Introduction

Top Abstract Introduction Experimental Procedures Results Discussion References

The best characterized, endothelium-derived relaxing factor nitric oxide (NO) (Furchgott and Zawadzki, 1980; Ignarro et al., 1987; Palmer et al., 1987) is synthesized in vascular endothelium from L-arginine by NO synthase (Palmer et al., 1988; Knowles et al., 1989; Moncada et al., 1989). Endothelium-derived NO is an important endogenous vasodilator that regulates vascular smooth muscle tone (Bult et al., 1990; Burnett et al., 1992). The cellular mechanisms of NO-induced vasodilation are not completely understood. They involve an increase in intracellular cGMP (Feelisch and Noack, 1987; Garthwaite, 1991; Miyoshi et al., 1994; Hussain et al., 1997; Olson et al., 1997) and membrane hyperpolarization (Tare et al., 1990), or Ca²⁺ storage in the sarcoplasmic reticulum (Khan et al., 1998). The study of the effects of NO has been limited by the rapid NO degradation in physiological systems (Broderick, 1995) that requires the synthesis of compounds capable of generating NO in situ (Feelisch, 1991). Nitrovasodilators are prodrugs that release NO and mainly mimic endogenous NO activities. Two representative compounds of this group are S-nitroso-glutathione (GSNO) and S-nitroso-N-acetylcysteine (NACysNO). All of the known S-nitrosothiols are unstable in solution, and although the decomposition of most of them has not been characterized in detail, it probably occurs by homolytic cleavage of the S-N bond, a process that is enhanced by heat, light, oxygen, and alkaline pH, resulting in the production of NO and the corresponding disulfide (Williams, 1983; Feelisch, 1991). It has been suggested that the biological activity of S-nitrosothiols is a direct consequence of their decomposition with NO production. However, experiments examining the relation between the chemical structure, stability, and biological activity of S-nitrosothiols indicate that decomposition of S-nitrosothiols in solution with production of NO cannot explain the biological activities of these compounds (Mathews and Kerr, 1993). Marks et al. (1995) demonstrated that NO formation from nitroprusside, 3-morpholinosydnonimine, and S-nitroso-N-acetylpenicillamine occurs concurrently with relaxation in rabbit aorta. In contrast, formation of NO from glyceryl trinitrate was only measurable after 5 min, at which time the relaxation induced by this compound was almost complete. In the present study, GSNO and NACysNO were used to examine the mechanism of relaxation in isolated rat aorta, and NO formation was determined by the amperometric method. We also evaluated the influence of changes in pH on the concentration of NO produced by GSNO and NACysNO.

The aim of the present study was to evaluate the mechanisms of vascular relaxation induced by NACysNO and GSNO in endothelium-denuded rats aorta, and to determine whether there is a relationship between the magnitude of NO formation and vasodilation.

	Experimental Procedures

Top Abstract Introduction Experimental Procedures Results Discussion References

Aortic Ring Preparation. Male Wistar rats (250-300 g) were killed by decapitation and their thoracic aortas were isolated. Aortic rings, 3 mm in length, were mounted between two steel hooks to measure the isometric tension. The aorta rings were mounted under a resting tension of 1.5 g in organ baths containing a modified Krebs' solution of the following composition: 130 mM NaCl, 4.7 mM KCl, 1.2 mM MgSO₄, 1.6 mM CaCl₂, 1.2 mM KH₂PO₄, 14.9 mM NaHCO₃, 5.5 mM glucose, and 0.03 mM CaNa₂ EDTA continuously bubbled with 95% O₂ and 5% CO₂, pH 7.4 and maintained at 37°C. The system was connected to an F-60 force-displacement transducer and the contractile responses were recorded on a polygraph (Narco Biosystems, Inc., Houston, TX). After an equilibration period of 60 min during which the Krebs' solution was changed every 20 min, the aortic rings were stimulated with 1 µM phenylephrine until reproducible contractile responses were obtained. Because the response to nitrosothiols does not require the presence of the endothelium, in the present study we examined the mechanism of relaxation induced by these compounds in endothelium-denuded arteries. The rings were mechanically denuded of endothelium and the presence of functional endothelium was assessed by the ability of 1 µM acetylcholine to induce relaxation in aortas precontracted with 1 µM phenylephrine. The following specific protocols were then performed.

