Selective Irreversible Inhibition of Neuronal and Inducible Nitric-oxide Synthase in the Combined Presence of Hydrogen Sulfide and Nitric Oxide

Background: NO and H2S signaling pathways may be interdependent. Results: Citrulline formation and NADPH oxidation by neuronal and inducible but not endothelial nitric-oxide synthase were inhibited irreversibly by H2S in the presence of NO. Conclusion: NO synthase is isoform-specifically inactivated by a reaction product of NO and H2S. Significance: Inhibition by NO/H2S may represent an autoregulatory feedback mechanism.

ginine (Arg), molecular oxygen (O 2 ), and NADPH-derived electrons in a reaction catalyzed by nitric-oxide synthase (NOS; EC 1.14. 13.39). NOS is only active as a dimer and exists in three isoforms, neuronal, endothelial, and inducible NOS (nNOS, 2 eNOS, and iNOS, respectively), that differ in tissue distribution and physiological function (4 -6). The constitutive isozymes nNOS and eNOS are activated by Ca 2ϩ /calmodulin (CaM), whereas the much higher affinity of iNOS for CaM renders its activity [Ca 2ϩ ]-independent under physiological conditions. Formation of NO requires the cofactor tetrahydrobiopterin (BH4), which couples NADPH oxidation to NO synthesis. In the absence of BH4, oxidation of NADPH results in O 2 . formation (5,7,8).
In the present study, we investigated whether H 2 S is able to directly affect NOS activity. We found that recombinant human nNOS and murine iNOS but not human eNOS were irreversibly inhibited by modest (ϳ10 Ϫ5 M) concentrations of H 2 S under conditions that allowed NO formation (i.e. ϩArg/ ϩBH4). In the absence of NO formation, inhibition required much higher H 2 S concentrations (ϳ10 Ϫ4 M) and was reversed by dilution. The results suggest that a product of the reaction between NO and H 2 S, possibly SSNO Ϫ , irreversibly inhibits nNOS and iNOS. The potential physiological relevance of these observations is discussed.
Enzyme Expression and Purification -Mouse macrophage iNOS was expressed in Escherichia coli and purified as described (20). Human eNOS was expressed in and purified from Pichia pastoris as described elsewhere (21). To subclone cDNA of human nNOS, the P. pastoris expression vector pPICZA was used (EasySelect Pichia expression kit). The plasmid pBBS230 containing cDNA for human nNOS was double digested with XbaI and NotI. The recessed 3Ј termini from the XbaI digest were filled by the Klenow fragment of E. coli DNA polymerase I in the presence of appropriate deoxynucleoside triphosphates. The vector was subsequently double digested with EcoRI and after filling the recessed 3Ј termini with NotI. The 4.3-kb insert was ligated to the restricted pPICZA. E. coli TOP10FЈ cells were transformed with the resulting ligation products and plated on LB/Zeocin medium (1% tryptone, 0.5% yeast extract, 0.5% NaCl, and 25 g/ml Zeocin at pH 7.5). The resulting transformants were tested by restriction analysis, and positive clones were amplified. The final DNA construct was linearized with PmeI, the DNA was transformed into P. pastoris GS115 (Mut ϩ ), and the cells were plated on YPDS/Zeocin medium (1% yeast extract, 2% peptone, 2% glucose, 1 M sorbitol, and 100 g/ml Zeocin) to select recombinants. A single colony of the best clone was grown for 36 h at 30°C in 50 ml of buffered minimal glycerol (BMGH) medium consisting of 100 mM potassium phosphate (pH 6.0), 13.4 g/liter yeast nitrogen base without amino acids, 400 g/liter biotin, 40 mg/liter L-histidine, and 1% (v/v) glycerol. The overnight culture was diluted in BMGH medium (1:200) and grown overnight at 30°C to an A 600 of 5-6. To induce nNOS expression, cells were harvested and resuspended in the presence of 4 mg/liter hemin chloride in buffered minimal methanol medium consisting of 100 mM potassium phosphate (pH 6.0), 13.4 g/liter yeast nitrogen base without amino acids, 400 g/liter biotin, 40 mg/liter L-histidine, and 0.5% methanol at an A 600 of ϳ1.
