Search tips
Search criteria 


Logo of scirepAboutEditorial BoardFor AuthorsScientific Reports
Sci Rep. 2017; 7: 43225.
Published online 2017 March 2. doi:  10.1038/srep43225
PMCID: PMC5333119

Abiotic and biotic factors responsible for antimonite oxidation in Agrobacterium tumefaciens GW4


Antimonite [Sb(III)]-oxidizing bacteria can transform the toxic Sb(III) into the less toxic antimonate [Sb(V)]. Recently, the cytoplasmic Sb(III)-oxidase AnoA and the periplasmic arsenite [As(III)] oxidase AioAB were shown to responsible for bacterial Sb(III) oxidation, however, disruption of each gene only partially decreased Sb(III) oxidation efficiency. This study showed that in Agrobacterium tumefaciens GW4, Sb(III) induced cellular H2O2 content and H2O2 degradation gene katA. Gene knock-out/complementation of katA, anoA, aioA and anoA/aioA and Sb(III) oxidation and growth experiments showed that katA, anoA and aioA were essential for Sb(III) oxidation and resistance and katA was also essential for H2O2 resistance. Furthermore, linear correlations were observed between cellular H2O2 and Sb(V) content in vivo and chemical H2O2 and Sb(V) content in vitro (R2 = 0.93 and 0.94, respectively). These results indicate that besides the biotic factors, the cellular H2O2 induced by Sb(III) also catalyzes bacterial Sb(III) oxidation as an abiotic oxidant. The data reveal a novel mechanism that bacterial Sb(III) oxidation is associated with abiotic (cellular H2O2) and biotic (AnoA and AioAB) factors and Sb(III) oxidation process consumes cellular H2O2 which contributes to microbial detoxification of both Sb(III) and cellular H2O2.

Antimony (Sb) is an element belonging to Group 15 of the Periodic Table and behaves similar to arsenic (As). Sb and its compounds are recognized as priority pollutants by the United States Environmental Protection Agency1 and European Union2. In recent years, the serious Sb pollution resulting from increased exploitation and industrial emission has aroused growing concern3,4,5. Among various oxidation states (−3, 0, 3, and 5), antimonite [Sb(III)] and antimonate [Sb(V)] are the most common forms6. Because microbial redox reactions can be used as a strategy for biochemical detoxification and can further affect the mobility, toxicity, and bioavailability of Sb in the environment7,8, a better comprehension of the microbe-Sb interactions is important for the bioremediation of Sb-contaminated environments and to understand the Sb biogeochemical cycle. In generally, Sb(III) is more toxic than Sb(V)5,9, thus, examining the mechanisms driving bacterial oxidation from Sb(III) to Sb(V) could be of significant value in this regard.

The understanding of microbial Sb transformation remains deficient compared to that of As10. It has been reported that the glycerol transporter GlpF and its homolog Fps1p are responsible for Sb(III) uptake, reflecting the structural similarities between Sb(OH)3 and glycerol11,12,13, while the As(III) efflux proteins ArsB and Acr3 can also function as Sb(III) efflux pumps14,15. Nevertheless, the pathway of Sb(V) entrance has not been found yet. In addition, the genes and enzymes involved in microbial Sb(V) reduction and Sb(III) methylation have not been identified, although these phenomena are environmentally widespread9.

In contrast with bacterial As(III) oxidation, which has been clarified for several decades, the mechanism of bacterial Sb(III) oxidation is of relatively recent interest. At present, about 60 Sb(III)-oxidizing bacteria, widely existing in various genera (e.g., Pseudomonas, Comamonas, Agrobacterium and Acinetobacter), have been reported16. Recently, we found that the As(III) oxidase AioAB is also function as an Sb(III) oxidase in Agrobacterium tumefaciens 5A17, and subsequently we found a novel Sb(III) oxidase AnoA belonging to the short-chain dehydrogenase/reductase family of enzymes in A. tumefaciens GW418. Compared with A. tumefaciens 5A17, strain GW4 has considerably higher Sb(III) resistance and Sb(III) oxidation efficiency18. However, deletion of each gene only partially influenced the Sb(III) oxidation efficiency of A. tumefaciens strains, indicating other unknown mechanisms.

Chemically, Sb(III) can be oxidized through several oxidants, such as H2O2, iodate, nature minerals (e.g., Fe and Mn oxyhydroxides) and humic acid under oxic conditions19,20,21,22. In bacterial cells, the aberrant electron flow under stress conditions from the electron transport chain or cellular redox enzymes to O2 results in the production of reactive oxygen species (ROS)23. The harmful ROS [e.g., superoxide (O2·), hydroxyl (OH·) and H2O2] can induce DNA damage and the oxidative deterioration of lipids and proteins24,25,26. Thus, bacteria have evolved defense mechanisms against the oxidative stress. Superoxide dismutase (Sod), which catalyzes the dismutation of O2· to H2O2 and O2, plays an important role in defense against ROS. The generated H2O2 is subsequently consumed by catalases and peroxidases23,27. In a recent work, we deleted the catalase gene katA in A. tumefaciens GW4 and observed that the Sb(III) oxidation efficiency of the mutant strain was significantly increased, and the phenotype of the complementary strain was recovered16. Moreover, the transcription of katA was induced by both H2O2 and Sb(III)16. Therefore, we proposed that the increased Sb(III) oxidation efficiency in the mutant strain might reflect the accumulation of H2O2 in bacterial cells. Nevertheless, there is no direct evidence of a correlation between H2O2 and Sb(III) oxidation.

In the present study, we performed gene knock-out/complementation of katA, anoA, aioA and anoA/aioA and in combination with the analyses of Sb(III) oxidation, cellular H2O2 content and resistance of Sb(III) and H2O2 in A. tumefaciens GW428. We provide the first evidence that besides the biotic factors (AnoA and AioAB), the Sb(III) induced cellular H2O2 also catalyzes bacterial Sb(III) oxidation as an abiotic oxidant. The present study documented the abiotic Sb(III) oxidation and clarified the relationship between Sb(III) resistance and bacterial oxidative stress. The results represent an important step toward unraveling the co-metabolism of bacterial Sb(III) oxidation.


