INTRODUCTION

Top Abstract Introduction Procedures Results Discussion References

Staphylococcus aureus is a Gram-positive opportunistic pathogen that is a major cause of nosocomial infections. S. aureus was the first organism identified as having resistance to penicillin and has since acquired resistance to most clinically approved antibiotics (1). Methicillin-resistant strains of S. aureus (MRSA)¹ are multiply drug-resistant and pose serious challenges to modern medicine and surgery. Current therapy for MRSA infections relies upon intravenous treatment with vancomycin. There is concern that resistance to vancomycin, which has emerged in Enterococcus faecalis, may transfer to MRSA, resulting in infections for which there may be no effective therapy (2). Efforts are now being focused to identify new cellular targets and to develop new chemotherapeutics effective against MRSA. We are investigating the unique thiol metabolism of S. aureus and other Gram-positive bacteria in hopes of facilitating this process.

Aerobic organisms maintain high levels of low molecular weight thiols that in combination with specific disulfide reductases support the reducing intracellular environment and the thiol/disulfide distribution of other thiol-containing molecules. Glutathione (GSH, -glutamyl-cysteinyl-glycine) is perhaps the best studied low molecular weight thiol. It protects these organisms from oxygen toxicity by functioning as a slowly autooxidizing reserve of cysteine and as a cofactor in the detoxification of peroxides, epoxides, and other products resulting from reaction with oxygen. It is also a cofactor in the reduction of disulfides and ribonucleotides and in the isomerization of protein disulfides (3-6). Glutathione reductase (GSR, EC 1.6.4.2) catalyzes the NADPH-dependent reduction of intracellular oxidized glutathione (GSSG) and thereby maintains the high intracellular GSH/GSSG ratio (GSH/GSSG > 100). Together, GSH, GSSG, and GSR make up the primary thiol/disulfide redox system of aerobic eukaryotes and Gram-negative bacteria (4, 5, 7, 8)

Many organisms do not produce GSH but instead employ alternative low molecular weight thiols that participate in unique thiol/disulfide redox systems (5, 7-12). It was recently shown that S.aureus, like many of the Gram-positive bacteria, produces no GSH but rather produces millimolar levels of reduced CoA as its predominant thiol (Fig. 1) (13). CoA, which is required as a cofactor in many acyl transfer reactions, likely has additional functions in S. aureus analogous to GSH. As a first step toward understanding this novel thiol metabolism, we set out to isolate and characterize the activity responsible for maintaining CoA in its reduced form. Other organisms that utilize alternative thiols produce an enzyme analogous to GSR whose preferred substrate is the disulfide of the predominant thiol in the cell (14-16). All such enzymes investigated belong to a widespread family of pyridine nucleotide-disulfide oxidoreductases that includes GSR, lipoamide dehydrogenase, and mercuric reductase. Most of these enzymes are homodimeric flavoproteins with M_r of ~100,000 that utilize a conserved active site disulfide bond to effect catalysis. We describe here the identification, purification, and characterization of a CoA disulfide reductase (CoADR) from S. aureus that catalyzes the NADPH-dependent reduction of CoA disulfide (Equation 1).

(Eq. 1)

In this and the following paper we highlight the similarities and the striking differences between CoADR and other members of the disulfide reductase superfamily.

View larger version (9K): [in this window] [in a new window]   Fig. 1.   Structures of GSH and CoA.

	EXPERIMENTAL PROCEDURES

Top Abstract Introduction Procedures Results Discussion References

Materials-- Reduced NADPH, riboflavin, flavin adenine mononucleotide, flavin adenine dinucleotide (FAD), coenzyme A disulfide, glutathionyl-coenzyme A mixed disulfide, 3'-dephospho-coenzyme A, 5,5'-dithio-bis(2-nitrobenzoic acid) (DTNB), lysostaphin, bicinchoninic acid solution, 4% copper sulfate pentahydrate solution, 2',5'-ADP-Sepharose, and phenylmethylsulfonyl fluoride (PMSF) were from Sigma (Mississauga, ON, Canada). 3'-dephospho-CoA was oxidized to the disulfide by incubation at room temperature overnight in Tris-HCl (20 mM), pH 9.0, containing copper (5 µM). EDTA was added to the solution (to 10 µM) chelate the copper. 4-Phosphopantetheine was formed by incubation of CoA with nucleotide pyrophosphatase. The thiol was oxidized as described above, and the disulfide was purified by HPLC. The purity of the disulfide was confirmed by HPLC and by thiol analysis upon reaction with monobromobimane (mBBr) (described below). All other chemicals were of reagent grade or better and were used without further purification.