After 24 h of growth at 30°C, cells were harvested by centrifugation at 2000 ϫ g for 5 min at room temperature and resuspended at a concentration equivalent to an A 600 of 125 (based on the A 600 of the culture) in 50 mM Tris (pH 7.4) containing 1 mM EDTA, 5% glycerol, 12 mM 2-mercaptoethanol (2-ME), 1 mM phenylmethylsulfonyl fluoride (PMSF), and 1 mM CHAPS. An equal volume of glass beads (0.5 mm) was added to the suspension, and the cells were broken by vigorous vortexing at 4°C for a total of 10 min in bursts of 30 s alternating with cooling on ice. The glass beads were separated by centrifugation at 800 ϫ g for 5 min. After a further clearing step at 1600 ϫ g for 5 min, the supernatant was centrifuged at 30,000 ϫ g for 15 min. The enzyme was purified from the resulting supernatant by affinity chromatography as described previously (22). Final elution was achieved with 20 mM Tris (pH 7.4), 150 mM NaCl, and 4 mM EGTA. After determination of the protein concentration according to Bradford (23) using bovine serum albumin as a standard, the enzyme was stored at Ϫ70°C in the presence of 1 mM CHAPS. Enzyme concentrations are expressed as the concentration of the monomer, assuming molecular masses of 160 (nNOS), 130 (iNOS), and 135 kDa (eNOS).
To test for irreversibility of inhibition, nNOS (250 g/ml; 1. NADPH oxidation was determined spectrophotometrically at 340 nm and 37°C as described elsewhere (25). Unless indicated otherwise, samples containing 10 g/ml nNOS (62.5 nM), 0.2 mM NADPH, 0.5 mM CaCl 2 , 0.2 mM CHAPS, 0.1 mM EDTA, 0.1 mM Arg, 10 M BH4, 30 M PROLI/NO, and Na 2 S as indicated in 50 mM TEA (pH 7.4) were incubated at 37°C. The reaction was initiated by the addition of 20 g/ml CaM and monitored for 5 min. Rates were corrected by subtraction of blank rates obtained in the absence of CaM.
Concentration-effect curves (Figs. 1; 3; 4A; 7, B and C; 8, A and B; and 9A) were fitted to the Hill equation Act ϭ Act ∞ ϩ (Act 0 Ϫ Act ∞ )/(1 ϩ ([I]/IC 50 ) h ) in which Act is the observed activity, [I] is the variable concentration of inhibitor (Na 2 S in Figs. 1, 3, 4, and 7; PROLI/NO in Fig. 8; and Angeli's salt in Fig.  9), Act 0 and Act ∞ are the respective activities at zero and infinite inhibitor concentration, IC 50 is the half-maximal inhibitory concentration, and h is the Hill coefficient. Values for IC 50 , h, and Act 0 and in Fig. 3 for Act ∞ were determined from the fits; Act ∞ was set to 0 in Figs. 1, 4, 7, 8, and 9. In Fig. 8B, h was set to 2.
The pH dependence of Fig. 4B was fitted to the equation IC 50 ϭ K i /(1 ϩ 10 pH Ϫ pKa ) where K i and pK a are the apparent inhibition constant of the protonated inhibitor and the corresponding acidity constant, respectively. This equation describes the dependence of the observed IC 50 on pH when inhibition involves the protonated form only. The time traces in the presence of CaM of Fig. 7A were fitted to single exponential functions.
UV/Visible Absorbance Spectroscopy-Spectra were measured with a Hewlett-Packard 8452A diode array spectrophotometer. For absorbance measurements, nNOS or eNOS samples were diluted to a final concentration of approximately 4 M in 50 mM TEA (pH 7.4) in the absence or presence of 5 mM NaHS.