Preparation of GSNO and NACysNO. S-Nitrosothiols were prepared by the method of Mathews and Kerr (1993) and stored in the dark below 0°C. Briefly, they were prepared by reacting equimolar amounts of NaNO₂ and the corresponding thiol (glutathione or N-acetylcysteine) in aqueous solution. The solutions were adjusted to pH 2.0 with 6 M HCl and incubated at 37°C for 5 min, a time sufficient for the development of characteristic red-orange compounds. The samples were neutralized to pH 7.0 with NaOH and then lyophilized.

Relaxation Induced by GSNO or NACysNO in Aortas Precontracted with Phenylephrine. Contractile responses were induced with 0.1 µM phenylephrine, and when they reached a plateau, cumulative concentrations of GSNO or NACysNO (from 10¹⁰ to 10⁴ M) were added to the bath to study the relaxation induced by these two compounds.

Relaxation Induced by GSNO or NACysNO in Aortas Precontracted with KCl. To determine whether K⁺ channels are involved in the relaxation induced by GSNO and NACysNO, the contractions were produced with a single concentration of 30, 60, or 90 mM KCl in Krebs' solution. High concentrations of KCl in Krebs' solution were prepared by replacing NaCl with an equivalent amount of KCl to maintain physiological osmolarity. At the plateau of the contractile response induced by KCl, cumulative concentrations of GSNO or NACysNO (from 10¹⁰ to 10⁴ M) were added until maximal relaxation was obtained.

Effect of ODQ on the Relaxation Induced by GSNO and NACysNO. To determine the involvement of guanylyl cyclase in the relaxation induced by GSNO and NACysNO, we studied the effect of 1H-[1,2,4]oxadiazolo[4,3-a]quinozalin-1-one (ODQ), a potent inhibitor of cytosolic guanylyl cyclase. Tissues were exposed to 10 µM ODQ for 30 min and the aorta rings were then stimulated with 0.1 µM phenylephrine. When the contractile response reached a plateau, cumulative concentration-effect curves for GSNO or NACysNO from 10¹⁰ to 10⁴ M were constructed.

Inhibitory Effects of K⁺ Channel Blockers. Aorta rings were incubated with the K⁺ channel blockers 1 h before the addition of 0.1 µM phenylephrine. When the contractile response reached a plateau, cumulative concentration-effect curves for GSNO or NACysNO (from 10¹⁰ to 10⁴ M) were constructed in the presence of 100 nM charybdotoxin, 500 nM apamin, 3 µM glibenclamide, or 5 mM 4-aminopyridine.

Effects of GSNO and NACysNO on the Contractile Response Induced by Phenylephrine in Ca²⁺-Free Solution. To determine whether the relaxation induced by GSNO or NACysNO occurs via Ca²⁺ uptake into the sarcoplasmic reticulum, we studied the effect of these two compounds on the Ca²⁺ release induced by the agonist phenylephrine in Ca²⁺-free solution. Aorta rings were stimulated with 10 µM phenylephrine to produce a maximal contractile response. When the contraction reached a plateau, aorta rings were rinsed until complete relaxation was achieved. The tissues were then placed in Ca²⁺-free solution for 1 min, and contractile responses were induced with 10 µM phenylephrine (control group). The aorta rings were then rinsed with Krebs' solution until complete relaxation was achieved and for an additional 30 min. After this period, the tissues were incubated for 5 min with different concentrations of NACysNO or GSNO (from 10¹⁰ to 10⁴ M) and placed again in Ca²⁺-free solution for 1 min. The contractile response to 10 µM phenylephrine was induced in the presence of the two compounds. One preparation was used for each concentration of GSNO or NACysNO (10¹⁰ to 10⁴ M).

Effects of Ryanodine and Thapsigargin on Relaxation Induced by GSNO and NACysNO. We also evaluated the effects of thapsigargin and ryanodine on the relaxation induced by GSNO and NACysNO. Tissues were pretreated with 30 µM ryanodine or 1 µM thapsigargin in Krebs' solution for 1 h at resting tension and then contracted with 10 µM phenylephrine in Krebs' solution, and cumulative concentrations of GSNO or NACysNO were added (10¹⁰ to 10⁴ M).

