Transformation of monothioarsenate by haloalkaliphilic, anoxygenic photosynthetic purple sulfur bacteria

Abstract

Purple sulfur bacteria from Mono Lake, CA were found to be able to grow on and transform the arsenic-sulfur compound monothioarsenate.

Open in new tabDownload slide

Introduction

Understanding arsenic biogeochemistry is important due to its toxicity and widespread distribution in groundwater as a result of geothermal processes and anthropogenic pollution. Biologically mediated transformations of the arsenic oxyanions arsenate (As(V)) and arsenite (As(III)) and of organoarsenicals have been well documented (Sharma & Sohn, 2009). In addition, recent reports have revealed the presence of thioarsenates in the environment, especially in alkaline and sulfidic waters (Stauder et al., 2005; Planer-Friedrich et al., 2007, 2009; Härtig & Planer-Friedrich, 2012) including the hypersaline and alkaline Mono Lake, CA, USA (Hollibaugh et al., 2005). In Mono Lake, periods of meromixis lead to high sulfide concentrations below the oxycline and to the accumulation of thioarsenates, which can contribute as much as half of the arsenic in the water column (Hollibaugh et al., 2005; Wallschlager & Stadey, 2007). Microbial transformations of arsenic compounds in Mono Lake have been examined extensively (Oremland et al., 2004, 2009), generally focusing on arsenite oxidation and arsenate reduction as energy-yielding reactions for bacteria (Oremland et al., 2002; Hoeft et al., 2004; Hollibaugh et al., 2006). Recent studies have indicated a potential biological role in the transformation of thioarsenates (Fisher et al., 2008; Planer-Friedrich et al., 2009; Härtig & Planer-Friedrich, 2012). Experiments in which arsenite and sulfide were added to Mono Lake water under oxidizing conditions revealed the formation of thioarsenates, followed by the production of arsenate. Abiotic controls showed no appreciable accumulation of arsenate (Fisher et al., 2008). Other work under oxidizing conditions in drainage channels from sulfidic, alkaline hot springs showed that trithioarsenate, the dominant arsenic species at the source, was biologically transformed to arsenate along the channel with intermediate accumulation of arsenite and monothioarsenate (Planer-Friedrich et al., 2009). Incubations of filter-sterilized geothermal water with and without the addition of filamentous microbial streamers (dominated by Thermocrinis ruber) showed that abiotic trithioarsenate transformation to arsenate was coupled to biotic sulfide oxidation, with an intermediate accumulation of arsenite. In sulfide-free medium, direct microbial transformation of monothioarsenate to arsenite or arsenate was shown (Härtig & Planer-Friedrich, 2012). A recent experiment using a pure culture confirmed that T. ruber is capable of aerobic growth using monothioarsenate as an electron donor, producing arsenate and sulfate (Härtig et al., 2014). The mechanism for the conversion of thioarsenates to arsenate and/or arsenite has been hypothesized to involve either direct biological conversion of thioarsenate to arsenate and sulfate (Eqn. ) by sulfur-oxidizing bacteria; abiotic conversion to arsenite and elemental sulfur (Eqn. ) with subsequent biological oxidation to arsenate (Eqn. ) and sulfate (Eqn. ); or abiotic decomposition (desulfidation) (Eqn. ) with subsequent biological oxidation of free sulfide (Eqn. ) driving the chemical equilibrium toward the products arsenate and finally sulfate (Fisher et al., 2008; Härtig & Planer-Friedrich, 2012).

Previous work with enrichment cultures from Mono Lake dominated by the chemolithotrophic sulfur-oxidizing bacterium Thioalkalivibrio sp. demonstrated the conversion of thioarsenates to arsenate (Fisher et al., 2008). This haloalkaliphilic member of the Gammaproteobacteria family Ectothiorhodospiraceae has been isolated from diverse soda lakes and bioreactors (Sorokin et al., 2008, 2013). Other members of the Ectothiorhodospiraceae include the phototrophic purple sulfur bacteria (Imhoff, 2006) and Alkalilimnicola ehrlichii MLHE-1, an arsenite-oxidizing chemolithotroph isolated from Mono Lake (Hoeft et al., 2007). As phototrophic purple sulfur bacteria have previously been shown to oxidize both reduced sulfur compounds (Dahl, 2008) and arsenite (Budinoff & Hollibaugh, 2008; Kulp et al., 2008), we hypothesized that they might play a role in thioarsenate transformation. To test this hypothesis, we established anaerobic enrichment cultures of anoxygenic phototrophs able to use thioarsenates for growth, using inocula from Mono Lake and Big Soda Lake, a meromictic, alkaline lake in Fallon, NV. In addition, we tested pure cultures of phototrophic purple sulfur bacteria for their ability to use thioarsenates for growth.

