Portal de Periódicos da CAPES

Redox Regulation of 3′-Phosphoadenylylsulfate Reductase from Escherichia coli by Glutathione and Glutaredoxins

2003; Elsevier BV; Volume: 278; Issue: 25 Linguagem: Inglês

10.1074/jbc.m302304200

1083-351X

Christopher Horst Lillig, Aristi Potamitou, J. D. Schwenn, Alexios Vlamis‐Gardikas, Arne Holmgren,

Redox biology and oxidative stress

Inorganic sulfate ( SO42−, S+VI) is reduced in vivo to sulfite ( SO32−, S+IV) via phosphoadenylylsulfate (PAPS) reductase. Escherichia coli lacking glutathione reductase and glutaredoxins (gor–grxA–grxB–grxC–) barely grows on sulfate. We found that incubation of PAPS reductase with oxidized glutathione leads to enzyme inactivation with simultaneous formation of a mixed disulfide between glutathione and the active site Cys-239. A newly developed method based on thiol-specific fluorescent alkylation and gel electrophoresis showed that glutathionylated PAPS reductase is reduced by glutaredoxins via a monothiol mechanism. This glutathionylated species was also observed in poorly growing gor–grxA–grxB–grxC– cells expressing inactive glutaredoxin 2 (Grx2) C9S/C12S. However, it was absent in better growing cells expressing monothiol Grx2 C12S or wild type Grx2. Reversible glutathionylation may thus regulate the activity of PAPS reductase in vivo. Inorganic sulfate ( SO42−, S+VI) is reduced in vivo to sulfite ( SO32−, S+IV) via phosphoadenylylsulfate (PAPS) reductase. Escherichia coli lacking glutathione reductase and glutaredoxins (gor–grxA–grxB–grxC–) barely grows on sulfate. We found that incubation of PAPS reductase with oxidized glutathione leads to enzyme inactivation with simultaneous formation of a mixed disulfide between glutathione and the active site Cys-239. A newly developed method based on thiol-specific fluorescent alkylation and gel electrophoresis showed that glutathionylated PAPS reductase is reduced by glutaredoxins via a monothiol mechanism. This glutathionylated species was also observed in poorly growing gor–grxA–grxB–grxC– cells expressing inactive glutaredoxin 2 (Grx2) C9S/C12S. However, it was absent in better growing cells expressing monothiol Grx2 C12S or wild type Grx2. Reversible glutathionylation may thus regulate the activity of PAPS reductase in vivo. Sulfur is an ingredient of all living organisms. The first, most common form of sulfur in nature is inorganic sulfate, which needs to be further reduced for incorporation in a living cell. Prototrophic bacteria, for example, use inorganic sulfate ( SO42−, S+ VI) as primary source for the biosynthesis of sulfur-containing amino acids and cofactors (1Schwenn J.D. Z. Naturforsch. 1994; 49c: 531-539Crossref Scopus (32) Google Scholar). Sulfate is first activated to adenylylsulfate (APS) 1The abbreviations used are: APS, adenylylsulfate; DTT, dithiothreitol; Grx, glutaredoxin; HED, 2-hydroxyethydisulfide; MALDI-TOF, matrix-assisted laser desorption ionization time-of-flight; PAP, adenosine-3′,5′-bisphosphate; PAPS, phosphoadenylylsulfate; PR, phosphoadenylylsulfate reductase; TR, thioredoxin reductase; Trx, thioredoxin. and then to 3′-phosphoadenylylsulfate (PAPS) by ATP sulfurylase and APS kinase. Subsequently, PAPS is reduced by PAPS reductase (PR) to sulfite ( SO32−, S+ IV) and adenosine-3′5′-bisphosphate (PAP). Sulfite is reduced to sulfide (S2–,S–II) by sulfite reductase, and, thereafter, is incorporated in O-acetyl serine (OAS) by OAS-(thiol)lyase to give the primary product of sulfate assimilation, cysteine. PAPS reductase (EC 1.8.99.4) is composed of two identical subunits of 28 kDa. It is devoid of chromophores and contains a single cysteine per subunit in a highly conserved ECGLH motif, which is identified as the redox-active center of the enzyme (2Berendt U. Haverkamp T. Prior A. Schwenn J.D. Eur. J. Biochem. 1995; 233: 347-356Crossref PubMed Scopus (62) Google Scholar). Reduction of sulfate to sulfite by PR requires two electrons during which the cysteines of PR are oxidized to a disulfide. The oxidized enzyme is inactive and needs to be reduced for the reduction of PAPS to continue. Kinetic data (2Berendt U. Haverkamp T. Prior A. Schwenn J.D. Eur. J. Biochem. 1995; 233: 347-356Crossref PubMed Scopus (62) Google Scholar, 3Krone F.A. Westphal G. Schwenn J.D. Mol. Gen. Genet. 1991; 225: 314-319Crossref PubMed Scopus (38) Google Scholar, 4Lillig C.H. Prior A. Schwenn J.D. Åslund F. Ritz D. Vlamis-Gardikas A. Holmgren A. J. Biol. Chem. 1999; 274: 7695-7698Abstract Full Text Full Text PDF PubMed Scopus (121) Google Scholar) and the crystal structure of PR (5Savage H. Montoya G. Svensson C. Schwenn J.D. Sinning I. Structure. 