Surface seawater was collected from Shin-Maiko Beach (Futsu, Chiba, Japan). An enrichment culture was prepared by inoculating 1 mL of seawater into 18 mL of minimal medium containing NH₄Cl (0.535 g), KH₂PO₄ (0.136 g), MgCl₂·6H₂O (0.204 g), CaCl₂·2H₂O (0.147 g), NaCl (20 g), trace mineral element solution (5 mL), vitamin solution (10 mL), and NaHCO₃ (2.52 g) per liter. The trace mineral element solution was based on that of DSM 318 medium (DSMZ, 1993), and the vitamin solution was based on that of DSM 141 medium (DSMZ, 1993). The minimal medium used in this study contained 25 μM of Fe(III) as FeCl₃·6H₂O instead of FeCl₂·6H₂O in the original DSM 318 medium. For anaerobic incubation, the minimal medium was dispensed into a 60-mL serum bottle under an N₂/CO₂ (80:20) gas stream. The bottle was sealed with a thick butyl rubber stopper and aluminum cap. After autoclaving at 121 °C for 20 min, sodium acetate and sodium bromate were added separately from sterile anaerobic stock solutions to achieve final concentrations of 20 mM and 250 μM, respectively. The final pH of the medium was 6.8–7.0. The bottles were subsequently incubated at 30 °C without shaking. Under microaerobic conditions, incubation was performed similarly to that under anaerobic conditions in a sealed serum bottle, but air substitution of the headspace and liquid phase with N₂/CO₂ gas was omitted. For microaerobic incubation, 0.174 g L⁻¹ of K₂HPO₄ was added to the minimal medium, but NaHCO₃ was excluded. The pH of the medium for microaerobic incubation was adjusted to 7.0 with NaOH. Under microaerobic conditions, oxygen in the medium is gradually consumed by bacteria, eventually resulting in anaerobic conditions (Tamai et al., 2016).

To isolate bromate-reducing bacteria, the enrichment culture sub-cultured five times under microaerobic conditions was serially diluted and spread on a solid minimal medium containing sodium acetate and 15 g L⁻¹ of agar. In some cases, Luria–Bertani (LB) agar medium was used instead of a solid minimal medium. After aerobic incubation at 30 °C, bacterial colonies were randomly selected and purified. Subsequently, their bromate-reducing capacity was evaluated using a liquid minimal medium under microaerobic conditions. Specifically, 25 isolates were cultured with 250 μM bromate for 3 to 30 days under microaerobic conditions and bromate concentration in the medium was determined.

Genomic DNA of the isolated bacteria was extracted as previously described (Hiraishi, 1992). The 16S rRNA gene was amplified via polymerase chain reaction (PCR) using the bacterial consensus primers, 10F (5′-AGAGTTTGATCCTGGCTCAG-3′) and 1500R (5′-GGTTACCTTGTTACGACTT-3′). PCR products were purified using the QIAquick PCR Purification Kit (Qiagen, Hilden, Germany) and sequenced using the BigDye Terminator Cycle Sequencing Kit (Applied Biosystems, Foster City, CA, USA) and ABI Prism 3100 Genetic Analyzer (Applied Biosystems) using appropriate sequencing primers (Weisburg et al., 1991). The obtained 16S rRNA gene sequences were subjected to a Basic Local Alignment Search Tool search¹ to determine the sequence similarity. The retrieved sequences were aligned using ClustalX version 2.0 (Larkin et al., 2007). Finally, a phylogenetic tree was constructed using the neighbor-joining method with the MEGA11 software package (Kumar et al., 2016).

Strain M-Br was first cultured aerobically in cultured in LB liquid medium. Cells were collected and washed twice with 20 mM Tris–HCl buffer (pH 7.0). The washed cells were then inoculated into a minimal medium containing 250 μM bromate, and cultured under anaerobic, microaerobic, and aerobic growth conditions. Sodium lactate was added as the electron donor and carbon source at a final concentration of 20 mM. Aerobic incubation was performed similarly to microaerobic incubation, except that the medium (20 mL) was dispensed into a 100-mL Erlenmeyer flask capped with a silicone plug, and the flask was incubated with shaking at 180 rpm.

