Inhibition of Glutathione Reductase by Dinitrosyl-Iron-Dithiolate Complex*

The biological signal molecule nitric oxide (NO) exists in a free and carrier-bound form. Since the structure of the carrier is likely to influence the interaction of NO with macromolecular targets, we assessed the interaction of a dinitrosyl-iron-dithiolate complex carrying different thiol ligands with glutathione reductase. The enzyme was irreversibly inhibited by dinitrosyl-iron-di-l-cysteine and dinitrosyl-iron-di-glutathione in a concentration- and time-dependent manner (IC50 30 and 3 μm, respectively). Evaluation of the inhibition kinetics according to Kitz-Wilson yielded a K i of 14 μm, and ak 3 of 1.3 × 10−3s−1. A participation of catalytic site thiols in the inhibitory mechanism was indicated by the findings that only the NADPH-reduced enzyme was inhibited by dinitrosyl-iron complex and that blockade of these thiols by Hg2+ afforded protection against irreversible inhibition. This inhibition was not accompanied by formation of a protein-bound dinitrosyl-iron complex and/orS-nitrosation of active site thiols (Cys-58 and Cys-63). However, one NO moiety exhibiting an acid lability similar to a secondary N-nitrosamine was present per mol of inhibited monomeric enzyme. These findings suggest specificallyN-nitrosation of glutathione reductase as a likely mechanism of inhibition elicited by dinitrosyl-iron complex and demonstrate in general that structural resemblance of an NO carrier with a natural ligand enhances NO+ transfer to the ligand-binding protein.

Nitrosyl transfer from endogenous nitric oxide (NO) 1 carriers such as S-nitrosoglutathione (1) and dinitrosyl-iron complex (DNIC) (2) to macromolecular targets is regarded as one major mechanism of biological NO signaling (3). Evidence has been provided that endogenous low mass S-nitrosothiols (predominantly S-nitrosoglutathione) and proteinaceous S-nitrosothiols (S-nitroso-hemoglobin, S-nitroso-serum albumin, and other yet unidentified proteins) exist in human erythrocytes (4), plasma (5) and bronchial secretion (6). A further NO adduct with GSH, GSNOH, was recently postulated as another transport form of NO (7). S-Nitrosation of protein thiols or subsequent reactions such as ADP-ribosylation (8), formation of protein disulfide (9), and cysteine sulfenic acid (10) may influence protein function by allosteric mechanisms. Furthermore, reversible S-nitrosation of cell membrane-bound proteins may be involved in transmembraneous NO transport (11).
A concept has been derived from established chemistry to account for the influence of the redox state of the NO moiety on biological NO transfer reactions. According to this concept nitrosation of nucleophilic targets occurs by attack of nitrosonium (NO ϩ )-like species assumed to be present in "NO carriers" such as N 2 O 3 , S-nitrosothiols, and certain iron-nitrosyl complexes (3,12). Less attention has been paid to the influence of the carrier structure on the interaction of NO with macromolecular targets. However, it is conceivable that the NO carrier will direct the NO moiety specifically to macromolecules recognizing the carrier structure, provided the NO adduct is sufficiently stable and the binding kinetics between the carrier and the macromolecules are rapid enough to outbalance the decomposition of the NO carrier adduct. There is also evidence that NO adducts may exhibit intrinsic bioactivity independent of NO release. Thus, the L-stereoisomer of S-nitrosocysteine was found to exhibit a significantly higher blood pressure lowering activity compared with the D-isomer, suggesting the existence of stereospecific S-nitroso-L-cysteine receptors in the cardiovascular system (13).
To assess the influence of the carrier structure on the interaction of NO with a given macromolecule we chose glutathione reductase (GR; EC 1.6.4.2) as a model target, and low mass dinitrosyl-iron complexes with L-cysteine and glutathione ligands as NO carriers. These complexes exhibit S-nitrosating activity toward serum albumin in vitro (2), and protein-bound forms exist in vivo in animal tissues expressing inducible NO synthase activity (14). The flavoenzyme GR catalyzes the NADPH dependent reduction of oxidized glutathione (GSSG) to maintain a high intracellular level of GSH. GR carries a redoxactive disulfide (Cys-58 -Cys-63) in its active site which is reduced by electron transfer from NADPH via the flavin (15). Recently it has been shown that GR is inhibited by certain NO carriers (S-nitrosoglutathione, sodium nitroprusside, S-nitroso-N-acetyl-DL-penicillamine) in millimolar concentrations (16), suggesting that GR is a potential target for nitrosation reactions.
