Introduction

Top Summary Introduction Materials & Methods Results Discussion References

Organic nitrates such as GTN are widely used for the treatment of coronary artery disease and heart failure. It has been shown that organic nitrates are prodrugs that undergo enzymatic bioactivation within the vascular wall to release NO, which is the pharmacologically active compound (Feelisch and Noack, 1987; Chung and Fung, 1990). GTN-induced vasorelaxation is preceded by vascular formation of the 1,2-GDN and 1,3-GDN and by activation of soluble guanylate cyclase producing cGMP (Brien et al., 1986). There is a general agreement that other organic nitrates, such as isosorbide dinitrate, isosorbide-5-nitrate, and pentaerythritol tetranitrate, undergo a similar bioactivation process as a prerequisite for their pharmacological activity (Ahlner et al., 1991). The second messenger cGMP activates the cGMP-dependent protein kinase and initiates several effects such as phosphorylation of myosin light chain, sequestration of intracellular calcium, reduction of calcium entry from the extracellular space, reduced release of intracellularly stored calcium, and inhibition of formation of inositol-1,4,5-triphosphate (Pfitzer et al., 1984; Collins et al., 1986; Twort and van Breemen, 1988; Lang and Lewis, 1989).

Among antianginal drugs used for therapy of coronary artery disease, organic nitrates elicit unique and favorable hemodynamic changes. The most striking difference from other antianginal drugs such as  blockers and calcium antagonists is the preferential venodilation causing preload reduction (Bassenge and Stuart, 1986). Selective reduction of preload has several advantages for patients with coronary artery disease; it reduces left ventricular end-diastolic pressure and systolic ventricular wall tension and increases cardiac output. Interestingly, the preferential reduction of preload also distinguishes organic nitrates from other nitrovasodilators such as sodium nitroprusside despite the presumed common generation of NO (Armstrong et al., 1975). It has been shown that sodium nitroprusside undergoes a completely different bioactivation process in the vascular wall (Bates et al., 1991; Kowaluk et al., 1992). Thus, it is likely that selective reduction of preload is related to the enzymatic bioactivation of organic nitrates.

One major difference between arteries and veins is the intensity of endogenous NO production in endothelial cells. Stimulation of venous endothelium results in a low production of NO as demonstrated by the weak endothelium-dependent vasorelaxation in veins of different species, including humans (De Mey and Vanhoutte, 1982; Lüscher et al., 1988; Kojda et al., 1994). The different intensity of endogenous NO production in arteries and veins might have an impact on vascular bioactivation of organic nitrates, leading to preferential venodilation. In accordance, it has been shown previously that endogenous NO production by the vascular endothelium reduces the vasodilator potency of organic nitrates such as GTN (Alheid et al., 1987; Moncada et al., 1991; Kojda et al., 1994). In this study, we sought to determine the influence on the activity of GTN of pretreatment of coronary arteries with NO. We measured the kinetics of NO release from the used NO donors, the vascular formation of 1,2-GDN and 1,3-GDN, alterations in the activity of isolated human soluble guanylate cyclase, vascular accumulation of cGMP, and vasorelaxation induced by GTN and by a specific stimulator of cGMP-dependent protein kinase. We demonstrate that continuous subjection of coronary arteries to micromolar concentrations of NO inhibits vascular bioactivation of GTN, accumulation of cGMP, and vasorelaxation. Pretreatment with NO also caused a desensitization of soluble guanylate cyclase.

	Materials and Methods

Top Summary Introduction Materials & Methods Results Discussion References

Measurement of NO release. Release of NO by GTN, DEA/NO, SNAP, and SPER/NO was measured at pH 7.4 and 37° in the presence of oxygen with a commercially available NO meter (ISO-NO; World Precision Instruments, Berlin, Germany) that works in a manner similar to that of the well known Clark-type electrode for oxygen. Calibration of the electrode was performed daily before the experiments. Volumes (10 µl) of aqueous KNO₂ solution (2 µM), used as a generator of NO, were added cumulatively (four times) to 300 µl of a mixture of KI and H₂SO₄ (0.1 M each). A typical plot of the measured signal (in pA) versus the NO concentration (in nM), calculated on the basis of a quantitative reaction of KNO₂ to NO according to the equation 2KNO₂ + 2KI + 2H₂SO₄2NO + I₂ + 2H₂O + 2K₂SO₄, yielded a linear relationship with a correlation coefficient of 0.999 and a slope of 0.87 nM NO/pA.