To determine the effect of Fe(III) on bromate reduction, strain M-Br was grown in a minimal medium containing 0–100 μM FeCl₃·6H₂O under microaerobic conditions. To control the final Fe(III) concentration in the medium, Fe(III) in the trace mineral element solution was excluded from this experiment.

S. putrefaciens CN-32 and S. oneidensis MR-1 were purchased from the American Type Culture Collection (ATCC BAA-453) and Japan Collection of Microorganisms (JCM 31522), respectively. Strain M-Br and these Shewanella spp. were grown in the minimum medium for 3 days under microaerobic conditions. The cells were collected, washed twice with 20 mM Tris–HCl buffer (pH 7.0), and resuspended in the same buffer to achieve an optical density at 600 nm of 0.5. Approximately 15 mL of the washed cell suspension was dispensed into a 100-mL Erlenmeyer flask or 60-mL serum bottle. Lactate, bromate, FeCl₃·6H₂O, and nitrilotriacetic acid (chelating reagent) were added to the suspension at final concentrations of 10 mM, 50 μM, 25 μM, and 335 μM, respectively. The flask was capped with a silicone plug and incubated with shaking at 180 rpm. In contrast, the serum bottle was degassed with N₂ gas, sealed with a butyl rubber stopper and an aluminum cap, and anaerobically incubated without shaking.

To determine the effects of endogenous and exogenous Fe(III) on bromate reduction, washed cell suspensions were prepared from the cells grown with or without Fe(III) and supplemented with lactate, bromate, FeCl₃·6H₂O, and nitrilotriacetic acid, as described above. In some cases, FeCl₃·6H₂O was not added to the cell suspensions. Then, the cell suspensions were anaerobically incubated in serum bottles, as described above.

To determine if other iron-reducing bacteria can reduce bromate, Geobacter sp. OR-1 (Ohtsuka et al., 2013) and Anaeromyxobacter sp. PSR-1 (Kudo et al., 2013) were used. They were grown anaerobically in the minimal medium containing 20 mM acetate and 20 mM Fe(III) as the electron donor and acceptor, respectively. Bromate was also added at a final concentration of 250 μM. In other cases, these strains were grown anaerobically with 20 mM acetate and 20 mM fumarate, which serves as an effective electron acceptor comparable to iron.

Bromate concentration was spectrophotometrically determined, as previously described (Tamai et al., 2016). Bromide was determined by the IC-2010 ion chromatography system (Tosoh, Tokyo, Japan) with the TSKgel SuperIC-Anion HR column (Tosoh) connected to the TSKgel SuperIC-A HS guard column (Tosoh). The mobile phase consisted of 2.2 mM NaHCO₃ and 2.7 mM Na₂CO₃ at a flow rate of 1.0 mL min⁻¹, and the column temperature was maintained at 40 °C.

Seawater was inoculated into a minimal medium and incubated with 250 μM bromate under anaerobic and microaerobic conditions (Supplementary Figure 1). Under anaerobic conditions, few bromate was reduced, even after incubation for 50 d. However, bromate was completely reduced in 14 d under microaerobic conditions. After five sub-cultures, bromate reduction was completed in 5 days. From this enrichment culture, multiple bacteria were randomly isolated and evaluated for their bromate-reducing capacity. Most isolated strains reduced bromate; however, their bromate reduction rates were very slow (data not shown). Among the tested strains, one exhibited the fastest bromate reduction rate, completely reducing 250 μM bromate in 4 days. This strain was selected as a novel bromate-reducing bacterium and designated as the strain M-Br.

Phylogenetic analysis via 16S rRNA gene sequencing revealed that the strain M-Br (NCBI accession number: LC900506) was closely related to the S. putrefaciens W3-18-1 and S. putrefaciens CN-32, with a sequence similarity of over 99% (Supplementary Figure 2). Other related bacteria included S. oneidensis MR-1, a representative dissimilatory Fe(III)-reducing bacterium.