We show here that low mass dinitrosyl-iron complexes in concentrations that may be present under pathophysiological conditions irreversibly inhibit GR, possibly via N-nitrosation. We furthermore demonstrate that the inhibitory potency of the dinitrosyl-iron moiety increases with structural resemblance of the NO carrier to the natural GR substrate, GSSG.

EXPERIMENTAL PROCEDURES
Materials-GR from bovine intestinal mucosa, fatty acid-free bovine serum albumin (BSA), 2,3-diaminonaphthalene, L-cysteine, glutathione * This work was supported by the Deutsche Forschungsgemeinschaft (SFB "Stickstoffmonoxid: Generator-und Effektorsysteme," Teilprojekt C3). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
Paramagnetic dinitrosyl-iron complexes of the type ((NO) 2 Fe(RS) 2 ) ϫ 18 RSH (RS ϭ L-cysteine, or GSH) were synthesized by mixing evacuated (5 min high vacuum) solutions of FeSO 4 (5 mg/ml) and neutralized thiols (72 mM) in a Thunberg-type reaction vessel under pure NO gas (P NO 500 mm Hg). NO was added 3 min before mixing. The solution immediately turned dark green and was evacuated after 1 min for a further 2 min to remove excessive NO. The solution bearing ((NO) 2 Fe(RS) 2 ) in Ͼ98% yield with respect to iron was immediately frozen and stored in liquid nitrogen.
Determination of S-Nitrosothiols and Nitrite-At concentrations Ͼ 1 M S-nitrosothiols were assessed by diazotization of sulfanilamide and azocoupling with N,N-ethylendiamine in the presence and absence of Hg 2ϩ ions (3 mM) according to Saville (18) as described recently (2). In the nanomolar range nitrite and S-nitrosothiols were quantified by an acid-catalyzed intramolecular diazotization reaction of 2,3-diaminonaphthalene with nitrite forming the highly fluorescent product 2,3diaminonaphthotriazole (19). The sample (570 l) was mixed with 30 l of 0.1 M potassium P i buffer, pH 7.4, and 84 l of freshly prepared 2,3-diaminonaphthalene (0.05 mg/ml in 1 M HCl). To release NO from S-nitrosothiols the buffer contained 20 mM HgCl 2 . Hg 2ϩ did not interfere with the fluorescent assay. After 10 min of incubation at 20°C in the dark the reaction was terminated by addition of 42 l of 4.5 M NaOH to maximize the intensity of the fluorescent signal (19). The fluorescence was measured with excitation at 375 nm and emission at 415 nm (Deltascan, Photon Technology). Emitted light was detected by a photon-counting photomultiplier (D-104, Photon Technology) and the photomultiplier digital output was collected by an IBM-compatible computer. The content of S-nitrosothiol was calculated by the difference of emission readings of Hg 2ϩ -containing versus Hg 2ϩ -free samples. A calibration curve was established in each experiment with freshly synthesized S-nitroso-L-cysteine and sodium nitrite as standards (0.02-2 M). The detection limit was 20 nM.
EPR Spectroscopy-EPR spectra were recorded on a Bruker EPR 300E spectrometer at 20°C on solutions (25 l) filled in a quartz capillary tube (1 mm, inner diameter). The measurements were performed with a modulation amplitude of 1 Gauss, a microwave frequency of 9.6 GHz, a microwave power of 20 mW, and a time constant of 0.2 s. The concentration of dinitrosyl-iron complex was calculated by comparison with the EPR signal of a standard low molecular mass dinitrosyliron complex based on double integration of the first derivative EPR signals.
Chemical Modifications of Glutathione Reductase-Histidine residues were carboxylated by adding a 100-fold molar excess of DEPC and subsequent incubation for 5 min at 20°C (20). Thiol groups were blocked by incubation of the protein (5 M protein in 0.1 M potassium phosphate buffer, pH 7.4) for 5 min at 20°C with Hg 2ϩ (15 M) in the presence of NADPH (1 mM). To further demonstrate the involvement of catalytic site thiols in DNIC-induced inhibition of GR the enzyme (5 M) was preincubated with Hg 2ϩ (15 M; 5 min at 20°C) prior to incubation with DNIC-GSH (30 M; 10 min at 37°C). Controls were performed without Hg 2ϩ and in the absence and presence of DNIC. The solutions (100 l) were desalted at 5°C by passing over a Sephadex G50 Nick® column (Pharmacia) equilibrated with assay buffer (see below) and then incubated with dithiothreitol (DTT) (10 mM, 37°C) for up to 70 min. After 10, 30 or 70 min of DTT-treatment the activity of GR was assessed in 1:250 diluted aliquots as described below.