Measurement of Nitric Oxide. The concentration of NO produced by GSNO and NACysNO was measured in the macroscopic experiment with an ISO-NO proprietary sensor (WPI, Sarasota, FL) with a sensitivity range of 1 nM to 20 µM NO, resolution of 1 nM, time constant of 18 s, and tip diameter of 2 mm. This sensor detects NO directly by an amperometric technique. The Iso-NO sensor was calibrated daily when in use by a quantitative NO generation reaction (note that while the half-life of NO is about 5 s in physiological preparations, in our buffer it was much longer), initiated by mixing NaNO₂ at the desired concentration with 0.1 M H₂SO₄ and 0.1 M KI, according to the equation 4H⁺ + 2I + 2NO  2H₂O + 2NO + I₂. The output of the sensor was recorded with a PC-based data acquisition system (Duo-18, WPI). All experiments were conducted at 37°C.

Effect of pH on Nitric Oxide Production. To study the effect of pH on NO production, we changed the pH in the Krebs' solution from pH 7.4 to pH 4.0 adjusted with 1 M HCl, and pH 9.0 adjusted with 1 M NaOH.

Statistical Analysis. Data are expressed as mean ± S.E.M. In each set of experiments, n indicates the number of rats studied. The concentration of agents producing a half-maximal relaxation amplitude (EC₅₀) was determined after logit transformation of the normalized concentration-response curves and is reported as the negative logarithm (pD₂) of the mean of individual values for each tissue using the GraphPad Prism, version 2.0 (GraphPad Software Corporation, San Diego, CA). The maximum effect was considered as the maximal amplitude response reached in the concentration-effect curves for relaxant agents. Statistical significance was tested by the Wilcoxon test for paired analysis or the Mann-Whitney test for unpaired analysis, and values of P < 0.05 were considered to be significant.

Materials. GSNO and NACysNO were synthesized in our laboratory as described under Experimental Procedures. ODQ (Calbiochem, La Jolla, CA), phenylephrine, thapsigargin, charybdotoxin, apamin, 4-aminopyridine, and glibenclamide were obtained from Sigma Chemical Co. (St. Louis, MO). Ryanodine was purchased from Research Biochemical International (Natick, MA). Drugs were dissolved in ultrapure water-Barnstead E-pure (GSNO, NACysNO, 4-aminopyridine, phenylephrine), in dimethyl sulfoxide (thapsigargin, ryanodine, glibenclamide), in 5% acetic acid (apamin), or in 150 mM NaCl (charybdotoxin). Dimethyl sulfoxide was used at concentrations of less than 0.1%, which did not affect the responses.

	Results

Top Abstract Introduction Experimental Procedures Results Discussion References

Relaxation Induced by GSNO and NACysNO in Aortas Precontracted with Phenylephrine. As shown in Fig. 1, the relaxation induced by both compounds was concentration-dependent in aortas precontracted with 0.1 µM phenylephrine or with high concentrations of KCl (30, 60, and 90 mM). In arteries precontracted with phenylephrine, the sensitivity was similar for GSNO and NACysNO as demonstrated by the pD₂ values of 6.76 ± 0.16 (n = 8) for GSNO and 6.64 ± 0.01 (n = 8) for NACysNO. Similarly, the maximum relaxation was 100% in all the preparations precontracted with phenylephrine, with no differences between the two compounds. In addition, the concentration-effect curves for GSNO and NACysNO in arteries precontracted with KCl were shifted to the right in relation to the phenylephrine-induced contractile response. When we compared the sensitivity to the two compounds in aortas precontracted with phenylephrine and KCl, we observed that the sensitivity was lower in contractions induced by 30 mM KCl, as shown by the pD₂ values for GSNO (5.02 ± 0.03, n = 6) and NACysNO (5.49 ± 0.04, n = 6) than in phenylephrine-precontracted arteries. However, no significant differences were observed among the pD₂ values obtained at the 30 mM, 60 mM, and 90 mM KCl concentrations used.

View larger version (20K): [in this window] [in a new window]   Fig. 1.   Effects of high extracellular K+ concentration on the relaxation induced by GSNO or NACysNO in rat aortic rings in the absence of endothelium. The arteries were precontracted with 0.1 µM phenylephrine (, n = 8), 30 mM KCl (, n = 6), 60 mM KCl (, n = 6), or 90 mM KCl (, n = 6) and cumulative concentrations of GSNO (A, 1010-104 M) or NACysNO (B, 1010-104 M) were added. Data are means ± S.E.M. of n experiments performed on preparations obtained from different animals.