Media, Assays, and Buffers-- S. aureus strain R8325-4, used in all studies, was from John Iandolo (Kansas State University, Manhattan, KS) and was grown in trypticase soy broth medium. Concentrations of FAD, NADPH, and CoA disulfide (and dephospho-CoA disulfide) were measured spectrophotometrically at 450 nm (₄₅₀ = 11,000 M¹ cm¹), 340 nm (₃₄₀ = 6220 M¹ cm¹), and 260 nm (₂₆₀ = 33,600 M¹ cm¹), respectively. DTNB assays were performed in Tris-HCl buffer (20 mM), pH 8.0, containing EDTA (1 mM) and were monitored at 412 nm for the thionitrobenzoate anion (₄₁₂ = 13,600 M¹ cm¹ (18)). Protein concentrations were determined by reaction with bicinchoninic acid and copper sulfate as described (19). During purification CoADR activity in crude extracts was monitored by the NADPH (1 mM) and CoA disulfide (1 mM)-dependent reduction of DTNB (1 mM). For kinetic analysis of purified CoADR, the oxidation of NADPH was measured as the decrease in absorbance at 340 nm (₃₄₀ = 6,220 M¹ cm¹) and was carried out in Tris-HCl buffer (50 mM), pH 7.8, containing NaCl (50 mM).

Analysis of Thiols from S. aureus-- An analysis of the thiols produced by S. aureus was carried out as described (20). Briefly, cell pellets (250 mg) were resuspended in 50% acetonitrile in Tris-HCl buffer (20 mM), pH 8.0, containing mBBr (2 mM) and incubated at 60 °C for 5 min in the dark. Control samples were pretreated with N-ethylmaleamide (NEM) (2 mM) under the same conditions before the addition of mBB (to 2 mM). The cellular debris was removed by centrifugation, and the samples were diluted in 10 mM aqueous methane sulfonic acid for reverse phase HPLC analysis. Thiol standards were prepared as described (20). To determine disulfide content, cell extracts were first treated with NEM (2 mM) to block the free thiols, then with dithiothreitol (3 mM) to reduce the disulfides and titrate the remaining NEM, and finally with mBBr to derivitize the resulting thiols. The thiol/disulfide ratio within the cell was then calculated.

Cu²⁺ Catalyzed Oxidation of Cysteine, GSH, and CoA-- To determine the relative stability of CoA to heavy metal catalyzed oxidation, the rates of Cu²⁺ catalyzed oxidation of cysteine, GSH, and CoA were compared. Each sample (2 ml) of thiol (1 mM) in Tris-HCl buffer (20 mM), pH 7.5, containing CuCl₂ (1 µM) was incubated at ambient temperature. Thiol determination was carried out at various times by adding aliquots (100 µl) from each sample to a solution (900 µl) of DTNB (1 mM) in Tris-HCl buffer (20 mM), pH 8.0, containing EDTA (1 mM). The absorbance of these samples at 412 nm were then measured, and the concentration of remaining thiol was determined.

