Active site of mercuric reductase resides at the subunit interface and requires Cys135 and Cys140 from one subunit and Cys558 and Cys559 from the adjacent subunit: evidence from in vivo and in vitro heterodimer formation

Mercuric reductase catalyzes the two-electron reduction of Hg(II) to Hg(0) using NADPH as the reductant; this reaction constitutes the molecular basis for detoxification of Hg(II) by bacteria. The enzyme is an alpha 2 homodimer and possesses two pairs of cysteine residues, Cys135 and Cys140 (redox-active pair) and Cys558 and Cys559 (C-terminal pair), which are known to be essential for catalysis. In the present study, we have obtained evidence for an intersubunit active site, consisting of a redox-active cysteine pair from one subunit and a C-terminal pair from the adjacent subunit, by reconstituting catalytic activity both in vivo and in vitro starting with two inactive, mutant enzymes, Ala135Ala140Cys558Cys559 (AACC) and Cys135Cys140Ala558Ala559 (CCAA). Genetic complementation studies were used to show that coexpression of AACC and CCAA in the same cell yielded an HgR phenotype, some 10(4)-fold more resistant than cells expressing only one mutant. Purification and catalytic characterization of a similarly coexpressed protein mixture showed the mixture to have activity levels ca. 25% those of wild type; this is the same as that statistically anticipated for a CCAA-AACC heterodimeric/homodimeric mixture with only one functional active site per heterodimer. Actual physical evidence for the formation of active mutant heterodimers was obtained by chaotrope-induced subunit interchange of inactive pure CCAA and AACC homodimers in vitro followed by electrophoretic separation of heterodimers from homodimers. Taken together, these data provide compelling evidence that the active site in mercuric reductase resides at the subunit interface and contains cysteine residues originating from separate polypeptide chains.     

Mutagenesis of the redox-active disulfide in mercuric ion reductase: catalysis by mutant enzymes restricted to flavin redox chemistry

Mercuric reductase, a flavoenzyme that possess a redox-active cystine, Cys135Cys140, catalyzes the reduction of Hg(II) to Hg(0) by NADPH. As a probe of mechanism, we have constructed mutants lacking a redox-active disulfide by eliminating Cys135 (Ala135Cys140), Cys140 (Cys135Ala140), or both (Ala135Ala140). Additionally, we have made double mutants that lack Cys135 (Ala135Cys139Cys140) or Cys140 (Cys135Cys139Ala140) but introduce a new Cys in place of Gly139 with the aim of constructing dithiol pairs in the active site that do not form a redox-active disulfide. The resulting mutant enzymes all lack redox-active disulfides and are hence restricted to FAD/FADH2 redox chemistry. Each mutant enzyme possesses unique physical and spectroscopic properties that reflect subtle differences in the FAD microenvironment. These differences are manifested in a 23-nm range in enzyme-bound FAD lambda max values, an 80-nm range in thiolate to flavin charge-transfer absorbance maxima, and a ca. 100-mV range in FAD reduction potential. Preliminary evidence for the Ala135Cys139Cys140 mutant enzyme suggests that this protein forms a disulfide between the two adjacent Cys residues. Hg(II) titration experiments that correlate the extent of charge-transfer quenching with Hg(II) binding indicate that the Ala135Cys140 protein binds Hg(II) with substantially less avidity than does the wild-type enzyme. All mutant mercuric reductases catalyze transhydrogenation and oxygen reduction reactions through obligatory reduced flavin intermediates at rates comparable to or greater than that of the wild-type enzyme. For these activities, there is a linear correlation between log kappa cat and enzyme-bound FAD reduction potential. In a sensitive Hg(II)-mediated enzyme-bound FADH2 reoxidation assay, all mutant enzymes were able to undergo at least one catalytic event at rates 50-1000-fold slower than that of the wild-type enzyme. We have also observed the reduction of Hg(II) by free FADH2. In multiple-turnover assays which monitored the production of Hg(0), two of the mutant enzymes were observed to proceed through at least 30 turnovers at rates ca. 1000-fold slower than that of wild-type mercuric reductase. We conclude that the Cys135 and Cys140 thiols serve as Hg(II) ligands that orient the Hg(II) for subsequent reduction by a reduced flavin intermediate.     