Preparation of guanylate cyclase. Preparation of human platelet guanylate cyclase was performed as reported previously (Kojda and Noack, 1993). Briefly, 1000 ml of human platelet-rich plasma was mixed with 50 ml of EDTA (0.1 M), and platelets were concentrated by centrifugation (1000 × g for 10 min). The platelets were washed twice with Tris buffer (50 mM, pH 7.6) containing 154 mM NaCl by repeated resuspension and centrifugation (500 × g). Washed platelets were resuspended in the Tris buffer (16 ml) and slowly cooled to 4°, which was the temperature for the next steps. Lysis of platelets was achieved by the addition of 100 ml of hypotonic Tris buffer (5 mM, pH 7.6) containing 0.05% leupeptin, 2 mM phenylmethylsulfonyl fluoride, and 1 mM DTT. During lysis, platelets were sonicated (50 W, 30 sec). The supernatant from centrifugation at 10,000 × g for 10 min was collected and recentrifuged at 105,000 × g for 1 hr. The obtained cytosolic fraction was loaded onto a diethylaminoethanol-Sepharose column (HiLoad 26/10 Q Sepharose HP; Pharmacia, Freiburg, Germany) after preequilibration with Tris buffer containing 1 mM DTT. A linear sodium chloride (0-0.4 M) gradient in the same buffer was started. Active fractions (cGMP accumulation >30% of the maximal value) were identified after stimulation with 500 µM SNAP in the presence of 1 mg/ml bovine serum albumin and were pooled and stored in aliquots at 80°. Protein content was measured according to the method of Bradford (1976) with bovine serum albumin as a standard.

Determination of guanylate cyclase activity. Specific activity of soluble guanylate cyclase was measured on the basis of the formation of [³²P]cGMP from [-³²P]GTP as described previously (Schulz and Böhme, 1984). Briefly, soluble guanylate cyclase of the single diethylaminoethanol fractions (20-40 µg of protein) was incubated in a total volume of 100 µl of a triethanolamine HCl buffer (50 mM, pH 7.4, 37°) containing 5 nM [-³²P]GTP (0.4 µCi), 100 µM GTP, 1 mM cGMP, 1 mM 3-isobutyl-1-methylxanthine, 1 mM MgCl₂, and 1 mM DTT in the presence of 500 µM SNAP or vehicle (0.25% DMSO). To determine the dose-dependent effects of SNAP, SPER/NO, DEA/NO, and GTN (with or without 5 mM cysteine), the assay volume contained concentrations of these drugs or vehicle as indicated in Results.

Preparation of isolated vessel segments. Right coronary arteries were obtained from the local slaughterhouse and taken from the hearts of freshly slaughtered female pigs (5-7 months old). Coronary arteries were prepared free from the aorta to the ramus interventricularis posterior and perfused with cold KH buffer, pH 7.4, containing 143.07 mM Na⁺, 5.87 mM K⁺, 1.6 mM Ca²⁺, 1.18 mM Mg²⁺, 125.96 mM Cl, 25.00 mM HCO₃, 1.18 mM H2PO₄, 1.18 mM SO₄², and 5.05 mM glucose. The arteries were cut from their muscle foundation, immediately stored in cooled KH buffer, and transferred into the laboratory, where they were carefully dissected free from all surrounding tissue. The proximal ends were rejected, and the remainder of the arteries was cut into rings (length, 5 mm). Great care was taken to preserve the intimal endothelium. In some cases, its function was controlled in separately performed organ bath studies. Two to four coronary segments were put into a polypropylene vial and equilibrated (37°) in modified and oxygenated (95% O₂/5% CO₂) KH buffer for 90 min. The buffer was exchanged every 15 min.

Determination of vascular metabolites of GTN. The porcine coronary artery rings were incubated in KH buffer after the application of vehicle [0.1% ethanol (v/v) or 0.01 M NaOH], 100 µM DEA/NO (four cumulative applications every 8 min), 100 µM SNAP, 200 µM SPER/NO, and 100 µM GTN. Incubation was terminated by repeated washout (two times at once and two times after 5 and 15 min). Then, [2-¹⁴C]GTN (specific activity, 55 mCi/mmol) was added. The amount of the radioactivity was 0.25 µCi (2.28 µM GTN). After 2 min, these coronary rings were flash frozen with liquid nitrogen and stored at 20°.

