Expression of mercuric ion reductase in Eastern cottonwood (Populus deltoides) confers mercuric ion reduction and resistance

Mercury is one of the most hazardous heavy metals and is a particular problem in aquatic ecosystems, where organic mercury is biomagnified in the food chain. Previous studies demonstrated that transgenic model plants expressing a modified mercuric ion reductase gene from bacteria could detoxify mercury by converting the more toxic and reductive ionic form [Hg(II)] to less toxic elemental mercury [Hg(0)]. To further investigate if a genetic engineering approach for mercury phytoremediation can be effective in trees with a greater potential in riparian ecosystems, we generated transgenic Eastern cottonwood (Populus deltoides) trees expressing modified merA9 and merA18 genes. Leaf sections from transgenic plantlets produced adventitious shoots in the presence of 50 microm Hg(II) supplied as HgCl2, which inhibited shoot induction from leaf explants of wild-type plantlets. Transgenic shoots cultured in a medium containing 25 microm Hg(II) showed normal growth and rooted, while wild-type shoots were killed. When the transgenic cottonwood plantlets were exposed to Hg(II), they evolved 2-4-fold the amount of Hg(0) relative to wild-type plantlets. Transgenic merA9 and merA18 plants accumulated significantly higher biomass than control plants on a Georgia Piedmont soil contaminated with 40 p.p.m. Hg(II). Our results indicate that Eastern cottonwood plants expressing the bacterial mercuric ion reductase gene have potential as candidates for in situ remediation of mercury-contaminated soils or wastewater.     

Biotoxicity of mercury as influenced by mercury(II) speciation

Integration of physicochemical procedures for studying mercury(II) speciation with microbiological procedures for studying the effects of mercury on bacterial growth allows evaluation of ionic factors (e.g., pH and ligand species and concentration) which affect biotoxicity. A Pseudomonas fluorescens strain capable of methylating inorganic Hg(II) was isolated from sediment samples collected at Buffalo Pound Lake in Saskatchewan, Canada. The effect of pH and ligand species on the toxic response (i.e., 50% inhibitory concentration [IC50]) of the P. fluorescens isolated to mercury were determined and related to the aqueous speciation of Hg(II). It was determined that the toxicities of different mercury salts were influenced by the nature of the co-ion. At a given pH level, mercuric acetate and mercuric nitrate yielded essentially the same IC50s; mercuric chloride, on the other hand, always produced lower IC50s. For each Hg salt, toxicity was greatest at pH 6.0 and decreased significantly (P = 0.05) at pH 7.0. Increasing the pH to 8.0 had no effect on the toxicity of mercuric acetate or mercuric nitrate but significantly (P = 0.05) reduced the toxicity of mercuric chloride. The aqueous speciation of Hg(II) in the synthetic growth medium M-IIY (a minimal salts medium amended to contain 0.1% yeast extract and 0.1% glycerol) was calculated by using the computer program GEOCHEM-PC with a modified data base. Results of the speciation calculations indicated that complexes of Hg(II) with histidine [Hg(H-HIS)HIS+ and Hg(H-HIS)2(2+)], chloride (HgCl+, HgCl2(0), HgClOH0, and HgCl3-), phosphate (HgHPO4(0), ammonia (HgNH3(2+), glycine [Hg(GLY)+], alanine [Hg(ALA)+], and hydroxyl ion (HgOH+) were the Hg species primarily responsible for toxicity in the M-IIY medium. The toxicity of mercuric nitrate at pH 8.0 was unaffected by the addition of citrate, enhanced by the addition of chloride, and reduced by the addition of cysteine. In the chloride-amended system, HgCl+, HgCl2(0), and HgClOH0 were the species primarily responsible for observed increases in toxicity. In the cysteine-amended system, formation of Hg(CYS)2(2-) was responsible for detoxification effects that were observed. The formation of Hg-citrate complexes was insignificant and had no effect on Hg toxicity.     

Biotoxicity of mercury as influenced by mercury(II) speciation.

