Microbial Antimony Biogeochemistry: Enzymes, Regulation, and Related Metabolic Pathways

Abstract

Antimony (Sb) is a toxic metalloid that occurs widely at trace concentrations in soil, aquatic systems, and the atmosphere. Nowadays, with the development of its new industrial applications and the corresponding expansion of antimony mining activities, the phenomenon of antimony pollution has become an increasingly serious concern. In recent years, research interest in Sb has been growing and reflects a fundamental scientific concern regarding Sb in the environment. In this review, we summarize the recent research on bacterial antimony transformations, especially those regarding antimony uptake, efflux, antimonite oxidation, and antimonate reduction. We conclude that our current understanding of antimony biochemistry and biogeochemistry is roughly equivalent to where that of arsenic was some 20 years ago. This portends the possibility of future discoveries with regard to the ability of microorganisms to conserve energy for their growth from antimony redox reactions and the isolation of new species of “antimonotrophs.”

INTRODUCTION

Antimony (Sb) occurs widely in soil and aquatic systems. It is a group 15 element in the periodic table, positioned directly below arsenic (As). It exists in four oxidation states (+V, +III, 0, and −III), of which pentavalent antimonate [Sb(V)] and trivalent antimonite [Sb(III)] are the prevalent forms in the environment (1). Being a strong chalcophilic element, Sb frequently cooccurs in sulfidic mineral phases, such as Sb₂S₃ (stibnite) (1, 2). In the aqueous environments at neutral pH, Sb(V) dominates as Sb(OH)₆⁻ under oxic conditions, while Sb(III) is more prevalent as Sb(OH)₃ in anoxic environments (3). Furthermore, Sb shares some chemical and toxicological properties with As (4). Antimony and its compounds are considered to be hazardous pollutants by both the U.S. Environmental Protection Agency (5) and the Council of the European Communities (6). In fact, the EPA drinking water standard for Sb is lower than that for As, reflecting its greater overall toxicity. The maximum contaminant level of Sb in drinking water is 6 μg/liter, according to USEPA (7), and the level established by the Council of the European Communities (CEC) is 5 μg/liter (8). Similar to most trace metals, Sb toxicity strongly depends upon its chemical speciation (9). The general order of toxicity for Sb species is greatest in Sb(III), followed by Sb(V) and then organoantimonials (10). Due to its affinity for the thiol groups of glutathione and proteins, exposure to antimony species can cause injury in many organ systems, such as the lungs, heart, liver, and kidney (11, 12).

Antimony contamination in the environment is caused by both natural and anthropogenic activities (1, 13). Natural sources of Sb to the environment include volcanism and the weathering of Sb-bearing crustal rocks and minerals (14, 15). Antimony is widely used in the manufacture of flame retardants, small-arms ammunition, semiconductors, batteries, alloys, pigments, and catalysts (1). For many years, Sb compounds have been used in the treatment of several tropical protozoan diseases, such as leishmaniasis (16). In addition, human activities, especially increased mining and industrial emissions, have significantly accelerated the release of Sb into the environment and the associated exposure of biota to Sb (17, 18).

The present world production and reserves of Sb are estimated at almost 160,000 and 1,800,000 tons, respectively, most of which is from deposits in China, Bolivia, Mexico, Russia, South Africa, and Tajikistan (19). However, the exploitation and utilization of Sb result in increasing Sb contamination in many countries (20–25). Currently, China is the largest producer of Sb, with more than 80% of the world's supply of Sb coming from the mines of Southwest China (20). In China, the Sb concentrations in water (up to 29.4 mg liter⁻¹), sediment (up to 1,163 mg kg⁻¹), and soil (>2 mg kg⁻¹) reported from the mining and smelting areas are extremely elevated compared to typical background concentrations (1 μg liters⁻¹, 0.800 to 3.00 mg kg⁻¹, and 0.57 mg kg⁻¹, respectively) (20). Antimony can be taken up by plants and photosynthetic biofilms and thereby enter the food chains of contaminated environments (2, 26). This can ultimately cause a series of human health risks (27); hence, the problem of Sb pollution demands global attention.

As was the case with As, microorganisms now appear to play an important role in Sb speciation, mobility, and bioavailability in nature (28, 29). For instance, microbial Sb(III) oxidation, which transforms Sb(III) to Sb(V), could be considered a means of environmental Sb bioremediation because Sb(V) could be stably immobilized (e.g., adsorbed) and safely disposed under an oxic environment (30). Some bacteria can utilize the energy generated from the microbial Sb redox reactions to support their growth (31, 32). Therefore, a better comprehension of the mechanisms driving microbe-Sb interactions is important to elucidate the Sb biogeochemical cycle and to further develop strategies for the bioremediation of Sb-contaminated environments. A search in the NCBI PubMed central database using the word “antimony” shows an exponentially growing number of publications over the past few years (Fig. 1A), suggesting an increased interest in this toxic metalloid. However, most of the reviews that have previously been published over the last decade focused on the behavior, bioavailability, and contamination of Sb in the environment. A detailed knowledge of the molecular mechanisms underpinning the interactions of microorganisms with antimony is still very limited compared with that for other toxic metalloids, like arsenic or selenium. Thus, this review covers the latest findings on microbial Sb transformations and describes our current understanding of the enzymes, regulatory mechanisms, and metabolic pathways involved in biogeochemical cycling of Sb.

FIG 1.

The exponential growth (red line) of the number of publications about Sb in NCBI PubMed (A) and the biotransformation pathways of different Sb species (B).

MICROBIAL ANTIMONY CYCLE

Microbial transformations of Sb influence the environmental fate and toxicity of this metalloid. Microbes have coped with the toxicity of Sb using various strategies to thrive in Sb-rich environments, such as Sb(III) efflux, Sb(V) reduction, Sb(III) methylation, and Sb(III) oxidation (29). These microbial Sb transformations mediate the conversion of Sb compounds among Sb(III), Sb(V), and organoantimonials (Fig. 1B).