Genomic information of aioA, anoA and katA in A. tumefaciens GW4

A. tumefaciens GW4 is an Sb(III)-oxidizing strain, and its genome sequence was previously published (Accession No. AWGV00000000)18. To investigate the abiotic factors of Sb(III) oxidation, two types of cellular oxidative stress-related genes were analyzed. The catalase gene katA, responsible to degrade H2O2 to H2O and O227, showed a 91% sequence identity with katA in A. tumefaciens C5829. We also analyzed the superoxide dismutase Sod, which could convert O2· to H2O2 and O2. The BlastN results showed that katA is a single-copy gene in the genome of strain GW4, while sod has two copies (sod1 and sod2). Thus, subsequent gene knock-out and complementation studies associated with the abiotic factors of Sb(III) oxidation were mainly focused on the katA gene. The arrangement of katA and its surrounding genes are shown in Fig. 1A. For biotic Sb(III) oxidation, the oxidoreductase gene anoA (Fig. 1B), identified as a novel Sb(III) oxidase, is conserved in the genomes of Agrobacterium, Sinorhizobium and Rhizobium strains18. Although several genes were annotated as “short chain dehydrogenase”, “oxidoreductase”, or “putative oxidoreductase” in the genome of strain GW4, the BlastN analyses indicated that the sequences of these genes showed no similarities with that of anoA. Moreover, BlastP analyses showed that the protein sequence of AnoA showed only ~30% similarity with those of other oxidoreductases. In addition, draft genome sequencing revealed that an arsenic gene island located in contig 215 contains the As(III) oxidase genes aioAB (Fig. 1C), and aioA encodes the large subunit of As(III) oxidase30.

Figure 1
Physical map of katA, anoA and aioA (A–C) and gene transcription of katA, sod1, sod2, anoA and aioA in A. tumefaciens GW4 (D). (A–C) Gene clusters of katA, anoA and aioA. (D) Quantitative reverse transcriptase-PCR analysis. Total RNA was ...

Sb(III) induces the transcription of katA, sod1, sod2 and anoA, but not aioA

To investigate the biotic and abiotic factors associated with Sb(III) oxidation in A. tumefaciens GW4, the transcription levels of genes katA, sod1, sod2, anoA and aioA were examined. The catalase KatA and superoxide dismutase Sod are involved in the bacterial oxidative stress response, and the Sb(III) oxidase AnoA and As(III) oxidase AioAB were both reported to catalyze Sb(III) oxidation in vitro16,17,18. The quantitative reverse transcriptase PCR assays indicated that the transcription levels of both katA and anoA were increased with the addition of Sb(III), consistent with the results of our previous studies16,17,18. In addition, the transcription of sod1 and sod2 were also induced by Sb(III). The transcription level of sod1 was much lower than sod2, suggesting that sod2 might play a more important role in dismutation of O2·. However, the transcription level of aioA was not induced by Sb(III) (Fig. 1D), consistent with the previous observations in A. tumefaciens 5A17.

Gene knock-out and complementation analyses showed effects of katA, anoA and aioA on Sb(III) oxidation

Previously, we showed that the disruption of H2O2 degradation gene katA increased Sb(III) oxidation efficiency in strain GW416. The successful deletion and complementation of katA were confirmed by diagnostic PCR shown in Fig. S1. Strains GW4-ΔaioA, GW4-ΔanoA and their complemented strains were obtained from previous studies17,18. Strains GW4-ΔaioA/anoA and GW4-ΔaioA/anoA-C were obtained from this study and diagnostic PCR and DNA sequencing were used to confirm the successful deletion and complementation (data not shown). Based on our previous studies16 and the growth tests in this study, all of the strains showed consistent growth profiles in CDM medium containing 50 μM Sb(III) (Fig. 2), indicating that the Sb(III) oxidation was not affected by the growth of the strains under 50 μM Sb(III). Based on our previous results16, we calculated that the Sb(III) oxidation efficiency of strain GW4-ΔkatA (~52%) was increased by ~80% compared with the wild-type strain GW4 (~29%). The catalase KatA is responsible for cellular H2O2 consumption27, thus we proposed that the high efficient Sb(III) oxidation in strain GW4-ΔkatA might be associated with the cellular H2O2 content. Moreover, the GW4-ΔanoA showed a ~30% decrease in the Sb(III) oxidation efficiency (Fig. 2B), which is similar to our previous study18. In contrast, deletion of aioA had no effect on the Sb(III) oxidation efficiency during the log phase, while the Sb(III) oxidation efficiency was slightly increased during the stationary phase (Fig. 2D). It has been suggested that in the stationary phase of bacterial growth, other Sb(III) oxidation mechanism(s) might exist and function more efficiently in the absence of aioA. The simultaneous deletion of aioA and anoA resulted in a phenotype of Sb(III) oxidation efficiency between GW4-ΔaioA and GW4-ΔanoA (~19% decreased) (Fig. 2F), and all of the complemented strains showed a Sb(III) oxidation efficiency similar to that of the wild-type strain GW4 (Fig. 2).

Figure 2
Growth and Sb(III) oxidation curves of A. tumefaciens strains.

