Reactivity of Glutaredoxins 1, 2, and 3 fromEscherichia coli Shows That Glutaredoxin 2 Is the Primary Hydrogen Donor to ArsC-catalyzed Arsenate Reduction*

In Escherichia coli ArsC catalyzes the reduction of arsenate to arsenite using GSH with glutaredoxin as electron donors. E. coli has three glutaredoxins: 1, 2, and 3, each with a classical -Cys-Pro-Tyr-Cys- active site. Glutaredoxin 2 is the major glutathione disulfide oxidoreductase in E. coli, but its function remains unknown. In this report glutaredoxin 2 is shown to be the most effective hydrogen donor for the reduction of arsenate by ArsC. Analysis of single or double cysteine-to-serine substitutions in the active site of the three glutaredoxins indicated that only the N-terminal cysteine residue is essential for activity. This suggests that, during the catalytic cycle, ArsC forms a mixed disulfide with GSH before being reduced by glutaredoxin to regenerate the active ArsC reductase.

Glutaredoxins (Grx) 1 are glutathione-dependent dithiol hydrogen donors for Escherichia coli enzymes such as ribonucleotide reductase, 3Ј-phosphoadenylsulfate reductase, and arsenate reductase. The product of the grxA gene, glutaredoxin 1 (Grx1), was originally identified as a hydrogen donor to ribonucleotide reductase (1). In searching for alternate reductants of ribonucleotide reductase, two new glutaredoxins, Grx2 and Grx3, were identified, both of which catalyzed GSH-disulfide reduction of 2-hydroxyethyldisulfide (HED) in a coupled system with NADPH and glutathione reductase (2). Grx2 was shown to be the predominant glutaredoxin in E. coli cells, with an intracellular concentration of 5 M compared with 0.2 M for Grx1 and 2.4 M for Grx3. Grx2 also has a higher turnover number than the other two glutaredoxins for reduction of a mixed disulfide of GSH and HED. As a consequence of its high concentration and turnover, Grx2 provides over 80% of the GSH-oxidoreductase activity in E. coli for reduction of GSH mixed disulfides. However, Grx2 is not a hydrogen donor to ribonucleotide reductase, and no function has yet been assigned for this protein (3).
Resistance to arsenicals and antimonials in E. coli is conferred by arsenic resistance (ars) operons (4). The arsC gene product of the ars operon of plasmid R773 is an arsenate reductase (5). ArsC reduces arsenate to arsenite, which is subsequently extruded from the cell, conferring resistance. In vitro reductase activity requires both GSH and Grx1 (6). Thioredoxin is not able to couple to the R773 arsenate reductase. In contrast, the unrelated arsenate reductase of Staphylococcus aureus plasmid pI258 requires thioredoxin and cannot use GSH and glutaredoxin (7).
In this study, we examined the relative efficiency of Grx1, Grx2, and Grx3 to serve as hydrogen donor for the reduction of arsenate by ArsC. Each glutaredoxin supported arsenite formation, with the relative efficiencies being Grx2 Ͼ Grx3 Ͼ Grx1. This is the first demonstration of a role of Grx2 in a physiological reaction. Glutaredoxins have two active site cysteine residues in the sequence Cys-Pro-Tyr-Cys. The N-terminal cysteine has been shown to be required for both protein disulfide reduction and reduction of mixed protein-glutathione disulfides (8). The other cysteine residue is required for the former but not for the latter. Here the codon for either of the two cysteine residues in grxA, grxB, and grxC was mutated to a serine codon, and the effect on arsenate reductase activity examined with the altered Grx1, Grx2, and Grx3. Mutation of the codon for the C-terminal active site cysteine of any of the three glutaredoxins had no effect on ArsC activity. In contrast, Grx1, Grx2, or Grx3 with a cysteine to serine substitution in the N-terminal residue were unable to serve as hydrogen donors to ArsC-catalyzed arsenate reduction. These results are consistent with a reaction cycle in which ArsC forms a mixed disulfide with glutathione, where the role of glutaredoxin would be to reduce the mixed disulfide, regenerating reduced arsenate reductase.