Materials and methods

Field site and sampling

Anoxic water was collected from Station 3 (37°58.97′N, 119°05.35′W) (Hollibaugh et al., 2001) at a depth of 25 m in Mono Lake, CA, USA, on 29 June 2011, and from Big Soda Lake, NV, USA (39°31.66′N, 118°52.62′W), at a depth of 20 m on 20 September 2011. The chemistry of these lakes has been described previously (Cloern et al., 1983; Hollibaugh et al., 2005). Vertical profiles of temperature, pressure, conductivity, photosynthetically active radiation (PAR), beam attenuation, in vivo fluorescence, and oxygen were obtained with a Sea-Bird SBE 19 Seacat CTD equipped with a C-Star transmissometer and WETstar fluorometer (Wet Labs), PAR sensor (Licor), and dissolved oxygen meter (SBE43, Sea-Bird). Vertical profiles of arsenic and sulfur speciation in Big Soda Lake were performed by collecting samples from discrete depths and analyzed as described below. See Supporting Information Fig. S1 for CTD and chemical depth profiles. Water was collected from a 5-L Niskin bottle into polycarbonate bottles flushed with 3 volumes of flowing water and capped with no head space as described previously (Hollibaugh et al., 2005). Bottles were transported and stored in the dark at 4 °C.

Chemicals and chemical analysis

All stock solutions were made using anoxic ultrapure water in an anaerobic chamber (5% H₂/95% N₂ atmosphere, Coy Laboratories). A mixture of thioarsenates was prepared by combining anoxic sodium sulfide (Na₂S·9H₂O) and sodium arsenite (NaAsO₂) stock solutions at a 4 : 1 S : As molar ratio and was used in experiments at a final concentration equivalent to 4 mM sulfide and 1 mM arsenite, unless otherwise noted. This mixture reacts spontaneously to form thioarsenic compounds, mostly trithioarsenite, which is extremely reactive and oxidizes to tri- and tetrathioarsenate. Thioarsenite oxidation can be induced by oxygen but also by elemental sulfur (e.g. found as polysulfide impurities in a sulfide standard). Tri- and tetrathioarsenate are both unstable and eventually decompose to a mixture of arsenite, arsenate, and di- and monothioarsenate (Planer-Friedrich et al., 2010). The thioarsenate mixture was prepared and added to medium immediately prior to inoculating cultures in all growth experiments. Monothioarsenate was synthesized as either Na₃AsO₃S·7H₂O (Suess et al., 2009) or Na₃AsO₃S·12H₂O (Brauer, 1963). Samples for chemical analysis were collected using sterile needles and syringes and filtered through 0.2-μm PTFE syringe filters in the anaerobic chamber. Samples for thioarsenate speciation were collected in 2-mL cryovials and immediately flash-frozen in liquid nitrogen, and stored at −80 °C prior to analysis, when they were thawed in an anaerobic chamber. This method has been previously shown to preserve thioarsenic species (Planer-Friedrich et al., 2010). Total arsenic and speciation of arsenite, arsenate, and thioarsenates was performed by ICP-MS and IC-ICP-MS, respectively, as described previously (Planer-Friedrich et al., 2007). Arsenic speciation in experiments with monothioarsenate amendments was analyzed by HPLC as described previously (Hoeft et al., 2004) using a Waters 626 HPLC with a 0.008 M H₂SO₄ eluent (pH c. 2). This protocol enables detection of arsenate, monothioarsenate, and arsenite (retention times of 12.8, 14, and 17.5 min, respectively) but not of the more thiolated arsenic species. Detection was by absorbance at 200 nm. Others (Hoeft et al., 2004) have measured arsenate and arsenite at 210 nm, but we obtained better sensitivity at 200 nm. Samples for sulfide analysis were collected by 0.2-μm filtration and precipitation with zinc acetate. Free sulfide was measured spectrophotometrically by the methylene blue method (Cline, 1969).