Determination of Glutathione Reductase Activity-The reduction of GSSG by GR was determined at 20°C by monitoring the oxidation of NADPH at 340 nm (⑀ 340 ϭ 6200 M Ϫ1 cm Ϫ1 ) (21). The enzyme was diluted in assay buffer (200 mM potassium chloride, 1 mM EDTA, 50 mM potassium phosphate, pH 6.9). To avoid an interference by NADPHoxidase activity both reference and sample cuvettes contained NADPH (0.38 mM) and GR (3.5-20 nM) in a final volume of 1 ml. The reaction was started by addition of GSSG (1 mM) to the sample cuvette. The enzyme activity was calculated from the initial rate of the absorbance decrease at 340 nm during 3 min of incubation. For establishment of concentration-response relationships, 1-5 M GR was preincubated with inhibitors for 30 min and then diluted 300 -1000-fold in the final assay mixture. For characterizing individual residues of the bovine GR, the numbering system of the well studied human enzyme (15,16) was used; the active site thiols are Cys-58 and Cys-63, and the catalytic imidazole is His-467.
Kinetics of Irreversible Inhibition-Irreversible inhibition kinetics were assessed according to Kitz and Wilson (22) based on the following reaction scheme (Equation 1).
If k 3 , the rate constant for transformation of the reversible enzymeinhibitor complex (E Q I) into the inhibited enzyme (E*), is relatively small, enzyme (E) and inhibitor (I) are in equilibrium with the reversible inhibitor enzyme complex. Also, in case of irreversible inhibition k 4 can be neglected. The integrated rate law derived from mass balances reads k 3 is the first order rate constant at high inhibitor concentrations ([I] Ͼ Ͼ K i ). At low inhibitor concentrations the kinetics are in accordance with a simple bimolecular mechanism (k app ϭ (k 3 /K i )⅐I), and the second order rate constant is (k 3 /K i ). K i and k 3 were derived from a double-reciprocal plot of the apparent first order rate constants k app versus concentration of inhibitor [I], i.e. dinitrosyl-iron complex. It should be noted that K i differs from the half-maximal inhibitory concentration (IC 50 ), since K i describes the reversible pre-equilibrium, while the IC 50 refers to the subsequent irreversible reaction. Measurement of Guanylyl Cyclase Activity-To assess the release of bioactive NO from inhibited GR soluble guanylyl cyclase (GC) purified to apparent homogeneity from bovine lung was used as a detector system (23). GC activity was measured by the formation of [ 32 P]cGMP from [␣- 32

Inhibition of GR by DNIC-
To study the influence of DNIC on the activity of isolated GR, the enzyme was incubated with different concentrations of DNIC-L-cysteine and DNIC-GSH in the presence of NADPH and substrate (GSSG). The GSSGdriven consumption of NADPH was monitored continuously by recording the decrease in absorbance at 340 nm (Fig. 1). During 3 min of reaction at 20°C, samples containing DNIC-GSH exhibited a nearly constant rate of decrease in absorbance (data not shown), which was inversely related to the concentration of DNIC used. In contrast, NADPH consumption in DNIC-L-cysteine containing reaction mixtures initially exhib-ited exponential kinetics (Fig. 1, A and B), indicating that the degree of inhibition of GR by this DNIC derivative progressively increased with time. At longer preincubation periods (Ͼ20 min) in the absence of GSSG the kinetics of inhibition by DNIC-L-cysteine became more linear (data not shown). Thus, it is conceivable that during incubation of DNIC-L-cysteine with GR in the presence of GSSG DNIC-GSH was formed by reaction with enzymatically generated GSH. In addition, during preincubation in the absence of GSSG an intrinsic slow inhibitory action of DNIC-L-cysteine was revealed (Fig. 1B). Therefore, to establish the dose-response relationship for inhibition of GR by DNIC, the enzyme (1 M) was preincubated at 20°C in the absence of GSSG with different concentrations of low molecular mass DNIC (6 -200 M) in buffer containing 1 mM NADPH. The enzyme activity was then measured after 30 min of preincubation. It decreased in a DNIC concentration-dependent manner (Fig. 2). The GSH complex was about 10-fold more potent than the L-cysteine complex, although both were equally efficacious. 50% inhibition was elicited by 3 Ϯ 1 M DNIC-GSH and by 30 Ϯ 3 M DNIC-L-cysteine. No direct oxidation of NADPH or reduction of NADP ϩ by DNIC could be observed.