Identification of a Coenzyme A Disulfide Reductase from S. aureus-- An overnight culture (10 ml) of S. aureus was centrifuged (5,000 × g for 10 min), resuspended in 3 ml of Tris-HCl buffer (20 mM), pH 8.0, containing lysostaphin (5 µg/ml), and incubated at 37 °C for 30 min until the suspension became viscous. Glass beads (1 g; 0.2 mm) and PMSF (to 1 mM) were added, and the mixture was vortexed for 2 min and then centrifuged (15,000 × g for 10 min) to remove the insoluble cellular debris. The resulting viscous lysate was dialyzed exhaustively (3,600 M_r cutoff) against Tris-HCl buffer (20 mM), pH 8.0, containing EDTA (1 mM). The dialysate (40 µl) was then assayed for the pyridine nucleotide (1 mM) and disulfide (1 mM)-dependent reduction of DTNB (1 mM). A typical assay was 800 µl.

Purification of CoADR from S. aureus-- An overnight culture of S. aureus was grown in trypticase soy broth (10 ml) at 37 °C and was used as an inoculum (0.4 ml) for each of ten 2-liter flasks containing trypticase soy broth (1 liter). These cells were shaken (180 rpm) for 12 h at 37 °C before being harvested by centrifugation (7000 × g for 15 min). All subsequent handling of the sample prior to chromatography was carried out at 4 °C. The cell pellet was resuspended in a minimum volume of Tris-HCl buffer (20 mM), pH 8.0, containing PMSF (1 mM) and lysostaphin (5.0 mg), incubated at 37 °C with agitation for 1 h (or until viscous), passed twice through a French pressure cell operating at 1,500 lb/in², and then centrifuged (15,000 × g for 20 min) to remove insoluble cellular debris. The supernatant was brought to 50% saturation with (NH₄)₂SO₄, stirred for 15 min, and centrifuged (15,000 × g, 10 min). The resulting supernatant was brought to 80% saturation with (NH₄)₂SO₄ and centrifuged, and the pellet containing the CoADR activity was dissolved in a minimum volume of TE buffer containing PMSF (1 mM). The resulting solution was dialyzed exhaustively (3,500 M_r cutoff) against TE buffer containing PMSF (1 mM).

Determination of Native Molecular Weight-- Purified CoADR (25 µg) was applied to a Superose 6 HR 10/30 gel exclusion column (0.5 ml/min) equilibrated with Tris-HCl buffer (20 mM), pH 8.0, containing NaCl (1 M) and then eluted isocratically in the same buffer. Fractions containing CoADR were identified by UV absorbance and by measuring enzyme activity. The native molecular weight of CoADR was calculated by comparison of its elution volume to that of protein standards as described (21).

Identification of Flavin Cofactor-- The visible absorption spectrum of purified CoADR is indicative of a flavoprotein. To identify the bound flavin cofactor, CoADR (5 µg or 100 pmol in 0.1 ml) was heated to 98 °C for 10 min and then centrifuged (15,000 × g, 10 min) to remove the denatured protein. The supernatant was then separated by reverse phase HPLC as described previously (15). The elution of the released cofactor was then compared with that of riboflavin, flavin adenine mononucleotide, and FAD (100 pmol each).

Quantitation of Active Site Thiols-- A solution of oxidized CoADR (9.5 µM) in Tris-HCl buffer (20 mM), pH 8.0, containing EDTA (1 mM) was incubated with NADPH (0.2 mM) for 10 min at ambient temperature before being diluted (1:1) with the same buffer containing urea (8 M) and DTNB (0.2 mM). The absorbance at 412 nm was then measured and compared with that of a similar reaction in which CoADR was not incubated with NADPH. The number of thiols liberated per FAD (₄₅₀ = 11,300 M¹ cm¹⁾ was then calculated. To determine the identity of the two thiols detected, oxidized CoADR (9.5 µM) in 100 µl of Tris-HCl buffer (20 mM), pH 8.0, containing NADPH (20 µM) and CoA disulfide (200 µM), 3'-dephospho-CoA disulfide (200 µM), or CoASSG (200 µM) was incubated (incubation 1) at 37 °C for 10 min, and then the reaction was filtered through a Centricon 30 (30,000 M_r cutoff) to separate CoADR from the low molecular weight reaction products. The extensively washed CoADR fraction was resuspended in 100 µl of Tris-HCl buffer (20 mM), pH 8.0, containing EDTA (1 mM) and NADPH (0.2 mM) and incubated for 10 min at 37 °C (incubation 2). The reaction product was either 1) filtered as above with the thiol content of the high and low molecular weight fraction determined by DTNB assay or 2) reacted directly with NEM and/or mBBr and then separated on HPLC as described above to identify any low molecular weight thiols present. The results were then compared with similar reactions in which oxidized CoADR was incubated alone or in the presence of either CoA disulfide or NADPH but not both.