The katA, anoA and aioA influence each other and further affect Sb(III) oxidation

To elucidate how katA, anoA and aioA influence each other and further affect Sb(III) oxidation, we detected the transcription level of these genes in each A. tumefaciens strain. Bacterial cells were cultivated in CDM medium, and samples were collected after 0.5 h of induction with 50 μM Sb(III). In strain GW4-ΔkatA, the transcription levels of aioA and anoA were not increased (Fig. 3A), suggesting that the reduced consumption of H2O2 might be responsible for the efficient Sb(III) oxidation. In addition, the transcription levels of aioA and katA in strain GW4-ΔanoA showed no significant difference with the wild-type strain, consistent with the phenotype of decreased Sb(III) oxidation efficiency (Fig. 3B). In contrast, the transcription levels of anoA and katA were up-regulated in strain GW4-ΔaioA (p < 0.01), indicating that the AnoA- and H2O2-catalyzed Sb(III) oxidation was enhanced relative than the wild-type strain (Fig. 3C). Therefore, the deletion of aioA increased the Sb(III) oxidation efficiency in strain GW4. In strain GW4-ΔaioA/anoA, although the transcription level of the katA was increased (p < 0.01), the loss function of AnoA could not be compensated (Fig. 3D). Expectedly, all of the complemented strains recovered the phenotype back to the wild-type strain GW4 (Fig. 3).

Figure 3
Quantitative reverse transcriptase-PCR analysis of the genes associated with Sb(III) oxidation in A. tumefaciens strains.

The katA, anoA and aioA are involved in Sb(III) resistance

Subsequent efforts focused on the Sb(III) resistance of the A. tumefaciens strains. After 48 h incubation in CDM medium without Sb(III) supplementation, all of the strains exhibited a similar amount of viable cell counts (Fig. 4A). Moreover, there was no significant difference between the viable cell counts of the strains cultured with or without 50 μM Sb(III), indicating that 50 μM Sb(III) has no effect on bacterial growth. However, the deletion of katA significantly inhibited the growth of strain GW4 with the addition of 100 or 200 μM Sb(III) (p < 0.05 and p < 0.01, respectively), suggesting that the reduced H2O2 consumption might associated with bacterial Sb(III) resistance. In addition, the growth of strains GW4-ΔanoA, GW4-ΔaioA and GW4-ΔaioA/anoA were also obviously constrained (p < 0.05) relative to strain GW4 in the presence of 100 or 200 μM Sb(III), and the phenotypes of the complemented strains were recovered to the wild-type strain (Fig. 4A). These results indicated that both biotic and abiotic factors of Sb(III) oxidation were essential for bacterial Sb(III) resistance.

Figure 4
Sb(III) (A) and H2O2 (B) resistance of A. tumefaciens strains. Bacterial cells were inoculated into 100 mL of CDM medium with the addition of different concentrations of Sb(III) or H2O2 (0, 50, 100 and 200 μM). After 24 h ...

The katA is involved in H2O2 resistance

Plate counting assays were also performed to evaluate the antibacterial activities of H2O2 against A. tumefaciens strains using the same culture conditions with Sb(III) resistance. H2O2 is a substantial component of cellular oxidative stress with a toxic effect on different types of macromolecules31,32. The growth of strain GW4 was not affected by 50, 100 and 200 μM H2O2, suggesting that the oxidative stress response in strain GW4 is efficient for the detoxification of such concentrations of H2O2. For strains GW4-ΔaioA, GW4-ΔanoA and GW4-ΔaioA/anoA and their complemented strains, the viable cell counts were consistent with wild-type strain GW4, indicating that the deletion of biotic factors in strain GW4 did not affect bacterial H2O2 resistance (Fig. 4B). However, the growth of GW4-ΔkatA was significantly inhibited by 50, 100 and 200 μM H2O2 (p < 0.01), because such concentrations of H2O2 could not be efficiently consumed without decomposition through KatA. The phenotype of the complemented strain GW4-ΔkatA-C was recovered (Fig. 4B). The results demonstrated that katA is essential for bacterial H2O2 resistance.

Correlation between H2O2 content and Sb(III) oxidation both in vivo and in vitro

To understand the relationship between the cellular H2O2 content and Sb(III) oxidation, we examined the H2O2 content in A. tumefaciens strains with or without the induction of 50 μM Sb(III). The results indicated the following: i) The generation of cellular H2O2 was induced by Sb(III), ii) The cellular H2O2 content was consistent with the transcription level of genes associated with abiotic Sb(III) oxidation, and iii) The content of H2O2 was proportional to the bacterial Sb(III) oxidation efficiency. The strain with a higher H2O2 content showed a faster Sb(III) oxidation efficiency (Figs 2 and and5A).5A). In addition, we measured the dynamic changes in the cellular H2O2 content and Sb(V) generation in strains GW4, GW4-ΔkatA and GW4-ΔkatA-C from 24 to 48 h cultivation with the addition of 50 μM Sb(III). A significant decrease in the residual H2O2 content was observed with incubation time, while the Sb(V) concentration correspondingly increased (Fig. 5B,C), indicating that the consumed H2O2 might catalyze bacterial Sb(III) oxidation. Moreover, the H2O2 content and the increased Sb(V) concentration were significantly linearly correlated, with a correlation coefficient of 0.93 (Fig. 5D).

Figure 5
The cellular H2O2 content is correlated with bacterial Sb(III) oxidation.

The dynamic changes in the H2O2 content and Sb(V) generation were also measured in CDM medium with the addition of 50 μM Sb(III) and different concentrations of H2O2. Figure 5E showed that Sb(III) was transformed to Sb(V) with the addition of H2O2, indicating that H2O2 could oxidize Sb(III) to Sb(V) in vitro. In addition, there is a correlation between the H2O2 and Sb(V) contents in vitro (R2 = 0.94) (Fig. 5F). Based on the in vivo and in vitro analyses, we proposed that H2O2 is responsible for bacterial Sb(III) oxidation as an abiotic oxidant in strain GW4.