EXPERIMENTAL PROCEDURES
Bacterial Strains, Plasmids, and Media-Strains and plasmids used in this study are described in Table I. Cells were grown in Luria-Bertani medium (9) at 37°C supplemented when necessary with 20 g/ml chloramphenicol, 10 g/ml tetracycline, or 50 g/ml kanamycin.
DNA Manipulations-All restriction enzymes and nucleic acid modifying enzymes were obtained from Life Technologies, Inc. Plasmid isolation, DNA restriction endonuclease analysis, ligation, and transformation were performed as described previously (9, 10) The Wizard TM Plus minipreps DNA purification system and Wizard TM DNA clean-up system from Promega were used to prepare plasmid DNA for restriction enzyme digestion and recovering DNA fragment from low melting agarose gels, respectively.
Cloning of grxC-The grxC gene from E. coli strain JM109 was amplified by PCR to introduce a NdeI site at the 5Ј end and EcoRI site at the 3Ј-end. The forward primer was 5Ј-CATATGGCCAATGTT-GAAATCTATACC-3, which introduced a NdeI site at the 5Ј-end of the fragment. The reverse primer was 5Ј-GAATTCTTATTTCAGCAGGG-GATCCAGTCC-3Ј. A 30-cycle PCR reaction (94°C for 1 min, 55°C for 0.5 min, and 72°C for 1 min) was run with DNA from E. coli strain JM109. The amplified product was cloned into pGEM-T vector and sequenced, then subcloned into pALTER-Ex2 and pET28a for mutagenesis or expression of Grx3, respectively.
Oligonucleotide-directed Mutagenesis-Mutations in grxA, grxB, and grxC were introduced by site-directed mutagenesis using the Altered Sites TM in vitro mutagenesis system (Promega). Plasmid pALTER-Grx1, pALTER-Grx2, and pALTER-Grx3 containing the grxA, grxB, and grxC genes were used as the template to obtain mutants ( Table I).
Construction of Vectors for the Overexpression of Grx2-The grxB gene was amplified using primers containing NdeI (G2-FNdeI primer: 5Ј-TGGAGGAGTCATATGAAGCTATAC-3Ј) and BamHI (G2-RCBa-mHI primer: 5Ј-CGCGGCGGGGGATCCTTAAATCGC-3Ј) sites at the 5Ј and 3Ј termini of the gene, respectively. The amplified fragment was digested with NdeI and BamHI and cloned into pET24aϩ (Novagen) and subsequently into pET28 for introduction of the six-histidine tag. For the construction of Grx2 C12S, grxB was initially amplified using primers G2-C12S (5Ј-ATGAAGCTATACATTTACGATCACTGCCCTT-ACAGCCTCAAAGC-3Ј). The amplified fragment was used as a template for a second PCR reaction using primers G2-FNdeI and G2-RCBamHI. The amplified fragment was digested with NdeI and BamHI and cloned into pET24aϩ and subsequently into pET28.
For the construction of the C9S/C12S derivative of Grx2, grxB was initially amplified using primers C9S-C12S (5Ј-ATGAAGCTATACATT-TACGATCACAGCCCTTACAGCCTCAAAGC-3Ј). The amplified fragment was used as a template for a second PCR reaction using primers G2-FNdeI and G2-RCBamHI. The amplified fragment was digested with NdeI and BamHI and cloned into pET24aϩ and subsequently into pET28.