Enrichment cultures

Water collected from Mono Lake as described above was diluted 1 : 10 in 0.2-μm filtered Mono Lake water amended with the 4 : 1 S : As thioarsenate mixture at a final concentration equivalent to 2 mM sulfide and 0.5 mM arsenite. This water was dispensed into 150-mL serum bottles and incubated anaerobically with incandescent light at room temperature (22–25 °C). Enrichments were transferred as 1 : 10 dilutions into the same medium monthly for 3 months and then transferred as either 1 : 10 or 1 : 100 dilutions into an artificial Mono Lake water (AMLW) medium containing the following (g L⁻¹): NaCl (60), (NH₄)₂SO₄ (0.5), MgSO₄^.7H₂O (0.25), KCl (1.7), Na₂B₄O₇·H₂O (2), Na₂SO₄ (16.5), KH₂PO₄ (0.25), K₂HPO₄ (0.5), and 900 mL ultrapure water. After the medium was autoclaved and cooled, 100 mL of a filter-sterilized solution of Na₂CO₃ (106 g L⁻¹) and NaHCO₃ (42 g L⁻¹) (150 mM final concentration in the AMLW medium), 10 mL L⁻¹ of vitamin solution (Oremland et al., 1994), and 1 mL L⁻¹ of unchelated trace elements SL-10 (Widdel et al., 1983) were added aseptically. The final pH of the medium was 9.3–9.5. The addition of basic sulfide and arsenite stock solutions to the AMLW did not raise the pH above 9.55 because of the carbonate–bicarbonate buffer in the medium. Anoxic media were obtained by storing the loosely capped container in the anaerobic chamber for a minimum of 48 h. Enrichment cultures were transferred every 2–4 weeks into fresh media amended with the thioarsenate mixture. Anoxic water from Big Soda Lake was amended with 0.1 mM sulfide and 0.1 mM arsenite. This enrichment culture was incubated at room temperature in the light for a month and then fed with the thioarsenate mixture. Once growth was apparent, the culture was transferred 1 : 1 to AMLW, and then 1 : 10 and 1 : 100 into AMLW, with the thioarsenate mixture added at each transfer. Growth was monitored by observing increases in pink to dark red pigmentation and an increase of cells monitored by epifluorescence microscopy (DAPI) (Porter & Feig, 1980) or optical density measurements at 600 nm (OD₆₀₀) (Eppendorf BioPhotometer^™).

Thioarsenate amendment experiments

All amendment experiments were set up and sampled in an anaerobic chamber using 150-mL serum bottles with black butyl rubber stoppers and aluminum crimp seals. All incubations were at 30 °C. Phototrophic organisms were incubated under fluorescent grow lights (three GE 49892 15W bulbs). An array of LEDs emitting at 850 nm (IR12-850 x 4, total c. 2.4 W, EnvironmentalLights.com) was added to the incubator for some experiments and experiments using this setup are noted. One of our Mono Lake enrichment cultures (EC8) was used for growth experiments. To compare growth of the enrichment culture on thioarsenates vs. sulfide alone, a late-log phase culture was diluted 1 : 25 into triplicate bottles of AMLW that were amended with either the thioarsenate mixture or 4 mM sulfide. Negative controls were prepared in triplicate as described above but inoculated with 0.2-μm-filtered, late-log phase culture medium; however, one replicate of the thioarsenate-amended negative controls was contaminated and discarded. A late-log phase culture was diluted 1 : 25 into triplicate bottles of AMLW that were amended with either 4 mM monothioarsenate or 4 mM thiosulfate ( ) to compare growth of the enrichment culture on monothioarsenate vs. thiosulfate. Triplicate negative controls were prepared as described above; however, one replicate of the monothioarsenate abiotic control was discarded due to contamination. We tested the dependence of growth and monothioarsenate oxidation on photoautotrophy by comparing replicates of a late-log phase culture diluted 1 : 25 into six bottles of AMLW. Two replicates were incubated in the dark and four replicates were incubated in light (grow lights plus IR LEDs). Growth was monitored by OD₆₀₀ and monothioarsenate conversion to arsenate was monitored by HPLC as described above. Once growth reached mid-exponential phase, two of the light-incubated cultures were placed in the dark for the duration of the experiment.