1997; 5: 895-906Abstract Full Text Full Text PDF PubMed Scopus (51) Google Scholar) implicate a ping-pong mechanism for its reduction. In the first step, shown here in Reaction 1, PRSHSH+PAPS→PRSS+PAP+SO32−Reaction 1 reduced PR catalyzes the reduction of PAPS, leading to oxidized PR, free sulfite, and PAP. Upon oxidation the enzyme dimer undergoes conformational changes leading to a significant decrease in apparent molecular weight and the inability to bind PAPS (2Berendt U. Haverkamp T. Prior A. Schwenn J.D. Eur. J. Biochem. 1995; 233: 347-356Crossref PubMed Scopus (62) Google Scholar, 3Krone F.A. Westphal G. Schwenn J.D. Mol. Gen. Genet. 1991; 225: 314-319Crossref PubMed Scopus (38) Google Scholar). In the second step, oxidized PR is reduced by thioredoxin as shown in Reaction 2, PRSS+TrxSHSH→PRSHSH+TrxSSReaction 2 or glutaredoxin, which is depicted in Reaction 3, PRSS+GrxSHSH→PRSHSH+GrxSSReaction 3 with electrons from thioredoxin reductase (TR) and NADPH or glutathione (GSH), glutathione reductase, and NADPH. Thioredoxins and glutaredoxins are small (9–14 kDa) ubiquitous proteins that utilize their two redox-active cysteines (CXXC motif) to catalyze the reduction of disulfides (6Holmgren A. J. Biol. Chem. 1989; 264: 13963-13966Abstract Full Text PDF PubMed Google Scholar). Whereas thioredoxins and glutaredoxins can reduce their substrates by using both active site cysteines (dithiol mechanism) (7Holmgren A. Structure. 1995; 3: 239-243Abstract Full Text Full Text PDF PubMed Scopus (382) Google Scholar), glutaredoxins can also utilize the thiols from GSH in solution together with the glutaredoxin N-terminal cysteine (monothiol mechanism) (8Bushweller J.H. Åslund F. Wüthrich K. Holmgren A. Biochemistry. 1992; 31: 9288-9293Crossref PubMed Scopus (204) Google Scholar). In addition to their ability to reduce intracellular disulfides, glutaredoxins may also reduce the mixed disulfides that form between a protein thiol and GSH. This is a reaction that is not catalyzed by thioredoxins. Escherichia coli contains two thioredoxins (Trx1 and Trx2) and three glutaredoxins (Grx1, Grx2, and Grx3) (9Vlamis-Gardikas A. Holmgren A. Methods Enzymol. 2002; 347: 286-296Crossref PubMed Scopus (112) Google Scholar). Trx1, Trx2, and Grx1 can reduce the disulfide that forms on ribonucleotide reductase 1a (RNR1a) upon the reduction of ribonucleotides with comparative efficiencies, whereas Grx3 is only a weak reductant in vitro (10Laurent T.C. Moore E.C. Reichard P. J. Biol. Chem. 1964; 239: 3436-3444Abstract Full Text PDF PubMed Google Scholar, 11Holmgren A. J. Biol. Chem. 1979; 254: 3664-3671Abstract Full Text PDF PubMed Google Scholar, 12Miranda-Vizuete A. Damdimopoulos A.E. Gustafsson J. Spyrou G. J. Biol. Chem. 1997; 272: 30841-30847Abstract Full Text Full Text PDF PubMed Scopus (124) Google Scholar, 13Åslund F. Ehn B. Miranda-Vizuete A. Pueyo C. Holmgren A. Proc. Natl. Acad. Sci. U. S. A. 1994; 91: 9813-9817Crossref PubMed Scopus (164) Google Scholar). Trx1, Trx2, and Grx1 participate in the in vitro reduction of PAPS by PR (4Lillig C.H. Prior A. Schwenn J.D. Åslund F. Ritz D. Vlamis-Gardikas A. Holmgren A. J. Biol. Chem. 1999; 274: 7695-7698Abstract Full Text Full Text PDF PubMed Scopus (121) Google Scholar). Other functions for Trx1 include the reduction of methionine sulfoxide via methionine sulfoxide reductase, whereas Trx2 participates in the antioxidant response as part of the OxyR regulon. Grx1 is also a member of the OxyR regulon, but Grx2 and Grx3 are not (14Potamitou A. Neubauer P. Holmgren A. Vlamis-Gardikas A. J. Biol. Chem. 2002; 277: 17775-17780Abstract Full Text Full Text PDF PubMed Scopus (24) Google Scholar). With levels at least 10-fold higher than those of Grx1, Grx2 and Grx3 are highly abundant proteins in E. coli (15Potamitou A. Holmgren A. Vlamis-Gardikas A. J. Biol. Chem. 2002; 277: 18561-18567Abstract Full Text Full Text PDF PubMed Scopus (58) Google Scholar) and contribute up to 98% of the GSH-dependent oxidoreductase activity, using the disulfide between β-mercaptoethanol and GSH as a substrate (HED assay) (11Holmgren A. J. Biol. Chem. 1979; 254: 3664-3671Abstract Full Text PDF PubMed Google Scholar). Grx2 is an atypical glutaredoxin with a molecular mass of 23.4 kDa and structural similarities to mammalian GSH S-transferases (16Xia B. Vlamis-Gardikas A. Holmgren A. Wright P.E. Dyson H.J. J. Mol. Biol. 2001; 310: 907-918Crossref PubMed Scopus (69) Google Scholar). Because of its high abundance (up to 1% of total soluble protein) and catalytic efficiency, it contributes to >80% of the cellular GSH-mixed disulfide reducing activities (in the HED assay) (17Vlamis-Gardikas A. Åslund F. Spyrou G. Bergmann T. Holmgren A. J. Biol. Chem. 1997; 272: 11236-11243Abstract Full Text Full Text PDF PubMed Scopus (76) Google Scholar, 18Lundström-Ljung J. Vlamis-Gardikas A. Åslund F. Holmgren A. FEBS Lett. 1999; 443: 