Gel Filtration-NOS dimerization was analyzed by gel filtration with a Superose 6 HR 10/30 column under the control of an ÄKTA chromatography system at 8°C. The flow rate was set to 0.3 ml⅐min Ϫ1 , and the elution buffer consisted of 20 mM TEA (pH 7.4), 150 mM NaCl, 5% (v/v) glycerol, and 0.5 mM diethylene triamine pentaacetic acid. Purified nNOS (250 g/ml; 1. Samples containing 50 ng of nNOS or eNOS were subjected to SDS-PAGE for 100 min at 100 V on discontinuous 4% SDS gels (1.5 mm) using the Mini-Protean II system from Bio-Rad. Gels and buffers were equilibrated at 4°C, and the buffer tank was cooled during electrophoresis in an ice bath. Separated proteins were transferred to nitrocellulose membranes (0.45 m) by electroblotting at 240 mA for 110 min followed by immunodetection with anti-nNOS or anti-eNOS antibodies (1:1000 or 1:2000 dilution, respectively; BD Transduction Laboratories) using horseradish peroxidaseconjugated anti-mouse IgG (1:5000; BD Transduction Laboratories) and ECL detection reagent (Biozym, Hessisch Oldendorf, Germany). Immunoreactive bands were quantified densitometrically using E.A.S.Y. 1.3 Win 32 (Herolab, Vienna, Austria) and ImageJ 1.46r software (Wayne Rasband, National Institutes of Health).

Results
Effect of Na 2 S on Citrulline Formation by nNOS, iNOS, and eNOS-To determine the effect of H 2 S on NOS activity, we measured citrulline formation by the NOS isoforms in the presence of varying concentrations of Na 2 S. As illustrated in Fig. 1, Na 2 S inhibited nNOS and iNOS with IC 50 values of (2.4 Ϯ 0.3)⅐10 Ϫ5 and (7.9 Ϯ 1.6)⅐10 Ϫ5 M, respectively, whereas eNOS was only marginally affected.
Effect of Na 2 S on the Optical Absorbance Spectra of nNOS-Because it has been demonstrated that DTT and other thiols inhibit NOS by binding to the heme (27), we measured the effect of Na 2 S on the UV/visible absorbance of nNOS and eNOS. We observed spectral changes typical of the conversion to a thiol complex (Fig. 2). However, the transition was slow (t1 ⁄2 ϭ 3.5 Ϯ 1.6 min for nNOS), incomplete (approximately 50%), and required high concentrations (5 mM) of the H 2 S donor, suggesting that binding of the thiol to the heme is not involved in NOS inhibition.
Effect of Substrate and Cofactor Concentration on Inhibition of Citrulline Formation by Na 2 S-To investigate whether H 2 S inhibition is competitive with substrates or cofactors, citrulline  Fig. 3), which indicates that H 2 S does not interfere with substrate or cofactor binding. . At time 0, NaHS (5 mM) was added, and spectra were measured at the indicated times. A and B show the absolute absorbance spectra of nNOS and eNOS, respectively. C and D show the corresponding difference spectra with the spectrum before Na 2 S addition subtracted from all other spectra.

Effect of Thiols on Na 2 S-induced Inhibition of Citrulline
Formation by nNOS-Because H 2 S can modulate enzyme function by sulfhydration of cysteine residues (9 -11, 13, 17, 28), it is conceivable that inhibition might be relieved in the presence of excess thiols. Therefore, we measured inhibition by Na 2 S in the presence of 2 mM DTT, 2 mM GSH, or 2.9 mM 2-ME. However, none of these thiols had any impact on IC 50 values (results not shown).