Effect of ODQ. In denuded arteries incubated with the guanylyl-cyclase inhibitor 10 µM ODQ, the concentration-effect curves were shifted to the right as shown in Fig. 2. pD₂ values were decreased for both GSNO, 4.09 ± 0.10 (n = 8) and NACysNO, 4.40 ± 0.05 (n = 8). The maximum 100% effect was obtained for the two compounds.

View larger version (19K): [in this window] [in a new window]   Fig. 2.   Effects of ODQ on the relaxation induced by GSNO or NACysNO in rat aortic rings in the absence of endothelium. The figure illustrates the effect of 10 µM ODQ on GSNO (A, control: , n = 6; ODQ: , n = 6)- or NACysNO (B, control: , n = 6; ODQ: , n = 6)-induced relaxation. Tissues were pretreated with ODQ for 30 min prior to application of 0.1 µM phenylephrine. Data are means ± S.E.M. of n experiments performed on preparations obtained from different animals.

Effect of K⁺ Channel Blockers. Incubation with the Ca²⁺-activated K⁺ channel blocker, charybdotoxin (100 nM), did not alter the relaxation induced by GSNO and NACysNO in denuded arteries precontracted with phenylephrine (data not shown). Similarly, the voltage-gated K⁺ channel blocker 4-aminopyridine (5 mM) had no inhibitory effect on the relaxation induced by GSNO and NACysNO (data not shown).

View larger version (17K): [in this window] [in a new window]   Fig. 3.   Effects of apamin on the relaxation induced by GSNO or NACysNO in rat aortic rings in the absence of endothelium. The figure illustrates the effect of 500 nM apamin on GSNO (A, control: , n = 6; apamin: , n = 6)- or NACysNO (B, control: , n = 6; apamin: , n = 6)-induced relaxation. Tissues were pretreated with apamin for 1 h prior to application of 0.1 µM phenylephrine. Data are mean ± S.E.M. of n experiments performed on preparations obtained from different animals.

View larger version (18K): [in this window] [in a new window]   Fig. 4.   Effects of glibenclamide on the relaxation induced by GSNO or NACysNO in rat aortic rings in the absence of endothelium. The figure illustrates the effect of 3 µM glibenclamide on GSNO (A, control: , n = 6; glibenclamide: , n = 7)- or NACysNO (B, control: , n = 6; glibenclamide: , n = 7)-induced relaxation. Tissues were pretreated with glibenclamide for 1 h prior to application of 0.1 µM phenylephrine. Data are means ± S.E.M. of n experiments performed on preparations obtained from different animals.

Effects of GSNO and NACysNO on the Contractile Response Induced by Phenylephrine in Ca²⁺-Free Solution. To investigate whether the relaxation induced by GSNO and NACysNO involves Ca²⁺ uptake by the sarcoplasmic reticulum, we studied the effect of increasing concentrations of GSNO and NACysNO on the contractile responses induced by phenylephrine in Ca²⁺-free solution. As shown in Fig. 5, we observed that at the concentrations of GSNO and NACysNO studied, there was a significant reduction in the amplitude of the phasic contractile response induced by phenylephrine compared with the phasic contractile response in the absence of the two compounds, and that this inhibitory effect was not concentration-dependent.

View larger version (48K): [in this window] [in a new window]   Fig. 5.   Effect of GSNO (A) and NACysNO (B) on Ca2+ release induced by the agonist phenylephrine. The tissues were placed in Ca2+-free Krebs' solution for 1 min, and contractile responses to 10 µM phenylephrine were induced (control group = open columns). Tissues were then rinsed with Krebs' solution until relaxation was obtained, then for an additional 30 min, and incubated for 5 min with different concentrations of GSNO or NACysNO (1010-104 M). The tissues were then placed again in Ca2+-free Krebs' solution for 1 min and contractile responses to 10 µM phenylephrine were induced in the presence of this compound. The closed columns represent a time control.

Effects of Ryanodine and Thapsigargin on Relaxation Induced by GSNO and NACysNO. When the aortas were incubated with 1 µM thapsigargin, the relaxation of arteries precontracted with phenylephrine induced by GSNO and NACysNO was not altered. The same was observed in the presence of 30 µM ryanodine, as shown in Figs. 6 and 7. Thapsigargin significantly increased resting tension by about 50% in comparison with 0.1 µM phenylephrine (data not shown). Similarly, pretreatment with ryanodine significantly increased resting tension by about 50% (data not shown).