Protein Purification-ArsC was purified as described previously (12). Purification of wild type and altered glutaredoxins was as following. E. coli strain BL21(DE3) bearing each grx gene in vector plasmid pET28a was grown in 1 liter of Luria-Bertani medium containing 50 g/ml kanamycin with shaking at 37°C. At an A 600 nm ϭ 0.5-0.6, isopropyl-1-thio-␤-D-galactopyranoside was added as inducer to a final concentration of 0.4 mM for Grx1 and its mutants, 0.5 mM for Grx2 and its mutants, or 0.1 mM for Grx3 and its mutants. The culture was shaken for an additional 3 h at 37°C. The cells were washed once with buffer A (50 mM MOPS, pH 7.5, containing 20 mM imidazole, 0.5 M NaCl, 8 mM ␤-mercaptoethanol, and 20% glycerol). The cells were suspended in buffer A at a ratio of 5 ml of buffer/g of cells and lysed by a single passing through a French pressure cell at 20,000 p.s.i. Diisopropylfluorophosphate (2.5 l/g of wet cells) was added to the lysate immediately after lysis. The lysate was centrifuged at 100,000 ϫ g for 60 min at 4°C, and the supernatant solution was loaded at a flow rate of 0.4 ml/min onto a 5 ml of Ni 2ϩ -nitrilotriacetic acid column pre-equilibrated with buffer A. The column was then washed with 100 ml of buffer A containing 0.02 M imidazole followed by elution with 100 ml of the buffer A containing 0.2 M inidazole. Protein-containing fractions were identified by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (13), pooled, and concentrated using a Millipore Ultrafree-15 BIOMAX-5K or -10K centrifugal filter (Millipore) at 2,000 ϫ g. All purified protein was  Assay of Reductase Activity-Arsenate reductase activity was assayed by a modification of a glutathione disulfide oxidoreductase-dependent assay for glutaredoxin (15). The assay buffer was 50 mM MOPS and 50 mM MES, pH 6.5, containing 0.1 mg/ml BSA, 0.25 mM NADPH, 15 nM yeast glutathione reductase, 1 mM GSH, 10 mM sodium arsenate, and 6.3 M ArsC. Glutaredoxin was added as indicated below. Reductase activity was monitored by the decrease in NADPH absorption at 340 nm and is expressed as nanomoles of NADPH oxidized per milligram of ArsC using a molar extinction coefficient of 6,200 for NADPH.

RESULTS AND DISCUSSION
Purification of Glutaredoxins-To allow for rapid purification of recombinant glutaredoxins, the sequence for 20 codons was added to the 5Ј-end of each glutaredoxin gene, including the sequence for a thrombin recognition site and six histidine codons, adding approximately 2 kDa to the molecular mass of each glutaredoxin species. Each protein was purified by chelate affinity chromatography. In saturating amounts, the three Histagged glutaredoxins reduced arsenate in rates identical to those of their wild type counterparts lacking the six histidine tag (data not shown). Because of the ease of purification, all subsequent assays were performed with the His-tagged glutaredoxins.
Grx2 Is the Major Hydrogen Donor for Arsenate Reduction-Each of the three glutaredoxins was able to serve as a hydrogen donor to ArsC-catalyzed arsenate reduction (Fig. 1). At saturating concentrations, the turnover numbers (k cat ) of the three proteins were within a factor of two to three of each other (Table II). In contrast, the K m of ArsC for the three glutaredoxins differed by 3 orders of magnitude. ArsC exhibited a 100-fold lower K m for Grx2 than Grx3 and more than 1,000-fold lower when compared with Grx1. Thus, the catalytic efficiency (k cat / K m ) for Grx2 is approximately 2 orders of magnitude greater than that for Grx3 and approximately 3 orders of magnitude greater than that for Grx1. These results suggest that Grx2 is most likely the major hydrogen donor for the reduction of arsenate to arsenite, the first credible physiological role found for Grx2. Since it is the major GSH-disulfide oxidoreductase in E. coli, the identification of an electron acceptor is important for understanding the function of Grx2.