85-88Crossref PubMed Scopus (25) Google Scholar). The enzyme is also highly active in the reduction of the mixed disulfide between glutathione and arsenate reductase (19Shi J. Vlamis-Gardikas A. Åslund F. Holmgren A. Rosen B.P. J. Biol. Chem. 1999; 274: 36039-36042Abstract Full Text Full Text PDF PubMed Scopus (124) Google Scholar). Grx2 is involved in the antioxidant response, as mutants lacking Grx2 have increased levels of carbonylation of their intracellular proteins after exposure to hydrogen peroxide. Glutaredoxins and thioredoxins are not only direct antioxidants, but may also participate in the signal transduction of redox-induced cellular responses (overviews in Refs. 20Cartmel-Harel O. Stortz G. Annu. Rev. Microbiol. 2000; 54: 439-461Crossref PubMed Scopus (576) Google Scholar and 21Holmgren A. Antioxid. Redox Signal. 2001; 2: 811-820Crossref Scopus (417) Google Scholar). Combined E. coli null mutants for glutathione reductase and the three glutaredoxins (gor–grxA–grxB–grxC–) barely grow on sulfate (S+VI) but grow normally on sulfite (S+IV) or methionine (S–II) (22Vlamis-Gardikas A. Potamitou A. Zarivach R. Hochman A. Holmgren A. J. Biol. Chem. 2002; 277: 10861-10868Abstract Full Text Full Text PDF PubMed Scopus (40) Google Scholar). Because these mutants contain sufficient amounts of thioredoxin to reduce PR (15Potamitou A. Holmgren A. Vlamis-Gardikas A. J. Biol. Chem. 2002; 277: 18561-18567Abstract Full Text Full Text PDF PubMed Scopus (58) Google Scholar), this disturbed growth must represent some sort of inhibition of PR activity not based on the reduction of the enzyme's disulfide that is formed upon the reduction of PAPS. Growth of gor–grxA–grxB–grxC– could be restored with monothiol or wild type Grx2 in trans but not with the inactive C9S/C12S species (22Vlamis-Gardikas A. Potamitou A. Zarivach R. Hochman A. Holmgren A. J. Biol. Chem. 2002; 277: 10861-10868Abstract Full Text Full Text PDF PubMed Scopus (40) Google Scholar). Because Grx2 cannot reduce the disulfide of oxidized PR (4Lillig C.H. Prior A. Schwenn J.D. Åslund F. Ritz D. Vlamis-Gardikas A. Holmgren A. J. Biol. Chem. 1999; 274: 7695-7698Abstract Full Text Full Text PDF PubMed Scopus (121) Google Scholar), this finding raises the possibility that the activity of PAPS reductase in vivo may be regulated by oxidized glutathione and glutaredoxins. General Methods—Materials, chemicals, and enzymes were purchased from different companies in the highest available purity. E. coli cells were transformed according to Hanahan (23Hanahan D. Glover D.M. DNA Cloning. 1. IRL Press, Oxford, UK1985: 109-135Google Scholar). The concentration of proteins in crude extracts was determined as described by Bradford (24Bradford M.M. Anal. Biochem. 1976; 72: 254-284Crossref Scopus (217389) Google Scholar). Pure proteins were quantified using the following molar absorbance coefficients at 280 nm: PAPS reductase, 52,630 m–1 cm–1 (monomer; Grx1, 10,810 m–1 cm–1; Grx2, 21,620 m–1 cm–1; Grx3, 3,840 m–1 cm–1; and Trx1, 15,220 m–1 cm–1. SDS-PAGE was performed using the Phast-Gel system (Amersham Biosciences) and the Ready-Gel system (Bio-Rad) according to each manufacturer's instructions. Strains and Plasmids—E. coli Bl21(DE3) (Novagen, Madison, WI) was used for the overexpression of PAPS reductase using plasmid pET16bcysH (4Lillig C.H. Prior A. Schwenn J.D. Åslund F. Ritz D. Vlamis-Gardikas A. Holmgren A. J. Biol. Chem. 1999; 274: 7695-7698Abstract Full Text Full Text PDF PubMed Scopus (121) Google Scholar). DHB4gor–grxA–grxB–grxC– and the arabinose promoter-based plasmids pISCGrx2, pISCGrx2C12S, and pISCGrx2C9S/C12S for the expression of wild type Grx2 and the mutants C12S and C9S/C12S, respectively, were first described and characterized in Ref. 22Vlamis-Gardikas A. Potamitou A. Zarivach R. Hochman A. Holmgren A. J. Biol. Chem. 2002; 277: 10861-10868Abstract Full Text Full Text PDF PubMed Scopus (40) Google Scholar. Protein Expression and Purification—PAPS reductase was expressed and purified as described (4Lillig C.H. Prior A. Schwenn J.D. Åslund F. Ritz D. Vlamis-Gardikas A. Holmgren A. J. Biol. Chem. 1999; 274: 7695-7698Abstract Full Text Full Text PDF PubMed Scopus (121) Google Scholar) using a 3-liter (Meredos, Göttingen, Germany) and a 25-liter fermenter (New Brunswick Scientific, Edison, NJ). Oxidized PAPS reductase was prepared by incubation with an excess of PAPS (15 min, 22 °C, 50 mm Tris/HCl, pH 8.0) essentially as described (25Schriek U. Schwenn J.D. Arch. Microbiol. 