDN [52]. Stochastic sampling of the proteins was filtered by removing the proteins without abundance value in at least two of the biological triplicates in at least one condition (no-As control or +As condition). Remaining missing data were imputed by random numbers drawn from a normal distribution (width = 0.3 and downshift = 2.8 using the Perseus software http://www.perseus-framework.org) [53]. The differentially abundant proteins were identified by Student’s t test method with adjusted q value ≤0.05. Proteins were further filtered using the absolute log2 fold change ≥1.

The isolation of the Paraclostridium sp. strain EML is described in SI.

Results

Arsenic methylation by soil-derived microbiomes

The first experimental set-up yielded samples for the metagenome, metaproteome and one of the metatranscriptomes (labelled metatranscriptome G for ‘growth in the presence of As’) (Supplementary Figs. S1, S2 and S3). The second set-up, assessing the microbiota’s short-term response to As(III), provided sample for the second metatranscriptome (labelled metatranscriptome R for ‘response to arsenic addition’) (Supplementary Figs. S1-A and S2-A). Both EA and TSB cultures exhibited As methylation, reaching an efficiency of As(III) transformation of 27.7% and 19.5%, respectively (Supplementary Figs. S1 and S2).

Microbiota composition

The taxonomic classification of 16S SSU rRNA sequences show that, although eukaryotic DNA was also identified, the main fraction of the communities was bacterial (>89.0 ± 0.8% for EA cultures and >98.5 ± 0.3% for TSB cultures, relative abundance) and was distributed amongst eight operational taxonomic units (OTUs) at the order level (Fig. 1 and Supplementary Tables S1S4). Statistically significant changes (unpaired Student’s t test and no significant difference considered when p value >0.05) in the OTUs relative abundances, +As condition versus no-As control, are described in SI and summarised in Supplementary Tables S5 and S6.

Fig. 1: Operational taxonomic units (OTUs) at order and genus level (with >1% relative abundance at genus level) identified from 16S SSU rRNA sequences from soil-derived cultures.
figure 1

Abbreviations: EA no As: EA culture no-As control, EA +As: EA culture +As condition, TSB no As: TSB culture no−As control, TSB +As: TSB culture +As condition. OTUs at the order level are indicated in bold in the legend. Plotted values are the average relative abundance and together with SD values and Student’s t test results are available in Supplementary Tables S5 and S6.

MAG selection

The contigs from the four metagenomes, EA (+As, no−As control) and TSB (+As, no−As control), were clustered separately into bins. High-quality (≥90% completeness and ≤5% contamination) bins were designated as MAGs [54]. For the +As condition, the parsing process led to a total of 36 MAGs (Table 1). Additionally, matching bins were sought in the bins from the no-As control cultures (Supplementary Tables S7 and S8). Only one of the 36 MAGs in the +As condition was left unpaired (TSB MAG 8).

Table 1 Metagenome-assembled genomes (MAGs) from EA (upper Table A) and TSB (lower Table B) cultures in the +As condition.

For each MAG, a lineage was assigned by CheckM, based on lineage-specific marker genes [40]. The MAGs identified belonged to the phyla: Firmicutes (orders Clostridiales, Selenomonadales and Lactobacillales, and the genus Clostridium), Proteobacteria (Enterobacteriaceae family and Deltaproteobacteria class) and Bacteroidetes (order Bacteroidales). Fifteen MAGs presented non-zero strain heterogeneity (Table 1), an index of the phylogenetic relatedness of binned contigs based on the amino acid identity of the encoded proteins. For ten MAGs, the value is ≥50%, suggesting some phylogenetic relation with the contaminating strains. Five MAGs had heterogeneity values ≤33.33%, suggesting contamination with microorganisms that are not closely related. In the remaining 21 MAGs, the strain heterogeneity is 0%, i.e., no strain heterogeneity or no contamination (Supplementary Tables S7 and S8).

Changes in the relative abundance of MAGs (no-As control vs. +As condition), relatedness of the +As EA and TSB microbial communities, along with the presence, transcription and translation of genes encoding key enzymes from major metabolic pathways of each MAG in the +As condition are included in SI.

Arsenic resistance genes

The metagenomic libraries from the +As condition of the EA and TSB cultures were mined for arsenic resistance (ars) genes and their encoded proteins (the pipeline is described in SI). A total of 309 and 282 genes were annotated as ars genes in the EA and TSB +As metagenomic libraries, respectively (Supplementary Tables S9 and S10). Of those, 255 and 226 were considered correctly annotated as ars genes based on BLAST and HMMER (refer to SI for pipeline), and 225 and 147 had above-threshold DNA abundances, respectively (Fig. 2) (refer to SI for abundance threshold values). Individual abundance values of ars genes, transcripts and proteins in the +As condition and the no-As control and their transcript and protein relative abundance values in the +As condition vs. the no-As control for each MAG group from the EA and TSB cultures are available in Supplementary Tables S11 and S12, respectively.

Fig. 2: Number of ars genes, encoded in the +As condition cultures and with increased expression in metatranscriptomes/metaproteome relative to no−As controls.
figure 2

Number of ars genes encoded in metagenomes and with increased expression in metatranscriptomes, R or G, or metaproteomes and the non-redundant overlap between genes with increased expression in metatranscriptomes and/or metaproteomes from +As condition EA (left panels) and TSB (right panels) cultures. Bar length and numbers on the right side of the bars correspond to the number of genes per ars gene category. Bold numbers on the lower left corner of each panel correspond to the sum of all ars genes per category.

The ars genes encode proteins involved in the detoxification of As oxyanions: arsB and acr3, encoding As(III)-efflux systems; arsA, encoding the ATPase energising the efflux of As(III) and As(III) chaperone; arsD, encoding a weak ars operon repressor [55]; arsC1 and arsC2, encoding As(V) reductases coupling As reduction to the oxidation of glutaredoxin or thioredoxin, respectively; and arsR genes encoding As(III)-regulated repressors (ArsR1, ArsR2, and ArsR3) classified based on the location of the As(III)-binding cysteine residues [56,57,58].