Structure and Dynamics of a Compact State of a Multidomain Protein,the Mercuric Ion Reductase

The functional efficacy of colocalized, linked protein domains is dependent on linker flexibility and system compaction. However, the detailed characterization of these properties in aqueous solution presents an enduring challenge. Here, we employ a novel, to our knowledge, combination of complementary techniques, including small-angle neutron scattering, neutron spin-echo spectroscopy, and all-atom molecular dynamics and coarse-grained simulation, to identify and characterize in detail the structure and dynamics of a compact form of mercuric ion reductase (MerA), an enzyme central to bacterial mercury resistance. MerA possesses metallochaperone-like N-terminal domains (NmerA) tethered to its catalytic core domain by linkers. The NmerA domains are found to interact principally through electrostatic interactions with the core, leashed by the linkers so as to subdiffuse on the surface over an area close to the core C-terminal Hg(II)-binding cysteines. How this compact, dynamical arrangement may facilitate delivery of Hg(II) from NmerA to the core domain is discussed.     

C-terminal cysteines of Tn501 mercuric ion reductase.

Mercuric ion reductase (MerA) catalyzes the reduction of Hg(II) to Hg(0) as the last step in the bacterial mercury detoxification pathway. A member of the flavin disulfide oxidoreductase family, MerA contains an FAD prosthetic group and redox-active disulfide in its active site. However, the presence of these two moieties is not sufficient for catalytic Hg(II) reduction, as other enzyme family members are potently inhibited by mercurials. We have previously identified a second pair of active site cysteines (Cys558 Cys559 in the Tn501 enzyme) unique to MerA, that are essential for high levels of mercuric ion reductase activity [Moore, M. J., & Walsh, C. T. (1989) Biochemistry 28, 1183; Miller, S. M., et al. (1989) Biochemistry 28, 1194]. In this paper, we have examined the individual roles of Cys558 and Cys559 by site-directed mutagenesis of each to alanine. Phenotypic analysis indicates that both merA mutations result in a total disruption of the Hg(II) detoxification pathway in vivo, while characterization of the purified mutant enzymes in vitro shows each to have differential effects on catalytic function. Compared to wild-type enzyme, the C558A mutant shows a 20-fold reduction in kcat and a 10-fold increase in Km, for an overall decrease in catalytic efficiency of 200-fold in kcat/Km. In contrast, mutation of Cys559 to alanine results in less than a 2-fold reduction in kcat and an increase in Km of only 4-5 fold for an overall decrease in catalytic efficiency of only ca. 10-fold in vitro. From these results, it appears that Cys558 plays a more important role in forming the reducible complex with Hg(II), while both Cys558 and Cys559 seem to be involved in efficient scavenging (i.e., tight binding) of Hg(II).     

Mutagenesis of the N- and C-terminal cysteine pairs of Tn501 mercuric ion reductase: consequences for bacterial detoxification of mercurials

Mercuric ion reductase (the merA gene product) is a unique member of the class of FAD and redox-active disulfide-containing oxidoreductases by virtue of its ability to reduce Hg(II) to Hg(0) as the last step in bacterial detoxification of mercurials. In addition to the active site redox-active disulfide, formed between Cys135 and Cys140 in Tn501 MerA, the protein products of the three merA gene sequences published to date have two additional conserved pairs of cysteines, one near the N-terminus (Cys10Cys13 in Tn501 MerA) and another near the C-terminus (Cys558Cys559 in Tn501 MerA). Neither of these pairs is found in other members of this enzyme family. To assess the possible roles of these peripheral cysteines in the Hg(II) detoxification pathway, we have constructed and characterized one single mutant, Cys10Ala13, and two double mutants, Ala10Ala13 and Ala558Ala559. The N-terminal mutants are fully functional in vivo as determined by HgCl2 resistance studies, showing the N-terminal cysteine pair to be dispensable. In contrast, the Ala558Ala559 mutant is defective for HgCl2 resistance in vivo and Hg(SR)2 reduction in vitro, thereby implicating Cys558 and/or Cys559 in Hg(II) reduction by the wild-type enzyme. Other activities, such as NADPH/thio-NADP+ transhydrogenation, NADPH oxidation, and DTNB reduction, are unimpaired in this mutant.     

Structure and Dynamics of a Compact State of a Multidomain Protein,the Mercuric Ion Reductase

The functional efficacy of colocalized, linked protein domains is dependent on linker flexibility and system compaction. However, the detailed characterization of these properties in aqueous solution presents an enduring challenge. Here, we employ a novel, to our knowledge, combination of complementary techniques, including small-angle neutron scattering, neutron spin-echo spectroscopy, and all-atom molecular dynamics and coarse-grained simulation, to identify and characterize in detail the structure and dynamics of a compact form of mercuric ion reductase (MerA), an enzyme central to bacterial mercury resistance. MerA possesses metallochaperone-like N-terminal domains (NmerA) tethered to its catalytic core domain by linkers. The NmerA domains are found to interact principally through electrostatic interactions with the core, leashed by the linkers so as to subdiffuse on the surface over an area close to the core C-terminal Hg(II)-binding cysteines. How this compact, dynamical arrangement may facilitate delivery of Hg(II) from NmerA to the core domain is discussed.     