1986; 145: 32-38Crossref PubMed Scopus (31) Google Scholar). The reduced protein was prepared by treatment with 10 mm dithiothreitol (DTT) and 1 μm Trx1 from E. coli (15 min, 22 °C, 50 mm Tris/HCl, pH 8.0). The glutathionylated form of PR was produced by incubating 200 μm reduced enzyme with 20 mm oxidized glutathione (GSSG, 30 min 22 °C, Tris/HCl, pH 8.0). Free nucleotides, reductants, or GSSG were removed using ultrafiltration (Amicon YM3, Millipore) and Sephadex G25-columns (Amersham Biosciences). The different isoforms were stored at –80 °C in 40 mm Tris/HCl, pH 8, 500 mm NaCl, and 10% glycerol. Grx1, Grx1C14S, Grx2, Grx2C12S, Grx3, Grx3C15S, and thioredoxin reductase from E. coli were expressed and purified as described previously (8Bushweller J.H. Åslund F. Wüthrich K. Holmgren A. Biochemistry. 1992; 31: 9288-9293Crossref PubMed Scopus (204) Google Scholar, 17Vlamis-Gardikas A. Åslund F. Spyrou G. Bergmann T. Holmgren A. J. Biol. Chem. 1997; 272: 11236-11243Abstract Full Text Full Text PDF PubMed Scopus (76) Google Scholar, 26Åslund F. Nordstrand K. Berndt K.D. Nikkola M. Bergman T. Ponstingl H. Jörnvall H. Otting G. Holmgren A. J. Biol. Chem. 1996; 271: 6736-6745Abstract Full Text Full Text PDF PubMed Scopus (59) Google Scholar, 27Luthman M. Holmgren A. Biochemistry. 1982; 21: 6628-6633Crossref PubMed Scopus (510) Google Scholar). Enzymatic Assays—PAPS reductase activity was measured as an acid-labile sulfite formation from [35S]PAPS at 30 °C (28Schwenn J.D. Schriek U. Z. Naturforsch. 1987; 42c: 93-102Crossref Scopus (26) Google Scholar). The assay mixture (100 μl) contained 50–500 ng ml–1 PAPS reductase, 100 mm Tris/HCl, pH 8, 10 mm Na2SO3, 100 μm [35S]PAPS, 100 μm Trx1, and 1 μm thioredoxin reductase from E. coli, and 5 mm NADPH. Oxidized glutathione, yeast glutathione reductase (Sigma), and glutaredoxins were added as indicated. For determination of the activity of glutaredoxin-treated oxidized, reduced, and glutathionylated enzyme, the assay mixtures (5 μg of PR, 50 mm Tris/HCl, pH8.0, and 100 mm NaCl) were desalted on Sephadex-G25, and the activity was determined in the absence of further reductants in a single turnover experiment (100 mm Tris/HCl pH 8, 10 mm Na2SO3, and 100 μm [35S]PAPS; 5-min reaction time). [35S]PAPS (1,700 Bq nmol–1) was prepared from [35S]sulfite (Amersham Biosciences) as described by Schriek and Schwenn (25Schriek U. Schwenn J.D. Arch. Microbiol. 1986; 145: 32-38Crossref PubMed Scopus (31) Google Scholar) using recombinant APS kinase from Arabidopsis thaliana (29Lillig C.H. Schiffmann S. Berndt C. Berken A. Tischka R. Schwenn J.D. Arch. Biochem. Biophys. 2001; 392: 303-310Crossref PubMed Scopus (39) Google Scholar). Synthesis and purity were monitored by high pressure liquid chromatography (30Schwenn J.D. Jender H.G. J. Chromatogr. 1980; 139: 285-290Crossref Scopus (30) Google Scholar). GSH-mixed disulfides were assayed in a reaction mixture (500 μl) containing 100 mm Tris/HCl, pH 8.0, 100 mm reduced glutathione, 100 μm NADPH, yeast glutathione reductase, 1 μm glutaredoxin, and glutathionylated PR as indicated. Glutaredoxins catalyze the reduction of mixed disulfides using GSH as electron donor. The resulting GSSG is reduced by glutathione reductase with electrons from NADPH. The reaction was initiated by the addition of glutaredoxin. The decrease in A340 was used for quantification of the GSH-moieties in the UV-2100 photometer (Shimadzu, Kyoto) at 25 °C. Fluorescent Experiments—5 μg of PAPS reductase in 50 mm Tris/HCl, pH 8.0, and 100 mm NaCl in a total volume of 20 μl was incubated with 5 mg of a reduced glutaredoxin for 30 s in the presence or absence of 0.5 mm reduced glutathione. The samples were alkylated and labeled with 0.5 mm 5-(iodoacetamido)-fluorescein (5-IAF, Sigma; solved in N,N-dimethylformamide) for 45 min at room temperature in the dark before they were separated by SDS-PAGE (8–16%) and analyzed on a UV table (Ultraviolet Products, San Gabriel, CA). Growth of E. coli—To determine the redox status of PR in cell-free extracts and for immunoprecipitation experiments, cells were grown overnight in LB, washed twice with cold M9 medium, and inoculated in fresh M9 medium with 33 mg liter–1 Leu and Ile, 100 mg liter–1 ampicillin, and 0.1% arabinose to an A600 of 0.14. Cells were grown in 300-ml cultures at 170 rpm and 37 °C until they reached stationary phase. The cells were then collected, incubated with 100 mm iodoacetamide for 20 min on ice to stop further reactions of thiol groups, harvested by centrifugation, resuspended in TE buffer (50 mm Tris/HCl, pH 8.0, and 1 mm EDTA), and frozen at –20 °C. Purification of Antibodies—Rabbit sera were adjusted with ammonium sulfate to 50% of saturation and left stirring at 4 °C overnight. The precipitated IgG fraction was resuspended in phosphate-buffered saline (PBS) and dialyzed extensively against PBS, pH 7.5. Affinity-purified antibodies for PAPS reductase were prepared using an Affi-Gel 10 column (Bio-Rad) on which 5 mg of PAPS reductase had been previously immobilized