Additional cultures were obtained from the sources listed in Supporting Information Table S1 and grown as described there. Cultures were given an electron donor (1–5 mM monothioarsenate or 4 : 1 S : As thioarsenate mixture as described in ) at the time of inoculation. We tested the ability of these cultures to grow on thioarsenates by inoculating (1 : 25 dilution) fresh medium containing the electron donor being tested with a culture grown to late exponential phase, then following substrate conversion (HPLC) while monitoring growth (OD₆₀₀). These cultures were incubated at 30 °C in the light as described above or in the dark with nitrate (5 mM) as the electron acceptor if the organism was not phototrophic. Thioalkalivibrio jannaschii was incubated aerobically.

Composition of enrichment cultures

Two sets of 16S rRNA gene clone libraries were prepared and sequenced to determine the composition of the enrichment cultures. Genomic DNA was extracted from enrichment culture Mono Lake EC8 after 6–7 transfers during the initial selection phase. Cells from 1.9 mL of the culture were harvested by centrifugation at 10 700 g for 5–10 min. The pellet was resuspended in 1.8 mL of lysis buffer, then DNA was extracted and purified as described previously (Ferrari & Hollibaugh, 1999). In addition, cells from Mono Lake (EC ‘ML Ecto’, EC8, and EC12) and Big Soda Lake (EC13) enrichment cultures were harvested as described above to compare the composition of the cultures after 20–30 transfers. We extracted and purified DNA from these cells using the PureLink^® Genomic DNA kit (Invitrogen) as per manufacturer's instructions for Gram-positive bacteria. PCR was performed in duplicate or triplicate (depending on the library) with OneTaq^® DNA Polymerase (NEB) using Bacteria-specific 27f (5′-AGAGTTTGATCMTGGCTCAG-3′) and universal 1492r (5′-GGTTACCTTGTTACGACTT-3′) 16S rRNA gene primers (Lane, 1991; Turner et al., 1999). The PCR program used for the initial EC8 clone libraries was as follows: initial denaturation at 94 °C for 30 s, 35 cycles of denaturation at 94 °C for 30 s, annealing at 62 °C for 30 s, and extension at 68 °C for 30 s, with a final extension of 45 min (Humayoun et al., 2003). The PCR program used for the second set of clone libraries increased the initial denaturation to 3 min, decreased the annealing temperature to 48 °C for decreased stringency (Frank et al., 2008), and decreased the final extension to only 10 min. PCR products were pooled and run on a 1% agarose gel. Bands corresponding to the expected product size were cut and DNA was extracted from the gel with QIAquick^® Gel Extraction Spin Kit (Qiagen). The amplicons were cloned into pCR^®-4-TOPO^® vector using the TOPO TA Cloning^® Kit for Sequencing (Invitrogen). Colonies containing cloned inserts were picked and grown in Luria–Bertani (LB) medium with ampicillin, then inserts were sequenced by GeneWiz, Inc. Sequences were checked for chimeras with DECIPHER (Wright et al., 2012) and Bellerophon (Huber et al., 2004) and chimeric sequences were discarded. Sequences were edited and aligned, then consensus sequences were generated for alignments that were > 99% similar using Geneious R7 (Biomatters, Ltd). Consensus and reference sequences were aligned using SINA (Pruesse et al., 2012). The SINA alignment was imported into Geneious and a neighbor-joining (Jukes-Cantor) tree was constructed with 100 bootstraps (Fig. S2). A blastn analysis (Altschul et al., 1990) against the NCBI nr and RefSeq databases was performed on the remaining sequences (Table S2). Sequences have been deposited in GenBank under accession numbers KM070827–KM070952.