Separation of GTN and the dinitrate metabolites by HPLC. After reconstitution of the evaporated dichloromethane phase (see above) in 55 µl of dichloromethane, 5 µl of a stock solution of GTN, 1,2-GDN, and 1,3-GDN (2 mM each) was added (final concentration, 167 µM), and the mixture was used directly for separation by HPLC. A liquid chromatograph was used (655A-11, LC-controller L 5000, chromatointegrator D-2000; Merck/Hitachi, Merck, Darmstadt, Germany), including a Li Chro Cart R 250-4 Superspher 100 RP-18 column combined with a Li Chro Cart R 4-4 filled with Li Chrosorb RP-18 (5 µm) as a precolumn. The columns were equilibrated with methanol/phosphate buffer [50 mM, pH 7.4; 4:6 (v/v)] at a flow rate of 0.5 ml/min (pressure, 193 kg/cm²) and room temperature (20-24°). Analysis was started by the injection of 50 µl of the reconstituted and spiked dichloromethane solution by means of an injection slope. A representative tracing of the separation is shown in Fig. 1. Collection of samples (24-sec steps for dinitrates and 60-sec steps for GTN) was done in 20-ml plastic tubes (Frac 100 fraction collector; Pharmacia) according to the peak pattern registered in parallel by UV detection at 210 nm (UV-Detector, 655 A variable wavelength; Merck Hitachi). After the addition of 10 ml of Rotiszint (eco plus; Carl Roth & Co., Karlsruhe, Germany), scintigraphic determination (in cpm) was done with a Beckman Instruments counter (LS 6500 or LS 5000 TD; Columbia, MD).

View larger version (15K): [in this window] [in a new window]   Fig. 1.   Separation of 1,2-GDN, 1,3-GDN, and GTN extracted from vascular tissue and reconstituted in dichloromethane. Radioactivity was measured in fractions collected at time intervals of 24 sec (8-16 min) and 60 sec (30-45 min) that eluted from the HPLC column. Incubations of isolated porcine coronary arteries were performed for 2 min with 2.28 µM [2-14C]GTN (specific activity, 55 mCi/mmol). Visualization of the peaks to simplify fraction collection was accomplished by spiking the dichloromethane extract to give concentrations of GTN and its metabolites (167 µM each) that were easily detected with a UV detector.

Determination of vascular cGMP accumulation. Porcine coronary artery rings were freshly prepared and cut to a length of 1 cm. These rings were equilibrated for 3 hr at 37° in polyethylene vials containing continuously oxygenated KH buffer, which was changed every 30 min. Then, the rings were incubated with GTN (100 µM), SNAP (100 µM), or vehicle (0.9% NaCl and 0.05% DMSO in KH buffer) for 30 min. Incubation was stopped by repeated (three times) washout with KH buffer (within 30 min). Thereafter, the rings were incubated again with GTN (10 µM), and after 5 min, rings were flash-frozen in liquid nitrogen and stored at 80°. Frozen artery rings were homogenized with a Polytron in 1 ml of ice-cold HClO₄ (10%) and then centrifuged at 4500 × g for 10 min. The pellet was used for protein determination (Bradford, 1976); 900 µl of supernatant was neutralized (pH 7.4) with K₃PO₄, centrifuged again, and used directly for determination of cGMP by radioimmunoassay with ¹²⁵I-cGMP as radiolabeled antigen. Preliminary experiments with this method yielded recovery rates for cGMP and protein of >90% (Kojda and Noack, 1993).

Organ bath studies. Porcine coronary arteries were cut into ring segments (4 mm) and fixed between stainless-steel hooks in a waterjacketed organ bath (37°) as described previously (Kojda et al., 1991). Resting tension was 2 g. After equilibration (1 hr), contractile function was tested by the addition of KCl (60 mM) and PGF₂ (0.1-100 µM) to reach a maximal tension of 5 g. The presence of intact endothelium was verified by complete, transient relaxation of PGF₂-precontracted (10 µM) segments after the application of 3 nM substance P (Cocks and Angus, 1983). Vasorelaxing activities of GTN (1 nM to 100 µM), DEA/NO (1 nM to 100 µM), and SNAP (1 nM to 100 µM) were evaluated by cumulative application after precontraction with PGF₂ (50 µM). To study the influence of NO on relaxant activity of GTN in these arteries, cumulative application of GTN was performed after washout (15 min) of a 30-min preincubation with either vehicle (0.01 M NaOH and 0.9% NaCl), GTN (100 µM), SNAP (100 µM), DEA/NO (100 µM once or 100 µM four times every 8 min), or SPER/NO (200 µM). In some experiments, this preincubation procedure was followed by the cumulative application of 8-pCPT-cGMP, a membrane-permeable activator of cGMP-dependent protein kinase (Sekhar et al., 1992).