Effect of Glutathione Persulfide on Citrulline Formation by nNOS-A potential complication is the facile formation of persulfides (RSS Ϫ ) from H 2 S in the presence of thiols (2, 9 -11, 17, 18). To study the possible involvement of persulfides in H 2 Smediated inhibition of nNOS, we determined the effect of glutathione persulfide (GSSH) on nNOS activity. GSSH, synthesized from Na 2 S and GSSG according to a published procedure (19), inhibited nNOS with lower affinity than Na 2 S (IC 50 ϭ (1.16 Ϯ 0.11)⅐10 Ϫ4 M, n ϭ 2; not shown). Because the conversion of GSSG and Na 2 S to GSSH by the applied method amounts to about 30 -40% (19), the observed inhibition was most likely due to the remaining H 2 S. This suggests that persulfides do not significantly contribute to nNOS inhibition.
Effect of pH on Na 2 S-induced Inhibition of Citrulline Formation by nNOS-To study the effect of pH on Na 2 S-induced inhibition of citrulline formation, we determined the activity of nNOS at pH 6.0, 7.4, and 8.0 (Fig. 4A). The IC 50 increased when the pH was raised from (1.5 Ϯ 0.2)⅐10 Ϫ5 M at pH 6.0 via (3.1 Ϯ 0.5)⅐10 Ϫ5 M at pH 7.4 to (8.3 Ϯ 1.2)⅐10 Ϫ5 M at pH 8.0. At first sight, these results suggest that Na 2 S-induced inhibition involves interaction of nNOS with H 2 S rather than with hydrogen sulfide anion (HS Ϫ ). However, from a plot of IC 50 against pH assuming inhibition by the low pH species only, we obtained a pK a value of 7.310 Ϯ 0.014 (Fig. 4B), which is considerably higher than the published pK a (6.76 at 37°C) of the H 2 S/HS Ϫ equilibrium (29). The pH profile of inhibition therefore appears to reflect the protonation state of another compound.
Irreversible Inhibition by Na 2 S under Turnover Conditions-As illustrated in Fig. 5, preincubation of nNOS with Na 2 S under turnover conditions decreased the activity after dilution (with  ϳ3.3 M Na 2 S remaining) by approximately 70%, suggesting that inhibition by H 2 S is irreversible. However, the activity of the diluted enzyme was not affected when CaM was omitted during preincubation, which suggests that irreversible inhibition requires the presence of both H 2 S and NO. To elucidate whether thiols could reverse inhibition, we added 2 mM DTT, 2 mM GSH, or 2.9 mM 2-ME to the activity assay. As shown in Fig.  6A, none of these thiols restored the activity. Similarly, neither bovine serum albumin (2 mg/ml; data not shown) nor thioredoxin/thioredoxin reductase (Fig. 6B) reversed inhibition.
Similar observations were made with iNOS (not shown): preincubation in the absence and presence of 0.5 mM Na 2 S yielded activities after dilution of 528 Ϯ 75 and 257 Ϯ 45 nmol of citrulline/mg/min when CaM was present during preincubation, whereas the corresponding activities were 569 Ϯ 65 and 617 Ϯ 90 nmol/mg/min when CaM was omitted. As with nNOS, virtually identical results were obtained when 2 mM GSH was added to the assay mixture.