Integration of physicochemical procedures for studying mercury(II) speciation with microbiological procedures for studying the effects of mercury on bacterial growth allows evaluation of ionic factors (e.g., pH and ligand species and concentration) which affect biotoxicity. A Pseudomonas fluorescens strain capable of methylating inorganic Hg(II) was isolated from sediment samples collected at Buffalo Pound Lake in Saskatchewan, Canada. The effect of pH and ligand species on the toxic response (i.e., 50% inhibitory concentration [IC50]) of the P. fluorescens isolated to mercury were determined and related to the aqueous speciation of Hg(II). It was determined that the toxicities of different mercury salts were influenced by the nature of the co-ion. At a given pH level, mercuric acetate and mercuric nitrate yielded essentially the same IC50s; mercuric chloride, on the other hand, always produced lower IC50s. For each Hg salt, toxicity was greatest at pH 6.0 and decreased significantly (P = 0.05) at pH 7.0. Increasing the pH to 8.0 had no effect on the toxicity of mercuric acetate or mercuric nitrate but significantly (P = 0.05) reduced the toxicity of mercuric chloride. The aqueous speciation of Hg(II) in the synthetic growth medium M-IIY (a minimal salts medium amended to contain 0.1% yeast extract and 0.1% glycerol) was calculated by using the computer program GEOCHEM-PC with a modified data base. Results of the speciation calculations indicated that complexes of Hg(II) with histidine [Hg(H-HIS)HIS+ and Hg(H-HIS)2(2+)], chloride (HgCl+, HgCl2(0), HgClOH0, and HgCl3-), phosphate (HgHPO4(0), ammonia (HgNH3(2+), glycine [Hg(GLY)+], alanine [Hg(ALA)+], and hydroxyl ion (HgOH+) were the Hg species primarily responsible for toxicity in the M-IIY medium. The toxicity of mercuric nitrate at pH 8.0 was unaffected by the addition of citrate, enhanced by the addition of chloride, and reduced by the addition of cysteine. In the chloride-amended system, HgCl+, HgCl2(0), and HgClOH0 were the species primarily responsible for observed increases in toxicity. In the cysteine-amended system, formation of Hg(CYS)2(2-) was responsible for detoxification effects that were observed. The formation of Hg-citrate complexes was insignificant and had no effect on Hg toxicity.     

Reduction of Hg(II) to Hg(0) by Biogenic Magnetite from two Magnetotactic Bacteria

Understanding the biogeochemical cycle of the highly toxic element mercury (Hg) is necessary to predict its fate and transport. In this study, we determined that biogenic magnetite isolated from Magnetospirillum gryphiswaldense MSR-1 and Magnetospirillum magnetotacticum MS-1 was capable of reducing inorganic mercury [Hg(II)] to elemental mercury [Hg(0)]. These two magnetotactic bacteria (MTB) lacked mercuric resistance operons in the genomes. However, they revealed high resistance to Hg(II) under atmospheric conditions and an even higher resistance under microaerobic conditions (1% O₂ and 99% N₂). Neither strain reduced Hg(II) to Hg(0) under atmospheric conditions. However, a slow rate (0.05–0.21 µM·d^?1) of Hg(II) loss occurred from late log phase to stationary phase in two MTBs' culture media under microaerobic conditions. Increased Hg(II) entered both cells under microaerobic conditions relative to atmospheric conditions. The majority of Hg(II) was still blocked by the cell membrane. Hg(II) reduction was more effective when biogenic magnetite was extracted out, with or without the magnetosome membrane envelope. When magnetosome membrane was present, 8.55–13.53% of 250 nM Hg(II) was reduced to Hg(0) by 250 mg/L biogenic magnetite suspension within 2 hours. This ratio increased to 55.07–64.70% while magnetosome membrane was removed. We concluded that two MTBs contributed to the reduction of Hg(II) to Hg(0) at a slow rate in vivo. Such reduction was more favorable to occur when biogenic magnetite is released from dead cells. It proposed a new biotic pathway for the formation of Hg(0) in aquatic systems.     

Analysis of mercuric reductase (merA) gene diversity in an anaerobic mercury-contaminated sediment enrichment

The reduction of ionic mercury to elemental mercury by the mercuric reductase (MerA) enzyme plays an important role in the biogeochemical cycling of mercury in contaminated environments by partitioning mercury to the atmosphere. This activity, common in aerobic environments, has rarely been examined in anoxic sediments where production of highly toxic methylmercury occurs. Novel degenerate PCR primers were developed which span the known diversity of merA genes in Gram-negative bacteria and amplify a 285 bp fragment at the 3' end of merA. These primers were used to create a clone library and to analyse merA diversity in an anaerobic sediment enrichment collected from a mercury-contaminated site in the Meadowlands, New Jersey. A total of 174 sequences were analysed, representing 71 merA phylotypes and four novel MerA clades. This first examination of merA diversity in anoxic environments suggests an untapped resource for novel merA sequences.     