Antimonite resistance.

Different strategies are employed by microbes to reduce the accumulation of toxic intracellular Sb(III), such as first inhibiting its entrance into the cell, promoting its active extrusion from the cell if it gains entry, or achieving its sequestration in a nontoxic form within the cell (33). Efflux of antimony is one of the most important mechanisms adopted by microorganisms to protect them from the toxicity of Sb. No specialized channel for antimony uptake has been identified, and possibly no such channel has evolved, because antimony is not an essential trace nutrient (11). At physiological pH, Sb is present as noncharged Sb(OH)₃ in solution, and because of its structural similarities to glycerol, the uptake of Sb(III) into prokaryotic and eukaryotic cells is often achieved by aquaglyceroporins (34). The glycerol facilitator GlpF in Escherichia coli was the first aquaglyceroporin identified to transport Sb(III) into bacterial cells (35, 36). Later, Fps1p, the yeast homologue of GlpF, was also found to mediate the uptake of Sb(III) into Saccharomyces cerevisiae (33). The deletion of fps1 improved the tolerance level of S. cerevisiae to Sb(III), while constitutive expression of this gene resulted in hypersensitivity. Interestingly, the expression of fps1 was repressed when cells were exposed to Sb(III), indicating a coordinated regulatory network to protect cells from the toxic effects of Sb(III) (33). In Leishmania species, Sb(III) entered cells primarily through an aquaglyceroporin named AQP1 (37). In general, the transcription level of AQP1 correlated well with the accumulation of, and the sensitivity level for, Sb(III) in Leishmania cells (38). It was suggested that the route of Sb(V) influx is different from Sb(III) (39). However, the uptake mechanism of Sb(V) remains unknown. Whether it enters the cells through the phosphate transport systems used by As(V), such as Pit or Pst, remains unresolved.

At present, two different transporter families have been shown to be responsible for prokaryotic Sb(III) efflux: the ArsB protein, which belongs to the ion transporter superfamily, and Acr3p, belonging to the arsenite carrier family. The ars operon that confers both arsenic and antimony resistance has been found on both plasmids and the chromosome (40, 41). The three-gene operon arsRBC was present in E. coli, Pseudomonas aeruginosa, Bacillus subtilis, and Staphylococcus aureus (42–44). The extended five-gene operon arsRDABC was first found in E. coli plasmids R773 and R46 and on Acidiphilium multivorum plasmid pKW301 (45, 46). Expression of the ars operons was induced in response to the presence of both As(III) and Sb(III) (43). The ArsR acts as a transcriptional repressor regulating the expression of itself and downstream genes of the ars operon. The arsD gene encodes an As chaperone that transfers As(III) and Sb(III) to ArsA, and ArsA acts as an ATPase, which binds to the As(III)/Sb(III) carrier protein ArsB to form an ATP-coupled efflux pump (47). Meng et al. (36) demonstrated that ArsB is a trivalent metalloid/H⁺ antiporter. In the presence of ArsA, ArsB catalyzes the extrusion of As(III)/Sb(III) by the hydrolysis of ATP, while it can extrude As(III) and Sb(III) by itself using the electrochemical proton gradient. ArsC was shown to be a cytoplasmic As(V) reductase, reducing As(V) to As(III), thereby enabling its efflux from the cell through the ArsAB pump. ArsC might be involved in the resistance to Sb(III) (47, 48), but it is not known whether it is directly related to the intracellular reduction of Sb(V).

Another trivalent metalloid/H⁺ antiporter, Acr3p, and its homolog, YqcL, which are mainly present in Actinobacteria and Alphaproteobacteria, can substitute for ArsB, also functioning as an Sb(III) efflux pump (28, 49). Acr3p is also found in archaea and eukaryotes (50), in which a three-gene cluster (acr1, acr2, and acr3) is responsible for Sb(III) resistance. Kang et al. (51) showed that deletion of acr3 in Agrobacterium tumefaciens 5A resulted in more sensitivity to Sb(III). In addition, it has been shown that the expression of acr3 is stimulated by Sb(III), and its gene product conferred Sb(III) tolerance in yeast (33, 52, 53). In S. cerevisiae, the cadmium factor protein Ycf1, which belongs to ABC transporter superfamily, is another system conferring Sb(III) tolerance through vacuolar sequestration (54). In Leishmania, the ABC transporter PGPR and ABCI4 were reported to be involved in Sb(III)-thiol extrusion (55).

Antimonate reduction.

Antimonate reduction appears to be widespread in the environment, and it is prone to occur under anaerobic conditions (3). It is known that Sb(V) can be abiotically reduced to Sb(III) by Fe(II)-containing minerals (56–58). A marine macroalga, Sargassum sp., was the first reported organism able to reduce Sb(V) in seawater (59). In the treatment of leishmaniasis, several studies suggested that Sb(V), when used as a prodrug medicine, might be reduced in both the vertebrate host and the parasites (60–63). However, the knowledge of bacterial Sb(V) reduction is limited.

Kulp et al. (64) reported anaerobic bacterial reduction of Sb(V) in anoxic sediments. Sb(V) reduction was coupled to a dissimilatory respiratory pathway, which utilized acetate or lactate as the electron donor. That same year, an Sb(V)-reducing bacterium, Bacillus sp. MLFW-2, was isolated and found to generate energy from anaerobic Sb(V) reduction (32), and another Sb(V)-respiring isolate was isolated from Sb-contaminated industrial sediments (65). The molecular mechanism of bacterial Sb(V) reduction remains unknown, and the enzymes involved in this reaction have not yet been identified.