Substances and solutions. SNAP was synthesized according to Field et al. (1978) as described previously (Kojda et al., 1996). GTN (4.404 mM in 154 mM NaCl, used directly as stock solution) was generously provided by Schwarz Pharma AG (Monheim, Germany). DEA/NO and SPER/NO was a gift from Dr. L. Keefer (National Cancer Institute, Frederick, MD). 8-pCPT-cGMP was obtained from Biolog (Bremen, Germany). [2-¹⁴C]GTN (specific activity, 55 mCi/mmol) was obtained from Biotrend (Köln, Germany). [-³²P]GTP (specific activity, 800 Ci/mmol) was obtained from Du Pont de Nemours (Bad Homburg, Germany). Leupeptin, phenylephrine, acetylcholine, 3-isobutyl-1-methylxanthine, PGF₂, and phenylmethylsulfonyl fluoride were obtained from Sigma Chemie (Deisenhofen, Germany). All other chemicals (analytical grade) were obtained from Merck.

Statistics. Vasorelaxation is expressed as remaining percentage of the contractile response achieved with PGF₂ (50 µM) at the beginning of the experiment. The concentrations for half-maximal inhibition of precontraction (IC₅₀) were calculated from the individual concentration-effect curves as proposed by Hafner et al. (1977). The pD₂ values, representing the negative logarithms of the half-maximal inhibiting concentrations, were taken to test for significant differences. All data were analyzed by one-way analysis of variance with subsequent Student-Newman-Keuls test (SAS PC Software 6.04, PROC ANOVA; SAS Institute, Cary, NC) and are expressed as mean ± standard error values. Significant differences were evaluated by using Student's t test, and a value of p < 0.05 was considered significant.

	Results

Top Summary Introduction Materials & Methods Results Discussion References

Kinetics of NO release by the different nitrovasodilators. Release of NO by GTN, SNAP, DEA/NO, and SPER/NO, which was measured under conditions that were present in the experiments with isolated enzymes and tissues (pH 7.4, 37°, presence of oxygen), showed substantial differences (Fig. 2). DEA/NO rapidly degraded, yielded the highest concentration of NO, and NO release was completed after 6-7 min. SNAP showed a similar time course of NO release but a much lower peak concentration of NO. SPER/NO degraded slowly. Maximal concentrations of NO were similar to those released by SNAP but occurred later. Maximal NO concentrations remained constant for 5 min and then slowly declined. Calculation of the area under curve resulted in similar values for DEA/NO (9,755 ± 1,404 nM × min) and SPER/NO (14,697 ± 1,770 nM × min), whereas the release of NO from SNAP was significantly lower (2,890 ± 311 nM × min, p < 0.01). There was no detectable NO release from 10 µM GTN in the absence and presence of 5 mM cysteine.

View larger version (22K): [in this window] [in a new window]   Fig. 2.   Kinetics of NO release from 10 µM of the spontaneous NO donors SPER/NO, DEA/NO, and SNAP. The experiments were performed in oxygenated buffer (37°, pH 7.4), and NO was measured polarographically. Plotted are the mean ± standard error values of NO concentration measured at each time point in three different experiments.

Activation of soluble guanylate cyclase from human platelets. The basal activity of the enzyme preparation was 70.2 ± 10.3 pmol cGMP/mg/min (16 experiments). The spontaneous NO donors SPER/NO, DEA/NO, and SNAP dose-dependently activated soluble guanylate cyclase partially purified from human platelets. DEA/NO and SPER/NO equieffectively activated the enzyme, but the activity of NO donor SNAP was 10-fold lower (data not shown). The specific activity of soluble guanylate cyclase after incubation with a maximally effective concentration of DEA/NO, SPER/NO, and SNAP was 5.18 ± 0.47 (six experiments), 6.78 ± 0.27 (six experiments), and 5.38 ± 0.31 (six experiments) nmol of cGMP/mg/min, respectively. In presence of equimolar concentrations of oxyhemoglobin, a scavenger of NO, the stimulating effect of 10 µM SNAP (382 ± 43 pmol of cGMP/mg/min, six experiments) was abolished completely.