using the procedure recommended by the manufacturer. Prior to the application of the IgG fraction, columns were equilibrated with 20 mm Tris-HCl, pH 7.5, followed by 20 mm Tris-HCl, pH 7.5, with 500 mm NaCl and, finally, 20 mm Tris-HCl, pH 7.5. After sample loading, columns were subsequently washed with the same buffers, and bound antibodies were eluted with a pulse of 0.1 m acetic acid-formic acid, pH 2.1. The eluate was immediately neutralized with 1 m Tris-HCl, pH 9, aliquoted and stored at –20 °C. Immunoprecipitation—Frozen cells were washed once with 40 mm Tris-HCl, pH 8.0, and 0.5 mm EDTA and resuspended in 3 ml of the same buffer. Cells were incubated for 45 min with 1 mg ml–1 lysozyme on ice. Cells were sonicated, treated with 1 mm phenylmethanesulfonyl fluoride (PMSF), and centrifuged for 1 h at 100,000 × g. The cell-free lysate supernatant was incubated for 2 h at 4 °C with 35 mg of affinity-purified polyclonal PAPS reductase antibodies while shaking, and then for another hour with 350 ml of 50% protein G-Sepharose (Amersham Biosciences). Cells were spun down, washed twice with 40 mm Tris-HCl, pH 8.0, and 0.5 mm EDTA, resuspended in 1% SDS, and boiled for 20 min before SDS-PAGE. Western Blotting—The BioRad system was used according to the manufacturer's protocol. After transfer, the nitrocellulose membrane was blocked for 20 min at room temperature with 2% bovine serum albumin in TBST (20 mm Tris-HCl, pH 7.5, 150 mm NaCl, and 0.05% Tween 20). It was further washed and incubated with the primary antibody overnight at 4 °C (1:500 for anti/PR-antibodies and 1:2000 for anti/GSH-antibodies). The membrane was washed with 150 mm NaCl for 20 min and then with TBST for additional 20 min, followed by incubation for 1 h with horseradish peroxidase-conjugated goat anti-rabbit antibodies (Dako) using a dilution of 1:4000 for PR antibodies and 1:2000 for GSH antibodies in TBST. The blots were developed by chemiluminescence using the Western Lightning kit from PerkinElmer Life Sciences and visualized using the MiltiImage Light cabinet (Alpha Innotech, San Leandro, CA). Mass Spectrometry—Mass spectrometry was performed on a PE Biosystems Voyager 6061 (Applied Biosystems) matrix-assisted laser desorption ionization time-of-flight (MALDI-TOF) system. Tryptic digestion was performed using sequencing grade-modified trypsin (Promega) according to the manufacturer's protocol. The peptides were diluted in 75% acetonitrile containing 1% trifluoroacetic acid and mixed with an equal amount of a saturated solution of α-cyano-4-hydroxycinnamic acid in 50% acetonitrile and 0.1% trifluoroacetic acid. 1 μl of this solution was allowed to crystallize on the applicator plate before ionization. Growth Properties of E. coli DHB4gor–grxA–grxB–grxC–— Combined E. coli null mutants for glutathione reductase and Grx1, Grx2, and Grx3 (DHB4gor–grxA–grxB–grxC–) barely grow in the presence of sulfate, but they grow well in the presence of sulfite, cysteine, or methionine (22Vlamis-Gardikas A. Potamitou A. Zarivach R. Hochman A. Holmgren A. J. Biol. Chem. 2002; 277: 10861-10868Abstract Full Text Full Text PDF PubMed Scopus (40) Google Scholar) (Fig. 1). Consistent with previous findings (22Vlamis-Gardikas A. Potamitou A. Zarivach R. Hochman A. Holmgren A. J. Biol. Chem. 2002; 277: 10861-10868Abstract Full Text Full Text PDF PubMed Scopus (40) Google Scholar), transformants with the monothiol Grx2 grew faster and reached the highest optical density (A600) at stationary phase (2.4). The wild type Grx2-containing cells reached an OD of 1.9, and the no-thiol Grx2-containing cells only an OD of 1.0 at stationary phase, as described previously for the non-transformed strain (22Vlamis-Gardikas A. Potamitou A. Zarivach R. Hochman A. Holmgren A. J. Biol. Chem. 2002; 277: 10861-10868Abstract Full Text Full Text PDF PubMed Scopus (40) Google Scholar). Redox Status of PAPS Reductase in Vivo—To investigate whether the limited growth of the null mutant was caused by an arrest of PR in the oxidized state, we determined the redox state of the enzyme in vivo. PR is a homodimeric enzyme whose active site is formed by an intermolecular dithiol-disulfide couple between the only cysteines at position 239 (2Berendt U. Haverkamp T. Prior A. Schwenn J.D. Eur. J. Biochem. 