Kinetic Characterization of the Substrate Specificity and Kinetic Mechanism of CoADR-- Kinetic measurements were performed in a 1-cm pathlength quartz cuvette maintained at 37 °C. Each assay (0.3 ml) was carried out in Tris-HCl buffer (50 mM), pH 7.8, containing NaCl (50 mM), CoADR (2-10 nM; subunit concentration determined using the calculated extinction coefficient for CoADR ₄₅₂ = 12,800 M¹ cm¹), NADPH (2-200 µM), and CoA disulfide (2-200 µM), 3'-dephospho-CoA A disulfide (10-500 µM), 4,4'-diphosphopantethine (2 µM to 2 mM), pantethine (10 µM to 100 mM), glutathione disulfide (10 µM to 100 mM), cystine (10 µM to 100 mM), CoA-glutathione mixed disulfide (10 µM to 1 mM), or H₂O₂ (10 µM to 10 mM) (EDTA (10 µM) was present in the stock solution of 3'dephospho-CoA disulfide and was thus present in the enzyme assays for this substrate up to 0.5 µM. We have found that EDTA up to 10 mM has no effect on the observed kinetics of the reaction catalyzed by CoADR.). Enzyme and NADPH were combined in buffer and equilibrated to 37 °C, and the reaction was initiated by the addition of the disulfide substrate or H₂O₂ (reduction of each substrate was monitored at 7 different substrate concentrations). The activity of CoADR was monitored at 340 nm as the decrease in absorbance resulting from the oxidation of NADPH. To study the kinetic mechanism of CoADR a two-substrate kinetic analysis was performed under identical conditions except for the concentrations of enzyme and substrate used: [CoADR] = 1.5 nM, [NADPH] = 4-30 µM (in reduction of CoA disulfide) and 6-10 µM (for reduction of 3'-dephospho-CoA; [CoA disulfide] = 4-80 µM; and [3'-dephospho-CoA] = 60-400 µM. All kinetic measurements were recorded in the linear range, and kinetic constants were calculated from a nonlinear least squares fit of the initial velocity data to the Michaelis-Menton equation using the program HYPERO (22).

	RESULTS

Top Abstract Introduction Procedures Results Discussion References

Analysis of Thiols from S. aureus-- Bimane derivatives of low molecular weight thiols separated by reverse phase HPLC are shown in Fig. 2. As described previously (13), S. aureus produces CoA, H₂S, and a small amount of cysteine. CoA is measured as the combination of 3'-dephospho-CoA and CoA, because the former is a hydrolysis product of CoA that arises during the acidic storage conditions prior to separation by the HPLC. Total CoA was estimated at 1.1 ± 0.1 µmol/g (dry residual weight). H₂S is commonly detected in bacterial extracts (13) and may derive from iron-sulfur proteins. The large peak running at 18 min was isolated and characterized by ¹H NMR as bismethylbimane (data not shown). This compound apparently arises from reaction of mBBr with an unknown cellular factor that is inactivated by NEM. If GSH was present, it was 0.02 µmol/g (dry residual weight). The predominant disulfide observed was CoA disulfide, and the ratio of reduced CoA/oxidized CoA was 450 ± 50. 

View larger version (29K): [in this window] [in a new window]   Fig. 2.   Reverse phase HPLC of S. aureus thiols. A, thiol sample, cell extracts were reacted with mBBr. B, control sample, cell extracts were reacted first with NEM and then with mBBr. Peaks that appear in chromatogram A but not chromatogram B are bimane derivatives of cellular thiols. C, bimane derivatives of standard thiols (Stds). BMB, bis methylbimane; MSH, mycothiol.