Location of the redox-active thiols of ribonucleotide reductase: sequence similarity between the Escherichia coli and Lactobacillus leichmannii enzymes

The redox-active thiols of Escherichia coli ribonucleoside diphosphate reductase and of Lactobacillus leichmannii ribonucleoside triphosphate reductase have been located by a procedure involving (1) prereduction of enzyme with dithiothreitol, (2) specific oxidation of the redox-active thiols by treatment with substrate in the absence of exogenous reductant, (3) alkylation of other thiols with iodoacetamide, and (4) reduction of the disulfides with dithiothreitol and alkylation with [1-14C]iodoacetamide. The dithiothreitol-reduced E. coli B1 subunit is able to convert 3 equiv of CDP to dCDP and is labeled with 5.4 equiv of 14C. Sequencing of tryptic peptides shows that 2.8 equiv of 14C is on cysteines-752 and -757 at the C-terminus of B1, while 1.0-1.5 equiv of 14C is on cysteines-222 and -227. It thus appears that two sets of redox-active dithiols are involved in substrate reduction. The L. leichmannii reductase is able to convert 1.1 equiv of CTP to dCTP and is labeled with 2.1 equiv of 14C. Sequencing of tryptic peptides shows that 1.4 equiv of 14C is located on the two cysteines of C-E-G-G-A-C-P-I-K. This peptide shows remarkable and unexpected similarity to the thiol-containing region of the C-terminal peptide of E. coli B1, C-E-S-G-A-C-K-I.     

Cell-free mercury volatilization activity from three marine caulobacter strains

Three mercury-resistant marine Caulobacter strains showed an inducible mercury volatilization activity. Cell-free mercury volatilization (mercuric reductase) from these three marine Caulobacter strains was characterized and compared with enzyme activities determined by plasmids of Escherichia coli and Staphylococcus aureus. The temperature sensitivity of the Caulobacter mercuric reductase was greater than that of mercuric reductase from other gram-negative sources. Cell-free enzyme activity required NADH or NADPH, with NADPH functioning much better at lower concentrations than NADH. The Km for the Caulobacter enzyme was 4 microM Hg2+. Ag+ was a competitive inhibitor of Caulobacter mercuric reductase (Ki = 0.2 microM Ag+), as with previously studied enzymes. Arsenite was a noncompetitive inhibitor of the Caulobacter enzyme with a Ki of 75 microM AsO2-.     

Cell-free mercury volatilization activity from three marine caulobacter strains.

Three mercury-resistant marine Caulobacter strains showed an inducible mercury volatilization activity. Cell-free mercury volatilization (mercuric reductase) from these three marine Caulobacter strains was characterized and compared with enzyme activities determined by plasmids of Escherichia coli and Staphylococcus aureus. The temperature sensitivity of the Caulobacter mercuric reductase was greater than that of mercuric reductase from other gram-negative sources. Cell-free enzyme activity required NADH or NADPH, with NADPH functioning much better at lower concentrations than NADH. The Km for the Caulobacter enzyme was 4 microM Hg2+. Ag+ was a competitive inhibitor of Caulobacter mercuric reductase (Ki = 0.2 microM Ag+), as with previously studied enzymes. Arsenite was a noncompetitive inhibitor of the Caulobacter enzyme with a Ki of 75 microM AsO2-.     

NMR structural studies reveal a novel protein fold for MerB, the organomercurial lyase involved in the bacterial mercury resistance system

Mercury resistant bacteria have developed a system of two enzymes (MerA and MerB), which allows them to efficiently detoxify both ionic and organomercurial compounds. The organomercurial lyase (MerB) catalyzes the protonolysis of the carbon-mercury bond resulting in the formation of ionic mercury and a reduced hydrocarbon. The ionic mercury [Hg(II)] is subsequently reduced to the less reactive elemental mercury [Hg(0)] by a specific mercuric reductase (MerA). To better understand MerB's unique enzymatic activity, we used nuclear magnetic resonance (NMR) spectroscopy to determine the structure of the free enzyme. MerB is characterized by a novel protein fold consisting of three noninteracting antiparallel beta-sheets surrounded by six alpha-helices. By comparing the NMR data of free MerB and the MerB/Hg/DTT complex, we identified a set of residues that likely define a Hg/DTT binding site. These residues cluster around two cysteines (C(96) and C(159)) that are crucial to MerB's catalytic activity. A detailed analysis of the structure revealed the presence of an extensive hydrophobic groove adjacent to this Hg/DTT binding site. This extensive hydrophobic groove has the potential to interact with the hydrocarbon moiety of a wide variety of substrates and may explain the broad substrate specificity of MerB.     