The most common ars genes in EA and TSB culture metagenomes were arsR, arsC, and arsP (Fig. 2). The first two genes are part of the canonical ars operon arsRBC [59], whilst arsP, encoding a recently discovered membrane transporter, has been found to be widely distributed in bacterial genomes [20]. Most of the surveyed arsP genes, 57% in EA and 50% in TSB, are encoded in putative ars operons, represented by ars genes contiguously encoded in the same contig (Supplementary Tables S11 and S12), supporting their As-related function and correct annotation. The next most abundant genes were those responsible for As(III) efflux (arsB, acr3, and arsA), typically found in organisms living in reducing environments in association with arsC [16, 60]. Finally, arsM and the two genes, arsI and arsH, encoding MMAs(III)-resistance mechanisms, were the least recurrent genes in the metagenomes. The results of gene and protein relative expression vs. the no-As control of the ars genes involved in the metabolism of inorganic As in the MAGs are described in SI.

Arsenic-methylating MAGs

The arsM gene can be expressed at similar, or slightly different levels in the absence or presence of As(III) in some organisms [61, 62], but expressed at significantly higher levels in the presence of As(III) in others [63,64,65,66]. Thus, we sought to identify arsM genes transcribed and ArsM proteins showing increased expression in the +As condition relative to the no−As control (Fig. 3) but also those simply exhibiting expression, not necessarily increased relative to the control (Supplementary Fig. S4).

Fig. 3: Distribution of ars genes involved in methylated arsenic metabolism encoded in MAGs from the +As condition and differentially expressed in metatranscriptomes/metaproteome relative to the no-As EA control.
figure 3

Each numbered box represents an ars gene. The number in each box corresponds to the “Numbering” column in Supplementary Tables S11 and S12 where individual gene abundance and fold change values can be found. Com. (%): community (%) as defined in caption from Table 1.

Sixteen phylogenetically distinct arsM genes were identified in the EA +As metagenome, but increased transcriptome reads or peptides (relative to the no-As control) were only detected for three genes (Fig. 3). The first is an arsM in Clostridiales EA MAG 8 classified by GhostKOALA as belonging to Paeniclostridium sordellii (EA MAG 8, arsM-1, psor type strain, in Supplementary Table S11). The second was found in Clostridium EA MAG 9, also detected in the metaproteome, and the GhostKOALA taxonomic classification of the corresponding gene (EA MAG 9, arsM-1 in Supplementary Table S11) revealed that it was attributed to the unclassified species Ruminococcaceae bacterium CPB6 (Fig. 3, Supplementary Table S11) [67]. Finally, the third arsM was obtained from transcriptomic data but not clustered in any EA MAG (EA unbinned, arsM-5 in Supplementary Table S11) and likewise classified as pertaining to Paeniclostridium sordellii.

In the TSB +As metagenome, nine distinct arsM genes were identified but none were detected in the metatranscriptome and only one exhibited increased expression in the metaproteome (Fig. 3). It corresponds to an arsM gene from MAG 9 (TSB MAG 9, arsM-2 in Supplementary Table S12). The expressed ArsM protein was assigned by GhostKOALA to a Clostridiales strain: Clostridium botulinum (cby type strain) (TSB MAG 9, arsM-2) (Fig. 3, Supplementary Table S12). Finally, there was one arsM expressed in the TSB +As metaproteome but with no increased expression relative to the no-As control, it was classified as Ruminococcaceae bacterium CPB6 (TSB MAG 9, arsM-1) (Supplementary Fig. S4), the same organism identified in the EA culture (EA MAG 9, arsM-1).

In addition to evidence for active As methylation, there was evidence for active detoxification of methylated arsenic. Indeed, the metagenome included genes encoding proteins involved in the metabolism of methylated As such as arsH, arsI, arsP, and arsR4 (Figs. 2 and 3). These genes encode proteins involved in the detoxification of methylated arsenic like MMAs(III) and roxarsone: the oxidase ArsH, responsible for the oxidation of trivalent methylated As to the less toxic pentavalent form [68]; the demethylase ArsI that removes methyl groups from the As atom [69]; and the transmembrane transporter ArsP, thought to efflux methylated As [70]. The arsR4 gene encodes an atypical MMAs(III)-responsive ArsR repressor, containing only two conserved cysteine residues [71]. The Enterobacteriaceae TSB MAG 14 exhibited activity of the oxygen-dependent ArsH protein [68] (Fig. 3). An arsR4, shown to induce expression of arsP in the presence of MMAs(III) [71], had increased transcription along with an arsP encoded in the same contig in the Selenomonadales TSB MAG 19 (Fig. 3, Supplementary Table S12). Both gene transcripts were <5 TPM-RNA (Supplementary Table S12) and thus, were not considered as transcribed in Supplementary Fig. S4. Finally, an ArsI protein, taxonomically related to class Clostridia ([Eubacterium] rectale), was expressed but encoded in an unbinned gene from the EA culture (Fig. 3, Supplementary Table S12).