Results

Comparison of growth on thioarsenates vs. sulfide

Analysis of arsenic speciation in a Mono Lake enrichment culture (EC8) amended with a mixture of 4 mM sulfide and 1 mM arsenite (thioarsenates mixture) or 4 mM sulfide only revealed that thioarsenates were present in the medium prior to the addition of the inoculum (t = −2 h) and that trithioarsenate and arsenate concentrations increased for the first 21 h after inoculation (Fig. 1a). Arsenite concentration decreased rapidly during the first 2 h prior to adding the inoculum then more slowly over the next 211 h. Trithioarsenate was the dominant thioarsenate compound formed, with mono-, di-, and tetrathioarsenate present as minor components. Concentrations of all arsenic compounds were relatively stable between 21 and 89 h. Trithioarsenate concentrations decreased between 89 and 211 h, coincident with increases in mono- and dithioarsenate. This was followed by the disappearance of thioarsenates between 211 and 259 h accompanied by a large increase in arsenate concentration. All added arsenite was completely converted to arsenate (1000 ± 35 μM) over the course of the experiment. Thioarsenate species formed initially in the abiotic control as described above, but speciation changed relatively little over the course of the incubation (Fig. 1b). Initial sulfide concentrations were lower in the thioarsenate amendment compared to the sulfide-only amendment (Fig. 2). Sulfide concentrations began to decrease steadily after 21 h in both the thioarsenate- and sulfide-only amendments. The time course suggests that the thioarsenate-amended culture was using free sulfide in the thioarsenate mixture preferentially during the first 68 h of the experiment, because the concentration of sulfide in the thioarsenate amendment decreased during the first 68 h (Fig. 3) of the incubation, whereas the concentration of thioarsenates remained constant during that time period (Fig. 1a). Sulfide and trithioarsenate concentrations decreased after 68 h, while di- and monothioarsenate concentrations increased until 211 h when di- and monothioarsenate began to be converted to arsenate. Sulfide concentrations in the sulfide-only amendment changed similarly to their pattern in the thioarsenates amendment, but did not decrease as quickly between 21 and 68 h. Sulfide concentrations decreased only slightly in abiotic controls. The growth rate of the enrichment culture was similar when amended with the thioarsenate mixture or with sulfide alone (Fig. 3).

Fig. 1

Thioarsenate speciation in (a) enrichment culture EC8 (triplicates) and (b) filtered abiotic controls (duplicates). Treatments were amended with a 4 : 1 S : As mixture of thioarsenates, and As speciation was determined by IC-ICP-MS. Points are means of replicates with error bars indicating standard deviation for triplicates and range for duplicates.

Open in new tabDownload slide

Fig. 2

Sulfide concentrations for the enrichment culture (EC8) and abiotic controls shown in Fig. 1. Data points are means and standard deviations, except for the thioarsenate mixture abiotic control where duplicates were analyzed and are shown as mean and range.

Open in new tabDownload slide

Fig. 3

Growth of the enrichment culture (EC8) shown in Fig. 1. Points are shown as means of triplicates with error bars indicating standard deviation. The y-axis is log scale.

Open in new tabDownload slide

Comparison of growth on monothioarsenate vs. thiosulfate

A Mono Lake enrichment culture (EC8) was amended with either 4 mM pure monothioarsenate or 4 mM thiosulfate in sulfide-free AMLW. The culture appeared to prefer thiosulfate to monothioarsenate, as there was a much shorter lag prior to the onset of exponential growth and the culture grew faster on thiosulfate than on monothioarsenate. The enrichment culture grew exponentially on monothioarsenate after an initial lag of 165 h. Monothioarsenate was completely converted to arsenate in 290 h (total incubation of 455 h; Fig. 4). There was no change in monothioarsenate or arsenate concentration in the abiotic controls during the incubation. Arsenite concentrations were initially < 65 μM in both the enrichment culture and abiotic controls and decreased to below the limit of detection over the course of the experiment.

Fig. 4

Conversion of monothioarsenate to arsenate by enrichment culture EC8 (closed symbols, solid lines) and comparison with abiotic controls (open symbols, solid line), and growth of the enrichment culture on monothioarsenate or thiosulfate (dashed lines). Experiments were run in triplicate with points shown as averages with error bars indicating standard deviation, except the monothioarsenate abiotic control in which duplicates were analyzed and error bars indicate range.