M-Br was grown with bromate under aerobic, microaerobic, and anaerobic conditions, and its growth and bromate reduction rate were analyzed (Figure 1). Lactate was used instead of acetate as the electron donor and carbon source because it is generally the preferred substrate for the growth of Shewanella species under aerobic and anaerobic conditions (Zhang et al., 2021). As shown in Figure 1A, M-Br grew well but only reduced a small amount of bromate under aerobic conditions. Notably, M-Br grew better under microaerobic conditions than under aerobic conditions, completely reducing 250 μM bromate in 4 days (Figure 1B). Interestingly, the strain did not grow under anaerobic conditions, but it reduced approximately half of the bromate in 8 days (Figure 1C).

Bromate reduction was much faster under microaerobic conditions than under aerobic and anaerobic conditions, suggesting that the strain M-Br reduces bromate preferably under oxygen-limited conditions but grows well under oxygen-rich conditions. To verify this hypothesis, a washed cell suspension of M-Br was prepared, and bromate reduction was examined under aerobic and anaerobic conditions (Figure 2). Indeed, no bromate was reduced under aerobic conditions, whereas 50 μM bromate was almost completely reduced in 3 h under anaerobic conditions.

We previously reported that Rhodococcus sp. Br-6 uses Fe(III) as a redox mediator for bromate reduction (Tamai et al., 2016). As Shewanella species are well-known dissimilatory Fe(III)-reducing bacteria, strain M-Br possibly also uses Fe(III) as a redox mediator for bromate reduction. To verify this, a minimal medium lacking Fe(III) was prepared, and M-Br was grown with or without 1–100 μM Fe(III) under microaerobic conditions. As shown in Figure 3A, M-Br did not reduce bromate in the absence of Fe(III). In contrast, Fe(III) addition accelerated bromate reduction in a dose-dependent manner. As shown in Figure 3B, bromate reduction rate (μM day⁻¹) of M-Br was positively correlated with the Fe(III) concentration in the medium.

Accelerated bromate reduction by Fe(III) is due to two possible reasons. First, Fe(III) functions as a redox mediator for bromate reduction, as in Rhodococcus sp. Br-6 (Tamai et al., 2016). Second, iron-associated proteins, such as c-type cytochromes, are involved in bromate reduction. To determine the specific mechanism, M-Br was pre-grown with or without Fe(III), and washed cells were prepared. The cells were incubated with bromate with or without exogenous Fe(III). As shown in Figure 4, bromate was reduced only when M-Br was pre-grown in the presence of Fe(III) and its washed cells were incubated with exogenous Fe(III). However, no significant bromate reduction was observed when M-Br was pre-grown in the absence of Fe(III) and its cells were incubated without exogenous Fe(III).

To determine whether Fe(III)-dependent bromate reduction is a general feature of Shewanella species, similar experiments were performed using S. putrefaciens CN-32 and S. oneidensis MR-1. As shown in Figure 5, bromate reduction by both bacteria was also Fe(III)-dependent, i.e., bromate reduction occurred only when the strains were pre-grown with Fe(III) and their washed cells were incubated with exogenous Fe(III).

To determine if other iron-reducing bacteria can reduce bromate, Geobacter sp. OR-1 (Ohtsuka et al., 2013) and Anaeromyxobacter sp. PSR-1 (Kudo et al., 2013) were grown anaerobically in the minimum medium containing 20 mM acetate and 20 mM Fe(III). Bromate was also added at a final concentration of 250 μM. However, due to the yellow color of the iron-containing culture supernatants, colorimetric quantification of bromate was difficult. In contrast, ion chromatographic analysis revealed that almost no bromide was produced (only up to 3 μM), indicating that bromate reduction had not proceeded. We next cultivated the strains anaerobically with 20 mM acetate and 20 mM fumarate [plus 25 μM Fe(III)], which serves as an effective electron acceptor comparable to iron. In the absence of bromate, both strains exhibited good growth, with OD₆₀₀ values ranging from 0.11 to 0.16 (Figures 6A,B). However, in the presence of 250 μM bromate, growth increased only up to a maximum OD₆₀₀ of 0.02. Quantification of bromate during this incubation showed that although a slight reduction of bromate occurred at the early stage of cultivation, no further reduction was observed thereafter (Figure 6C).