Isolation of an arsenic-methylating anaerobic microorganism

Based on the analysis of the active metabolic activity from the EA MAG 8, expressing an ArsM (Supplementary Fig. S5), an appropriate selective medium was identified for its isolation. We utilised the fact that this MAG harbours and expresses the anaerobic assimilatory sulfite reductase encoded by the asrABC operon which is responsible for the NADH-dependent reduction of sulfite to sulfide [72,73,74] in sulfite-reducing Clostridia (SRC). From the nine Clostridia MAGs, only two expressed this capability in the EA microbiome (Supplementary Fig. S5). Thus, the isolation relied on growing the EA culture on agar medium selective for the SRC phenotype. In TSC agar, designed for the enumeration of Clostridium perfringens in food [75], the colonies from SRC are black, as the ammonium ferric citrate forms iron sulfide during sulfite reduction. Additionally, D-cycloserine acts a selective agent for the isolation of Clostridia strains [76] while inhibiting facultative anaerobes [75]. Finally, the bromocresol purple contained in the agar allows the identification of sucrose fermenters, resulting in a change of colour from purple to yellow. As none of the genes involved in sucrose transport or hydrolysis were binned in EA Clostridiales MAG 8 (Supplementary Fig. S5), only non-sucrose fermenting black colonies were considered. Those colonies were selected and using a colony PCR screen specifically targeting the arsM gene of EA MAG 8, we isolated a Clostridiales strain encoding the gene of the expressed ArsM in the EA MAG 8 (protein id k119_30669_28, Supplementary Table S11) (Supplementary Fig. S6).

The isolate consists of non-sucrose-fermenting, rod-shaped and spore-forming bacteria forming convex and circular black colonies on TSC agar (Supplementary Figs. S7 and S8). The BLAST (NCBI) search of the 16S rRNA sequence gives >99% identity to Paraclostridium strains (Supplementary Table S13). On the basis of the 16S rRNA sequence, we assign the following name to the bacterium: “Paraclostridium species str. EML”. Strain EML was tested for As methylation under anaerobic conditions with 25 μM As(III). The growth of strain EML was hindered by As(III) (Fig. 4A) and starting from ~4 h, the isolate transformed As(III) to monomethylated soluble As representing 48.3 ± 1.5% of the soluble arsenic in the culture after 83 h (panels B and C from Fig. 4). A fraction (14.7 ± 0.6 μM) of the arsenic was found associated with biomass almost exclusively as inorganic As (Fig. 4D).

Fig. 4: Isolate Paraclostridium sp. EML.
figure 4

A Growth as OD600 with 25 μM As(III) and without, B proportion of soluble arsenic species in filtered medium containing 25 μM As(III), C concentration of arsenic species soluble in filtered medium containing 25 μM As(III) (solid lines) and biomass-bound (dashed lines) and D proportion of biomass-bound arsenic species. Data points and bars represent the mean value and error bars, plus and minus one standard deviation. Individual values for each measurement and biological replicate are available in Supplementary Tables S23 and S24.

Discussion

Our results demonstrate the successful translation of multi-omic information to a specific strategy for targeted microbial isolation. The metagenomes from the anaerobic soil-derived cultures identified the potential for As methylation in microorganisms from diverse taxa. While there were a large number of ars genes in the metagenomes, only a small proportion was transcribed or translated in the presence of As when compared to the no-As control (Fig. 2). This contrast was particularly evident for the gene responsible for As methylation, arsM. The post-genomic approaches of community gene and protein expression in TSB as in EA soil-derived microbiomes clearly pointed to the active As-methylating role of various fermenting bacteria from the order Clostridiales. This information paved the way for the identification of As-methylating microorganisms and the successful isolation of an anaerobic As methylator.

The TSB and EA media were chosen to selectively enrich for putative As methylators from the microbial soil community based on the study from Bright et al., in which lake sediments enriched in TSB medium, either sulfate-amended or unamended, were shown to have greater As methylation rates than in iron- or manganese-reducing TSB cultures [26]. The selected media caused a great shift in the original soil microbial diversity [25] along with the loss of putative As-methylating microorganisms. Nonetheless, the As-methylating TSB and EA soil-derived cultures offered the opportunity to study active As methylation from paddy-soil microbiota in an environment that is less complex than soil but that remains environmentally relevant. In contrast to soil slurries, the absence of soil minerals in the soil-derived cultures facilitated the detection of soluble methylarsenicals and the extraction of DNA, RNA and proteins. The multi-omic approach made it possible to identify putative microorganisms driving As methylation and their metabolism. Targeting a specific arsM gene rather than the synthesis of methylarsenicals greatly accelerated colony screening, as colony PCR could be employed instead of analytical detection by HPLC-ICP-MS.

Had only the metagenomic approach been implemented, the data would have pointed to SRB MAGs as putative As methylators, as they harboured the most abundant arsM genes (Fig. 5). Indeed, SRB have been proposed as drivers of As methylation in rice paddy soils based on the correlation in the abundance of arsM and dissimilatory sulfite reductase (dsr) genes [77] and RNA transcripts [78], and a decrease in As methylation by the addition chemical inhibitors of dissimilatory sulfate reduction (DSR) [77, 78]. Additionally, the use of degenerate primers for arsM amplification may underestimate arsM phylogenetic diversity, a drawback overcome by metagenomic and metatranscriptomic sequencing. In the present findings, the SRB Deltaproteobacteria MAGs, although actively reducing sulfate (Supplementary Figs. S5 and S9), did not exhibit As-methylating activity as their arsM genes were neither transcribed nor translated (Fig. 5). Desulfovibrio MAGs were metabolically active in both cultures, but amongst all their encoded ars genes, only an arsR3 exhibited increased expression in the presence of As(III), providing strong evidence for their lack of involvement in As methylation in the TSB and EA cultures.

Fig. 5: Gene abundance of arsM genes in MAGs from the +As condition cultures.
figure 5

Coloured bars correspond to arsM genes with increased expression in the metaproteome (blue-green) or in the metatranscriptome G (purple) from +As condition relative to the no-As control in EA (left panel) and TSB (right panel) cultures. The taxonomic classification shown on the right side of the error bars for selected arsM genes corresponds to the individual gene classification assigned by GhostKOALA - “Genus” column in Supplementary Tables S11 and S12. Columns with matching symbols on the right side of the error bars, correspond to matching arsM genes between the cultures. Individual gene abundance can be found in Supplementary Tables S11 and S12. Numbers inside bar and bar length represent mean and error bars one standard deviation.