Bacterial Sb(V) reduction is not only a respiratory pathway but also a promising bioremediation strategy, since Sb(III) can readily precipitate with sulfide or be strongly absorbed by Fe phases in a reducing environment (64, 66, 67). A study by Hockmann et al. (18) indicated that Sb(V) could be rapidly reduced to Sb(III) in anaerobic calcareous soil by the indigenous microorganisms. The generated Sb(III) subsequently bound to the surface of iron (hydr)oxides, which led to the immobilization of Sb. In addition, sulfate-reducing bacteria (SRB) were employed to remove Sb(V) from Sb mine drainage (68). The SRB converted sulfate ions into sulfide that reduced Sb(V) to Sb(III) and resulted in the precipitation of stibnite (Sb₂S₃). Moreover, a chemoautotrophic microorganism belonging to the Rhizobium genus was found to be able to use H₂ as the sole electron donor for the reduction of Sb(V), producing an Sb(III) precipitate in the form of Sb₂O₃ (69). Therefore, bacterial Sb(V) reduction holds promise for the anaerobic biotreatment of wastewater containing toxic Sb(V).

Antimonite methylation.

Methylation of inorganic Sb can influence the environmental mobility, toxicity, and bioaccumulation of Sb (Fig. 1B) (70). The presence of stibine (STB; SbH₃), monomethylstibine (MMSb), and dimethylstibine (DMSb) was first reported in natural waters by Andreae et al. (71). Subsequently, the presence of Sb volatile and methylated species was further observed in freshwater, seawater, geothermal waters, sewage, soils, sediments, and landfill gas (71–74). In addition, methylantimony species have been found in plants, such as pondweed (Potamogeton pectinatus), moss (Drepanocladus sp.), and liverwort (75–77).

In contrast to arsenic biomethylation, which has been known for several decades (78), the biomethylation of Sb is of relatively recent interest. So far, Sb biomethylation has been detected in strains of fungi, methanogenic archaea, and bacteria. The filamentous fungi Scopulariopsis brevicaulis and Phaeolus schweinitzii have been found to generate STB, DMSb, TMSb, and some nonvolatile methylantimony species during aerobic growth (79–82). In addition, the biovolatilization and bioaccumulation of antimony by S. brevicaulis were recently quantified (83). The aerobic yeast Cryptococcus humicolus was also reported to biomethylate both inorganic Sb(III) and Sb(V) (84). It has been suggested that the toxic gases generated from Sb methylation by the fungus S. brevicaulis in crib mattresses might be a cause for sudden infant death syndrome (SIDS) (85). However, further studies indicated that a mix of common environmental Bacillus species strains in crib mattress contributed to the formation of TMSb and some nonvolatile methylantimony species, but no causal relation to SIDS was proven (86, 87).

In the process of the anaerobic digestion of sewage sludge, three methanogenic archaea (Methanobacterium formicicum, Methanosarcina barkeri, and Methanobacterium thermoautotrophicum), a sulfate-reducing bacterium (SRB) (Desulfovibrio vulgaris), and a peptolytic bacterium (Clostridium collagenovorans) were shown to produce TMSb in their culture headspaces; among these, M. formicicum displayed strong methylating activity and could also produce STB, MMSb, and DMSb (88). Another study demonstrated that biomethylation of Sb was stimulated by strains of methanogenic archaea and SRB (89). In addition, a Gram-positive strain, Clostridium glycolicum ASI-1, could convert inorganic Sb into the volatile derivatives STB, DMSb, and TMSb (90). Despite using organic Sb as a substrate, the production of TMSb could only be accomplished by the transformation of trimethyldibromoantimony in pure culture of strain Pseudomonas fluorescens K27 (91). Low yields of MMSb, DMSb, and TMSb by an aerobic Flavobacterium sp. strain suggested that Sb methylation may be a fortuitous process rather than a primary resistance mechanism (70). This is consistent with the results of studies with S. brevicaulis and C. humicolus (80, 81, 92). However, the molecular mechanisms of Sb methylation have not been clarified, and it appears that Sb is methylated much more slowly than is arsenic (29, 93).

Because of the similarity in physicochemical properties and the cooccurrence of arsenic and Sb in the natural environment, it is important to understand the effect of arsenic on Sb biomethylation. It was found that Sb biomethylation by S. brevicaulis, Flavobacterium sp., and Cryptococcus humicolus was enhanced in the presence of arsenic (70, 92, 94), while conversely, arsenic biomethylation was significantly inhibited by the presence of Sb (94). In contrast, Hartmann et al. (92) reported that the addition of arsenic not only enhanced the biomethylation of Sb by C. humicolus but also influenced the speciation of Sb. In addition, Sb biomethylation has been shown to utilize the methylation pathway proposed by Challenger in a study that utilized isotopically labeled antimonite (Sb¹²³) (89), whereby antimony methylation demonstrates a stepwise reduction process of monomethyl-, dimethyl-, and trimethylantimony (78). Therefore, Sb biomethylation probably occurs via similar or identical mechanisms to arsenic and is catalyzed, at least in part, by arsenic methyltransferase (78, 89, 95). However, no studies have directly identified genes and enzymes that are involved in arsenic methylation that are also responsible for Sb biomethylation.

Antimonite oxidation.

Although little is known about the geochemical properties of Sb, some studies have indicated that Sb(III) adsorbs more strongly to surfaces and over a wider range of pH than does Sb(V) (67, 96), a situation that is reversed for arsenic oxyanions (e.g., see reference 97). Thus, Sb(III) oxidation may critically affect the hydrologic mobility of Sb in the environment (Fig. 1B) (67). In nature, Sb(III) is thermodynamically predicted to be dominant in anoxic environments, while Sb(V) is dominant in oxic environments (72, 98). Thus, any Sb(III) molecules that enter oxic environments tend to be oxidized to Sb(V), a situation that is similar to As(III).