It should be pointed out, however, that disruption of the gene for any of the three glutaredoxins does not eliminate arsenate resistance conferred by the ars operon (data not shown). This is not surprising in light of the redundancy of glutaredoxins and glutaredoxin-like proteins (16) and suggests the presence of an additional yet unidentified glutaredoxin activity. Similarly, in vivo reduction of ribonucleotides by ribonucleotide reductase is not affected by single or multiple disruptions of the trxA or grxA genes. However, cells lacking both Trx1 and Grx1 have highly induced ribonucleotide reductase (17). Later work resulted in the isolation of a second thioredoxin (Trx2) encoded by the trxC gene (18,19). On the other hand, 3Ј-phosphoadenylsulfate reduction is more specifically coupled to Trx1, Trx2 or Grx1 (20), and a double disruption of the trxA and grxA genes resulted in cysteine auxotrophy (21).
Role of Glutaredoxin in Arsenate Reduction-Glutaredoxins can reduce either intramolecular disulfides (e.g. ribonucleotide reductase) or mixed disulfides between a thiol compound and GSH (e.g. the complex between HED and GSH). Grx1, Grx2, and Grx3 each have the conserved active site sequence Cys-Pro-Tyr-Cys. Both cysteine residues are required for protein disulfide reduction (8). For reducing glutathione-containing mixed disulfides, however, the N-terminal cysteine is sufficient (8,22). In the monothiol mechanism, the N-terminal cysteine of glutaredoxin reacts with the glutathionyl moiety of the mixed disulfide, forming the mixed disulfide intermediate GrxS-SG. Simultaneously, the non-glutathione component is released in its reduced form. The GrxS-SG can be further reduced by GSH to give reduced glutaredoxin and GSSG.
To determine which glutaredoxin mode was involved in ArsC-catalyzed arsenate reduction, mutant glutaredoxin genes encoding single and double cysteine-to-serine substitutions of  the three E. coli glutaredoxins were constructed. Glutaredoxins with their N-terminal cysteine changed to serine (Grx1 C11S, Grx2 C9S, or Grx3 C12S) were unable to serve as hydrogen donors for the reduction of arsenate (Fig. 2). The three double substitutions (Grx1 C11/14S, Grx2 C9/12S, or Grx3 C12/15S) exhibited the same phenotype as the single substitutions. In contrast, glutaredoxins with intact N-terminal cysteines but serine substitutions in the C-terminal cysteines (Grx1 C14S, Grx2 C12S, or Grx3 C15S) retained nearly complete wild type activity. Thus, for ArsC catalysis, Cys 11 for Grx1, Cys 9 for Grx2, or Cys 12 for Grx3 was sufficient. These results are consistent with formation of a mixed protein-SG disulfide during the reaction cycle, as indicated in the following proposed reaction scheme (where ES Ϫ represents the Cys 12 thiolate of ArsC).
In this proposed minimal reaction scheme, Reaction 1 shows noncovalent binding of arsenate to ArsC. Since ArsC is competitively inhibited by other tetrahedral oxyanions such as phosphate and sulfate (6), the first step represents binding at an oxyanion binding site. In Reaction 2, reduction of As(V) to As(III) is achieved by one electron donated by the thiolate of Cys 12 of ArsC and another electron derived from the thiolate of GSH. Thus, Reaction 2 is likely to be the sum of several steps, one possibly involving a protein thiyl radical. In Reaction 3 Grx2 is indicated as the hydrogen donor for reduction of the ArsCS-SG mixed disulfide. This is followed by regeneration of reduced Grx2 by GSH, forming GSSG in reaction (4). The GSSG will finally be reduced to 2 mol of GSH by NADPH and glutathione reductase in Reaction 5.
Since Grx2 has a catalytic efficiency that is 1 or 2 orders of magnitude greater than Grx3 or Grx1, what governs the glu-taredoxin specificity of arsenate reduction? Since the three E. coli glutaredoxins have similar activities in an assay using reduction of the mixed disulfide between 2-hydroxyethyldisulfide and GSH (23), it is reasonable to speculate that specific protein-protein interactions between ArsC and each of the three glutaredoxins are involved, and the efficiency of those interactions determines the rate of reaction (3). Thus, the role of Grx2 in the ArsC reduction system is the first example of a monothiol mechanism in a substrate reduction involving electron transport by a glutaredoxin.