Previous work had identified another As-methylating Clostridiales strain, Clostridium sp. BXM [24], that performed fermentation and DSR but that is no longer available. The sole attribution of As-methylating activity to fermenting Firmicutes in that work, along with the isolation of the present sulfite-reducing fermenter, point to a key role for fermenting Clostridiales microorganisms harbouring sulfur-related metabolism in As methylation. Other studies have reported an increase in As methylation efficiency after the amendment of sulfate [79] or organic matter to soil [15, 80,81,82], or after the increase in dissolved organic carbon in soil [83]. The positive impact of sulfate amendment on As methylation was interpreted as pointing to the role of SRB in As methylation [79]. Here, we offer an alternative explanation, supported by examples of organic amendments enhancing As methylation [15, 80,81,82]. The sulfate amendment could have indirectly increased the availability of short-chain fatty acids through DSR, providing fermentable substrates. Thus, we propose that direct or indirect organic amendments result in the enrichment of fermenting communities, and consequently, in an increase in As methylation.

It was previously proposed that the As-methylating activity of anaerobic microorganisms may be limited by efficient efflux of intracellular As(III) [22], or that it may function as a defensive response against nutrient competition [18]. Indeed, the identification of MAGs exhibiting a detoxification response to methylarsenicals supports the hypothesis of the role of monomethylated As as an arsenic-bearing antibiotic. Although the expression of ArsI and ArsH, catalysing oxygen-dependent MMAs(III)-resistance mechanisms (Fig. 3), is difficult to reconcile with anoxic conditions, it is conceivable that these proteins are capable of additional functions in the absence of O2. Up until now, the lack of available anaerobic microbial isolates able to methylate As in vitro [22] precluded the investigation of the hypotheses raised above. This work represents the first study applying a combination of three meta-omic techniques in order to characterise As metabolism in microbial communities and to perform meta-omics-aided isolation of a microorganism [84, 85]. The successful isolation of Paraclostridium sp. EML is part of the “new era of omics information-guided microbial cultivation technology” described by Gutleben et al. [84] and represents a milestone to obtain novel targeted microbial isolates from the environment and to elucidate the controls on anaerobic As methylation.

Further work is needed to elucidate why ArsM expression was restricted to members of Clostridiales fermenters and did not occur in other organisms harbouring arsM genes. The availability of As-methylating anaerobes will allow investigation of why the arsM gene evolved under an anoxic atmosphere [86], of the controls on the production of toxic methylated As species in flooded rice paddies, and the development of microbially-mediated remediation technologies for As-contaminated soils via the synthesis of volatile methylarsenicals [87, 88].

Data availability

Metagenomic and metatranscriptomic raw sequencing reads are available at the National Centre for Biotechnology Information (NCBI) Sequence Read Archive (SRA), BioProject PRJNA714492. Data from the meta-omic analyses and source data from figures are available in Zenodo data repository (10.5281/zenodo.4605527).

References

  1. Whitman WB, Coleman DC, Wiebe WJ. Prokaryotes: the unseen majority. Proc Natl Acad Sci. 1998;95:6578–83.

    CAS  PubMed  PubMed Central  Article  Google Scholar 

  2. Bardgett RD, Van Der Putten WH. Belowground biodiversity and ecosystem functioning. Nature. 2014;515:505–11.

    CAS  PubMed  Article  Google Scholar 

  3. Roesch LFW, Fulthorpe RR, Riva A, Casella G, Hadwin AKM, Kent AD, et al. Pyrosequencing enumerates and contrasts soil microbial diversity. ISME J. 2007;1:283–90.

    CAS  PubMed  Article  Google Scholar 

  4. Daniel R. The soil metagenome - a rich resource for the discovery of novel natural products. Curr Opin Biotechnol. 2004;15:199–204.

    CAS  PubMed  Article  Google Scholar 

  5. Bodor A, Bounedjoum N, Vincze GE, Erdeiné Kis Á, Laczi K, Bende G, et al. Challenges of unculturable bacteria: environmental perspectives. Rev Environ Sci Biotechnol. 2020;19:1–22.

    CAS  Article  Google Scholar 

  6. Houle D, Govindaraju DR, Omholt S. Phenomics: the next challenge. Nat Rev Genet. 2010;11:855–66.

    CAS  PubMed  Article  Google Scholar 

  7. Ajees AA, Marapakala K, Packianathan C, Sankaran B, Rosen BP. Structure of an As(III) S-adenosylmethionine methyltransferase: Insights into the mechanism of arsenic biotransformation. Biochemistry. 2012;51:5476–85.

    CAS  PubMed  Article  Google Scholar 

  8. Mori K, Kamagata Y. The challenges of studying the anaerobic microbial world. Microbes Environ. 2014;29:335–7.

    PubMed  PubMed Central  Article  Google Scholar 

  9. Dong H, Xu W, Pillai JK, Packianathan C, Rosen BP. High-throughput screening-compatible assays of As(III) S-adenosylmethionine methyltransferase activity. Anal Biochem. 2015;480:67–73.

    CAS  PubMed  PubMed Central  Article  Google Scholar 

  10. Lomax C, Liu WJ, Wu L, Xue K, Xiong J, Zhou J, et al. Methylated arsenic species in plants originate from soil microorganisms. N Phytol. 2012;193:665–72.

    CAS  Article  Google Scholar 

  11. Zhao FJ, Zhu YG, Meharg AA. Methylated arsenic species in rice: geographical variation, origin, and uptake mechanisms. Environ Sci Technol. 2013;47:3957–66.