Cu²⁺ Catalyzed Oxidation of Cysteine, GSH, and CoA-- Biological thiols found as predominant cellular thiols, such as GSH, -glutamylcysteine, and mycothiol have proved to be kinetically resistant to metal catalyzed auto oxidation (12, 15). The relative resistance to Cu²⁺ catalyzed air oxidation for the intracellular thiols cysteine and CoA were compared with that for GSH (Fig. 3). CoA oxidized 4-fold less rapidly than GSH, whereas GSH oxidized 180-fold less rapidly than cysteine. Cysteine proved to be the least stable.

View larger version (25K): [in this window] [in a new window]   Fig. 3.   Cu2+ catalyzed autooxidation of cysteine, glutathione, and coenzyme A. Each thiol (1 mM) was incubated with CuCl2 (1 µM). At various times samples were removed, and total free thiol was measured by reaction with DTNB.

Identification and Purification of CoADR from S. aureus-- S. aureus grows aerobically, yet maintains millimolar levels of reduced CoA and an intracellular CoASH/CoA disulfide ratio of >100. A comparison of enzyme activity from an S. aureus extract incubated with various disulfide substrates and pyridine nucleotide cofactors was performed to determine whether any specific disulfide reductases were present. The extract catalyzed the reduction of DTNB slowly in the presence of either NADH or NADPH. However, in the presence of both NADPH and CoA disulfide this reduction was 3-fold greater than that observed for NADPH alone and 8-fold greater than that observed for NADH alone. No disulfide enhanced the rate of NADH-dependent reduction of DTNB, nor did GSSG, cystine, or pantethine enhance the NADPH-dependent reduction of DTNB. These results suggested that S. aureus produces a CoADR that catalyzes specifically the reduction of CoA disulfide by NADPH and does not produce detectable glutathione reductase.

View larger version (52K): [in this window] [in a new window]   Fig. 4.   Purification of CoADR from S. aureus. A, chromatogram for the elution of CoADR from a 2',5'-ADP-Sepharose column. B, chromatogram for the elution of CoADR from the MonoQ column. The arrows mark where the CoADR activity eluted. C, silver-stained SDS-PAGE analysis of different fractions containing CoADR activity: I, 50-80% (NH4)2SO4 fraction; II, 2',5' ADP active eluate; and III, MonoQ active eluate. D, the native molecular weight of CoADR was determined by gel filtration chromatography and a plot of the parameter K versus the log of the molecular weight of various protein standards. The parameter, K, is defined as (Ve  Vo)/Vs, where Ve is the elution volume, Vo is the column void volume, and Vs is the volume of the stationary phase (21).

Physical Characterization-- The known pyridine nucleotide-disulfide oxidoreductases are multimeric, most being homodimers, and have a FAD noncovalently bound to each subunit (14, 16, 24, 25). Purified CoADR migrates as a single polypeptide of ~50,000 apparent molecular weight according to SDS-PAGE (Fig. 4C). The native molecular weight of CoADR was estimated by its relative elution through a Superose 6 HR 10/30 gel filtration column. A plot of the parameter, K, verses the log of the molecular weight of various protein standards is shown in Fig. 4D. The parameter, K, is defined as (V_e  V_o)/V_s, where V_e is the elution volume, V_o is the column void volume, and V_s is the volume of the stationary phase (21). Native CoADR elutes between chicken ovalbumin (44 kDa) and bovine gamma globulin (158 kDa) from the Superose gel filtration column, and the K value calculated for CoADR (0.49) can be extrapolated from the standard plot to a M_r of approximately 90,000. This suggests that native CoADR is a homodimer.

1

1

1

1

(Eq. 2)

View larger version (20K): [in this window] [in a new window]   Fig. 5.   A, the absorption spectrum of purified CoADR. CoADR shows absorbance peaks typical of flavoproteins (max at 276, 375, and 452 nm). B, identification of the flavin cofactor of CoADR by reverse phase HPLC. a, flavin standards: riboflavin, flavin adenine mononucleotide (FMN), and FAD. b, the flavin released from thermally denatured CoADR.