Effect of the Enzyme Concentration on Inhibition of nNOS and eNOS by H 2 S under Turnover Conditions-Because eNOS has lower turnover than nNOS and iNOS, the lack of inhibition of eNOS might be caused by the lower NO formation rate of that isoform. To explore that possibility, we determined the effect of the concentration of nNOS and eNOS (between 0.5 and 15.0 g/ml and between 2.0 and 30.0 g/ml, respectively) on the inhibition by 0.5 mM Na 2 S (Table 1). Both isoforms exhibited constant specific activities over the studied concentration range. However, whereas nNOS activity was almost completely (ϳ90%) blocked at all concentrations, eNOS activity was hardly affected even though the estimated NO formation rate (in the absence of Na 2 S) for the highest concentration of eNOS was  Experimental conditions were the same as for Fig. 5 except that thiol (2 mM DTT, 2 mM GSH, or 2.9 mM 2-ME) was added to the 50-fold diluted reaction mixture (see "Experimental Procedures"). Note that the final mixtures contained ϳ3 M Na 2 S carried over from the preincubation mixture and ϳ0.7 mM DTT, ϳ0.7 mM GSH, or ϳ1 mM 2-ME carried over from the 50-fold diluted samples. Data (n ϭ 2) are presented as mean values ϮS.E. (error bars). B shows the effect of inclusion of thioredoxin/thioredoxin reductase on the reversibility of inhibition of nNOS by Na 2 S. Citrulline formation by nNOS was determined after preincubation in the absence or presence of Na 2 S. Experimental conditions were as for Fig. 5 except for the presence of thioredoxin reductase (TRXR) (6 M) and thioredoxin-1 (TRX) (5 M) during the final assay. Note that the final assay samples also contained ϳ3 M Na 2 S carried over from preincubation mixture. Data (n ϭ 2) are presented as mean values ϮS.E. (error bars). Ctrl, control. OCTOBER 9, 2015 • VOLUME 290 • NUMBER 41

JOURNAL OF BIOLOGICAL CHEMISTRY 24937
20ϫ as high as that for the lowest concentration of nNOS. In the presence of Na 2 S, 30 g/ml eNOS produced ϳ160ϫ as much NO as 0.5 g/ml nNOS did. These results clearly demonstrate that the lack of inhibition of eNOS is not due to its lower intrinsic activity.
Effect of Na 2 S on NADPH Oxidation by nNOS-To examine whether inhibition of citrulline formation was accompanied by NOS uncoupling, we determined the effect of Na 2 S on the rate of NADPH oxidation in the absence or presence of Arg and/or BH4 (Fig. 7, A and B). The NADPH oxidation rate under control conditions in the presence of Arg and BH4 (714 Ϯ 23 nmol⅐mg Ϫ1 ⅐min Ϫ1 ) corresponds to a NADP ϩ /citrulline stoichiometry of 1.51 Ϯ 0.06, indicative of strong coupling (7). Na 2 S completely blocked NADPH oxidation with an IC 50 of (1.9 Ϯ 0.3)⅐10 Ϫ5 M in good accordance with the value observed for citrulline formation (Fig. 1). This indicates that inhibition targets NADPH oxidation without any sign of uncoupling. Interestingly, when Arg and/or BH4 were omitted, conditions under which no irreversible inhibition of citrulline formation occurs (see above), NADPH oxidation was still blocked but at considerably higher concentrations of Na 2 S (IC 50 values: ϩArg/ ϪBH4, ϪArg/ϪBH4, (9.5 Ϯ 1.4)⅐10 Ϫ5 M). These observations suggest that H 2 S alone inhibits nNOS reversibly with an IC 50 of ϳ0.1-0.6 mM but that inhibition becomes more pronounced and irreversible in the presence of NO.
To confirm this, we repeated the experiments in which Arg and/or BH4 was omitted in the presence of the NO donor PROLI/NO (Fig. 7C). Under these conditions, 30 M PROLI/NO lowered the IC 50  These results confirm that inhibition by H 2 S is potentiated by NO.
Effect of the NO Concentration on Inhibition of nNOS and eNOS by Na 2 S-To study the effect of the NO concentration, we measured the rate of NADPH formation at varying PROLI/NO concentrations in the presence of Arg but in the absence of BH4 under which conditions the enzyme does not produce NO. Determination of the effect of the NO concentration is complicated by the fact that Na 2 S alone already inhibits the enzyme (see Fig. 7B). Moreover, NO alone will also inhibit NOS activity by binding to the heme (30 -32). Therefore we determined the effect of the NO concentration in the absence and presence of 10 M Na 2 S, a concentration that does not by itself inhibit NADPH oxidation but that becomes inhibitory in the presence of PROLI/NO (Fig. 7, B and C, blue traces). As illustrated by Fig. 8A, PROLI/NO inhibited NADPH activity with an IC 50 of ϳ(8.0 Ϯ 0.5)⅐10 Ϫ5 M in the absence of Na 2 S, which is most likely caused by binding of NO to the heme. In the presence of Na 2 S, the IC 50 shifted leftward to (1.1 Ϯ 0.2)⅐10 Ϫ5 M, probably reflecting the (irreversible) effect of NO on H 2 Sinduced inhibition.