Abiotic dark oxidation of Sb(III) with O₂ is extremely slow, with a half-life of 170 years at pH 8.5 in homogeneous solutions (99). In contrast, H₂O₂-linked oxidation is much faster, with a half-life of 118 days for 1 μM H₂O₂ held at pH 8.0 (99, 100). Several other oxidants also have the potential to oxidize Sb(III) to Sb(V), including natural and synthetic Fe and Mn oxyhydroxides (101, 102), humic acids (103), and iodate (104). In addition, the oxidizing capacities of Fe and Mn oxyhydroxides can transform As species (105, 106), and the amorphous Fe and Mn oxyhydroxides present in natural water and sediment also play a detoxifying role by adsorbing and oxidizing Sb(III). It was found that amorphous Fe and Mn oxyhydroxides, either natural or synthetic, could effectively oxidize Sb(III) under different pH conditions (101).

Antimony(III) can also be oxidized via photo-induced oxidation in natural surface waters, especially when adsorbed to goethite (107). In experiments with seawater, the Sb(III) photo-oxidation rate was increased in the presence of various live phytoplankton species (e.g., Chlorella autotrophica, Dunaliella salina, Nannochloropsis sp., and Tetraselmis subcordiformis), and the oxidation rate increased with higher cell densities (108), although it was not clear if this was a direct metabolic effect or that of an interaction with cellular exudates. In the case of humic acids, the Sb(III) oxidation rate was 9,000 times faster in the light than in the dark (103). It was reported that Sb(III) bound to natural organic matter and mineral particles is oxidized by photo-oxidants more readily because of the change in electron density that results from adsorption (67, 103). Moreover, in the case of Sb(III) adsorbed onto goethite, oxidation only occurs in the light and was pH dependent, increasing at pH values of >5 (107). Since the Sb(III)-oxidizing bacteria reported so far have all been isolated from neutral-pH-range environments (31, 109–115), it appears that microbes may play a major role of Sb(III) oxidation at circumneutral pH.

The first report that bacteria in ore deposits could oxidize Sb(III) to Sb(V) came from a Russian scientist (31), yet beyond that benchmark work, the field remained dormant for nearly 40 years. Although the amount of research currently being conducted on Sb is growing, our understanding of the role that microbial Sb(III) oxidation plays in the biogeochemical cycle of Sb remains far from complete. Nonetheless, the results reported by several research groups in recent years suggest that microbial processes play an important role in the Sb cycle.

Diversity of antimonite-oxidizing bacteria.

An Sb(III)-oxidizing bacterium, Stibiobacter senarmontii, was found to be able to use the energy produced by Sb(III) oxidation with O₂ to support chemoautotrophic growth (31). The Sb(III)-oxidizing strains that have been described since that early study generally oxidize Sb(III) during heterotrophic growth, suggesting that this process may serve as a cellular detoxification mechanism rather than one whereby energy is conserved from the oxidation to support the biochemical incorporation of CO₂ into the cell's organic matrix. Due to an increased focus on bacterial Sb(III) oxidation, >60 Sb(III)-oxidizing strains were isolated from mining soil (112–114) and contaminated sediments (115–117). Chemoautotrophy, as defined by Sb(III)-dependent growth and inorganic carbon fixation, is more difficult to achieve experimentally. Nonetheless, Terry et al. (115) conducted growth experiments with Variovorax paradoxus IDSBO-4 using radiolabeled [¹⁴C]bicarbonate and demonstrated that aerobic Sb(III) oxidation in that strain was coupled to the fixation of CO₂ in an apparent chemoautotrophic process, thereby reinforcing the earlier observations made with S. senarmontii and opening the possibility of its wider occurrence in nature.

A compilation of the genera found and a phylogenetic tree of the bacterial strains shown to oxidize Sb(III) are given in Fig. 2. The Sb(III)-oxidizing strains identified thus far belong to 17 genera, including Pseudomonas (22 strains), Comamonas (10 strains), Agrobacterium (8 strains), Acinetobacter (7 strains), Stenotrophomonas (3 strains), Variovorax (3 strains), Paracoccus (2 strains), Sphingopyxis (2 strains), Aminobacter (1 strain), Arthrobacter (1 strain), Bacillus (1 strain), Janibacter (1 strain), Stibiobacter (1 strain), Thiobacillus (1 strain), Hydrogenophaga (1 strain), Cupriavidus (1 strain), and Sinorhizobium (1 strain) (31, 109, 111–116). Among all of these Sb(III)-oxidizing strains, Pseudomonas, Comamonas, Agrobacterium, and Acinetobacter are four major genera that make up 34%, 15%, 12%, and 11% of known Sb(III)-oxidizing strains, respectively (Fig. 2A). Of the 65 strains listed in this tally, only two thus far appear to be lithoautotrophs, one of which (S. senarmontii) has been lost. Unlike the case for As(III) (118), to date, there are no examples of anaerobes that can oxidize Sb(III) by using it as an electron donor to support anoxygenic photosynthesis.

FIG 2.