Activation of mercuric reductase by the substrate NADPH

Preincubation of the oxidized form of the flavoenzyme mercuric reductase with the reducing substrate, NADPH, or with a high concentration of cysteine (30 mM) results in a substantial increase of the catalytic activity as measured in a standard spectrophotometric assay. Also NADH has some activating effect but NADP+ or EDTA have no effect. In the presence of 1 mM cysteine only one equivalent of NADPH per FAD seems to be required for full activation which occurs after an incubation time of about 10 min. Activated mercuric reductase appears to be stable under anaerobic conditions but eventually returns to the original level of activity in the presence of oxygen. The activated state seems to be stabilized by 1 mM cysteine. Activation of mercuric reductase does not seem to be correlated with a change in the number of reactive thiol groups. The chemical nature of the activation process is not yet understood. Stopped-flow studies have shown that the nonactivated enzyme is practically inactive prior to contact with the substrates. The enzyme is gradually activated during the assay. The kinetics of activation of the 'native' enzyme is biphasic but 'clipped' enzyme, lacking an 85-residue N-terminal domain, is activated in a single first-order process. The progress curves obtained with preactivated enzyme are approximately exponential even at saturating concentrations of NADPH (Km = 0.4 microM at 25 degrees C, pH 7.3) and Hg2+ (Km = 3.2 microM in the presence of 1 mM cysteine). The initial rates yield kcat values of about 13 s-1 per FAD molecule (25 degrees C, pH 7.3). We find no evidence for a thiol-dependent change from a rapid to a slow kinetic phase. The shape of the progress curves presumably depends on product inhibition, but NADP+ is not a sufficiently effective inhibitor to explain the effect fully.     

A stable mercury-containing complex of the organomercurial lyase MerB: catalysis, product release, and direct transfer to MerA

Bacteria isolated from organic mercury-contaminated sites have developed a system of two enzymes that allows them to efficiently convert both ionic and organic mercury compounds to the less toxic elemental mercury. Both enzymes are encoded on the mer operon and require sulfhydryl-bound substrates. The first enzyme is an organomercurial lyase (MerB), and the second enzyme is a mercuric ion reductase (MerA). MerB catalyzes the protonolysis of the carbon-mercury bond, resulting in the formation of a reduced carbon compound and inorganic ionic mercury. Of several mercury-containing MerB complexes that we attempted to prepare, the most stable was a complex consisting of the organomercurial lyase (MerB), a mercuric ion, and a molecule of the MerB inhibitor dithiothreitol (DTT). Nuclear magnetic resonance (NMR) spectroscopy and extended X-ray absorption fine structure spectroscopy of the MerB/Hg/DTT complex have shown that the ligands to the mercuric ion in the complex consist of both sulfurs from the DTT molecule and one cysteine ligand, C96, from the protein. The stability of the MerB/Hg/DTT complex, even in the presence of a large excess of competing cysteine, has been demonstrated by NMR and dialysis. We used an enzyme buffering test to determine that the MerB/Hg/DTT complex acts as a substrate for the mercuric reductase MerA. The observed MerA activity is higher than the expected activity assuming free diffusion of the mercuric ion from MerB to MerA. This suggests that the mercuric ion can be transferred between the two enzymes by a direct transfer mechanism.     

Properties of the cysteine residues and the iron-sulfur cluster of the assimilatory 5'-adenylyl sulfate reductase from Enteromorpha intestinalis

The 5'-adenylyl sulfate (APS) reductase from the marine macrophytic green alga Enteromorpha intestinalis uses reduced glutathione as the electron donor for the reduction of APS to 5'-AMP and sulfite. The E. intestinalis enzyme (EiAPR) is composed of a reductase domain and a glutaredoxin-like C-terminal domain. The enzyme contains a single [4Fe-4S] cluster as its sole prosthetic group. Three of the enzyme's eight cysteine residues (Cys166, Cys257, and Cys260) serve as ligands to the iron-sulfur cluster. Site-directed mutagenesis experiments and resonance Raman spectroscopy are consistent with the presence of a cluster in which only three of the four ligands to the cluster irons contributed by the protein are cysteine residues. Site-directed mutagenesis experiments suggest that the thiol group of Cys250, a residue found only in algal APS reductases, is not an absolute requirement for activity. The other four cysteines that do not serve as cluster ligands, all of which are required for activity, are involved in the formation of two redox-active disulfide/dithiol couples. The couple involving Cys342 and Cys345 has an E(m) value at pH 7.0 of -140 mV, and the one involving Cys165 and Cys285 has an E(m) value at pH 7.0 of -290 mV. The C-terminal portion of EiAPR, expressed separately, exhibits the cystine reductase activity characteristic of glutaredoxins. It is proposed that the Cys342-Cys345 disulfide provides the site for entry of electrons from reduced glutathione and that the Cys166-Cys285 disulfide may serve as a structural element that is essential for keeping the enzyme in the catalytically active conformation.     