For comparison, we also looked into the effect of the PROLI/NO concentration in the presence of Na 2 S on eNOS activity. Unlike nNOS, eNOS exhibits greatly reduced NADPH oxidation when either BH4 or Arg is omitted (33). Therefore we decided to include Arg and BH4 in the reaction mixture, and as a consequence, the enzyme already produces NO in the absence of PROLI/NO. We also applied a much higher Na 2 S concentration (1 mM) because at that concentration we earlier observed moderate inhibition of eNOS activity (see Fig. 1). Fig. 8B shows that in this case too NADPH oxidation was inhibited by high concentrations of PROLI/NO (IC 50 ϭ (1.0 Ϯ 0.2)⅐10 Ϫ4 M). Fig.  8B also confirms the moderate effect of 1 mM Na 2 S on NADPH oxidation. However, this weak inhibitory effect was not potentiated by PROLI/NO (IC 50 ϭ (1.17 Ϯ 0.16)⅐10 Ϫ4 ). Taken together, these results demonstrate that NO concentrationdependently potentiates the inhibition by Na 2 S of nNOS but not of eNOS.
In the experiments described above, the effect of NO on H 2 Sinduced inhibition of nNOS was clearly concentration-dependent. By contrast, in the experiments of Table 1, similar inhibition was observed at all nNOS concentrations and therefore at all NO concentrations. This suggests that in the studied enzyme concentration range (which corresponded to NO concentrations after 10 min between 2 and 50 M in the absence of Na 2 S and approximately 10-fold lower concentrations in the presence of Na 2 S) the potentiation by NO of H 2 S-induced inhibition is not affected by its concentration. To corroborate this finding, we determined the effect of the nNOS concentration on the IC 50 value of Na 2 S. Variation of the nNOS concentration did not affect the IC 50 values (22 Ϯ 5, 25 Ϯ 6, and 37 Ϯ 3 M at 1, 5, and 15 g/ml, respectively), confirming that in this concentration range, which corresponds to uninhibited NO formation rates between 0.37 and 4.3 M/min, inhibition does not depend on the NO concentration.
Effect of Slow H 2 S and NO Donors on nNOS Activity-To study the effect of the rate of H 2 S generation on the inhibition of nNOS, we replaced Na 2 S, which releases H 2 S almost instantaneously, by the slow H 2 S-releasing agent GYY4137 (t1 ⁄ 2 ϳ 415

Effect of the enzyme concentration on H 2 S-induced inhibition of nNOS and eNOS
Conc is the enzyme concentration; Act 0 and Act H2S are the specific activities measured as citrulline formation in the absence and presence of 0. 5 Fig. 1).
To study the effect of the NO release rate, we replaced PROLI/NO (t 1/2 ϳ 1-2 s) by SPER/NO (t 1/2 ϳ 1800 s) (35). In the absence of BH4, i.e. when the enzyme does not produce NO, SPER/NO potentiated the inhibitory effect of Na 2 S on NADPH formation after 10 min with an apparent IC 50 value of (1.06 Ϯ 0.17)⅐10 Ϫ4 M (not shown). At this concentration, SPER/NO will release approximately 24 M NO in 10 min in good agreement with the values obtained with PROLI/NO (see Fig. 7C).
Inhibition of nNOS by HNO-According to a recent report, the combination of NO and H 2 S regulates vascular tone by the intermediate formation of HNO (36), and it has been reported in the past that HNO is a stronger inhibitor of nNOS than NO (37). To investigate the potential involvement of HNO in the inhibition observed here, we determined the effect of the HNO donor Angeli's salt on citrulline formation by nNOS. As illustrated by Fig. 9A, Angeli's salt inhibited nNOS, but the IC 50 value of (1.9 Ϯ 0.4)⅐10 Ϫ4 M was considerably higher than that of Na 2 S (see Fig. 1). More importantly, unlike the effect of Na 2 S, inhibition by HNO was completely reversed in the presence of thiols (Fig. 9B).