Comparison of Sb(III) oxidation between A. tumefaciens GW4 and A. tumefaciens 5A

Previous studies have shown that A. tumefaciens 5A, which has a 16S rRNA homology of 99% compared with A. tumefaciens GW4, also oxidizes Sb(III) to Sb(V)17,33. AioAB is responsible for Sb(III) oxidation in strain 5A, as the deletion of aioA decreased the Sb(III) oxidation efficiency, in contrast with the phenotype of strain GW4-ΔaioA17. To clarify the different effects of aioA on Sb(III) oxidation between strain GW4 and 5A, we also investigated the aioA mutant in strain 5A under the same culture conditions of strain GW4. The growth of strains GW4 and 5A were not affected by disruption of aioA in CDM medium supplemented with 50 μM Sb(III) (Fig. S2A,B). However, the Sb(III) oxidation efficiency was increased in strain GW4-ΔaioA (Fig. S2C), and the transcription of anoA and katA and the cellular H2O2 content were increased when aioA was deleted (Figs S3A and S4A). In contrast, the Sb(III) oxidation efficiency was decreased in strain 5A-ΔaioA (Fig. S2D), and no increased transcription of katA was observed, moreover, the transcription level of anoA was only slightly increased (Fig. S3B). Although Sb(III) also stimulated the generation of H2O2, this process was not affected by the disruption of aioA in strain 5A (Fig. S4B).


Currently, studies have shown that bacterial Sb(III) oxidation is catalyzed through AioAB or AnoA with a certain percentage of contribution17,18, indicating the existence of other new bacterial Sb(III) oxidation mechanisms. The present study documents a non-enzymatic basis for microbial Sb(III) oxidation, according to the following observations: i) The transcription of katA, sod1, sod2 and the cellular H2O2 content were induced by Sb(III); ii) the Sb(III) oxidation efficiency was consistent with the cellular H2O2 content in A. tumefaciens strains; and iii) The cellular H2O2 content in the katA mutant was remarkably linearly correlated with the Sb(V) concentration. Thus, we concluded that the cellular H2O2 acts as an abiotic factor in bacterial Sb(III) oxidation. The cellular H2O2 mediated bacterial Sb(III) oxidation is an smart detoxification process of Sb(III)-oxidizing bacteria through the “using poison against poison” strategy, which could transform the toxic Sb(III) to the much less toxic Sb(V) and consume the toxic cellular H2O2 simultaneously.

In addition, the Sb(III) resistance mechanism associated with bacterial oxidative stress has not been well clarified so far. It has been reported that H2O2 induces the death of E. coli, primarily reflecting DNA damage via the Fenton reaction34,35,36. Recently, the bacterial oxidative stress was found to associate with Sb(III) oxidation in Pseudomonas stutzeri TS4437. In this study, the deletion of katA significantly decreased H2O2 resistance, reflecting the disruption of the release of the oxidative stress response in strain GW4. A large number of literatures have shown that heavy metals (e.g. Cr, Cd)38,39, transition metals (e.g. Fe, Cu)23 and metalloid (As)40 could induce bacterial oxidative stress response due to their toxic effects. As a most common toxic heavy metal, the production of H2O2 could be the primary response of bacterial to Sb(III). It appears that the katA mutant is more tolerant to Sb(III) than H2O2 because Sb(III) oxidation consumed the cellular H2O2 even the katA was disrupted. In the presence of 50 μM Sb(III), the amount of H2O2 was consumed by Sb(III) oxidation and the cellular H2O2 was not toxic enough to inhibit bacterial growth. However, in the presence of high concentrations (e.g. 100 or 200 μM) of Sb(III), the high amount of H2O2 induced by Sb(III) in the katA mutant has a higher toxic effect on bacterial cells, even though the Sb(III) oxidation also consumed some of the H2O2. Thus, the Sb(III) resistant level in strain GW4-ΔkatA was still lower than the wild-type strain GW4.

To understand bacterial Sb(III) oxidation and the contribution of each cellular oxidative factor comprehensively, we also investigated the effects of biotic factors on Sb(III) oxidation in A. tumefaciens GW4. The deletion of anoA led to a ~ 30% decrease in the Sb(III) oxidation efficiency and the abiotic Sb(III) oxidation was not enhanced, indicating that the decreased Sb(III) oxidation efficiency reflected the contribution of AnoA. However, the deletion of aioA increased Sb(III) oxidation efficiency, reflecting the increased expression of AnoA and the generation of more H2O2. Thus, the contribution of AioAB to Sb(III) oxidation in strain GW4 was not obvious, however, AioAB is indeed related to Sb(III) oxidation, since it affects the expressions of AnoA and KatA. In addition, the results of a kinetic analysis in our previous study indicated that AnoA tends to catalyze the Sb(III) oxidation more efficiently than As(III) oxidation, while AioAB is prone to catalyze As(III) oxidation16. These results suggested that the effect of AnoA on Sb(III) oxidation may higher than that of AioAB. The contribution of H2O2-catalyzed abiotic Sb(III) oxidation may higher than that of enzymatic catalysis in strain GW4 since the disruption of KatA significantly increased Sb(III) oxidation efficiency16.

The results of a previous study demonstrated that AioAB was responsible for bacterial Sb(III) oxidation in A. tumefaciens 5A, suggesting that the effect of aioA on Sb(III) oxidation between strain 5A and GW4 was different. In strain 5A, AioAB has a substantial contribution to Sb(III) oxidation efficiency (approximately 25%)17. However, it appears that AioAB has no positive effect on Sb(III) oxidation in strain GW4, potentially reflecting the different characteristics between these two strains. The Sb(III) MIC of strain 5A is 0.3 mM, while strain GW4 is a highly Sb(III) resistance bacterium with a 8 mM MIC of Sb(III) (data not shown). In addition, strain 5A had a longer lag phase than that of strain GW4 with the addition of 50 μM Sb(III), indicating that Sb(III) might has a more toxic effect on strain 5A. Nonetheless, we cannot exclude the possibility that AioA might catalyze Sb(III) oxidation along with AnoA in strain GW4 because complex regulatory mechanism(s) might be involved in the compensation of the Sb(III) oxidation efficiency in the aioA mutant, which needs to be further studied.