The percentages (A) and a neighbor-joining (NJ) phylogenetic tree (B) based on 16S rRNA gene sequences of the published Sb(III)-oxidizing strains. The Pseudomonas spp. include 22 Sb(III)-oxidizing strains (DA2, DC5, DF12, DF11, DA5, DF3, DF9, DC8, DC7, DS4, DF7, TC13, JC11, DS7, DF8, DF5, DA4, NL6, IK-S1, NL10, NL2, and NL5). The Acinetobacter spp. include seven Sb(III)-oxidizing strains (DC2, LH3, LH4, JL7, DS2, NL1, and NL12). The Comamonas spp. include 10 Sb(III)-oxidizing strains (JL25, JL40, DF1, DS1, DF2, JL13, JL12, JC9, S44, and NL11). The Agrobacterium spp. include eight Sb(III)-oxidizing strains (C58, 5A, GW4, C13, LY4, TS43, TS45, and D14) (see Table S1 in the supplemental material). Among these, the Sb(III) oxidation capabilities of six Agrobacterium (strains C58, C13, LY4, TS43, TS45, and D14) and Sinorhizobium sp. GW3 are unpublished. All of the type strains are used for taxonomic determination without knowing their Sb(III) oxidation abilities. Bootstrap values (>50%) are shown at nodes as percentages of 1,000 replicates. Bar, 0.02 substitutions per nucleotide position.

All of these Sb(III)-oxidizing strains could be classified into Alphaproteobacteria, Betaproteobacteria, Gammaproteobacteria, and Actinobacteria (Fig. 2B). Among all of these Sb(III)-oxidizing strains, 49% belong to Gammaproteobacteria, and among these, Pseudomonas and Acinetobacter are two of the most common species (Fig. 2A). Comamonas strains belong to Betaproteobacteria, while Agrobacterium strains are members of the Alphaproteobacteria. The strains belonging to Betaproteobacteria showed the highest Sb(III) oxidation rate; for example, Comamonas testosteroni S44 could completely oxidize 50 μM Sb(III) to Sb(V) within 3 days (114). It is interesting to note that C. testosteroni S44 could not oxidize As(III) (119), indicating that the molecular mechanism of Sb(III) oxidation may at times be different from As(III) oxidation (116). Sb-dependent chemoautotrophic growth of V. paradoxus strain IDSBO-4 was able to oxidize ∼500 μM Sb(III) to Sb(V) over a 10-day incubation period (115). Studies that attempt to show clear Sb-dependent growth are difficult, as they must employ higher concentrations (millimolar range) of this toxic electron donor to elicit significant increases in cell density over the incubation period.

Biotic antimonite oxidation mediated by AioA and AnoA.

Bacterial As(III) oxidation involves the As(III) oxidase AioBA or ArxAB (120, 121). AioBA functions as an aerobic As(III) oxidase (122), while ArxAB catalyzes the anaerobic oxidation of As(III) (121). Based on the similar chemistries between As and Sb, it has been considered that they may share the same resistance and oxidation mechanisms. Indeed, the ars operon conferring arsenic resistance can be induced by Sb(III) and can also transport Sb(III) out of the cell (36). In contrast, Sb(III) does not induce the transcription of aioBA (123). Even though there is a previous study that reported that the oxidation of Sb(III) and As(III) required different biochemical pathways (116), more recent literature has shown that AioBA is able to oxidize Sb(III) both in vivo and in vitro, although the aioBA gene is expressed only by the presence of As(III) and not Sb(III) (123). However, novel Sb(III) oxidation biochemical pathways were also implicated because the disruption of aioA gene only reduced the Sb(III) oxidation rate, but did not eliminate it, thereby implying the cooccurrence of another mechanism (123). Previous studies found that the Comamonas strains could only oxidize Sb(III) but not As(III) (114, 119, 124, 125). Furthermore, Hydrogenophaga taeniospiralis IDSBO-1, isolated by Terry et al. (115), was shown to possess the aioA gene and oxidized As(III), but not Sb(III), under aerobic conditions. That strain also exhibited anaerobic Sb(III) oxidation coupled to the reduction of nitrate via an unknown enzymatic pathway.

By using a proteomics approach, Li et al. (117) discovered an oxidoreductase (AnoA) in A. tumefaciens GW4 that was induced by the presence of Sb(III). The disruption of anoA reduced resistance to Sb(III) and also decreased the Sb(III) oxidation rate by ∼27% compared with that of the wild-type strain, while the overexpression of anoA increased the Sb(III) oxidation rate by ∼34%. In addition, heterologous expression of AnoA significantly increased the Sb(III) oxidation rate in E. coli (117). Acting as a novel Sb(III) oxidase, AnoA could also oxidize As(III) in vitro, with a K_m of 103.2 ± 17.7 μM and a maximum rate of metabolism (V_max) of 88.23 ± 6 nmol min⁻¹ mg⁻¹ (Fig. 3A). Using Sb(III) as a substrate, AnoA yielded K_m and V_max values of 64 ± 10 μM and 150 ± 7 nmol min⁻¹ mg⁻¹, respectively (Fig. 3C). In contrast, based on published data (123), AioBA yielded K_m and V_max values for As(III) of 9.3 ± 1.5 μM and 120.4 ± 6 μmol min⁻¹ mg⁻¹, respectively (Fig. 3B). The addition of Sb(III) yielded K_m and V_max values of 163 ± 8 nM and 18.4 ± 1.2 nmol min⁻¹ mg⁻¹, respectively (Fig. 3D). These results indicated that AnoA tends to catalyze the Sb(III) oxidation more efficiently than As(III) oxidation, while AioBA is likely to favor oxidation of As(III), although both enzymes could oxidize both As(III) and Sb(III). The existence of the novel Sb(III) oxidase AnoA may explain the occurrence of discernible Sb(III) oxidation in bacteria that lack the As(III) oxidase AioBA (114).

FIG 3.

The comparison of Michaelis-Menten kinetics of AnoA and AioBA for As(III) and Sb(III), and the influence of H₂O₂ concentration on bacterial Sb(III) oxidation efficiency of A. tumefaciens GW4. (A and B) Kinetic data for As(III). (C and D) Kinetic data for Sb(III). The data from panel A are from our unpublished data, the data from panels B and D are from reference 123, and the data from panel C are from reference 117. (E) The growth curves of strain GW4, the katA mutant strain, and the katA complementary strain. (F) Sb(III) oxidation profiles of the three strains as in panel E (shown with the same symbols in panel E). (G and H) The lacZ reporter assays of katA gene with the addition of H₂O₂ and Sb(III), respectively (panels G and H have the same symbols). The data are shown as the mean of the results from three replicates, with the error bars representing the standard deviation (SD). v, volume; OD₆₀₀, optical density at 600 nm; d, days; β-gal, β-galactosidase; MU, Miller units.