Open in new tabDownload slide

Light dependence of photoautotrophic monothioarsenate oxidation

We compared growth of the Mono Lake enrichment culture (EC8) in light and dark incubations to verify that the oxidation of monothioarsenate was a phototrophic process. Monothioarsenate was only oxidized when enrichment cultures were illuminated. The concentration of monothioarsenate did not change when cultures were incubated in the dark (Fig. 5a). The concentration of monothioarsenate decreased, with concomitant production of arsenate (Fig. 5b) and growth (Fig. 5c), when cultures were incubated in the light. Monothioarsenate oxidation, arsenate production and growth ceased when light-incubated cultures were placed in the dark.

Fig. 5

Change in monothioarsenate concentration (a), arsenate concentration (b), and growth (c) for the enrichment culture (EC8) amended with monothioarsenate (4 mM) and grown in the light, dark, or grown in light until the time point indicated by the arrow on the graph, then transferred to the dark.

Open in new tabDownload slide

Monothioarsenate oxidation by additional Mono Lake and Big Soda Lake enrichment cultures

We tested additional enrichment cultures from Mono Lake (EC12) and Big Soda Lake (EC13) that had been selected for growth on thioarsenates, for their ability to grow on monothioarsenate only. We also tested the enrichment culture containing Ectothiorhodospira sp. ML Ecto (designated EC ‘ML Ecto’), established and maintained in our laboratory (Budinoff & Hollibaugh, 2008) and originally selected for its ability to grow photoautotrophically on arsenite. All of the enrichments were able to grow photoautotrophically by oxidizing sulfide, thiosulfate, monothioarsenate, and a mixture of thioarsenates (Table 1). All of the enrichment cultures, including EC8, were also able to grow by oxidizing arsenite photoautotrophically, but growth was very slow (weeks to months) with a long lag before growth and oxidation began (data not shown), as first described for EC ‘ML Ecto’ (Budinoff & Hollibaugh, 2008; Kulp et al., 2008).

Composition of enrichment cultures

Ninety-four cloned 16S rRNA gene amplicons from an early transfer of EC8 were sequenced yielding 83 usable sequences. Forty-six (55.4%) of these shared > 99% identity to the Ectothiorhodospira variabilis strain WN22 (Gorlenko et al., 2009) 16S rRNA gene (NR_042700). The remaining sequences consisted of Firmicutes (28, 33.7%), Bacteroidetes (8, 9.6%) and Spirochetes (1, 1.2%). Their top blastn hits are listed in Table S1. Additional libraries from EC8, constructed after the experiments shown in Fig. 1 (21 transfers) and Fig. 4 (27 transfers), were conducted, contained only sequences that were > 99% identical to the E. variabilis WN22 16S rRNA gene. Additional libraries from other Mono Lake (EC12 and EC ‘ML Ecto’) and Big Soda Lake (EC13) enrichment cultures also contained Ectothiorhodospira sp. as the dominant organism. The phylogenetic relationships of the dominant Ectothiorhodospira sp. in the enrichment cultures to pure cultures of members of the family Ectothiorhodospiraceae are shown in Fig. S2.

Growth of pure cultures on monothioarsenate

We tested a number of pure cultures from the family Ectothiorhodospiraceae and one member of the Chromatiaceae (see Table S1) for their ability to grow autotrophically by oxidizing monothioarsenate. This capability appears to be widespread among the Ectothiorhodospiraceae, but not universal (Table 1). Halorhodospira halophila, Ectothiorhodospira shaposhnikovii, Ectothiorhodospira sp. PHS-1, and Ectothiorhodospira sp. Bogoria Red were all able to grow in media containing only monothioarsenate as an electron donor, while converting monothioarsenate to arsenate. The chemolithotrophic Thioalkalivibrio jannaschii oxidized monothioarsenate aerobically but did not appear to grow (no increase in OD₆₀₀). The chemolithotrophic Alkalilimnicola ehrlichii did not oxidize monothioarsenate or thiosulfate with nitrate as an electron acceptor, and Halorhodospira abdelmalekii was the lone phototrophic Ectothiorhodospiraceae tested that was not able to oxidize monothioarsenate or thiosulfate photoautotrophically. Additionally, we tested the well-studied purple sulfur bacterium from the family Chromatiaceae, Allochromatium vinosum, for its ability to oxidize thioarsenates. When amended with the thioarsenate mixture, there was a decrease in total sulfur and an increase in monothioarsenate and arsenate, and growth. However, A. vinosum was not able to grow on monothioarsenate or to oxidize it to arsenate (Table 1).