1995; 233: 347-356Crossref PubMed Scopus (62) Google Scholar). As there are no other covalent links between the two subunits, the reduced and oxidized forms of the enzyme can be separated by nonreducing SDS-PAGE, where the reduced enzyme corresponds to an apparent Mr of 30 kDa, and the oxidized enzyme to 60 kDa. No oxidized PR could be detected in the null mutant transformed with the no-thiol Grx2 (Fig. 2, lanes 2 and 3) or in any other strain (data not shown). Therefore, the inhibition of cell growth in the particular strain was not caused by an arrest of PR in its oxidized conformation. Reversible Inhibition of PAPS Reductase by Oxidized Glutathione—As the GSH/GSSG ratio in the gor– strain would be expected to shift more toward oxidized glutathione, we investigated whether GSSG affects PR activity. The enzyme was incubated with different amounts of GSSG before reduction of PAPS was performed with electrons delivered from Trx1, Trx reductase (TrxR), and NADPH (Fig. 3). Following incubation with GSSG, the activity of PR decreased exponentially until no PAPS reduction was detectable. This inhibition pattern is characteristic for pseudo first order kinetics and suggested a covalent modification of PAPS reductase by GSSG as the basis for the inactivation, which, as shown in Reaction 4, E+GSSG→E−SG+GSHReaction 4 is most likely due to the formation of a mixed disulfide between the enzyme and glutathione. The non-linear curve fitting of these results was made assuming pseudo-first order kinetics. Calculated from the first-order rates obtained for the different GSSG-concentrations, the second-order rate constant was 80.4 ± 5.6 m–1 min–1, indicating a rapid reaction between PAPS reductase and GSSG. As the inhibition of PR was likely due to formation of a mixed disulfide between the active site thiol and glutathione, we tried to restore enzymatic activity by the addition of reductants. The addition of DTT or glutathione reductase could not restore the enzymatic activity of PR. When an additional glutaredoxin was added to the reaction mixture, the activity of the enzyme was restored to the former extent in less than 3 min (data not shown). All E. coli glutaredoxins (Grx1, Grx2, and Grx3) as well as their monothiol (CXXS)-mutants (Grx1C14S, Grx2C12S, and Grx3C15S) were capable of reactivating PR in that time period. Glutathionylation of Cysteine 239 —To confirm the glutathionylation of the active site cysteine 239, reduced and GSSG-treated PAPS reductase were analyzed by MALDI-TOF (Fig. 4). Tryptic digestion of the reduced protein generated a peptide mass of 900.413 ± 0.017 (n = 2) corresponding to the partly cleaved C-terminal fragment Arg-Glu-Cys239-Gly-Leu-His-Glu-Gly with a calculated molecular weight of 900.399. GSSG-treatment induced a signal with the size of m/z = 1206.455, which is compatible with the formation of a disulfide between Cys239 and glutathione (calculated m/z 1206.474). These results indicated that Cys239 can form a mixed disulfide with glutathione. Oxidation and Reduction of Glutathionylated PR—Highly purified oxidized and glutathionylated PR (1.93 ± 0.08 GSH per PR) were analyzed using thiol-specific alkylation with fluorescent mm 5-(iodoacetamido)-fluorescein and non-reducing SDS-PAGE (Fig. 5). Oxidized and glutathionylated PR exhibited virtually no fluorescence or activity (Fig. 5, lanes 1 and 2), whereas the Cys239 thiol of reduced PR was accessible for alkylation (Fig. 5, lane 3) and active in single turnover experiments (Fig. 5, panel I). The remaining portion of the 30-kDa protein in the oxidized protein (lane 1) corresponds to redox-inactive C-terminal truncated protein without any cysteine (5Savage H. Montoya G. Svensson C. Schwenn J.D. Sinning I. Structure. 1997; 5: 895-906Abstract Full Text Full Text PDF PubMed Scopus (51) Google Scholar). Reduced Grx1 reduced both the PR disulfide and the mixed disulfide between PR and GSH in the presence and absence of GSH (Fig. 5, lanes 4 and 5). Reduced monothiol Grx1C14S reduced the two PR species in the presence of GSH, but formed a stable mixed disulfide with PR in the absence of GSH (Fig. 5, lanes 6 and 7). Grx3, as well as its monothiol mutant Grx3C15S, reduced the glutathionylated form of PR in the presence of GSH (Fig. 5A, lane 9 and 11) but not the intramolecular PR disulfide (Fig. 5A, lanes 8 and 10). As seen in the reactivation assays, DTT or GSH alone could not reduce glutathionylated PR, as they did not increase the amount of free thiols of the glutathionylated form (Fig. 5, lanes 2 and 12). Remarkably, in all preparations the existence of free thiols in PR corresponded to protein in the active conformation (sections I and II). Glutathionylation of PR in Vivo—Was glutathionylated PR the reason for the limited growth of the gor–grxA–grxB–grxC– strain in M9 media free of reduced sulfur? Antibodies raised against GSH-moieties on bovine serum albumin (31Hjelle O.P. Chaudhry F.A. Ottersen O.P. Eur. J. Neurosci. 