Cloning, sequencing and expression of the glutamine synthetase structural gene (glnA) from the obligate methanotroph Methylococcus capsulatus (Bath)

The structural gene (glnA) encoding the ammonia-assimulation enzyme glutamine synthetase (GS) has been cloned from the obligate methanotroph Methylococcus capsulatus (Bath). Complementation of Escherichia coli glnA mutants was demonstrated. In vitro expression analysis revealed that the cloned glnA gene coded for a polypeptide of apparent Mr 60,000, as determined by PAGE. Expression of the M. capsulatus (Bath) glnA gene in E. coli was regulated by nitrogen levels in an Ntr+ but not an Ntr- background. The nucleotide sequence of the M. capsulatus (Bath) glnA gene and flanking sequences was determined. This gene, of 1407 bp, encoded a polypeptide of Mr 51717 containing 468 amino acids. The 5' leader region contained three putative promoters. Promoters P1 and P3 resembled the canonical -10 -35 E. coli-type promoter. Promoter P2, which was located between P1 and P3, resembled the NtrA-dependent promoters of enteric organisms. A potential NtrC-binding site was also determined, flanking the Pribnow box at P1. Comparisons of nucleotide-derived amino acid sequences of GS enzymes from prokaryotes and eukaryotes with GS from M. capsulatus are made.     

Transformation of peanut using a modified bacterial mercuric ion reductase gene driven by an actin promoter from Arabidopsis thaliana

In order to test an alternative selectable marker system for the production of transgenic peanut plants (Arachis hypogaea), the bacterial mercuric ion reductase gene, merA, was introduced into embryogenic cultures via microprojectile bombardment. MerA reduces toxic Hg(II) to the volatile and less toxic metallic mercury molecule, Hg(0), and renders its source Gram-negative bacterium mercury resistant. A codon-modified version of the merA gene, MerApe9, was cloned into a plant expression cassette containing the ACT2 promoter from Arabidopsis thaliana and the NOS terminator. The expression cassette also was inserted into a second vector containing the hygromycin resistance gene driven by the UBI3 promoter from potato. Stable transgenic plants were recovered through hygromycin-based selection from somatic embryo tissues bombarded with the plasmid containing both genes. However, no transgenic somatic embryos were recovered from selection on 50-100 micromol/L HgCl2. Expression of merA as mRNA was detected by Northern blot analysis in leaf tissues of transgenic peanut, but not in somatic embryos. Western blot analysis showed the production of the mercuric ion reductase protein in leaf tissues. Differential responses to HgCl2 of embryo-derived explants from segregating R1 seeds of one transgenic line also were observed.     

Mercuric reductase. Purification and characterization of a transposon-encoded flavoprotein containing an oxidation-reduction-active disulfide

The flavoprotein mercuric reductase catalyzes the two-electron reduction of mercuric ions to elemental mercury using NADPH as an electron donor. It has now been purified from Pseudomonas aeruginosa PAO9501 carrying the plasmid pVS1. In this plasmid system, where the mer operon is on the transposon Tn501, mercuric reductase comprises up to 6% of the soluble cellular protein upon induction with mercurials. The purification is a rapid (two-step), high yield (80%) procedure. Anaerobic titrations of mercuric reductase with dithionite revealed the formation of a charge transfer complex with an absorbance maximum around 540 nm. Striking spectroscopic similarities to lipoamide dehydrogenase and glutathione reductase were observed. These two enzymes, which catalyze the transfer of electrons between pyridine nucleotides and disulfides, are flavoproteins which contain an oxidation-reduction-active cysteine residue at the active site. The expectation that mercuric reductase contains a similar electron acceptor was confirmed when it was shown that mercuric reductase has the capacity to accept four electrons per FAD-containing subunit, and that two thiols become kinetically titrable by 5,5'-dithiobis-(2-nitrobenzoate) upon reduction with NADPH. These are characteristic features of the disulfide reductase class of flavoproteins. Further similarities with at least one of these enzymes, lipoamide dehydrogenase, include the E/EH2 midpoint potential (-269 mV), fluorescence properties, and extinction coefficients of E and EH2. Preliminary observations relevant to an understanding of the mechanism of mercuric reductase are discussed.     