In nature, H2O2 is a strong oxidant and it can oxidize not only Sb(III), but also other metalloids, such as As(III). However, the efficiency of bacterial As(III) oxidation catalyzed by H2O2 is not as obvious as Sb(III) oxidation and the abiotic As(III) oxidation could be hardly observed in vivo (data not shown). So far, bacterial As(III) oxidation has been found to be an enzymatic reaction which is primarily catalyzed through AioAB in most As(III)-oxidizing bacteria41. However, the mechanism of bacterial Sb(III) oxidation is a co-metabolism process catalyzed by AioAB, AnoA and H2O2, which is different from bacterial As(III) oxidation (at least in A. tumefaciens GW4). Although previous studies have shown that AnoA could also oxidize As(III) in vitro16, the expression of anoA was not induced by As(III)18, and the deletion of anoA did not affect the As(III) oxidation efficiency in strain GW4 (data not shown). Thus, the effect of AnoA on bacterial As(III) oxidation was hardly observed. In addition, H2O2 is an important and effective oxidant responsible for Sb(III) oxidation in alkaline aqueous environments42,43, and the Sb(III) oxidation rate is much faster than that of As(III)44,45,46. In the present study, the pH of the cultures increased from the initial 6.5 to approximately 8.0 following exposure to Sb(III), indicating that the culture conditions are suitable for H2O2 to catalyze Sb(III) oxidation (Fig. S3E,F). However, the culture pH decreased with the increasing incubation time during As(III) oxidation of strain GW4 (data not shown), and this pH might not be suitable for abiotic oxidation mediated through H2O2. Therefore, the abiotic oxidation is more effective on Sb(III) in strain GW4.

Based on the observations of the present study, we proposed that microbial Sb(III) oxidation is a co-metabolism process in strain GW4 (Fig. 6): (i) AioAB might be responsible for Sb(III) oxidation in the periplasm17; (ii) AnoA catalyzes cytoplasmic Sb(III) oxidation with NADP+ as a co-factor18; (iii) Sb(III) induces the bacterial oxidative stress response, leading to the production of ROS37 and H2O2; iv) the disruption of AioAB increases the expression of AnoA; (v) the disruption of AioAB increases the cellular H2O2 content and expression of KatA; (vi) the induced H2O2 oxidizes Sb(III) to Sb(V); and (vii) the redundant H2O2 is partially consumed by KatA.

Figure 6
A proposed model of bacterial Sb(III) oxidation based on the present study and literatures.

In summary, the present study provides novel evidences that microbial antimonite oxidation contains both abiotic and biotic mechanisms and elucidates the contribution of each oxidative factor. We show that Sb(III) causes oxidative stress to bacterial cells and further induces the generation of cellular H2O2. Sb(III) oxidation is a detoxification process by transforming the toxic Sb(III) to the much less toxic Sb(V). Meanwhile, since the cellular H2O2 is consumed by Sb(III) oxidation process, the Sb(III) oxidation also contributes to against the toxic H2O2. Such co-mechanism may be widely exist in other Sb(III)-oxidizing microorganisms. The relationship among the biotic and abiotic factors may be further studied by changing Sb(III) concentration, environmental and nutritious conditions.

Materials and Methods

Strains and genomic analysis

Bacterial strains and plasmids used in the present study are listed in Table S1. A. tumefaciens strains were grown in a chemically defined medium (CDM)47 containing 0 or 50 μM K2Sb2(C4H2O6)2 [Sb(III)] with aeration through shaking at 28 °C. E. coli strains were cultured at 37 °C in Luria-Bertani (LB) medium. When required, ampicillin (Amp, 100 mg/mL), kanamycin (Kan, 50 mg/mL), tetracycline (Tet, 5 mg/mL), gentamicin (Gen, 50 mg/mL) or chloromycetin (Cm, 50 mg/mL) were added. The genomic analyses of aioA, anoA, katA and sod were conducted through blastn and blastp in the genome of A. tumefaciens GW4 on the NCBI website (

Constructions A. tumefaciens GW4 mutant strains and complemented strains

An in-frame deletion in katA was constructed in strain GW4 using crossover PCR48. The primers used for construction of the deletion are listed in Table S2. The PCR products were double digested with BamHI and XbaI and subsequently cloned into pJQ200SK digested with the same restriction enzymes. The final construct pJQ-katA was mobilized into strain GW4 via conjugation with E. coli S17-1. Single-crossover mutants were identified on LB agar plates containing 100 μg/mL Amp and 50 μg/mL Gen, which were subsequently screened on CDM agar containing 20% sucrose49. SucroseR and GenSen transconjugants were screened using PCR and DNA sequencing to verify the katA deletion. For GW4-ΔaioA/anoA, an in-frame deletion in anoA was constructed in the mutant strain GW4-ΔaioA using the method described above. The GW4-ΔaioA and GW4-ΔanoA mutants and their complementary strains were obtained from previous works17,18.

The construction of GW4-ΔkatA complementation was accomplished using plasmid pCPP30. The complete katA coding region was PCR-cloned into BamHI - PstI double-digested pCPP30. The resulting plasmid pCPP30-katA was subsequently mobilized into strain GW4-ΔkatA via E. coli S17-1. TetR and AmpR transconjugants were screened on LB agar plates, yielding the complementary strain GW4-ΔkatA-C. The complementation of GW4-ΔaioA/anoA was performed using two plasmids, pCPP30 and pCT-Zori50. The aioAB genes along with the upstream RpoN binding site and the complete anoA coding region were PCR-cloned into BamHI - PstI double-digested pCT-Zori and pCPP30, respectively. The resulting plasmid pCT-Zori-aioAB and pCPP30-anoA were simultaneously mobilized into strain GW4-ΔaioA/anoA via E. coli S17-1. Subsequently, TetR, CmR and AmpR transconjugants were screened on LB agar plates. The complementary strains were verified through PCR and DNA sequencing.