Discussion

Thioarsenates as an electron donor for anoxygenic photosynthesis

An enrichment culture from Mono Lake was shown to be able to grow on both a mixture of thioarsenates (Figs ^0001- 0003) and solely on monothioarsenate (Fig. 4). Comparison of the enrichment culture with the abiotic control shows that decomposition of thioarsenates is driven by bacterial activity, as there is relatively little change in speciation of the thioarsenates in the abiotic control. Growth of the enrichment culture on sulfide and on the thioarsenate mixture was comparable (Fig. 3), with a larger final OD in the sulfide-grown culture, potentially due to a difference in cell morphology or cell clumping. Sulfide concentrations were higher in enrichment cultures amended with sulfide only vs. thioarsenates. The difference between treatments may reflect rapid formation of stable monothioarsenate or transient formation of insoluble elemental sulfur or arsenic–sulfur (e.g. orpiment) compounds that would have been removed by filtration prior to sulfide analysis. This experiment showed that the photoautotrophic enrichment culture was responsible for the conversion of thioarsenates to arsenate, similar to previous results from experiments with an aerobic, nonphototrophic enrichment culture (Fisher et al., 2008).

In the first set of experiments, it was unclear whether the enrichment culture was oxidizing thioarsenates directly or whether it was oxidizing free sulfide present in the culture, driving the equilibrium toward the abiotic desulfidation of the thioarsenates, which (except for monothioarsenate) are unstable (Planer-Friedrich et al., 2010; Härtig & Planer-Friedrich, 2012). Thus, we examined growth of the enrichment culture on a single thioarsenate compound (monothioarsenate) and compared the time course of growth on this substrate with growth on thiosulfate because of the similarities in the structures of the two sulfur-containing compounds (Table 2). Our results indicate direct transformation of monothioarsenate to arsenate. Previous experiments of monothioarsenate with T. ruber under aerobic conditions and high temperature (80 °C) had shown that abiotic decomposition of monothioarsenate was relevant (Härtig et al., 2014). However, abiotic decomposition of monothioarsenate did not occur under the conditions of our incubations. In addition, abiotic oxidation of monothioarsenate to arsenite and elemental sulfur (followed by the oxidation of those compounds to arsenate and sulfate) was observed in the T. ruber experiments. We saw no increase in arsenite concentrations in both biotic and abiotic treatments in our experiments.

Potential mechanism of phototrophic growth on thioarsenates

Our data suggest that direct oxidation of monothioarsenate may be facilitated by the thiosulfate oxidation (sox) pathway. This pathway has been well studied in a related member of the family Chromatiaceae, A. vinosum. The pathway involves oxidation of thiosulfate and attachment to the SoxYZ enzyme by SoxXAK, followed by hydrolysis and release of sulfate from the SoxYZ complex by the SoxB enzyme, and transfer of the sulfane sulfur from SoxYZ to sulfur globules (Dahl, 2008; Weissgerber et al., 2011). Further conversion of the sulfur globules to sulfate proceeds by a pathway discussed elsewhere (Dahl, 2008). Due to the similarity of the structures of monothioarsenate and thiosulfate (Table 2), we hypothesized that Sox enzymes could cleave the thiol group from monothioarsenate. Among the Ectothiorhodospiraceae with sequenced genomes, Halorhodospira halophila (Dahl, 2008), Ectothiorhodospira sp. PHS-1 (Zargar et al., 2012), Ectothiorhodospira haloalkaliphila (JGI), Ectothiorhodospira shaposhnikovii DSM 2111 (T. Meyer, pers. commun.), and a number of Thioalkalivibrio species (JGI), all contain sox genes and many have been shown here to be able to oxidize monothioarsenate. Of the strains we tested that could not oxidize monothioarsenate, Alkalilimnicola ehrlichii only contains soxY and soxZ genes (Hoeft et al., 2007). We used degenerate PCR primers (Meyer et al., 2007) to test for the presence of soxB in the enrichment cultures and the strains for which we had no genome sequences. Only Halorhodospira abdelmalekii did not yield a PCR product (data not shown), and this organism was unable to oxidize monothioarsenate. However, if a complete sox pathway is all that is required, presumably A. vinosum should have been able to oxidize monothioarsenate. Other factors, such as resistance to arsenic toxicity or the differences in pH and salt requirements between A. vinosum and the Ectothiorhodospira strains we tested, could play a role in their ability to oxidize monothioarsenate.