GR carries a redox-active disulfide (Cys-58 -Cys-63) at its catalytic site, which is reduced after binding of NADPH. To investigate whether the inhibition by DNIC depends on the redox state of the enzyme, GR (5 M) was incubated with DNIC-GSH (30 M) at 37°C in the presence and absence of NADPH (1 mM). The activity of GR decreased to 5 Ϯ 2% (n ϭ 4) of control in the presence of NADPH, but did not change in its absence (100 Ϯ 3%, n ϭ 4). This finding suggests that only the reduced enzyme is susceptible to inhibition by DNIC. Since the homodimeric form is essential for catalytic function of GR (15) we assessed whether DNIC influenced the aggregation state of the native enzyme by gel permeation chromatography (Superose 200, Pharmacia). DNIC-treated and untreated GR migrated as single peaks at identical positions with an apparent molecular mass of 100 kDa. Therefore DNIC does not inhibit GR by promoting dissociation of the homodimeric enzyme.
NO-mediated oxidation and S-nitrosation of protein thiols usually is reversible by an excess of low molecular weight thiols (1,24,25). To test the reversibility of the GR modification by DNIC, the complex-inhibited GR was treated with 5 mM DTT or dimercaptopropanol (30 min, 20°C); neither reducing agent was able to restore enzyme activity. Moreover, dilution (1000fold) of the inhibited enzyme in thiol-free or thiol-containing buffer failed to restore activity even after 24 h. To test the accessibility of low mass thiols to the catalytic site thiols in GR, we analyzed the reversibility of Hg 2ϩ -elicited inhibition of GR by dithiols. Hg 2ϩ exhibits a high affinity for thiol groups. GR (5 M) was inhibited by the treatment with Hg 2ϩ (15 M) within 60 s at 20°C. DTT and dimercaptopropanol regenerated the initial activity (5 mM, 10 min, 20°C) of the enzyme, indicating that the Hg 2ϩ -bound thiol groups at the active center were accessible to both agents (see also Williams (15)). Therefore the dithiols used should be able to interact with the thiol groups in the complex-inhibited enzyme without sterical hindrance. Altogether these findings show that inhibition of GR by DNIC is irreversible.
To assess the involvement of cysteine-thiols within the catalytic site in DNIC-induced inhibition of GR we examined whether or not pretreatment of GR by Hg 2ϩ , which inhibits GR in a thiol-reversible manner (see above), affords protection against the irreversible inhibition of GR by DNIC-GSH. Since GR contains 3 cysteine thiols per subunit, NADPH-reduced GR was pretreated with a 2-fold molar excess of Hg 2ϩ prior to incubation with a maximally inhibitory concentration of DNIC-GSH (30 M). Half of the original GR activity was restored within 10 min following incubation of Hg 2ϩ /DNIC-and Hg 2ϩtreated GR with DTT. DTT, however, failed to restore catalytic activity to GR treated with DNIC only. The degree of inhibition after 10 min of DTT treatment was: DNIC-treated GR, 85 Ϯ 10%; Hg 2ϩ -treated GR, 45 Ϯ 4%; Hg 2ϩ /DNIC-treated GR, 52 Ϯ 7% (mean Ϯ S.E.; n ϭ 3). This inhibition was not significantly altered after either 30 or 70 min of treatment with DTT. Thus, Hg 2ϩ pretreatment protects GR from inhibition by DNIC. Consequently, thiols within the catalytic site of GR appear to be the main targets of DNIC and are involved in the irreversible inhibition.
Kinetics of Inhibition-To study the kinetics of GR inhibition by DNIC the enzyme (1 M) was exposed at 20°C to 0, 6, 10, 15, Chemical Characterization of DNIC-modified GR-The following experiments were performed to reveal the molecular mechanism of inhibition of GR by DNIC. The dinitrosyl-iron moiety of low molecular DNIC binds to free thiol groups of proteins due to a thiol-ligand exchange reaction (2,26). To assess whether or not the dinitrosyl-iron group was attached to the inhibited protein, GR (5 M) was incubated with different concentrations of DNIC-GSH or DNIC-Lcysteine (5-50 M) at 20 or 37°C. After 3, 10, and 60 min the mixture was analyzed by EPR spectroscopy at 20°C. Recording EPR signals at room temperature allows discrimination between DNIC bound to low and high molecular mass ligands, since the former exhibits an isotropic signal at g av 2.03 with 13-line hyperfine structure (Fig. 5a), while the latter is characterized by an anisotropic signal at g Ќ 2.04 and g ʈ 2.01 (Fig. 5c). The initial reaction mixtures exhibited exclusively the EPR signal of the low molecular weight DNIC (Fig. 5b). After 60 min of reaction this signal completely disappeared (Fig. 5d) because of decomposition of low mass DNIC. The characteristic signal of the proteinbound DNIC (serum albumin-DNIC; Fig. 5c) was not detectable at any time, though the enzyme was completely inhibited after 30 min of incubation. These findings show that inhibition of GR does not involve formation of a stable DNIC-protein linkage.