Based on genome analysis, the antimonite oxidase gene anoA (117) exists in all of the arsenite-oxidizing Agrobacterium and Comamonas strains tested thus far, and the gene exists in other as-yet-untested bacterial strains. To understand the phylogenetic relationship among AnoA in different bacteria, we performed phylogenetic analysis based on amino acid sequences of the putative AnoA from 10 Agrobacterium strains, together with five Rhizobium strains, three Sinorhizobium strains, and two Comamonas strains (see Fig. S1 in the supplemental material). The AnoA can be classified into two main groups. One group contains the AnoA from strains of Agrobacterium, Rhizobium, and Sinorhizobium, and another group contains AnoA of Comamonas strains. This indicates that AnoA in Comamonas has a distant phylogenetic relationship from that in Agrobacterium, Rhizobium, and Sinorhizobium strains. Based on the literature published to date, the deletion of either aioBA or anoA only reduces the Sb(III) oxidation rate but does not completely eliminate Sb(III) oxidation, implying (but not proving) the possible existence of another mechanism(s) of Sb(III) oxidation (117, 123).

Abiotic antimonite oxidation is mediated by H₂O₂ and possible regulatory mechanisms involved.

In a variety of naturally occurring surface waters, hydrogen peroxide (H₂O₂) is present at concentrations exceeding 10⁻⁷ mol liter⁻¹ (i.e., 10 nM) and is thought to play a key role in the redox chemistry of a number of trace elements in aquatic environments (126–129). Sb(III) oxidation by H₂O₂ has been studied over a wide range of pH values, ionic strengths, and temperatures, and it may be relevant in surface water with elevated H₂O₂ with alkaline pH values, such as seawater (100). H₂O₂ is not only widespread in natural surface water and rainwater but also exists within bacterial cells.

Aberrant electron flow especially under stress conditions from the electron transport chain or cellular redox enzymes to O₂ leads to the production of reactive oxygen species (ROS) in bacterial cells (130). The harmful ROS, including superoxides (O₂⁻), hydrogen peroxide (H₂O₂), and hydroxyl radicals (·OH), can cause damage to [Fe-S] clusters, protein carbonylation, membrane lipid peroxidation, and DNA damage (131). Superoxide dismutase (Sod), which catalyzes the dismutation of O₂⁻ to H₂O₂ and O₂, is part of a first-line defense against these ROS and commonly occurs in nearly all aerobic bacteria. Catalases and peroxidases represent the second line of defense against ROS by being able to consume H₂O₂ (131). It has been reported that KatA is a major catalase that can be detected during all phases of growth (132). Therefore, the katA gene in the Sb(III)-oxidizing strain A. tumefaciens GW4 was disrupted, and the mutant strain GW4-ΔkatA and its complemented strain GW4-ΔkatA-C were created. The wild-type strain, GW4-ΔkatA, and GW4-ΔkatA-C showed consistent growth profiles in chemically defined medium (CDM) after the addition of 50 μM Sb(III) (Fig. 3E). The disruption of katA significantly increased the Sb(III) oxidation rate (Fig. 3F), which may have been caused by the increasing cellular H₂O₂ concentration (130). Consistent with a previous study by Khakimova et al. (133), H₂O₂ could induce the expression of KatA in strain GW4 (Fig. 3G). Moreover, the addition of Sb(III) was also able to stimulate the expression of katA (Fig. 3H). These results indicated that H₂O₂ may act as a chemical oxidant in Sb(III) oxidation, along with the aforementioned enzymatic reactions.

In addition to catalase, glutathione was considered to be a vital component of the bacterial oxidative stress response (134). A previous study with C. testosteroni S44 showed that the [Fe-S] assembly transcription factor IscR could positively contribute to glutathione (GSH) formation, possibly through the regulation of IscS-mediated cysteine desulfurization (134). The expression of iscR was induced by Sb(III), and the deletion of iscR decreased the cellular GSH content. These results suggested that bacterial Sb(III) oxidation was partly under the regulation of IscR (134). The hypothetical model of IscR's regulation of bacterial Sb(III) oxidation may be summarized by the following: (i) Sb(III) can induce the bacterial oxidative stress response, leading to the production of H₂O₂; (ii) the induced H₂O₂ oxidizes Sb(III) to Sb(V) under alkaline conditions; and (iii) IscR is involved in the regulation of GSH formation. Then, H₂O₂ is consumed by KatA and GSH, which might also affect bacterial Sb(III) oxidation (Fig. 4A). Although there are other regulators of the bacterial oxidative response, such as the Mer-like redox sensor SoxR and the LysR regulator OxyR (131), their function(s) with respect to Sb(III) oxidation has not been determined.

FIG 4.

Overview of mechanisms of bacterial antimonite resistance and oxidation. (A) A hypothetical model of IscR's regulation of bacterial Sb(III) oxidation. (i) Sb(III) induced the production of H₂O₂ via the bacterial oxidative stress response and subsequently H₂O₂ oxidized Sb(III) to Sb(V). (ii) H₂O₂ was partially consumed by catalase KatA. (iii) Sb(III) induced the expression of [Fe-S] assembly transcription factor IscR, which could positively contribute to GSH formation. Then, H₂O₂ was partially consumed by GSH. (B) Cellular events are represented on this model according to the published literature. Sb(III) is taken up through glycerol channel and extruded from the cell by Acr3 and ArsAB, and transportation of Sb(V) remains unknown. Bacteria obtained Sb(III) resistance by the ars operon. In addition, Sb(III) oxidation, Sb(V) reduction, and Sb(III) methylation were also involved in bacterial Sb detoxification. For energy generation, Sb(III) could induce activation of the TCA cycle and produce energy to protect against the toxicity of Sb.