Thiol Titration of Active Site-- The deduced amino acid sequence of CoADR contains two cysteine residues, and only one of these (Cys⁴³) lies in the putative active site region (36). To determine whether CoADR utilizes a thiol-disulfide exchange mechanism, a thiol titration of the active site of the enzyme was performed. CoADR in its oxidized state and reduced with NADPH was denatured in the presence of DTNB. The denatured oxidized enzyme had 0.9 ± 0.1 reactive thiol/subunit, whereas denatured CoADR that had been reduced with NADPH had 3.2 ± 0.2 reactive thiols/subunit. This suggests that CoADR in its oxidized state has one reduced cysteine and that upon reduction with NADPH a disulfide bond was reduced. However, the 3.2 thiols/subunit measured also suggested that one of the thiols produced upon reduction with NADPH was not an enzymic cysteine (because CoADR only has two cysteines/subunit) (36).

Kinetic Characterization of CoADR-- CoADR is highly specific for its physiological substrates, CoA disulfide and NADPH, and at the optimum pH 7.8 is saturated by micromolar concentrations of each. The K_m for NADPH, at a saturating CoA disulfide concentration, was 2 µM, and the K_m for CoA disulfide, at a saturating NADPH concentration, was 11 µM. The results of the kinetic analysis of the CoADR catalyzed reduction of various disulfide substrates by NADPH are shown in Table III. Although CoADR reduces CoA disulfide preferentially, it also reduces 3'-dephospho-CoA disulfide (K_m = 40 µM), 4,4'-diphosphopantethine (K_m = 80 µM), and CoASSG (K_m = 1100 µM). Activity of CoADR upon pantethine, cystine, GSSG, or peroxide was below the detectability of the assay (A₃₄₀  0.0001 min¹) and thus has an upper limit of 0.1%, the activity observed for CoA disulfide. A two-substrate kinetic analysis of the NADPH-dependent reduction of CoA disulfide and 3'-dephospho-CoA disulfide by CoADR suggests that the reaction proceeds by a sequential rather than a simple ping-pong kinetic mechanism (Fig. 6).

View this table: [in this window] [in a new window]   Table III Steady-state kinetic analysis of the substrate specificity of CoADR The rate of NADPH oxidation catalyzed by CoADR in the presence of various disulfide substrates was spectrophotometrically followed. The reactions were carried out in Tris-HCl buffer (50 mM), pH 7.8, containing NaCl (50 mM), and NADPH (200 µM). CoA disulfide was maintained at 120 µM to determine the kinetic parameters for NADPH. Kinetic parameters for reactions having rates below that detectable by the assay (A260  0.0001/min) were not determined (ND).

View larger version (38K): [in this window] [in a new window]   Fig. 6.   Two-substrate analysis of the kinetic mechanism of CoADR. Enzyme and NADPH were incubated at 37 °C, and the reactions were initiated by the addition of the disulfide substrate. The kinetic parameters for each set of data were determined by fitting the initial velocity data to the Michaelis-Menton equation using HYPERO (22). The equation for each line in these figures is simply the reciprocal of the Michaelis-Menton equation carrying the measured kinetic parameters. Each line results from varying the concentration of CoA disulfide at constant NADPH concentrations: 1 (), 6 (), 8 (), 10 (), and 30 µM ().

	DISCUSSION

Top Abstract Introduction Procedures Results Discussion References

Although cysteine is ubiquitous, it rarely is the predominant cellular thiol of aerobic organisms. High intracellular levels of cysteine apparently are toxic because rapid metal catalyzed autooxidation results in the production of cystine and hydrogen peroxide (27). Glutathione is resistant to metal catalyzed oxidation (Fig. 3) (10) and provides cells with a reserve of cysteine and a stable intracellular redox environment. It was shown that S. aureus does not produce GSH but instead produces CoA as its predominant cellular thiol (Fig. 2) (13). Like GSH, CoA is stable to metal catalyzed oxidation (Fig. 3), is found at millimolar levels in the cell and is maintained in its reduced form (CoASH/CoASSCoA  450). This thiol/disulfide ratio is likely an underestimate because the calculated disulfide content would be contaminated by cellular thioesters, which liberate thiol upon treatment with dithiothreitol (28). CoA thus represents a stable intracellular thiol/disulfide redox buffer in S. aureus. The current understanding of cysteine biosynthesis, however, does not provide an efficient route for conversion of CoA to cysteine. If indeed aerobes store cysteine in a form that is resistant to oxidation, the manner by which S. aureus stores cysteine is not obvious.