Since free cysteine thiols within proteins can be nitrosated by low molecular weight DNIC (2), we assessed whether this covalent modification accounts for inhibition of GR by DNIC. Therefore DNIC-inactivated enzyme (5 M, 95 Ϯ 3% inhibited) was passed through a desalting column (Sephadex G-25) to remove excess inhibitor. The protein was concentrated by centrifugation (Ultrafree 30-kDa cut-off, Millipore) and assayed for NO x and S-nitrosothiol (18). Neither chromatography nor centrifugation reversed inactivation of GR. In the presence of Hg 2ϩ 0.76 Ϯ 0.04 mol nitrite/mol inactive GR was detected (n ϭ 3), but a similar amount of nitrite was found in the absence of Hg 2ϩ (0.75 Ϯ 0.08). This indicates that inhibited GR contains a Griess-reactive NO moiety, which is not bound to a thiol group.
Hence it was investigated whether a N-or C-nitrosation of GR by DNIC accounts for inhibition. N-Nitrosopyrrolidine, a nitrosamine of a cyclic secondary amine, also released NO independently of Hg 2ϩ under the acid conditions of the Griess reaction (0.25 M HCl). After 30 min 50 Ϯ 4 M nitrite was generated by this agent (100 M). In contrast, the C-nitroso compound 1-nitroso-2-hydroxynaphthalene-3,6-disulfonic acid (50 M, 60 min incubation), failed to give a positive Griess reaction, either in the presence or absence of Hg 2ϩ . To increase the sensitivity of the Griess reaction the inhibited GR was also assessed for S-nitrosothiols by the 2,3-diaminonaphthalene assay. Under the mildly acid conditions (0.1 M HCl) of this assay neither nitrite nor S-nitrosothiol could be detected. However when the inhibited GR was preincubated in 0.25 M HCl for 30 min at 37°C and then neutralized with NaOH, 0.83 Ϯ 0.05 mol nitrite/mol enzyme were found by the 2,3-diaminonaphthalene assay. This finding reveals that the NO moiety bound to GR exhibits a peculiar acid lability. A similar acid lability was exhibited by N-nitrosopyrrolidine. The N-NO bond was split in 0.25 M HCl but not in 0.1 M HCl.
To further substantiate that the NO moiety is firmly bound to DNIC-inhibited GR under neutral conditions, purified soluble guanylyl cyclase (GC) was used as a sensitive NO detector. DNIC-inhibited GR was desalted and then incubated with GC for assessment of cGMP formation. The basal GC activity was not influenced by DNIC-inhibited GR (400 nM). In contrast, the reference S-nitroso protein S-nitroso-serum albumin at 10-fold lower concentration (40 nM) stimulated GC activity 17-fold. N-Nitrosopyrrolidine (4 M) and 1-nitroso-2-hydroxynaphthalene-3,6-disulfonic acid (4 M) did not enhance GC activity (data not shown). These findings tend to exclude the possibility that DNIC-inhibited GR carries a labile S-nitrosothiol moiety, but favor the concept of the formation of a stable N-nitroso group.
According to the tertiary structure of GR derived from x-ray diffraction analysis, a histidine residue (His-467) is located in the direct vicinity of Cys-58 (35). During catalysis of GSSG reduction His-467 takes a proton from Cys-58, thereby facili- tating the nucleophilic attack of the resulting thiolate anion on one sulfur atom of the substrate GSSG (27). Consequently one imidazole-nitrogen of His-467 could function as an NO ϩ acceptor. To demonstrate the key function of His-467 in catalysis, the histidine residues of GR (2 M) were carboxylated by DEPC (200 M) at 20°C. DEPC treatment completely inactivated GR within 60 s (n ϭ 3; data not shown). No direct oxidation of NADPH or reduction of NADP ϩ by DEPC could be observed, indicating that the modification of histidine residues accounted for inhibition of the enzyme. This finding supports the notion that covalent modification of a histidine, presumably His-467, by N-nitrosation at an imidazole nitrogen could account for inhibition of GR by DNIC. DISCUSSION One objective of the present study was to demonstrate that the structure of NO-carrier complexes influences the interaction of NO with macromolecular targets. Since GR was previously shown to be inhibited by S-nitrosothiols (16), we chose this enzyme as a model target to study its interaction with another class of biological NO carriers, DNIC. Two different types of DNIC were used with L-cysteine and GSH as thiol ligands, and the inhibitory reaction of GR with these compounds was studied in detail to reveal the molecular mechanism underlying the inhibition of GR by NO carriers.