Quantitative RT-PCR analysis

To investigate the expression of the genes associated with Sb(III) oxidation in A. tumefaciens strains, overnight cultures of these strains were each inoculated into 100 mL of CDM at 28 °C with 120 rpm shaking. When the OD600 reached 0.2–0.3, 0 or 50 μM Sb(III) was added to the cultures. After 0.5 h of induction, the bacterial cells were harvested for total RNA extraction using Trizol reagent (Invitrogen) and treated with RNase-free DNase I (Takara) according to the manufacturer’s instructions (Invitrogen, Grand Island, NY, USA). The quality and quantity of the RNA were monitored using a spectrophotometer (NanoDrop 2000, Thermo). Reverse transcription was performed using the RevertAid First Strand cDNA Synthesis Kit (Thermo) with 300 ng total RNA for each sample51. Subsequently, the obtained cDNA was diluted 10-fold and used as a template for further analysis. Quantitative RT-PCR was carried out by ABI VIIA7 in 0.1 mL Fast Optical 96-well Reaction Plate (ABI) using SYBR® Green Real-time PCR Master Mix (Toyobo) and the primers listed in Table S2. To eliminate error, three technical and biological replicates were established for each reaction. The A. tumefaciens GW4 16S rRNA gene was used as an internal control and the expression data of the genes were normalized to 16S rRNA without Sb(III) using the formula 2−(ΔCT−CT,16SrRNA,zero Sb), which was modified from the 2−ΔΔCT method52,53,54.

Growth, Sb(III) oxidation and sensitivity assays

A. tumefaciens strains were each inoculated into 5 mL of CDM with the addition of 50 μM Sb(III) and incubated at 28 °C with shaking at 120 rpm. When the OD600 reached 0.5–0.6, the strains were each inoculated into 100 mL of CDM in the presence of 50 μM Sb(III). Culture samples were collected every 8 h for measuring OD600 by spectrophotometry (DU800, Beckman). In addition, the samples were centrifuged (13,400 × g) and subsequently filtered (0.22 μm filter) to monitor the Sb(III)/Sb(V) concentrations through HPLC-HG-AFS (Beijing Titan Instruments Co., Ltd., China) according to Li et al.55.

To determine the Sb(III) and H2O2 resistance of A. tumefaciens strains, the viable plate counting method was employed. The strains were each inoculated into 100 mL of CDM medium with the addition of different concentrations of Sb(III) or H2O2 (0 μM, 50 μM, 100 μM and 200 μM) respectively, and incubated at 28 °C with shaking at 120 rpm. After cultivation for 48 h, the samples were collected for gradient dilution and spread onto solid LB medium, respectively. The plates were incubated at 28 °C and counted after 2–3 days until colonies formed.

H2O2 content and Sb(III) oxidation assays

To assess the H2O2 contents, A. tumefaciens strains were cultured as described above. After incubation for 0.5 h with 50 μM Sb(III), the bacterial cells (2 mL) were harvested through centrifugation (13,400 × g for 5 min at 4 °C) and washed twice with 50 mmol/L K3PO4 (pH 7.8). Subsequently, the cells were resuspended in 1 mL K3PO4 (pH 7.8) and sonicated on ice. The supernatants were obtained through centrifugation (13 400 × g, 10 min, 4 °C) to remove cell debris and subsequently mixed with 50 μL amplex red (AR) (Chemical Co., St. Louis, MO, USA) and 50 μL horseradish peroxidase (HRP) (F. Hoffmann-La Roche Ltd, Shanghai, China)56. After incubation at 37 °C for 15 min, fluorescence (530 ex/587 em) was measured using an EnVision® Multimode Plate Reader (Perkin Elmer).

To determine the dynamic variations in the H2O2 and Sb(V) contents in vivo, strains GW4, GW4-ΔkatA and GW4-ΔkatA-C were each inoculated into 100 mL of CDM medium supplemented with 50 μM Sb(III) and incubated at 28 °C for 48 h with shaking at 120 rpm. At designated times, the culture samples were collected to assess the H2O2 contents and monitor the Sb(V) contents as described above. For in vitro dynamic changes in the H2O2 and Sb(V) contents, at designated times, the measurement of Sb(V) concentrations was performed in CDM medium with the addition of 50 μM Sb(III) and different concentrations of H2O2 (5 μM, 10 μM, 15 μM and 20 μM).

Additional Information

How to cite this article: Li, J. et al. Abiotic and biotic factors responsible for antimonite oxidation in Agrobacterium tumefaciens GW4. Sci. Rep. 7, 43225; doi: 10.1038/srep43225 (2017).

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

Supplementary Material

Supplementary Information:


This work was financially supported through a grant from the National Natural Science Foundation of China (31470226) to G.W.


The authors declare no competing financial interests.

Author Contributions J.L. designed and performed the experiments and drafted the manuscript. B.Y., M.S., K.Y., W.G. and Q.W. performed the experiments. G.W. designed the study and revised the manuscript. All authors read and approved the final manuscript.