View larger version (18K): [in this window] [in a new window]   Fig. 6.   Effects of thapsigargin on the relaxation induced by GSNO or NACysNO in rat aortic rings in the absence of endothelium. The figure illustrates the effect of 1 µM thapsigargin on GSNO (A, control: , n = 6; thapsigargin: , n = 6)- or NACysNO (B, control: , n = 6; thapsigargin: , n = 6)-induced relaxation. Tissues were pretreated with thapsigargin for 1 h prior to application of 10 µM phenylephrine. Data are means ± S.E.M. of n experiments performed on preparations obtained from different animals.

View larger version (17K): [in this window] [in a new window]   Fig. 7.   Effects of ryanodine on the relaxation induced by GSNO or NACysNO in rat aortic rings in the absence of endothelium. The figure illustrates the effect of 30 µM ryanodine on GSNO (A, control: , n = 6; ryanodine: , n = 6)- or NACysNO (B, control: , n = 6; ryanodine: , n = 6)-induced relaxation. Tissues were pretreated with ryanodine for 1 h prior to application of 10 µM phenylephrine. Data are means ± S.E.M. of n experiments performed on preparations obtained from different animals.

Effect of pH on NO Production. NO production was dependent on the concentration of GSNO and NACysNO present in the incubation bath, and maximal production was obtained at 1 mM in both cases, independently of the pH studied. We also observed that only above 100 µM GSNO or NACysNO was it possible to detect NO production. Figure 8 shows the effect of changes to acid pH (pH 4.0) or alkaline pH (pH 9.0) on the NO production induced by several concentrations of GSNO and NACysNO. At pH 4.0, GSNO at all concentrations used except 1 mM produced a higher concentration of NO than at pH 7.4. In contrast, at pH 9.0, NO production induced by GSNO was significantly increased only at 1 mM. At pH 4.0, the NO produced from NACysNO was not different at any of the concentrations studied except 1 mM NACysNO. At this concentration, NO production was significantly reduced in relation to pH 7.4. At pH 9.0, NO production by NACysNO was not altered at any concentration studied.

View larger version (24K): [in this window] [in a new window]   Fig. 8.   Effect of pH on nitric oxide concentration produced by GSNO (A) and NACysNO (B). Data are means ± S.E.M. of at least six experiments. The results are expressed as NO concentration (nM) produced by different concentrations of GSNO (A, 105-103 M) or NACysNO (B, 105-103 M) at the pH values indicated.

	Discussion

Top Abstract Introduction Experimental Procedures Results Discussion References

The present study demonstrated that the nitrosothiols GSNO and NACysNO induced a complete relaxation of aortas precontracted with phenylephrine and a marked relaxation of aortas precontracted with KCl. One of the most intriguing observations was that there was no correlation between the magnitude of NO formation and the magnitude of relaxation. The inhibition of the relaxation induced by GSNO and NACysNO when the aortas were precontracted with high K⁺ in comparison to the arteries precontracted with phenylephrine indicates the involvement of membrane hyperpolarization. The dependence of the relaxation induced by GSNO and NACysNO on a membrane K⁺ gradient suggests that K⁺ efflux through K⁺ channels is involved in these effects. Similarly, Zhao et al. (1997) demonstrated that both the authentic 0.3 µM NO and 10 µM nitroprusside produced significant relaxation in isolated pulmonary arteries precontracted with KCl. Further elevation of K⁺ concentration from 20 to 60 mM resulted in a significant increase in contraction but caused a marked decline in NO- and nitroprusside-mediated pulmonary artery relaxation.