Evidence for direct interactions between the mercuric ion transporter (MerT) and mercuric reductase (MerA) from the Tn<Emphasis Type="Italic">501</Emphasis><Emphasis Type="Italic">mer</Emphasis> operon

Mercuric ion resistance in bacteria requires transport of mercuric ions (Hg²⁺) into the cytoplasmic compartment where they are reduced to the less toxic metallic mercury (Hg⁰) by mercuric reductase (MR). The long-established model for the resistance mechanism predicts interactions between the inner membrane mercuric ion transporter, MerT, and the N-terminal domain of cytoplasmic MR, but attempts to demonstrate this interaction have thus far been unsuccessful. A recently developed bacterial two-hybrid protein interaction detection system was used to show that the N-terminal region of MR interacts with the cytoplasmic face of MerT. We also show that the cysteine residues on the cytoplasmic face of the MerT protein are required for maximal mercuric ion transport but not for the interaction with mercuric reductase.     

The ribulose-1,5-bisphosphate carboxylase/oxygenase gene cluster of Methylococcus capsulatus (Bath)

The genes encoding the ribulose-1,5-bisphosphate carboxylase/oxygenase (Rubisco) from Methylococcus capsulatus (Bath) were localised to an 8.3-kb EcoRI fragment of the genome. Genes encoding the large subunit ( cbbL), small subunit ( cbbS) and putative regulatory gene ( cbbQ) were shown to be located on one cluster. Surprisingly, cbbO, a second putative regulatory gene, was not located in the remaining 1.2-kb downstream (3') of cbbQ. However, probing of the M. capsulatus (Bath) genome with cbbO from Nitrosomonas europaea demonstrated that a cbbO homologue was contained within a separate 3.0-kb EcoRI fragment. Instead of a cbbR ORF being located upstream (5') of cbbL, there was a moxR-like ORF that was transcribed in the opposite direction to cbbL. There were three additional ORFs within the large 8.3-kb EcoRI fragment: a pyrE-like ORF, an rnr-like ORF and an incomplete ORF with no sequence similarity to any known protein. Phylogenetic analysis of cbbL from M. capsulatus (Bath) placed it within clade A of the green-type Form 1 Rubisco. cbbL was expressed in M. capsulatus (Bath) when grown with methane as a sole carbon and energy source under both copper-replete and copper-limited conditions. M. capsulatus (Bath) was capable of autotrophic growth on solid medium but not in liquid medium. Preliminarily investigations suggested that other methanotrophs may also be capable of autotrophic growth. Rubisco genes were also identified, by PCR, in Methylococcus-like strains and Methylocaldum species; however, no Rubisco genes were found in Methylomicrobium album BG8, Methylomonas methanica S1, Methylomonas rubra, Methylosinus trichosporium OB3b or Methylocystis parvus OBBP.     

Mercury adaptation among bacteria from a deep-sea hydrothermal vent

Since deep-sea hydrothermal vent fluids are enriched with toxic metals, it was hypothesized that (i) the biota in the vicinity of a vent is adapted to life in the presence of toxic metals and (ii) metal toxicity is modulated by the steep physical-chemical gradients that occur when anoxic, hot fluids are mixed with cold oxygenated seawater. We collected bacterial biomass at different distances from a diffuse flow vent at 9 degrees N on the East Pacific Rise and tested these hypotheses by examining the effect of mercuric mercury [Hg(II)] on vent bacteria. Four of six moderate thermophiles, most of which were vent isolates belonging to the genus Alcanivorax, and six of eight mesophiles from the vent plume were resistant to >10 microM Hg(II) and reduced it to elemental mercury [Hg(0)]. However, four psychrophiles that were isolated from a nearby inactive sulfide structure were Hg(II) sensitive. A neighbor-joining tree constructed from the deduced amino acids of a PCR-amplified fragment of merA, the gene encoding the mercuric reductase (MR), showed that sequences obtained from the vent moderate thermophiles formed a unique cluster (bootstrap value, 100) in the MR phylogenetic tree, which expanded the known diversity of this locus. The temperature optimum for Hg(II) reduction by resting cells and MR activity in crude cell extracts of a vent moderate thermophile corresponded to its optimal growth temperature, 45 degrees C. However, the optimal temperature for activity of the MR encoded by transposon Tn501 was found to be 55 to 65 degrees C, suggesting that, in spite of its original isolation from a mesophile, this MR is a thermophilic enzyme that may represent a relic of early evolution in high-temperature environments. Results showing that there is enrichment of Hg(II) resistance among vent bacteria suggest that these bacteria have an ecological role in mercury detoxification in the vent environment and, together with the thermophilicity of MR, point to geothermal environments as a likely niche for the evolution of bacterial mercury resistance.     