  • United States Environmental Protection Agency. Water related fate of the 129 priority pollutants, vol 1, EP-440/4-79029A. US EPA, Washington, DC (1979).
  • Council of the European Communities. Pollution Caused by Certain Dangerous Substances Discharged into the Aquatic Environment of the Community. 76/464/EEC 23−29 (1976).
  • Vásquez L. et al. . Pharmacokinetics of experimental pentavalent antimony after intramuscular administration in adult volunteers. Curr Ther Res Clin Exp 67, 193−203 (2006). [PMC free article] [PubMed]
  • Wilson S. C., Lockwood P. V., Ashley P. M. & Tighe M. The chemistry and behaviour of antimony in the soil environment with comparisons to arsenic: A critical review. Environ Poll 158, 1169−1181 (2010). [PubMed]
  • He M., Wang X., Wu F. & Fu Z. Antimony pollution in China. Sci Total Environ 421–422, 41−50 (2012). [PubMed]
  • Filella M., Belzile N. & Chen Y. W. Antimony in the environment: a review focused on natural waters. I. Occurrence. Earth-Sci Rev 57, 125−176 (2002a).
  • Filella M., Belzile N. & Lett M. C. Antimony in the environment: A review focused on natural waters. III. Microbiota relevant interactions. Earth-Sci Rev 80, 195−217 (2007).
  • Kruger M. C., Bertin P. N., Heipieper H. J. & Arsène-Ploetze F. Bacterial metabolism of environmental arsenic-mechanisms and biotechnological applications. Appl Microbiol Biotechnol 97, 3827−3841 (2013). [PubMed]
  • Filella M., Belzile N. & Chen Y. W. Antimony in the environment: a review focused on natural waters. II. Relevant solution chemistry. Earth-Sci Rev 59, 265−285 (2002b).
  • Stolz J. F., Basu P., Santini J. M. & Oremland R. S. Arsenic and selenium in microbial metabolism. Annu Rev Microbiol 60, 107−130 (2006). [PubMed]
  • Sanders O. I., Rensing C., Kuroda M., Mitra B. & Rosen B. P. Antimonite is accumulated by the glycerol facilitator GlpF in Escherichia coli. J. Bacteriol 179, 3365−3367 (1997). [PMC free article] [PubMed]
  • Wysocki R. et al. . The glycerol channel Fps1p mediates the uptake of arsenite and antimonite in Saccharomyces cerevisiae. Mol Microbiol 40, 1391−1401 (2001). [PubMed]
  • Meng Y. L., Liu Z. & Rosen B. P. As(III) and Sb(III) uptake by GlpF and efflux by ArsB in Escherichia coli. J Biol Chem 279, 18334−18341 (2004). [PubMed]
  • Rosen B. P. Biochemistry of arsenic detoxification. FEBS Lett 529, 86−92 (2002). [PubMed]
  • Kang Y. S., Shi Z., Bothner B., Wang G. & McDermott T. R. Involvement of the Acr3 and DctA anti-porters in arsenite oxidation in Agrobacterium tumefaciens 5A. Environ Microbiol 17, 1950−1962 (2015). [PubMed]
  • Li J. et al. . Microbial antimony biogeochemistry–enzymes, regulation and related metabolic pathways. Appl Environ Microbiol 82, 5482−5495 (2016). [PMC free article] [PubMed]
  • Wang Q. et al. . Arsenite oxidase also functions as an antimonite oxidase. Appl Environ Microbiol 81, 1959−1965 (2015). [PMC free article] [PubMed]
  • Li J. et al. . Proteomics and genetics for identification of a bacterial antimonite oxidase in Agrobacterium tumefaciens. Environ Sci Technol 49, 5980−5989 (2015). [PubMed]
  • Belzile N., Chen Y. W. & Wang Z. Oxidation of antimony (III) by amorphous iron and manganese oxyhydroxides. Chemical Geology 174, 379−387 (2001).
  • Leuz A. K. & Johnsona C. A. Oxidation of Sb(III) to Sb(V) by O2 and H2O2 in aqueous solutions. Geochimica et Cosmochimica Acta 69, 1165−1172 (2005).
  • Buschmann J. S., Canonica S. & Sigg L. Photo induced oxidation of antimony(III) in the presence of humic acids. Environ Sci Technol 39, 5335−5341 (2005). [PubMed]
  • Quentel F., Filella M., Elleouet C. & Madec C. L. Sb(III) oxidation by iodate in seawater: A cautionary tale. Sci Total Environ 355, 259−263 (2006). [PubMed]
  • Ma J. F. et al. . Bacterioferritin a modulate catalase A (KatA) activity and resistance to hydrogen peroxide in Pseudomonas aeruginosa. J Bacteriol 12, 3730−3742 (1999). [PMC free article] [PubMed]
  • Cabiscol E., Tamarit J. & Ros J. Oxidative stress in bacteria and protein damage by reactive oxygen species. Int Microbiol 3, 3−8 (2000). [PubMed]
  • Cornelis P., Wei Q., Andrews S. C. & Vinckx T. Iron homeostasis and management of oxidative stress response in bacteria. Metallomics 3, 540−549 (2011). [PubMed]
  • Yoon J. H., Lee M. S. & Kang J. H. Reaction of ferritin with hydrogen peroxide induces lipid peroxidation. BMB Rep 43, 219−224 (2010). [PubMed]
  • Hassett D. J., Charniga L., Bean K. A., Ohman D. E. & Cohen M. S. Antioxidant defense mechanisms in Pseudomonas aeruginosa: resistance to the redox-active antibiotic pyocyanin and demonstration of a manganese-cofactored superoxide dismutase. Infect Immun 60, 328−336 (1992). [PMC free article] [PubMed]
  • Fan H. et al. . Sedimentary arsenite-oxidizing and arsenate-reducing bacteria associated with high arsenic groundwater from Shanyin, Northwestern China. J Appl Microbiol 105, 529−539 (2008). [PubMed]
  • Wood D. W. et al. . The genome of the natural genetic engineer Agrobacterium tumefaciens C58. Science 294, 2317−2323 (2001). [PubMed]