1994; 6: 793-804Crossref PubMed Scopus (102) Google Scholar) reacted specifically with glutathionylated PR and showed no cross reactivity with reduced or oxidized PR (Fig. 6B, lanes 1–3). No glutathionylated PR was detected in extracts from gor–grxA–grxB–grxC– transformed with Grx2C12S (Fig. 6, lane 6). The mixed disulfide species was detected in extracts from transformants encoding wild type Grx2 or the no thiol Grx2C9S/C12S mutant (Fig. 6, lanes 4 and 5). From the density of the bands on the blot, 20% of the PR in the pISCGrx2 transformants was glutathionylated (2.8 of 14 ng detected). 40% was the estimation for the pISCGrx2C9S/C12S-transformants (6.2 of 15.9 ng detected). Regulation of biological activity by the reversible modification of protein thiols is a growing concept in cell signaling. A fine tuning in the DNA binding properties of the transcription factor OxyR centers on Cys199, which can be hydroxylated, nitrosylated, or glutathionylated, with each modification resulting in differential binding of the protein to DNA; glutathionylated OxyR has the highest transcriptional activity (32Kim S.O. Merchant K. Nudelman R. Beyer Jr., W.F. Keng T. DeAngelo J. Hausladen A. Stamler J.S. Cell. 2002; 109: 383-396Abstract Full Text Full Text PDF PubMed Scopus (391) Google Scholar). Glutathionylation of Cys62 of the eukaryotic NF-κB subunit p50 lead to loss of DNA binding activity (33Pineda-Molina E. Klatt P. Vazquez J. Martina A. de Lacoba M.G. Perez-Sala D. Lamas S. Biochemistry. 2001; 40: 14134-14142Crossref PubMed Scopus (343) Google Scholar), whereas nitric oxide treatment of c-Jun lead to nitrosylation/glutathionylation of Cys269 with concomitant loss of DNA binding activity (34Klatt P. Pineda-Molina E. de Lacoba M.G. Padilla A.C. Martinez-Galisteo E. Barzena J.A. Lamas S. FASEB J. 1999; 13: 1481-1490Crossref PubMed Scopus (251) Google Scholar). Deglutathionylation of Cys374 in G-actin resulted in a 6-fold increase in the rate of its polymerization (35Wang J. Boja E.S. Tan W. Tekle E. Fales H.M. English S. Mieyal J.J. Chock P.B. J. Biol. Chem. 2001; 276: 47763-47766Abstract Full Text Full Text PDF PubMed Scopus (287) Google Scholar). The activity of tyrosine hydroxylase, the rate-limiting enzyme for the biosynthesis of dopamine, was inhibited by reversible glutathionylation (36Borges C.R. Geddes T.J. Watson J.T. Kuhn D.M. J. Biol. Chem. 2002; 277: 48295-48302Abstract Full Text Full Text PDF PubMed Scopus (87) Google Scholar). The same has been reported for protein kinase Cα and protein tyrosine phosphatase 1B (37Barett W.C. DeGnore J.P. König S. Fales H.M. Keng Y.F. Zhang Z.Y. Yim M.B. Chock P.B. Biochemistry. 1999; 38: 6699-6705Crossref PubMed Scopus (432) Google Scholar). Nitrosylated mammalian thioredoxin (Cys69) has antiapoptotic properties (38Haendeler J. Hoffmann J. Tischler V. Berk B.C. Zeiher A.M. Dimmeler S. Nat. Cell Biol. 2002; 4: 743-749Crossref PubMed Scopus (344) Google Scholar), whereas its in vitro reducing activity to insulin disulfides is abolished by glutathionylation of the non-active site Cys72 (39Casagrande S. Bonetto V. Fratelli M. Gianazza E. Eberini I. Massignan T. Salmona M. Chang G. Holmgren A. Ghezzi P. Proc. Natl. Acad. Sci. U. S. A. 2002; 99: 9745-9749Crossref PubMed Scopus (303) Google Scholar). Consistent with the concept of glutathionylation in modifying biological activity, we propose that PR is inhibited by the formation of a mixed disulfide between cysteine 239 and GSH. E. coli glutaredoxins can reverse this inhibition by reducing the mixed disulfide. As precedent for this hypothesis, human glutaredoxin can reduce the mixed disulfide for both G-actin and protein tyrosine phosphatase 1B, thus reversing the effects of glutathionylation and restoring biological activity (35Wang J. Boja E.S. Tan W. Tekle E. Fales H.M. English S. Mieyal J.J. Chock P.B. J. Biol. Chem. 2001; 276: 47763-47766Abstract Full Text Full Text PDF PubMed Scopus (287) Google Scholar, 37Barett W.C. DeGnore J.P. König S. Fales H.M. Keng Y.F. Zhang Z.Y. Yim M.B. Chock P.B. Biochemistry. 1999; 38: 6699-6705Crossref PubMed Scopus (432) Google Scholar). The gene encoding PR (cysH) is located in the cysJIH operon and transcribed in a coordinated way with other genes, all belonging to the cys regulon (reviewed in Ref. 40Kredich N.M. Neidhardt F.C. Escherichia coli and Salmonella: Cellular and Molecular Biology. ASM Press, Washington, D. C.1996: 514-527Google Scholar). Genes within this regulon are only transcribed when sulfur is limiting and no forms of reduced sulfur are available for the cell. This regulation is controlled by the negatively auto-regulated transcription activator CysB and the inducer N-acetyl serine, which is derived from the cysteine precursor O-acetyl serine. No other form of kinetic regulation has been reported previously for the sulfate reduction pathway. This was demonstrated by a mutant that cannot repress this pathway and accumulates large amounts of sulfide (41Borum P.R. Monty K.J. J. Bacteriol. 