Inhibition of GR by DNIC-Both types of DNIC readily accomplished inhibition of GR in the presence of NADPH. The potency of the inhibitors depended on the nature of the thiol ligand, DNIC-GSH being a more potent inhibitor (IC 50 ϭ 3 M) than DNIC-L-cysteine (IC 50 ϭ 30 M). In comparison, the IC 50 of other NO donors such as sodium nitroprusside, S-nitrosoglutathione, or S-nitroso-N-acetyl-DL-penicillamine are at least 100-fold higher (16), and also the nitrosourea diethyl [1-[3-(2chloroethyl)-3-nitrosoureido]ethyl]phosphonate (Fotemustine) exhibits an IC 50 of 1.5 mM (28). Therefore, DNIC are the most potent inhibitors of GR reported to date. Assuming that these complexes attack the catalytic center of GR, the higher inhibitory potency of DNIC-GSH compared with DNIC-L-cysteine may be explained by its similarity to the natural substrate GSSG (see Structure 1), which favors the interaction of DNIC-GSH with the active center of the enzyme. Since inhibition of the enzyme was not reversible by dilution (1000-fold) and progressed exponentially with time, inhibition kinetics could be evaluated according to Kitz and Wilson (22). The apparent rate constants of inhibition correlated with the concentration of DNIC-GSH applied. A dissociation constant of the reversible complex K i of 14 M and a rate constant of the conversion of the reversible enzyme-inhibitor complex to the irreversibly inhibited enzyme k 3 of 1.3 ϫ 10 Ϫ3 sec Ϫ1 were derived from the Kitz-Wilson replot. Thus DNIC-GSH was bound with high affinity in a rapid equilibrium preceding the irreversible inhibitory reaction.
Chemical Characterization of DNIC-inactivated GR-The irreversible inhibition of GR suggests a covalent modification of the enzyme by DNIC. A likely target site is the catalytic center of GR, which carries a redox-active disulfide (Cys-58 -Cys-63). The requirement for reduced thiols at the catalytic site was evidenced by the finding that DNIC led to inhibition of GR only in the presence of the coenzyme NADPH, and that GR pretreated with Hg 2ϩ was protected against inhibition by DNIC. Other agents which inhibit GR by carbamoylation (1,3-bis(2chloroethyl)-1-nitrosourea) or alkylation (1-(2-chloroethyl)-3-(2-hydroxyethyl)-1-nitrosourea) of Cys-58 also influence the enzyme activity only after two-electron reduction of the enzyme by NADPH (28,29). Therefore inhibition of GR by DNIC could involve a stable attachment of the dinitrosyl-iron moiety to one or both cysteines via a ligand exchange reaction (30) or a S-nitrosation of one or both cysteines, as has been shown for the interaction of DNIC with serum albumin (2).
Low molecular weight DNIC did not react with GR to form a protein-bound DNIC, as assessed by EPR spectroscopy (Fig. 5). This finding excludes that GR is inhibited by a linkage of the dinitrosyl-iron group to the active site cysteine residues. We next assessed the formation of a S-nitroso group in DNICinhibited GR by the Saville reaction. No S-nitrosation was detectable, although 1 mol of the DNIC-inhibited enzyme contained about 0.8 mol/subunit Griess-positive NO x released from the enzyme by 0.25 M HCl, but not by 0.1 M HCl. Similar acid lability of the NO moiety was observed with N-nitrosopyrrolidine, a nitrosamine of a cyclic secondary amine, whereas the C-nitroso bond of 1-nitroso-2-hydroxynaphthalene-3,6-disulfonic acid was acid-resistant. This result indicates that DNICinhibited GR bears a N-nitroso moiety.
A similar conclusion was derived from the comparison of the guanylyl cyclase-stimulating activity of DNIC-inhibited GR,