METABOLIC PATHWAYS ASSOCIATED WITH ANTIMONITE RESISTANCE

A proteomics approach was used to study Sb resistance and oxidation in Leishmania spp. (135–138), Miscanthus sinensis (139), and Agrobacterium tumefaciens (117). The proteomics analysis in Leishmania spp. showed that its mechanism of antimony resistance is complex, incorporating aspects of protein folding/chaperones, stress response, antioxidant/detoxification, diverse metabolic processes, RNA/DNA processing, and de novo protein biosynthesis (135–138). The proteomic study of A. tumefaciens GW4 revealed that Sb(III) could influence the Ars resistance, the Sb(III) oxidase AnoA, phosphate metabolism, carbohydrate transport and metabolism, and the metabolism of lipids, purines, and amino acids (Table 1).

TABLE 1.

Proteins induced by the addition of Sb(III) in A. tumefaciens GW4^a

Gene name	Protein name	Accession no.	Upregulated ratio [zero Sb(III):50 μM Sb(III)]b
Antimony resistance
arsC1	Arsenate reductase	AFM38847	1.0:2.9
arsC2	Arsenate reductase	AFM38848	1.0:4.7
ohr	Organic hydroperoxide resistance protein	KDR90118	1.0:3.2
Antimonite oxidation
anoA	Oxidoreductase	KDR88348	1.0:4.1
Phosphate metabolism
pstS2	Phosphate-binding protein	KDR86346	1.0:2.0
ppa	Putative phosphatase	KDR90647	0:11.2
phnM	Metal-dependent hydrolase involved in phosphonate metabolism	KDR86941	1.0:2.8
phnI	Putative enzyme of phosphonate metabolism	KDR86951	1.0:4.1
afuA	ABC transporter, substrate-binding protein	KDR89957	1.0:2.7
Carbohydrate transport and metabolism
pdhB	Pyruvate dehydrogenase E1 component, beta subunit	KDR89057	1.0:2.7
pfp	Pyrophosphate fructose 6-phosphate 1-phosphotransferase	KDR87902	1.0:2.8
ugpB1	Periplasmic glycerol-3-phosphate-binding protein	KDR87393	0:5.3
ugpB2	Periplasmic glycerol-3-phosphate-binding protein	KDR89469	0:8.7
acnAb	Aconitate hydratase	KDR88332	1.0:2.1
sdhAb	Succinate dehydrogenase	KDR89039	1.0:3.3
fumCb	Fumarate hydratase	KDR89425	1.0:1.7
Lipid transport and metabolism
sitA	Manganese ABC transporter, periplasmic-binding protein	KDR90951	1.0:2.1
Purine metabolism
cpdP	3′,5′-Cyclic-nucleotide phosphodiesterase	KDR86320	1.0:25.9
Amino acid metabolism
trpB	Trp repressor-binding protein	KDR87480	1.0:3.9

^aCertain gene/protein and ratio data are from reference 117.

^bGenes tested by qRT-PCR.

The arsenic resistance system (ars) was shown to catalyze the efflux of both As(III) and Sb(III) and was induced by Sb(III) (36). Consistent with these results, the upregulation of the Ars resistance system by Sb(III) could be considered for use as a positive control for the validity of this proteomic analysis in A. tumefaciens GW4. The phosphate system and the proteins involved in phosphate and phosphonate metabolism were both upregulated with the addition of Sb(III) (Table 1). The induction of PstS2 in the presence of Sb(III) (Table 1) implies that Pst2 may have an effect on bacterial Sb(III) oxidation (117).

In comparative proteomic and genomic analyses in the presence or absence of As(III), the enzymes involved in the tricarboxylic acid (TCA) cycle were upregulated after the addition of As(III), indicating the bacteria needed a large amount of energy to resist As(III) toxicity (140–142). In contrast, the TCA cycle in A. tumefaciens GW4 was downregulated by the addition of As(III). Interestingly, this strain could use the energy generated from As(III) oxidation to support growth (Q. Wang, K. Shi, X. Wang, Y. Han, J. Li, L. Wang, J. He, M. Li, and G. Wang, unpublished data), indicating that As(III) is not simply a toxic element to strain GW4 and that it has an energy-linked chemolithotrophic metabolic facet. It is significant that the expression of the enzymes associated with carbohydrate metabolism in A. tumefaciens GW4 was increased by the presence of Sb(III) (Table 1) (117), indicating that strain GW4 possesses totally different resistance mechanisms for As(III) and Sb(III). In addition, reverse transcription-quantitative PCR (qRT-PCR) analysis showed that the genes involved in the TCA cycle were induced by Sb(III), suggesting that strain GW4 may require increased energy to tolerate Sb(III) (Table 1; see also Fig. S2 in the supplemental material).

CONCLUSION AND PERSPECTIVE

This review highlights the recent advances in our understanding of microbial Sb transformations (Fig. 4B). Due to the similar chemical characteristics between As and Sb, the biochemical pathways of Sb(III) oxidation as well as the pathway for dissimilatory Sb(V) reduction were predicted to be shared with As(III) and As(V). However, based on recent literature and our published work, we propose that microbial Sb transformation proceeds by some unique biochemical mechanisms compared with As(III). In the future, particular issues that require attention may include the following points.