The relationships of Hg(II) volatilization from a freshwater pond to the abundance ofmer genes in the gene pool of the indigenous microbial community

The role of biological activities in the reduction and volatilization of Hg(II) from a polluted pond was investigated. Elemental mercury was evolved from pond water immediately following spiking with²⁰³Hg(NO₃)₂, whereas an acclimation period of 36 hours was required in control samples collected from a nearby, unpolluted river before onset of volatilization. Genes encoding the bacterial mercuric reductase enzyme (mer genes) were abundant in DNA fractions extracted from biomass of the pond microbial community, but not in samples extracted from control communities. Thus, evolution of Hg⁰ was probably due to activities mediated by the bacterial mercuric reductase. Of four characterizedmer operons, the system encoded by transposon 501 (mer(Tn501)) dominated and likely contributed to the majority of the observed Hg(II) volatilization. Thus,mer-mediated reduction and volatilization could be used to reduce Hg(II) concentrations in polluted waters, in turn decreasing rates of methylmercury formation by limiting substrate availability.     

细菌抗汞分子机制与进化的研究及应用

微生物中存在一类抗汞细菌,操纵子Mer中的MerRTPA参与细菌抗汞的调控、转运及还原。汞通过MerTP所表达的蛋白由细胞外转运至细胞内,经还原酶MerA将其还原为毒性小的可挥发的金属汞。细菌抗汞基因的形成有着古老的起源,基因间的整合、转移进化形成了Mer操纵子结构与功能的多样性。抗汞细菌对汞的吸附具有高选择性及专一性,可利用此特性对汞污染环境进行修复,也可作为分子遗传操作中稳定的抗性筛选标记。     

Redox properties of the hydroxylase component of methane monooxygenase from Methylococcus capsulatus (Bath). Effects of protein B, reductase, and substrate.

The reduction potentials of the hydroxylase component of the soluble methane monooxygenase from Methylococcus capsulatus (Bath) have been investigated through potentiometric titrations. The potentials were determined by EPR spectroscopic quantitation of the mixed valent hydroxylase as a function of added sodium dithionite in the presence of appropriate mediators. The reduction of the oxidized Fe(III).Fe(III) form to the mixed valent Fe(II).Fe(III) form occurs at 48 mV versus NHE while the potential for the formation of the fully reduced Fe(II).Fe(II) species from the mixed valent form was determined to be -135 mV. Addition of the substrate propylene to the hydroxylase did not have a major effect on the reduction potentials. Introduction of the protein B and the reductase components, however, completely inhibited reduction of the hydroxylase at potentials as far negative as -200 mV. Addition of propylene to all three methane monooxygenase components greatly facilitated hydroxylase reduction. Under these conditions, the fully reduced form of the protein was obtained at potentials of greater than 150 mV. This high redox potential indicates that the oxidized form of the protein is highly reactive, as required for methane oxidation. The present results reveal aspects of how both protein B and substrate can regulate electron transfer into and out of the hydroxylase component of methane monooxygenase.     

Model of the molecular basis for hydroxylamine oxidation and nitrous oxide production in methanotrophic bacteria