    CAS  PubMed  Article  Google Scholar 

  12. Zheng MZ, Li G, Sun GX, Shim H, Cai C. Differential toxicity and accumulation of inorganic and methylated arsenic in rice. Plant Soil. 2013;365:227–38.

    CAS  Article  Google Scholar 

  13. Abedin MJ, Feldmann J, Meharg AA. Uptake kinetics of arsenic species in rice plants. Plant Physiol. 2002;128:1120–8.

    CAS  PubMed  PubMed Central  Article  Google Scholar 

  14. Dunivin TK, Yeh SY, Shade A. A global survey of arsenic-related genes in soil microbiomes. BMC Biol. 2019;17:45.

    CAS  PubMed  PubMed Central  Article  Google Scholar 

  15. Jia Y, Huang H, Zhong M, Wang F-H, Zhang L-M, Zhu Y-G. Microbial arsenic methylation in soil and rice rhizosphere. Environ Sci Technol. 2013;47:3141–8.

    CAS  PubMed  Article  Google Scholar 

  16. Xiao KQ, Li LG, Ma LP, Zhang SY, Bao P, Zhang T, et al. Metagenomic analysis revealed highly diverse microbial arsenic metabolism genes in paddy soils with low-arsenic contents. Environ Pollut. 2016;211:1–8.

    CAS  PubMed  Article  Google Scholar 

  17. Zhang SY, Su JQ, Sun GX, Yang Y, Zhao Y, Ding J, et al. Land scale biogeography of arsenic biotransformation genes in estuarine wetland. Environ Microbiol. 2017;19:2468–82.

    CAS  PubMed  Article  Google Scholar 

  18. Chen J, Rosen BP. The arsenic methylation cycle: How microbial communities adapted methylarsenicals for use as weapons in the continuing war for dominance. Front Environ Sci. 2020;8:43.

    CAS  Article  Google Scholar 

  19. Ping LiY, Ben Fekih I, Chi Fru E, Moraleda-Munoz A, Li X, Rosen BP, et al. Antimicrobial activity of metals and metalloids. Annu Rev Microbiol. 2021;75:175–97.

    Article  CAS  Google Scholar 

  20. Yang Y, Wu S, Lilley RM, Zhang R. The diversity of membrane transporters encoded in bacterial arsenic-resistance operons. PeerJ. 2015;3:e943.

    PubMed  PubMed Central  Article  Google Scholar 

  21. Yang P, Ke C, Zhao C, Kuang Q, Liu B, Xue X, et al. ArsM-mediated arsenite volatilization is limited by efflux catalyzed by As efflux transporters. Chemosphere. 2020;239:124822.

    CAS  PubMed  Article  Google Scholar 

  22. Viacava K, Meibom KL, Ortega D, Dyer S, Gelb A, Falquet L, et al. Variability in arsenic methylation efficiency across aerobic and anaerobic microorganisms. Environ Sci Technol. 2020;54:14343–51.

    CAS  PubMed  Article  Google Scholar 

  23. Thomas F, Diaz-Bone RA, Wuerfel O, Huber B, Weidenbach K, Schmitz RA, et al. Connection between multimetal(loid) methylation In methanoarchaea and central intermediates of methanogenesis. Appl Environ Microbiol. 2011;77:8669–75.

    CAS  PubMed  PubMed Central  Article  Google Scholar 

  24. Wang PP, Bao P, Sun GX. Identification and catalytic residues of the arsenite methyltransferase from a sulfate-reducing bacterium, Clostridium sp. BXM. FEMS Microbiol Lett. 2015;362:1–8.

    CAS  PubMed  Article  Google Scholar 

  25. Reid MC, Maillard J, Bagnoud A, Falquet L, Le Vo P, Bernier-Latmani R. Arsenic methylation dynamics in a rice paddy soil anaerobic enrichment culture. Environ Sci Technol. 2017;51:10546–54.

    CAS  PubMed  Article  Google Scholar 

  26. Bright DA, Brock S, Reimer KJ, Cullen WR, Hewitt GM, Jafaar J. Methylation of arsenic by anaerobic microbial consortia isolated from lake sediment. Appl Organomet Chem. 1994;8:415–22.

    CAS  Article  Google Scholar 

  27. Andrews S. FastQC: a quality control tool for high throughput sequence data. http://www.bioinformatics.babraham.ac.uk/projects/fastqc/.

  28. Xu H, Luo X, Qian J, Pang X, Song J, Qian G, et al. FastUniq: a fast de novo duplicates removal tool for paired short reads. PLoS One. 2012;7:e52249.

    CAS  PubMed  PubMed Central  Article  Google Scholar 

  29. Li D, Liu CM, Luo R, Sadakane K, Lam TW. MEGAHIT: an ultra-fast single-node solution for large and complex metagenomics assembly via succinct de Bruijn graph. Bioinformatics. 2015;31:1674–6.

    CAS  Article  PubMed  Google Scholar 

  30. Bray NL, Pimentel H, Melsted P, Pachter L. Near-optimal probabilistic RNA-seq quantification. Nat Biotechnol. 2016;34:525–7.

    CAS  Article  PubMed  Google Scholar 

  31. Schaeffer L, Pimentel H, Bray N, Melsted P, Pachter L. Pseudoalignment for metagenomic read assignment. Bioinformatics. 2017;33:2082–8.

    CAS  PubMed  PubMed Central  Article  Google Scholar 

  32. Wagner GP, Kin K, Lynch VJ. Measurement of mRNA abundance using RNA-seq data: RPKM measure is inconsistent among samples. Theory Biosci. 2012;131:281–5.

    CAS  PubMed  Article  Google Scholar 

  33. Hyatt D, Chen GL, LoCascio PF, Land ML, Larimer FW, Hauser LJ. Prodigal: prokaryotic gene recognition and translation initiation site identification. BMC Bioinforma. 2010;11:119.