  • Wang Q. et al. . Fate of arsenate following arsenite oxidation in Agrobacterium tumefaciens GW4. Environ Microbiol 17, 1926−1940 (2015). [PubMed]
  • Henkler F., Brinkmann J. & Luch A. The role of oxidative stress in carcinogenesis induced by metals and xenobiotics. Cancers 2, 376–396 (2010). [PMC free article] [PubMed]
  • Bayr H. Reactive oxygen species. Crit Care Med 33, 498−501 (2005). [PubMed]
  • Lehr C. R., Kashyap D. R. & McDermott T. R. New insights into microbial oxidation of antimony and arsenic. Appl Environ Microbiol 73, 2386−2389 (2007). [PMC free article] [PubMed]
  • Imlay J. A. & Linn S. Bimodal pattern of killing of DNA-repairdefective or anoxically grown Escherichia coli by hydrogen peroxide. J Bacteriol 166, 519–527 (1986). [PMC free article] [PubMed]
  • Imlay J. A., Chin S. M. & Linn S. Toxic DNA damage by hydrogen peroxide through the Fenton reaction in vivo and in vitro. Science 240, 640–642 (1988). [PubMed]
  • Park S. & Imlay J. A. High levels of intracellular cysteine promote oxidative DNA damage by driving the fenton reaction. J Bacteriol 185, 1942–1950 (2003). [PMC free article] [PubMed]
  • Wang D. et al. . Disrupting ROS-protection mechanism allows hydrogen peroxide to accumulate and oxidize Sb(III) to Sb(V) In Pseudomonas stutzeri TS44. BMC Microbiol 16, 279 (2016). [PMC free article] [PubMed]
  • Bagchi D., Stohs S. J., Downs B. W., Bagchi M. & Preuss H. G. Cytotoxicity and oxidative mechanisms of different forms of chromium. Toxicology 180, 5–22 (2002). [PubMed]
  • Jia H. et al. . Hydrogen sulfide–cysteine cycle system enhances cadmium tolerance through alleviating cadmium-induced oxidative stress and ion toxicity in Arabidopsis roots. Sci Rep 6, 39702 (2016). [PMC free article] [PubMed]
  • Weiss S. et al. . Enhanced structural and functional genome elucidation of the arsenite-oxidizing strain Herminiimonas arsenicoxydans by proteomics data. Biochimie 91, 192−203 (2009). [PubMed]
  • Silver S. & Phung L. T. Genes and enzymes involved in bacterial oxidation and reduction of inorganic arsenic. Appl Environ Microbiol 71, 599−608 (2005). [PMC free article] [PubMed]
  • Leuz A. K., Hug S. J., Wehrli B. & Johnson C. A. Iron-mediated oxidation of antimony(III) by oxygen and hydrogen peroxide compared to arsenic(III) oxidation. Environ Sci Technol 40, 2565−2571 (2006). [PubMed]
  • Kong L., Hu X. & He M. Mechanisms of Sb(III) oxidation by pyrite-induced hydroxyl radicals and hydrogen peroxide. Environ Sci Technol 49, 3499−505 (2015). [PubMed]
  • Pettine M., Campanella L. & Millero F. J. Arsenite oxidation by H2O2 in aqueous solutions. Geochim Cosmochim Acta 63, 2727−2735 (1999).
  • Quentel F., Filella M., Elleouet C. & Madec C. L. Kinetic studies on Sb(III) oxidation by hydrogen peroxide in aqueous solution. Environ Sci Technol 38, 2843−2848 (2004). [PubMed]
  • Kim D. H., Bokare A. D., Koo Ms. & Choi W. Heterogeneous Catalytic Oxidation of As(III) on Nonferrous Metal Oxides in the Presence of H2O2. Environ Sci Technol 49, 3506−3513 (2015). [PubMed]
  • Weeger W. et al. . Oxidation of arsenite to arsenate by a bacterium isolated from an aquatic environment. Biometals 12, 141−149 (1999). [PubMed]
  • Link A. J., Phillips D. & Church G. M. Methods for generating precise deletions and insertions in the genome of wild-type Escherichia coli: Application to open reading frame characterization. J Bacteriol 179, 6228−6237 (1997). [PMC free article] [PubMed]
  • Pelicic V., Reyrat J. M. & Gicquel B. Expression of the Bacillus subtilis sacB gene confers sucrose sensitivity on mycobacteria. J Bacteriol 178, 1197−1199 (1996). [PMC free article] [PubMed]
  • Chen F. et al. . Regulation of arsenite oxidation by the phosphate two-component system PhoBR in Halomonas sp. HAL1. Front Microbiol 6, 923 (2015). [PMC free article] [PubMed]
  • Wang Q., Lei Y., Xu X., Wang G. & Chen L. Theoretical prediction and experimental verification of protein-coding genes in plant pathogen genome Agrobacterium tumefaciens strain C58. PLoS One 7, 43176, (2012).10.1371/journal.pone.0043176 [PMC free article] [PubMed] [Cross Ref]
  • López-Contreras A. J., Ramos-Molina B., Cremades A. & Peñafiel R. Antizyme inhibitor 2 (AZIN2/ODCp) stimulates polyamine uptake in mammalian cells. J Biol Chem 283, 20761–20769 (2008). [PMC free article] [PubMed]
  • Moorthi A. et al. . Expression of microRNA-30c and its target genes in human osteoblastic cells by nano-bioglass ceramic-treatment. Int J Biol Macromol 56, 181–185 (2013). [PMC free article] [PubMed]
  • Livak K. J. & Schmittgen T. D. Analysis of relative gene expression data using real-time quantitative PCR and the 2−ΔΔCT Method. Methods 25, 402–408 (2001). [PubMed]
  • Li J., Wang Q., Zhang S., Qin D. & Wang G. Phylogenetic and genome analyses of antimony-oxidizing bacteria isolated from antimony mined soil. Int Biodet Biodegr 76, 76–80 (2013).
  • Seaver L. C. & Imlay J. A. Alkyl hydroperoxide reductase is the primary scavenger of endogenous hydrogen peroxide in Escherichia coli. J Bacteriol 183, 7173–7181 (2001). [PMC free article] [PubMed]

Articles from Scientific Reports are provided here courtesy of Nature Publishing Group