Recent studies have suggested that cGMP may be involved in vascular smooth muscle relaxation in response to vasodilator drugs. Hussain et al. (1997) demonstrated that the selective inhibitor of guanylyl cyclase, 10 µM ODQ, abolished the relaxation of rabbit aortas induced by 30 µM carbon monoxide, whereas only a partial attenuation of NO-induced relaxation was achieved with the same concentration of ODQ. This supports the idea that ODQ is less potent in inhibiting relaxation induced by NO and thereby implicates a component of NO-induced relaxation that is independent of soluble guanylyl cyclase/cGMP. In our studies, 10 µM ODQ significantly attenuated but did not abolish the relaxation induced by high concentrations of nitrosothiols (above 10⁴ M). These high concentrations of nitrosothiols could overcome the guanylyl cyclase inhibition by ODQ, and cGMP was probably generated. In addition, at these nitrosothiol concentrations there is production of NO at pH 7.4, as shown in Fig. 8. This increase in NO production may be responsible for overcoming the inhibitory effect of ODQ on the relaxation induced by the compounds. In contrast, Onoue and Katusic (1998) demonstrated that DEA-NONOate, which in turn is a potent activator of guanylyl cyclase, caused concentration-dependent relaxation. These authors demonstrated that the highest concentration (3 µM) of ODQ abolished the production of cGMP in response to DEA-NONOate, but relaxation induced by DEA-NONOate was almost complete, suggesting that mechanisms independent of the production of cGMP may mediate the relaxation induced by high concentrations of the NO donor. Yet Onoue and Katusic (1998) showed in cerebral arteries that the K_Ca channel blocker charybdotoxin significantly reduced the relaxation to DEA-NONOate that is resistant to ODQ, supporting the idea that NO may activate K⁺ channels independently of cGMP.

According to Archer et al. (1994), in rat pulmonary arteries the relaxation induced by NO that is cGMP-dependent, but not cGMP-independent, is inhibited by tetraethylammonium, a classical K⁺ channel blocker, and by charybdotoxin, an inhibitor of K_Ca channels. Increasing extracellular K⁺ concentration also inhibited the relaxation induced by NO that is dependent on cGMP without reducing the levels of cGMP in vascular smooth muscle. The failure of NO to relax arteries precontracted with KCl was due to the elimination of the chemical gradient for K⁺ efflux, and not to impaired cGMP synthesis, since cGMP levels were similar in arteries precontracted with norepinephrine and KCl. In whole-cell patch-clamp experiments, NO and cGMP increased whole-cell K⁺ current by activating K_Ca channels, an effect mimicked by intracellular administration of (Sp)-guanosine cyclic 3',5'-phosphorothioate, a preferential cGMP-dependent protein kinase activator. Thus, NO and cGMP relax vascular smooth muscle by a cGMP-dependent protein kinase that is dependent on the activation of K⁺ channels. In contrast, in our studies charybdotoxin did not alter the relaxation induced by GSNO or NACysNO. The same tendency was observed with the voltage-gated K⁺ channel blocker 4-aminopyridine. The present study differs from that of Zhao et al. (1997), who showed that authentic 0.3 µM NO and 10 µM sodium nitroprusside produce a significant relaxation in pulmonary arteries that is significantly inhibited by 4-aminopyridine and by charybdotoxin. Zhao et al. (1997) also demonstrated that the ATP-sensitive K⁺ channel blocker glibenclamide had no effect on the relaxation, whereas we demonstrated that glibenclamide blocked the relaxation induced by GSNO and NACysNO. We also demonstrated that apamin, a low-conductance Ca²⁺-activated K⁺ channel blocker, inhibited the relaxation induced by GSNO and NACysNO. Khan et al. (1998), investigating the relaxation induced by nitroglycerin, observed that this relaxation was significantly inhibited by charybdotoxin, but not by glibenclamide or apamin. Our results indicate that GSNO and NACysNO partially relax the rat aorta by activating K_ATP channels and partially by activating low-conductance Ca²⁺-activated K⁺ channels in rat aorta. The differences between our results and the data reported by Archer et al. (1994), Zhao et al. (1997), and Khan et al. (1998) may be explained by the fact that pulmonary artery and aorta appear to have different expression patterns of K⁺ channels and Ca²⁺ channels.

One mechanism proposed to explain the relaxation of vascular smooth muscle cells produced by cytoplasmic Ca²⁺ reduction is the activation of sarcoplasmic reticulum Ca²⁺-ATPase by protein kinase G stimulated with cGMP (Raeymaekers et al., 1988). Several studies have demonstrated that thapsigargin inhibits repletion of Ca²⁺ stores that have been previously depleted by phenylephrine in arterial rings (Baró and Eisner, 1992; Ceron and Bendhack, 1998).