    Article  CAS  Google Scholar 

  34. Kanehisa M, Sato Y, Morishima K. BlastKOALA and GhostKOALA: KEGG tools for functional characterization of genome and metagenome sequences. J Mol Biol. 2016;428:726–31.

    CAS  PubMed  Google Scholar 

  35. Bengtsson-Palme J, Hartmann M, Eriksson KM, Pal C, Thorell K, Larsson DGJ, et al. Metaxa2: improved identification and taxonomic classification of small and large subunit rRNA in metagenomic data. Mol Ecol Resour. 2015;15:1403–14.

    CAS  PubMed  Article  Google Scholar 

  36. Alneberg J, Bjarnason BS, De Bruijn I, Schirmer M, Quick J, Ijaz UZ, et al. Binning metagenomic contigs by coverage and composition. Nat Methods. 2014;11:1144–6.

    CAS  PubMed  Article  Google Scholar 

  37. Kang DD, Li F, Kirton E, Thomas A, Egan R, An H, et al. MetaBAT 2: an adaptive binning algorithm for robust and efficient genome reconstruction from metagenome assemblies. PeerJ. 2019;7:e7359.

  38. Wu YW, Simmons BA, Singer SW. MaxBin 2.0: an automated binning algorithm to recover genomes from multiple metagenomic datasets. Bioinformatics. 2016;32:605–7.

    CAS  PubMed  Article  Google Scholar 

  39. Uritskiy GV, Diruggiero J, Taylor J. MetaWRAP - a flexible pipeline for genome-resolved metagenomic data analysis. Microbiome. 2018;6:158.

    PubMed  PubMed Central  Article  Google Scholar 

  40. Parks DH, Imelfort M, Skennerton CT, Hugenholtz P, Tyson GW. CheckM: assessing the quality of microbial genomes recovered from isolates, single cells, and metagenomes. Genome Res. 2015;25:1043–55.

    CAS  PubMed  PubMed Central  Article  Google Scholar 

  41. Olm MR, Brown CT, Brooks B, Banfield JF. DRep: a tool for fast and accurate genomic comparisons that enables improved genome recovery from metagenomes through de-replication. ISME J. 2017;11:2864–8.

    CAS  PubMed  PubMed Central  Article  Google Scholar 

  42. Bolger AM, Lohse M, Usadel B. Trimmomatic: a flexible trimmer for Illumina sequence data. Bioinformatics. 2014;30:2114–20.

    CAS  PubMed  PubMed Central  Article  Google Scholar 

  43. Kopylova E, Noé L, Touzet H. SortMeRNA: fast and accurate filtering of ribosomal RNAs in metatranscriptomic data. Bioinformatics. 2012;28:3211–7.

    CAS  PubMed  Article  Google Scholar 

  44. Langmead B, Salzberg SL. Fast gapped-read alignment with Bowtie 2. Nat Methods. 2012;9:357–9.

    CAS  PubMed  PubMed Central  Article  Google Scholar 

  45. Liao Y, Smyth GK, Shi W. featureCounts: an efficient general purpose program for assigning sequence reads to genomic features. Bioinformatics. 2014;30:923–30.

    CAS  PubMed  Article  Google Scholar 

  46. Love MI, Huber W, Anders S. Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2. Genome Biol. 2014;15:550.

    PubMed  PubMed Central  Article  CAS  Google Scholar 

  47. Washburn MP, Wolters D, Yates JR. Large-scale analysis of the yeast proteome by multidimensional protein identification technology. Nat Biotechnol. 2001;19:242–7.

    CAS  PubMed  Article  Google Scholar 

  48. McDonald WH, Ohi R, Miyamoto DT, Mitchison TJ, Yates JR. Comparison of three directly coupled HPLC MS/MS strategies for identification of proteins from complex mixtures: single-dimension LC-MS/MS, 2-phase MudPIT, and 3-phase MudPIT. Int J Mass Spectrom. 2002;219:245–51.

    CAS  Article  Google Scholar 

  49. Clarkson SM, Giannone RJ, Kridelbaugh DM, Elkins JG, Guss AM, Michenera JK. Construction and optimization of a heterologous pathway for protocatechuate catabolism in Escherichia coli enables bioconversion of model aromatic compounds. Appl Environ Microbiol. 2017; 83:e01313-17.

  50. Dorfer V, Pichler P, Stranzl T, Stadlmann J, Taus T, Winkler S, et al. MS Amanda, a universal identification algorithm optimized for high accuracy tandem mass spectra. J Proteome Res. 2014;13:3679–84.

    CAS  PubMed  PubMed Central  Article  Google Scholar 

  51. Käll L, Canterbury JD, Weston J, Noble WS, MacCoss MJ. Semi-supervised learning for peptide identification from shotgun proteomics datasets. Nat Methods. 2007;4:923–5.

    PubMed  Article  CAS  Google Scholar 

  52. Polpitiya AD, Qian WJ, Jaitly N, Petyuk VA, Adkins JN, Camp DG, et al. DAnTE: a statistical tool for quantitative analysis of -omics data. Bioinformatics. 2008;24:1556–8.

    CAS  PubMed  Article  Google Scholar 

  53. Tyanova S, Temu T, Sinitcyn P, Carlson A, Hein MY, Geiger T, et al. The Perseus computational platform for comprehensive analysis of (prote)omics data. Nat Methods. 2016;13:731–40.

    CAS  PubMed  Article  Google Scholar 

  54. Bowers RM, Kyrpides NC, Stepanauskas R, Harmon-Smith M, Doud D, Reddy TBK, et al. Minimum information about a single amplified genome (MISAG) and a metagenome-assembled genome (MIMAG) of bacteria and archaea. Nat Biotechnol. 2017;35:725–31.