Effect of Na 2 S on Dimeric Structure of nNOS and eNOS in the Absence or Presence of NO Synthesis-To investigate the effect of H 2 S on the dimer content of nNOS, we performed gel filtration chromatography after preincubation under various conditions. Because dimer stability is affected by Arg and BH4 but not by CaM (26), both Arg and BH4 were included in all preincubations, and the effect of NO formation was instead determined by omitting or including CaM. As shown in Fig. 10, after preincubation in the absence of CaM and Na 2 S, the enzyme was mostly (ϳ70%) dimeric. Preincubation in the presence of CaM or Na 2 S appeared to cause a slight decrease in dimer content, whereas a somewhat larger decrease was observed when CaM and Na 2 S were both present. However, ϳ55% of the enzyme was still dimeric even after preincubation under full-turnover conditions. Similar observations were made with low temperature PAGE followed by Western blotting analysis. As shown in Fig. 11, A and B, Na 2 S alone did not affect dimer stability (33.2 Ϯ 2.6 versus 34.0 Ϯ 1.8%), whereas the combination of  OCTOBER 9, 2015 • VOLUME 290 • NUMBER 41

Irreversible Inhibition of Neuronal NO Synthase by H 2 S/NO
CaM and Na 2 S reduced the amount of SDS-resistant dimers by more than half (11.0 Ϯ 1.6 versus 25.6 Ϯ 1.5%). The dimer/ monomer ratio of eNOS was not affected by Na 2 S at all (Fig.  11C). Although these results suggest some correlation between dimer strength and NO/H 2 S-induced inhibition, nNOS remained mainly dimeric under conditions that resulted in complete loss of activity, indicating that the main mechanism for inhibition does not involve monomerization.

Discussion
Inhibition of NOS by H 2 S has been reported previously by Kubo et al. (15,16). In those studies, all three isoforms were inhibited with comparably low potencies (IC 50 values between 0.13 and 0.21 mM). Furthermore (contrary to what was stated in Ref. 16), inhibition was partly or completely countered by  increasing the NADPH concentration and except for iNOS by increasing the BH4 concentration. The reason for these and other discrepancies, the study also reported an IC 50 for inhibition of nNOS by DTT of 13.2 mM where we previously found 0.16 mM (27), cannot be resolved here but may be related to the use of commercial enzyme preparations with low activity and different experimental conditions (for instance, extremely low concentrations of Arg and CaM and very long incubation times).
The respective mechanisms of inhibition by H 2 S and by NO/H 2 S are unclear. Inhibition is not competitive with any

Irreversible Inhibition of Neuronal NO Synthase by H 2 S/NO
of the substrates or cofactors. Furthermore, inhibition by NO/H 2 S but not by H 2 S alone is irreversible under the present conditions. A puzzling aspect of the present study is the apparent difference in the inhibitory efficiency of nNOS-and PROLI/ NO-generated NO. No [NO] dependence was observable down to concentrations as low as 0.5 g/ml nNOS, which corresponds to a production of NO of ϳ0.3 M after 10 min in the presence of Na 2 S. In contrast, PROLI/NO exhibited an apparent IC 50 of ϳ10 Ϫ5 M. Tentatively, one may ascribe this remarkable difference to the close proximity of the sites of NO formation and H 2 S inhibition in the case of endogenously produced NO.