The thiol/disulfide ratio of thiols in equilibrium with air is ~10¹⁶, yet aerobes maintain a cytoplasmic thiol/disulfide ratio of 10²-10³ (29). S. aureus maintains this ratio, at least in part, using the enzyme CoADR, which catalyzes specifically the NADPH-dependent reduction of CoA disulfide. CoADR is purified over 3000-fold from S. aureus by 2',5'-ADP affinity and anion exchange chromatography, resulting in a 25% recovery of the active enzyme (Table I). Chromatography with 2',5'-ADP-Sepharose, which mimics the adenyldiphosphate moiety of NADPH, is the most effective step in the purification of CoADR, and the concentration of NaCl (~2 M) necessary to elute CoADR from this column reflects the high affinity CoADR has for this cofactor (K_m (NADPH) = 2 µM). Native CoADR has an apparent M_r of 90,000 ± 10,000 according to gel filtration chromatography (Fig. 4D) and runs as a single 49-kDa band on an SDS-PAGE gel (Fig. 4C). These data are consistent with a subunit M_r of 49,200 calculated from the deduced primary structure of CoADR (36) and suggest that CoADR is a homodimer. The absorption spectrum (Fig. 5A) and HPLC analysis (Fig. 5B) show CoADR to be a flavoprotein that noncovalently binds a single FAD to each 49-kDa subunit.

The known pyridine nucleotide-disulfide oxidoreductases utilize a pair of active site cysteine residues to catalyze a thiol disulfide exchange reaction with the disulfide substrate (23, 24). In human erythrocyte GSR, Cys⁵⁸, the interchange cysteine, reduces the substrate disulfide resulting in the formation of a transient mixed disulfide intermediate. Cys⁶³, the charge transfer cysteine, is adjacent to the enzyme bound FAD and reduces the mixed disulfide effecting the release of substrate and the formation of cystine 58-63. The deduced primary structure of CoADR does not have the conserved CXXXXC motif observed for other disulfide oxidoreductases (36). CoADR instead has a single cysteine residue (Cys⁴³) that is homologous to the active site cysteine residue of the SFXXC motif of NADH oxidase and NADH peroxidase from E. faecalis (30). These enzymes, also members of the pyridine nucleotide-disulfide oxidoreductase superfamily, utilize this single catalytic cysteine to form a stable sulfenic acid (Cys-SOH) during the reduction of O₂ and H₂O₂. We have shown that the NADPH reduction of oxidized CoADR results in the production of two thiols/enzyme subunit, of which half originates from an enzymic cysteine residue (presumably Cys⁴³) and half originates from the disulfide substrate (Table II). These results suggest that reduction of disulfides by CoADR occurs by a thiol-disulfide exchange reaction but involves only a single cysteine residue that forms a stable mixed disulfide (Enz-SSR) analogous to the sulfenic acid formed by NADH peroxidase. A proposed catalytic mechanism for CoADR is shown in Equation 2.

(Eq. 3)

(Eq. 4)

In the reduced enzyme Cys⁴³ is a free thiol. Upon encounter with CoA disulfide the Cys⁴³ thiolate anion attacks one sulfur atom of the substrate, displacing a molecule of CoA and forming a Cys⁴³-CoA mixed disulfide. This oxidized state is stable and is the enzyme form isolated from cells (reduction of this sample with NADPH releases CoA; Table II). NADPH binds the oxidized enzyme and reduces the mixed disulfide via the FAD cofactor, regenerating the reduced enzyme and another molecule of CoA. However, the sequential kinetics observed from a two-substrate kinetic analysis (Fig. 6) suggests that CoADR does not utilize a simple two-site ping-pong mechanism. The specific sequence of binding events and the flow of electrons cannot be addressed from the present data but will have to await future stopped flow kinetic studies.