    CAS  PubMed  PubMed Central  Article  Google Scholar 

  55. Lin Y-F, Walmsley AR, Rosen BP. An arsenic metallochaperone for an arsenic detoxification pump. PNAS. 2006;103:15617–22.

    CAS  PubMed  PubMed Central  Article  Google Scholar 

  56. Shi W, Wu J, Rosen BP. Identification of a putative metal binding site in a new family of metalloregulatory proteins. J Biol Chem. 1994;269:19826–9.

    CAS  PubMed  Article  Google Scholar 

  57. Murphy JN, Saltikov CW. The ArsR repressor mediates arsenite-dependent regulation of arsenate respiration and detoxification operons of Shewanella sp. strain ANA-3. J Bacteriol. 2009;191:6722–31.

    CAS  PubMed  PubMed Central  Article  Google Scholar 

  58. Santha S, Pandaranayaka EPJ, Rosen BP, Thiyagarajan S. Purification, crystallization and preliminary X-ray diffraction studies of the arsenic repressor ArsR from Corynebacterium glutamicum. Acta Crystallogr Sect F Struct Biol Cryst Commun. 2011;67:1616–8.

    CAS  PubMed  PubMed Central  Article  Google Scholar 

  59. Fekih IB, Zhang C, Li YP, Zhao Y, Alwathnani HA, Saquib Q, et al. Distribution of arsenic resistance genes in prokaryotes. Front Microbiol. 2018;9:2473.

    PubMed  PubMed Central  Article  Google Scholar 

  60. Cai L, Yu K, Yang Y, Chen BW, Li XD, Zhang T. Metagenomic exploration reveals high levels of microbial arsenic metabolism genes in activated sludge and coastal sediments. Appl Microbiol Biotechnol. 2013;97:9579–88.

    CAS  PubMed  Article  Google Scholar 

  61. Huang K, Xu Y, Packianathan C, Gao F, Chen C, Zhang J, et al. Arsenic methylation by a novel ArsM As(III) S-adenosylmethionine methyltransferase that requires only two conserved cysteine residues. Mol Microbiol. 2018;107:265–76.

    CAS  PubMed  Article  Google Scholar 

  62. Zhang J, Cao T, Tang Z, Shen Q, Rosen BP, Zhao FJ. Arsenic methylation and volatilization by arsenite S-adenosylmethionine methyltransferase in Pseudomonas alcaligenes NBRC14159. Appl Environ Microbiol. 2015;81:2852–60.

    CAS  PubMed  PubMed Central  Article  Google Scholar 

  63. Huang K, Chen C, Zhang J, Tang Z, Shen Q, Rosen BP, et al. Efficient arsenic methylation and volatilization mediated by a novel bacterium from an arsenic-contaminated paddy soil. Environ Sci Technol. 2016;50:6389–96.

    CAS  PubMed  PubMed Central  Article  Google Scholar 

  64. Yin XX, Chen J, Qin J, Sun GX, Rosen BP, Zhu YG. Biotransformation and volatilization of arsenic by three photosynthetic cyanobacteria. Plant Physiol. 2011;156:1631–8.

    CAS  PubMed  PubMed Central  Article  Google Scholar 

  65. Zhao C, Zhang Y, Chan Z, Chen S, Yang S. Insights into arsenic multi-operons expression and resistance mechanisms in Rhodopseudomonas palustris CGA009. Front Microbiol. 2015;6:986.

    PubMed  PubMed Central  Google Scholar 

  66. Wang G, Kennedy SP, Fasiludeen S, Rensing C, DasSarma S. Arsenic resistance in Halobacterium sp. strain NRC-1 examined by using an improved gene knockout system. J Bacteriol. 2004;186:3187–94.

    CAS  PubMed  PubMed Central  Article  Google Scholar 

  67. Tao Y, Zhu X, Wang H, Wang Y, Li X, Jin H, et al. Complete genome sequence of Ruminococcaceae bacterium CPB6: a newly isolated culture for efficient n-caproic acid production from lactate. J Biotechnol. 2017;259:91–94.

    CAS  PubMed  Article  Google Scholar 

  68. Chen J, Bhattacharjee H, Rosen BP. ArsH is an organoarsenical oxidase that confers resistance to trivalent forms of the herbicide monosodium methylarsenate and the poultry growth promoter roxarsone. Mol Microbiol. 2015;96:1042–52.

    CAS  PubMed  PubMed Central  Article  Google Scholar 

  69. Yoshinaga M, Rosen BP. A C-As lyase for degradation of environmental organoarsenical herbicides and animal husbandry growth promoters. PNAS. 2014;111:7701–6.

    CAS  PubMed  PubMed Central  Article  Google Scholar 

  70. Chen J, Madegowda M, Bhattacharjee H, Rosen BP. ArsP: a methylarsenite efflux permease. Mol Microbiol. 2015;98:625–35.

    CAS  PubMed  PubMed Central  Article  Google Scholar 

  71. Chen J, Nadar VS, Rosen BP. A novel MAs(III)-selective ArsR transcriptional repressor. Mol Microbiol. 2017;106:469–78.

    CAS  PubMed  PubMed Central  Article  Google Scholar 

  72. Doyle CJ, O’Toole PW, Cotter PD. Genomic characterization of sulphite reducing bacteria isolated from the dairy production chain. Front Microbiol. 2018;9:1507.

    PubMed  PubMed Central  Article  Google Scholar 

  73. Czyzewski BK, Wang DN. Identification and characterization of a bacterial hydrosulphide ion channel. Nature. 2012;483:494–7.

    CAS  PubMed  PubMed Central  Article  Google Scholar 

  74. Huang CJ, Barrett EL. Sequence analysis and expression of the Salmonella typhimurium asr operon encoding production of hydrogen sulfide from sulfite. J Bacteriol. 1991;173:1544–53.