Cysteinyl side chains are the most likely targets for inhibition by H 2 S. In the presence of an electron acceptor, H 2 S may cause protein S-sulfhydration (17,28). Indeed, S-sulfhydration of NOS has been reported recently (38). However, in that study, eNOS was stimulated rather then inhibited by sulfhydration. There have also been several reports on eNOS glutathionylation, which blocked NO synthesis but not NADPH oxidation, resulting in uncoupled catalysis (39 -41). By contrast, we are not aware of any study on the glutathionylation or sulfhydration of the neuronal and inducible isoforms. All three NOS isoforms are also inhibited by S-nitrosation (42)(43)(44)(45), which targets the cysteinyl side chains coordinating the zinc cation that stabilizes the NOS dimeric structure (46,47). In addition to modification of cysteinyl side chains, it is conceivable that H 2 S directly interferes with NOS zinc binding as has been proposed as a potential inhibitory mechanism in the case of angiotensinconverting enzyme and phosphodiesterase (48,49). If the interdomain zinc cation or its cysteinyl ligands are indeed the target for inhibition by H 2 S, this would offer a tentative explanation for the remarkable resistance of eNOS to inhibition. Of the three isoforms, eNOS has by far the greatest dimer stability (50). Although the present results indicate that inhibition is not caused by NOS monomerization, it is conceivable that the same forces that stabilize the eNOS dimer also protect the zinc site against inhibition by H 2 S.
Whereas sulfhydration of specific cysteinyl residues might be causing the low affinity reversible inhibition, irreversible inhibition in the presence of NO may involve a product of the reaction between H 2 S and NO. We recently demonstrated efficient nitrosation of GSH and other thiols by NO at submicromolar concentrations (35,51). A similar reaction with H 2 S would yield HSNO, which in principal might inhibit NOS by transnitrosation of one of the cysteinyl zinc ligands. However, the observation that inhibition was not reversed by thiols or thioredoxin/ thioredoxin reductase argues against that possibility. For the same reason, the involvement of HNO can be ruled out as well.
In the presence of excess H 2 S, the highly unstable HSNO is rapidly transformed to nitrosopersulfide (SSNO Ϫ ) (17,18,52). Conceivably, it is this compound that is responsible for irreversible inhibition of nNOS and iNOS. Although as far as we are aware the pK a of nitrosopersulfide has not been reported, it is tempting to ascribe the value of 7.3 that we observed for nNOS inhibition to the HSSNO/SSNO Ϫ equilibrium. Alternatively, NO or an NO-derived compound may react with the sulfhydrated protein formed by H 2 S in the absence of NO, which would possibly explain the absence of an effect of the NO con-centration on inhibition (although not the higher potency of NO/H 2 S compared with H 2 S alone). Clearly, elucidation of the inhibitory mechanism must await identification and characterization of the inhibitory site. To this end, we are currently performing mass spectrometric analysis of the modification of nNOS by NO/H 2 S. Preliminary results suggest that a specific cysteine residue in the reductase domain (Cys 1231 ) becomes sulfinated in the presence of H 2 S under turnover conditions. 3 However, additional studies are required to confirm or refute these observations.
The present results demonstrate that H 2 S completely blocks nNOS activity (coupled and uncoupled) at moderately high concentrations. Importantly, inhibition gets stronger and becomes irreversible under conditions of coupled turnover or when NO is co-administered. Similar effects were observed for iNOS but not for eNOS, demonstrating that inhibition by NO/H 2 S is isoform-specific. There is controversy in the literature on the physiological levels of H 2 S with earlier reports suggesting unrealistically high values (for a review, see Ref. 53), whereas more recent estimates seem to converge on values in the submicromolar or even low nanomolar range (2,10,54,55). Whereas the higher estimates would render the effects observed here physiologically relevant, inhibition by H 2 S alone would be too weak to play a significant role if the lower estimates apply. However, because of its apparent irreversible nature, inhibition by NO/H 2 S might still be relevant. One may speculate that such inhibition could serve a protective role as a negative feedback mechanism in the case of excessive NO/H 2 S production. It will therefore be important to establish whether inhibition by NO/H 2 S remains irreversible in an in vivo setting.
In summary, we have observed inhibition of nNOS but not of eNOS that may be physiologically relevant provided that the irreversible character observed here persists under (patho) physiological conditions. If so, these observations may help resolve some of the controversies concerning the impact of H 2 S on NO signaling where both stimulatory and inhibitory effects have been reported.