Properties of the cysteine residues and iron-sulfur cluster of the assimilatory 5'-adenylyl sulfate reductase from Pseudomonas aeruginosa

APS reductase from Pseudomonas aeruginosa has been shown to contain a [4Fe-4S] cluster. Thiol determinations and site-directed mutagenesis studies indicate that the single [4Fe-4S] cluster contains only three cysteine ligands, instead of the more typical arrangement in which clusters are bound to the protein by four cysteines. Resonance Raman studies in the Fe-S stretching region are also consistent with the presence of a redox-inert [4Fe-4S](2+) cluster with three cysteinate ligands and indicate that the fourth ligand is likely to be an oxygen-containing species. This conclusion is supported by resonance Raman and electron paramagnetic resonance (EPR) evidence for near stoichiometric conversion of the cluster to a [3Fe-4S](+) form by treatment with a 3-fold excess of ferricyanide. Site-directed mutagenesis experiments have identified Cys139, Cys228, and Cys231 as ligands to the cluster. The remaining two cysteines present in the enzyme, Cys140 and Cys256, form a redox-active disulfide/dithiol couple (E(m) = -300 mV at pH 7.0) that appears to play a role in the catalytic mechanism of the enzyme.     

A model for mercuric ion reduction in recombinant Escherichia coli

Plasmid-encoded mercuric reduction involves transfer of Hg(2+) across the cellular envelope and reduction to Hg(0) by the cytoplasmic mercuric reductase using NADPH. A mathematical model was developed for the binding and transfer of Hg(2+) by transport proteins and the subsequent reduction of Hg(2+). The values of the model parameters were determined using experimental data. The derived rate expressions were similar to the previously experimentally determined ones. The model predicted that a differential amplification of the transport protein relative to mercuric reductase expression levels may enhance the Hg(2+) reduction rate in whole cells.     

Attachment of the N-terminal domain of Salmonella typhimurium AhpF to Escherichia coli thioredoxin reductase confers AhpC reductase activity but does not affect thioredoxin reductase activity

AhpF of Salmonella typhimurium, the flavoprotein reductase required for catalytic turnover of AhpC with hydroperoxide substrates in the alkyl hydroperoxide reductase system, is a 57 kDa protein with homology to thioredoxin reductase (TrR) from Escherichia coli. Like TrR, AhpF employs tightly bound FAD and redox-active disulfide center(s) in catalyzing electron transfer from reduced pyridine nucleotides to the disulfide bond of its protein substrate. Homology of AhpF to the smaller (35 kDa) TrR protein occurs in the C-terminal part of AhpF; a stretch of about 200 amino acids at the N-terminus of AhpF contains an additional redox-active disulfide center and is required for catalysis of AhpC reduction. We have demonstrated that fusion of the N-terminal 207 amino acids of AhpF to full-length TrR results in a chimeric protein (Nt-TrR) with essentially the same catalytic efficiency (k(cat)/K(m)) as AhpF in AhpC reductase assays; both k(cat) and the K(m) for AhpC are decreased about 3-4-fold for Nt-TrR compared with AhpF. In addition, Nt-TrR retains essentially full TrR activity. Based on results from two mutants of Nt-TrR (C129, 132S and C342,345S), AhpC reductase activity requires both centers while TrR activity requires only the C-terminal-most disulfide center in Nt-TrR. The high catalytic efficiency with which Nt-TrR can reduce thioredoxin implies that the attached N-terminal domain does not block access of thioredoxin to the TrR-derived Cys342-Cys345 center of Nt-TrR nor does it impede the putative conformational changes that this part of Nt-TrR is proposed to undergo during catalysis. These studies indicate that the C-terminal part of AhpF and bacterial TrR have very similar mechanistic properties. These findings also confirm that the N-terminal domain of AhpF plays a direct role in AhpC reduction.