1976; 125: 53-55Crossref Google Scholar). Regulation of PR activity by glutathionylation introduces another level of complexity in the overall regulation of sulfite biosynthesis. What could be the biological relevance of this control? Reduction of PAPS is linked to loss of electrons, which means loss of reducing equivalents. Under conditions favoring protein glutathionylation (e.g. oxidative stress), the unhindered continuation of electron flow via the PR pathway would further deteriorate cell homeostasis by using electrons that would have otherwise been used to reverse undesired oxidations. Therefore, stopping the activity of PR under conditions favoring its glutathionylation could be considered an adaptation of the cell to severe oxidative stress. Neurons constitute another example in which a multilevel regulation in the production of an oxidant and the responses against it could be regulated in a coordinated manner at many levels. The activity of tyrosine hydroxylase, the rate-limiting enzyme for the biosynthesis of dopamine, a potent oxidant, is inhibited by reversible nitrosylation of a structural cysteine (36Borges C.R. Geddes T.J. Watson J.T. Kuhn D.M. J. Biol. Chem. 2002; 277: 48295-48302Abstract Full Text Full Text PDF PubMed Scopus (87) Google Scholar). At the same time, dopamine-induced oxidative stress leading to the apoptosis of rat neurons may be offset by glutaredoxin activity, which activates NF-κB via Ref1 (42Daily D. Vlamis-Gardikas A. Offen D. Mittelman L. Melamed E. Holmgren A. Barzilai A. J. Biol. Chem. 2001; 276: 1335-41334Abstract Full Text Full Text PDF PubMed Scopus (101) Google Scholar). The glutaredoxin-driven signaling pathways can include both the Ras phosphoinositide 3-kinase signaling cascade and the c-Jun N-terminal kinase pathway (43Daily D. Vlamis-Gardikas A. Offen D. Mittelman L. Melamed E. Holmgren A. Barzilai A. J. Biol. Chem. 2001; 276: 21618-21626Abstract Full Text Full Text PDF PubMed Scopus (68) Google Scholar). In view of the findings of this work, a more complete picture emerges for the catalytic mechanism of PR (Fig. 7). In a "normal" reducing cell environment, reduced PR is a homodimer. PAPS can bind to this reduced form (Fig. 7, R), to yield sulfite, PAP, and oxidized PR with an intra-molecular disulfide bridge between the active site cysteines (Cys239) (Fig. 7, reaction 1). Dimeric oxidized PR migrates on SDS-PAGE with an apparent molecular mass (60 kDa) higher than that of the reduced form (30 kDa). The disulfide of the oxidized enzyme can be reduced by Trx1, Trx2, or Grx1 but not the other glutaredoxins (Fig. 7, reaction 2) (4Lillig C.H. Prior A. Schwenn J.D. Åslund F. Ritz D. Vlamis-Gardikas A. Holmgren A. J. Biol. Chem. 1999; 274: 7695-7698Abstract Full Text Full Text PDF PubMed Scopus (121) Google Scholar, 44Russel M. Model P. Holmgren A. J. Bacteriol. 1990; 172: 1923-1929Crossref PubMed Scopus (105) Google Scholar). If the intracellular environment is somewhat oxidizing (e.g. the gor– strain), a mixed disulfide may form between Cys239 and glutathione, rendering the enzyme inactive. All glutaredoxins can catalyze the reduction of this mixed disulfide (Fig. 7, reaction 4). Formation of protein-glutathione-mixed disulfides is of physiological relevance for E. coli. Up to 2% of the total glutathione content (10–20 μm) is in the form of protein-mixed disulfides, and this value can be increased, as for example in trxA–grxA– mutants (5–7%) (45Miranda-Vizuete A. Rodriguez-Ariza A. Toribio F. Holmgren A. Lopez-Barea J. Pueyo C. J. Biol. Chem. 1996; 271: 19099-19103Abstract Full Text Full Text PDF PubMed Scopus (56) Google Scholar). In mammalian cells, extensive glutathionylation of protein substrates has been identified to include chaperons, cytoskeletal proteins, cell cycle regulators, and enzymes participating in the intermediary metabolism (46Lind C. Gerdes R. Hamnell Y. Schuppe-Koistinen I. von Löwenhielm H. Holmgren A. Cotgreave I. Arch. Biochem. Biophys. 2002; 406: 229-240Crossref PubMed Scopus (280) Google Scholar). Such a study has not, to date, been performed for E. coli. Glutaredoxins and their monothiol activity would be the molecules expected to key regulate deglutathionylation reactions and reverse related changes in biological activity. The identification and characterization of further proteins that undergo reversible S-glutathionylation and are specifically related to the glutaredoxin species will be necessary for a deeper understanding of cellular redox regulation and signaling. We are grateful to Professor Ole Petter Ottersen from the University of Oslo for the antibody against GSH. We thank Dr. Jochen Kruip from the Ruhr-University Bochum for help with mass spectrometry, Mathias Lundberg from our department for valuable suggestions, and Lena Ringdén for excellent secretarial work.