Relevance of thioarsenate oxidation to Mono Lake and other sulfidic environments

Biological oxidation (as well as reduction, which we have not explored here) of monothioarsenate and other thioarsenates is relevant to arsenic geochemistry as it is an underexplored transformation of environmentally relevant arsenic and sulfur compounds that could affect their toxicity (Planer-Friedrich et al., 2008) or mobility (Stucker et al., 2013). Humayoun et al. (2003) detected many potentially sulfur-oxidizing Gammaproteobacteria related to Thioalkalivibrio throughout the water column of Mono Lake. The primers used in that study only cover about 25% of the diversity of the Ectothiorhodospira [determined by examining 16S rRNA gene reference sequences with TestPrime using the Silva SSU r117 database (Klindworth et al., 2013)], so it is possible that Ectothiorhodospira are an even larger portion of the microbial community in Mono Lake than previously indicated (Humayoun et al., 2003). Our work has shown that both chemolithotrophic and photoautotrophic processes may impact the concentrations of thioarsenic compounds in natural waters.

Conceptual model of thioarsenate oxidation to arsenate

Aerobic oxidation of monothioarsenate (Eqn. ), may occur in the oxic upper layers of Mono Lake mediated by aerobic sulfide-oxidizing bacteria as previously described (Fisher et al., 2008). Anaerobic oxidation of monothioarsenate by purple sulfur bacteria would be more likely in deeper and more sulfidic anoxic waters. Our interpretation of the oxidation of the thioarsenate mixture by the enrichment culture is that biological oxidation of free sulfide occurs preferentially to thioarsenate oxidation. As free sulfide is oxidized biotically, the medium becomes sulfide-deficient, leading to abiotic desulfidation of tetra-, tri-, and dithioarsenates, leaving monothioarsenate which is more stable under these conditions (Planer-Friedrich et al., 2010; Härtig & Planer-Friedrich, 2012). Monothioarsenate oxidation then occurs directly (Eqn. ), potentially due to direct cleavage by an enzyme in the thiosulfate oxidation pathway (sox) or by a similar mechanism involving an unknown enzyme or pathway. An alternative hypothesis is that monothioarsenate decomposes abiotically to arsenite and elemental sulfur (Eqns 2–4), as shown in experiments with T. ruber (Härtig et al., 2014) where arsenite and elemental sulfur are oxidized sequentially. More rapid growth of enrichment cultures and pure Ectothiorhodospira strains on thiosulfate and monothioarsenate vs. arsenite suggests that they prefer sulfur-containing compounds over arsenite (data not shown). In addition, we did not see an increase in arsenite concentration in any of our experiments. Therefore, decomposition of monothioarsenate seems less likely than direct oxidation of the sulfur group on monothioarsenate.

In conclusion, we have shown that many Ectothiorhodospiraceae are capable of photoautotrophic (and perhaps chemoautotrophic) growth on thioarsenic compounds (especially monothioarsenate). This process could play an important role in the transformation of arsenic and sulfur compounds in sulfidic environments.

Acknowledgements

We would like to thank Chris Abin for synthesis of monothioarsenate and for fruitful discussions. The assistance of Robert Jellison and others at Sierra Nevada Aquatic Research Laboratory in sampling Mono Lake is greatly appreciated. We thank Ron Oremland, Chad Saltikov, and others at USGS for providing field support for sampling Big Soda Lake. Ron Oremland, Christiane Dahl, Chad Saltikov, and Mike Madigan supplied some of the strains used in this study. We would also like to thank the helpful comments from two anonymous reviewers that helped improved the manuscript. This research was supported by an award from the US National Science Foundation (NSF EAR 09-52271 to JTH) and an Emmy Noether grant from the German Research Foundation to Britta Planer-Friedrich (PL 302/3-1).

References

Author notes