Many methane-oxidizing bacteria (MOB) have been shown to aerobically oxidize ammonia and hydroxylamine (NH(2)OH) to produce nitrite and nitrous oxide (N(2)O). Genome sequences of alphaproteobacterial, gammaproteobacterial, and verrucomicrobial methanotrophs revealed the presence of haoAB, cytL, cytS, nirS or nirK, and norCB genes that may be responsible for N(2)O production, and additional haoAB genes were sequenced from two strains of Methylomicrobium album. The haoAB genes of M. album ATCC 33003 were inducible by ammonia and NH(2)OH, similar to haoAB induction by ammonia in Methylococcus capsulatus Bath. Increased expression of genes encoding nitric oxide reductase (cNOR; norCB) was measured upon exposure of M. capsulatus Bath to NaNO(2) and NO-releasing sodium nitroprusside. Only incubations of M. capsulatus Bath with methane, ammonia, and nitrite produced N(2)O. The data suggest a possible pathway of nitrite reduction to NO by reversely operating NH(2)OH oxidoreductase and NO reduction to N(2)O by cNOR independently or in conjunction with ammonia-induced enzymes (i.e. HAO or cytochrome c'-β). Results of this study show that MOB likely have diverse mechanisms for nitrogen oxide metabolism and detoxification of NH(2)OH that involve conventional and unconventional enzymes.     

The metal centers of particulate methane monooxygenase from Methylosinus trichosporium OB3b

Particulate methane monooxygenase (pMMO) is a membrane-bound metalloenzyme that oxidizes methane to methanol in methanotrophic bacteria. The nature of the pMMO active site and the overall metal content are controversial, with spectroscopic and crystallographic data suggesting the presence of a mononuclear copper center, a dinuclear copper center, a trinuclear center, and a diiron center or combinations thereof. Most studies have focused on pMMO from Methylococcus capsulatus (Bath). pMMO from a second organism, Methylosinus trichosporium OB3b, has been purified and characterized by spectroscopic and crystallographic methods. Purified M. trichosporium OB3b pMMO contains approximately 2 copper ions per 100 kDa protomer. Electron paramagnetic resonance (EPR) spectroscopic parameters indicate that type 2 Cu(II) is present as two distinct species. Extended X-ray absorption fine structure (EXAFS) data are best fit with oxygen/nitrogen ligands and reveal a Cu-Cu interaction at 2.52 A. Correspondingly, X-ray crystallography of M. trichosporium OB3b pMMO shows a dinuclear copper center, similar to that observed previously in the crystal structure of M. capsulatus (Bath) pMMO. There are, however, significant differences between the pMMO structures from the two organisms. A mononuclear copper center present in M. capsulatus (Bath) pMMO is absent in M. trichosporium OB3b pMMO, whereas a metal center occupied by zinc in the M. capsulatus (Bath) pMMO structure is occupied by copper in M. trichosporium OB3b pMMO. These findings extend previous work on pMMO from M. capsulatus (Bath) and provide new insight into the functional importance of the different metal centers.     

Mercury Adaptation among Bacteria from a Deep-Sea Hydrothermal Vent

Since deep-sea hydrothermal vent fluids are enriched with toxic metals, it was hypothesized that (i) the biota in the vicinity of a vent is adapted to life in the presence of toxic metals and (ii) metal toxicity is modulated by the steep physical-chemical gradients that occur when anoxic, hot fluids are mixed with cold oxygenated seawater. We collected bacterial biomass at different distances from a diffuse flow vent at 9°N on the East Pacific Rise and tested these hypotheses by examining the effect of mercuric mercury [Hg(II)] on vent bacteria. Four of six moderate thermophiles, most of which were vent isolates belonging to the genus Alcanivorax, and six of eight mesophiles from the vent plume were resistant to >10 μM Hg(II) and reduced it to elemental mercury [Hg(0)]. However, four psychrophiles that were isolated from a nearby inactive sulfide structure were Hg(II) sensitive. A neighbor-joining tree constructed from the deduced amino acids of a PCR-amplified fragment of merA, the gene encoding the mercuric reductase (MR), showed that sequences obtained from the vent moderate thermophiles formed a unique cluster (bootstrap value, 100) in the MR phylogenetic tree, which expanded the known diversity of this locus. The temperature optimum for Hg(II) reduction by resting cells and MR activity in crude cell extracts of a vent moderate thermophile corresponded to its optimal growth temperature, 45°C. However, the optimal temperature for activity of the MR encoded by transposon Tn501 was found to be 55 to 65°C, suggesting that, in spite of its original isolation from a mesophile, this MR is a thermophilic enzyme that may represent a relic of early evolution in high-temperature environments. Results showing that there is enrichment of Hg(II) resistance among vent bacteria suggest that these bacteria have an ecological role in mercury detoxification in the vent environment and, together with the thermophilicity of MR, point to geothermal environments as a likely niche for the evolution of bacterial mercury resistance.