(i) It is known that Sb(III) is extruded by the As(III) transporter Acr3 or ArsB. However, the mechanisms of Sb(V) transportation into cells remain unknown. A proteomics study (117) hinted that, as in the case of As(V), a phosphate transport system may be involved in Sb(V) transport; further research to identify the underlying molecular mechanism of Sb(V) importation still needs to be conducted.

(ii) In addition to biologically driven Sb(III) oxidation, some abiotic factors, such as H₂O₂, are also responsible for bacterial Sb(III) oxidation. The regulation of AnoA's expression and other factors related to abiotic Sb(III) oxidation require further study.

(iii) Dissimilatory microbial Sb(V) reduction using organic substrates as electron donors has been reported. While it is conceivable that the respiratory As(V) reductase ArrAB might be involved in all or part of dissimilatory Sb(V) reduction, this has not been proven, and it remains an open subject for future research scrutiny. Likewise, cytoplasmic Sb(V) reduction, analogous to internal cellular As(V) resistance, reduction, and export, has not been investigated thus far.

(iv) Sb(III) methylation has been described in some microorganisms. To better understand the environmental mobility, toxicity, and biogeochemical cycle of Sb, it is important to clarify mechanisms of Sb(III) methylation and the roles of Sb(III) transmethylase. The volatile methyl and hydride derivatives generated from microbial Sb(III) methylation may represent a significant environmental hazard.

(v) Because of significant and growing Sb environment contamination problems, it is necessary to develop Sb biosensors and bioremediation tools. The expression of AnoA in environmental microbial communities may be a potential biosensor for the monitoring of Sb(III). Anaerobic bacterial Sb(V) reduction holds the possibility of being applied for the remediation of Sb-contaminated environments by producing an immobilized Sb(III) phase. Antimony(III)-oxidizing bacteria may also be applied to enhance phytoremediation effects in combination with plants.

Biographies

Jingxin Li received her B.S. in Biology from Huazhong Agriculture University in 2012, and now she is a Ph.D. student of the College of Life Science and Technology at Huazhong Agriculture University, China. Her major is environmental microbiology, and she does research at the State Key Laboratory of Agricultural Microbiology. In recent years, she mainly studied the molecular mechanism of microbial transformation of toxic metalloids (Sb and As), especially the genes and enzymes responsible for bacterial antimonite oxidation. She has been working in this area for five years.

Qian Wang received her B.S. in Bioengineering from Huazhong Agriculture University in 2007, and got her Ph.D. in Science from Huazhong Agriculture University in 2013. After that, she worked as a postdoc at Huazhong Agriculture University. For up to 10 years, she studied microbial transformation mechanisms of arsenic and antimony and soil arsenic bioremediation using poplar and arsenite-resistant bacteria. Since 2016, she has been a postdoc researcher at Montana State University working on arsenite oxidation regulation and metagenomics of the Yellowstone microbial ecosystem.

Ronald S. Oremland received his B.S. in Biology from Rensselaer Polytechnic Institute in Troy, NY, and his Ph.D. in Marine Sciences from the Rosenstiel School of Marine and Atmospheric Sciences of the University of Miami. He did an NRC postdoc at the NASA Ames Research Center in Moffett Field, CA, after which he joined the U.S. Geological Survey in Menlo Park, CA, where he has been since 1977, currently as a Senior Research Scientist. His research interests lie in the realm of microbial metabolism of toxic metalloids (As, Sb, Se, and Te) and of noxious gases (e.g., methane, methyl bromide, acetylene), as well as microbes and microbial processes occurring in extreme environments, especially soda lakes of the western United States. He is a Fellow of the following societies: ASM, AGU, ISEB, and AAAS.

Thomas R. Kulp received his degrees in Geology from Juniata College in Huntingdon, PA (B.S.), East Carolina University in Greenville, NC (M.S.), and Indiana University in Bloomington, IN (Ph.D.). He was an NRC postdoctoral fellow based at the U.S. Geological Survey in Menlo Park, CA, and later worked there as a research scientist. He joined the faculty at Binghamton University, State University of New York, in 2011 in the Department of Geological Sciences and Environmental Studies. His research investigates the geomicrobiological cycling of toxic metalloids, including As, Sb, and Se. He works with microorganisms from a range of different environmental settings that cycle these metalloids, including contaminated freshwater ecosystems, ground water aquifers, hot springs, and soda lakes. He is a member of AGU, GSA, and ASM.

Christopher Rensing obtained his Ph.D. at the Freie Universität Berlin and later at Martin-Luther Universität Halle-Wittenberg. He was a postdoc working with Barry P. Rosen at Wayne State University Medical School in Detroit, MI. Then, he joined the University of Arizona as an Assistant Professor in 1999 and was promoted to Associate Professor in 2007. In 2011, he joined RTI International as the Director of Research into Food Safety at the Center for Agricultural and Environmental Biotechnology, but shortly after that, he accepted a position as a full professor at the University of Copenhagen in 2012. He holds a rank of full professor at the Institute of Urban Environment, Chinese Academy of Sciences, Xiamen, and became an Adjunct Scientist at the J. Craig Venter Institute in 2016. For his main research, concerning metal-microbe interactions, he is actively investigating mechanisms of such from the cellular to the ecosystem levels.

Gejiao Wang received a B.S. in Plant Physiology and Biochemistry from China Agriculture University in 1984. After that, she worked for five years in the Institute of Microbiology, Chinese Academy of Sciences. She got her Ph.D. in Genetics from the University of Pavia, Italy, in 1995 and was a postdoctoral fellow at Clemson University and the University of Arizona, USA. Since 2004, she has been working as a full professor at Huazhong Agriculture University, China. She devoted herself to the study of microbial transformation of metal(loids) (Sb, As, Cr, and Se) for 15 years. Currently, she is an editorial board member of Applied and Environmental Microbiology and Frontiers in Microbiology.