CoADR is a very specific for NADPH (K_m = 2 µM) and CoA disulfide (K_m = 11 µM), suggesting that these compounds are natural substrates for this enzyme. A kinetic survey of disulfide substrates (Table III) demonstrates the specificity of CoADR and the contribution of the 3'-phosphoryl, the adenyl, and the 4-phosphoryl groups to the binding and turnover of CoA disulfide. This analysis shows that CoADR recognizes only substrates containing the 4-phosphopantethiene moiety and is most proficient in the reduction of CoA disulfide. A comparison of the kinetic constants for 3'-dephospho-CoA disulfide reveals that loss of the 3' phosphate moieties from CoA disulfide results in a 10-fold increase in K_m with no detectable change in k_cat. A similar comparison for 4,4' diphosphopantethine and 3' dephospho-CoA disulfide demonstrates that loss of the adenyl moiety results in a 2-3-fold decrease in k_cat and little change in K_m. The 4 and/or 4' phosphate moieties are essential for substrate binding and turnover because reduction of pantethine by CoADR is not detectable. CoADR does reduce CoASSG, although poorly (K_m = 1.4 mM), but does not reduce GSSG, cystine, O₂, or H₂O₂ (despite the homology to E. faecalis NADH oxidase and peroxidase; Ref. 36). It should be noted that upon reduction of CoASSG CoADR selectively forms a mixed disulfide with CoA (Table II). Thus, the active site is designed to form a mixed disulfide with the more tightly bound portion of the substrate.

Although CoADR apparently belongs to a new separate subfamily of disulfide reductases (36), CoADR is physically and functionally similar to the pyridine nucleotide-disulfide oxidoreductase superfamily. Indeed, all of the low molecular weight disulfide reductases appear to belong to this family. Enzymes studied include trypanothione reductase (14, 31), bis--glutamyl-cysteine reductase (15), and 4,4'-diphosphopantethine reductase (16). Although each appears to catalyze the same chemical reaction, the NAD(P)H reduction of a disulfide bond, each enzyme is specific for the disulfide of the predominant thiol in the cell. CoADR is related functionally to the Bacillus megaterium 4',4'-diphosphopantethine reductase (4PPR). Although both CoADR and 4PPR each prefer substrates having the 4'-phosphopantethienyl moiety, these enzymes are quite different. 4PPR has a K_m approximately 10-fold higher for CoA disulfide (7.6 mM) than for its preferred substrate 4,4'-diphosphopantethine (0.67 mM) (16), which itself is 2 orders of magnitude higher than the K_m for turnover of CoA disulfide by CoADR. 4PPR is also 20% larger. It is interesting that B. megaterium produces an enzyme of such specificity, because B. megaterium likely has a thiol content similar to that of B. subtilis, which like S. aureus produces CoA as its predominant thiol. The B. megaterium enzyme was identified and isolated from spores and may have evolved to function in this dormant state. If so, there may be an additional enzyme such as CoADR that is expressed in vegetative cells.

GSH plays many different roles in cellular metabolism. How these functions occur in organisms lacking GSH represents an enigma in the field of thiol biology. In organisms like S. aureus, which produce CoA as a predominant thiol, many GSH-mediated functions are presumably mediated by CoA and enzymes that utilize CoA as cofactor. Enzymes present in pathogens that utilize cofactors structurally distinct from their eukaryotic counterparts may be potential drug targets because cofactor discrimination could be exploited in the design of selective drugs. Indeed, successes in this area have been reported in the design of inhibitors to trypanothione reductase (32-35). The disparate disulfide specificities, structures (36), and catalytic mechanisms of CoADR and GSR suggest that CoADR could be a useful target for developing new drugs. The study of bacterial thiol metabolism will likely unveil many such promising targets.