In our studies, we observed that at all the GSNO and NACysNO concentrations studied, there was a significant reduction of the phasic contractile response induced by phenylephrine in Ca²⁺-free solution. Further investigations indicated that relaxation induced by GSNO and NACysNO was not altered by inhibition of the sarcoplasmic reticulum Ca²⁺ release channel by ryanodine or by the inhibition of sarcoplasmic reticulum Ca²⁺-ATPase by thapsigargin. Taken together, these data suggest that GSNO and NACysNO relaxation does not involve reduction of Ca²⁺ levels by uptake by the sarcoplasmic reticulum. However, Khan et al. (1998) observed a significant attenuation by thapsigargin of the relaxation induced by nitroglycerin, indicating the participation of Ca²⁺ uptake by the sarcoplasmic reticulum in relaxation induced by nitroglycerin.

The release of NO from S-nitrosothiols is considered necessary for their pharmacological and potential biological activity. Because S-nitrosothiols are chemically unstable, it has been generally assumed that spontaneously released NO mediates the vascular relaxant activity, although S-nitrosothiols stability was found not to be related to biological activity in several assays (Mathews and Kerr, 1993). The stability of nitrosothiols shows remarkable variation ranging from seconds to hours. First-order half-lives are 159 h for GSNO and 93.9 h for NACysNO (as calculated on the basis of first-order decomposition kinetics in 0.5 mM physiological salt solution, pH 7.4, 37°C), indicating that these compounds are stable. Mathews and Kerr (1993) demonstrated that the starting thiols are inactive, and that the -S-N==O functional group is responsible for the biological activity of the nitrosothiols. Accordingly, Ignarro et al. (1981) showed that S-nitrosothiols are active intermediates involved in the biological activities of organic nitrates and nitrites. There is evidence that the decomposition of S-nitrothiols and spontaneous release of NO in biological solutions are not responsible for the vasodilation induced by these compounds (Kowaluk and Fung, 1990; Mathews and Kerr, 1993). As shown by Travis et al. (1996), S-Nitroso-,-dimethylcysteine does not generate detectable amounts of NO although it generates significant amounts of cGMP. These authors suggested that the extracellular or intracellular generation of NO is not the only mechanism by which the nitrosothiols generate cGMP in vascular smooth muscle.

In the present study, we observed that GSNO and NACysNO release NO, but the amount of NO produced was detectable only with high concentrations (0.1 mM) of both compounds. One way to explain the dissociation between relaxation and NO production, as suggested by Hoque et al. (1999) and Travis et al. (2000), is that there is a nitrosothiol-binding protein that acts like a receptor to mediate the actions of vascular smooth muscle cells. There are recognition sites that may be a novel class of receptors which specifically recognize structurally similar S-nitrosothiols such as L-S-nitroso-,-dimethylcysteine but not larger molecular weight compound S-nitrosothiols such as L-S-nitrosoglutathione.

In the current studies, the maximal relaxation was obtained at pH 7.4 with 10 µM GSNO and NACysNO, and the generation of NO was detectable only from 0.1 mM GSNO and NACysNO. Ignarro et al. (1981) demonstrated that the stability of S-nitrosothiols in solution varied markedly depending on temperature, pH, and the presence of O₂. They observed that pH 8.8 profoundly decreased their stability. Similarly, Feelisch (1991) reported that S-nitrosothiols release NO via homolytic cleavage of the S-N bond, a process that is enhanced by alkaline pH. In contrast, in the procedure for the calibration of the NO ISONOP sensor (WPI) performed in our laboratory according to the manufacturer's instructions, each mole of S-nitroso-N-acetylpenicillamine generates 0.539 mol of NO in the proposed experimental parameters. A pivotal parameter for this measurement is pH; at alkaline pH (9.0) the decomposition to NO is negligible, and at pH 4.0 the rate is as reported above. Since we did not observe a correlation between relaxation and NO production, and considering the controversial findings in the literature, we studied the effect of pH changes on NO production. Our results agreed with the above instructions for GSNO, with increased NO formation at pH 4.0 and no changes at pH 9.0. However, formation of NO from NACysNO was not altered at the pH levels studied.

In summary, relaxation induced by GSNO and NACysNO is partially due to activation of K_ATP channel and partially due to activation of low-conductance Ca²⁺-activated K⁺ channels in rat aorta. However, the ability of the nitrosothiol compounds to overcome the inhibitory effect of high extracellular K⁺ concentrations suggests another mechanism of relaxation contributing to the nitrosothiol response. The relaxation induced by GSNO and NACysNO was not related to the NO produced. Our results suggest that NACysNO and GSNO can produce relaxation by themselves by direct activation of the K_ATP and K_Ca channels and/or by stimulation of cGMP production or another mechanism.