    CAS  PubMed  PubMed Central  Article  Google Scholar 

  75. Harmon SM, Kautter DA, Peeler JT. Improved medium for enumeration of Clostridium perfringens. Appl Microbiol. 1971;22:688–92.

    CAS  PubMed  PubMed Central  Article  Google Scholar 

  76. George WL, Sutter VL, Finegold SM. Toxigenicity and antimicrobial susceptibility of Clostridium difficile, a cause of antimicrobial agent-associated colitis. Curr Microbiol. 1978;1:55–58.

    CAS  Article  Google Scholar 

  77. Wang M, Tang Z, Chen XP, Wang X, Zhou WX, Tang Z, et al. Water management impacts the soil microbial communities and total arsenic and methylated arsenicals in rice grains. Environ Pollut. 2019;247:736–44.

    CAS  PubMed  Article  Google Scholar 

  78. Chen C, Li L, Huang K, Zhang J, Xie WY, Lu Y, et al. Sulfate-reducing bacteria and methanogens are involved in arsenic methylation and demethylation in paddy soils. ISME J. 2019;13:2523–35.

    CAS  PubMed  PubMed Central  Article  Google Scholar 

  79. Zeng X, Jiang Y, Fan X, Chao S, Yang Y, Liu J, et al. Effects of sulfate application on inhibiting accumulation and alleviating toxicity of arsenic in Panax notoginseng grown in arsenic-polluted soil. Water Air Soil Pollut. 2016; 227:148.

  80. Mestrot A, Feldmann J, Krupp EM, Hossain MS, Roman-Ross G, Meharg AA. Field fluxes and speciation of arsines emanating from soils. Environ Sci Technol. 2011;45:1798–804.

    CAS  PubMed  Article  Google Scholar 

  81. Huang H, Jia Y, Sun G-X, Zhu Y-G. Arsenic speciation and volatilization from flooded paddy soils amended with different organic matters. Environ Sci Technol. 2012;46:2163–8.

    CAS  PubMed  Article  Google Scholar 

  82. Seyfferth AL, Morris AH, Gill R, Kearns KA, Mann JN, Paukett M, et al. Soil incorporation of silica-rich rice husk decreases inorganic arsenic in rice grain. J Agric Food Chem. 2016;64:3760–6.

    CAS  PubMed  Article  Google Scholar 

  83. Zhao FJ, Harris E, Yan J, Ma J, Wu L, Liu W, et al. Arsenic methylation in soils and its relationship with microbial arsM abundance and diversity, and As speciation in rice. Environ Sci Technol. 2013;47:7147–54.

  84. Gutleben J, Chaib De Mares M, van Elsas JD, Smidt H, Overmann J, Sipkema D. The multi-omics promise in context: From sequence to microbial isolate. Crit Rev Microbiol. 2018;44:212–29.

    CAS  PubMed  Article  Google Scholar 

  85. Chandran H, Meena M, Sharma K. Microbial biodiversity and bioremediation assessment through omics approaches. Front Environ Chem. 2020;1:9.

    Article  Google Scholar 

  86. Chen S-C, Sun G-X, Yan Y, Konstantinidis KT, Zhang S-Y, Deng Y, et al. The Great Oxidation Event expanded the genetic repertoire of arsenic metabolism and cycling. PNAS. 2020;117:10414–21.

    CAS  PubMed  PubMed Central  Article  Google Scholar 

  87. Liu S, Zhang F, Chen J, Sun G. Arsenic removal from contaminated soil via biovolatilization by genetically engineered bacteria under laboratory conditions. J Environ Sci. 2011;23:1544–50.

    CAS  Article  Google Scholar 

  88. Chen J, Sun GX, Wang XX, Lorenzo VD, Rosen BP, Zhu YG. Volatilization of arsenic from polluted soil by Pseudomonas putida engineered for expression of the arsM arsenic(III) S-adenosine methyltransferase gene. Environ Sci Technol. 2014;48:10337–44.

    CAS  PubMed  PubMed Central  Article  Google Scholar 

Download references

Acknowledgements

The work was funded by the Swiss National Science Foundation (SNSF) grant 310030_176146-1 and the SNSF NCCR Microbiomes (grant # 51NF40_180575). The work at ORNL was conducted under the Plant-Microbe Interface Science Focus Area, as supported by the U.S. Department of Energy, Office of Science, Office of Biological and Environmental Research, Genomic Science Program.

Funding

Open access funding provided by EPFL Lausanne.

Author information

Authors and Affiliations

Authors

Contributions

RB-L conceived and supervised the study. KV conducted the experimental work. MR produced the soil microbiomes. AJ and NJ conducted all bioinformatic analyses with guidance from KV and RB-L. SP performed metaproteomic analyses, assisted by HS and supervised by RH. KV interpreted the meta-omics results and designed the experimental pipeline for the isolation of Paraclostridium sp. strain EML. JQ isolated and characterised strain EML. KV wrote the manuscript with revisions and edits by KLM and RB-L. Funding was acquired by RH and RB-L.

Corresponding author

Correspondence to Rizlan Bernier-Latmani.

Ethics declarations

Competing interests

The authors declare no competing interests.

Additional information

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary information

Supplementary Information (SI)

41396_2022_1220_MOESM2_ESM.xlsx

Supplementary Information (SI) Tables for the research article: Meta-omics-aided isolation of an elusive anaerobic arsenic-methylating soil bacterium

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/.

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Viacava, K., Qiao, J., Janowczyk, A. et al. Meta-omics-aided isolation of an elusive anaerobic arsenic-methylating soil bacterium. ISME J (2022). https://doi.org/10.1038/s41396-022-01220-z

Download citation

  • Received:

  • Revised:

  • Accepted:

  • Published:

  • DOI: https://doi.org/10.1038/s41396-022-01220-z

Search

Quick links