Desensitization of soluble guanylate cyclase from human platelets. To determine whether prolonged subjection of soluble guanylate cyclase with NO results in a change of enzyme activity, maximal stimulation of the enzyme with 500 µM SNAP was investigated after preincubation with either vehicle (0.05% DMSO) or 100 µM SNAP for different time periods (Fig. 3). The significantly lower maximal stimulation after preincubation with SNAP for 30 and 45 min indicates a desensitization of the enzyme by NO. A similar preincubation procedure using 10 µM SNAP had no effect on the activity of the enzyme (data not shown).

View larger version (26K): [in this window] [in a new window]   Fig. 3.   Influence of a preincubation with either vehicle (0.05% DMSO) or 100 µM SNAP for different time periods on maximal activation of soluble guanylate cyclase isolated from human platelets. Reactions to measure maximal activation were started after the preincubation periods by simultaneous addition of the radioactive substrate (-32P-GTP) and 500 µM SNAP. Plotted are mean ± standard error values of specific activity measured in four different experiments for each time point. *, Significant difference compared with control conditions, p < 0.05.

Generation of cGMP in isolated arteries. Accumulation of cGMP was determined in porcine coronary artery rings stimulated with GTN. To determine whether prolonged subjection of these rings with NO results in a change on cGMP accumulation, stimulation of the arteries with 10 µM GTN was preceded by preincubation with either SNAP (100 µM) or GTN (100 µM) for 30 min. After preincubation with vehicle, a GTN-induced (10 µM) cGMP accumulation of 8.2 ± 1.4 pmol/mg was observed. Preincubation with both SNAP and GTN significantly diminished GTN-induced (10 µM) accumulation of cGMP to 5.4 ± 1.2 pmol/mg (p < 0.05) and 1.3 ± 0.4 pmol/mg (p < 0.01), respectively.

Vascular formation of 1,2-GDN and 1,3-GDN. The total radioactivity recovered by extraction of porcine coronary rings after a 2-min incubation period with ¹⁴C-GTN was separated into three fractions (¹⁴C-1,3-GDN, ¹⁴C-1,2-GDN, and ¹⁴C-GTN) by HPLC and corrected for protein content (Fig. 1, Table 1). A 5-min incubation period increased these values only slightly for GTN (from 5732 ± 977 to 6779 ± 384 cpm/mg), 1,2-GDN (from 5705 ± 572 to 6761 ± 192 cpm/mg), and 1,3-GDN (from 6535 ± 532 to 7762 ± 256 cpm/mg). Thus, all experiments on vascular metabolization of GTN were done using a 2-min incubation with ¹⁴C-GTN.

View larger version (22K): [in this window] [in a new window]   Fig. 4.   Inhibition of GTN metabolization in the vascular wall of porcine coronary arteries after pretreatment with the different spontaneous NO donors SNAP, DEA/NO, and SPER/NO and by GTN itself compared with controls (preincubation with vehicle, set to 100%). Plotted are mean ± standard error values of the percentage changes in 1,2-GDN, 1,3-GDN, and GTN that were detected in one or two coronary rings of four or five different animals (for absolute values, see Table 1). The 30-min pretreatment period was followed by a 15-min washout period and subsequent subjection of the rings with 2.28 µM [2-14C]GTN for 2 min. GTN and the GDN metabolites were extracted and separated by HPLC as shown in Fig. 1. *, Significant difference compared with control conditions, p < 0.05. Preincubation with any of the NO donors substantially decreased bioactivation of GTN as indicated by the reduced vascular formation of the GDNs.

Relaxation of porcine coronary arteries. The vasorelaxing potency of GTN was determined in isolated ring segments of porcine right coronary artery that had been preincubated for 30 min with vehicle, GTN, or a NO donor (DEA/NO, SPER/NO, or SNAP). The different vasorelaxing potencies of GTN are listed in Table 2. Preincubation with any spontaneous NO donor resulted in a comparable shift to the right of the dose-response curve of GTN (Fig. 5), indicating a desensitization of the vessel segments to the relaxant effects of GTN. The decrease in respective pD₂ values was approximately one order of magnitude (Table 2). Preincubation with GTN, which was performed as a control experiment, resulted in the most pronounced rightward shift of the concentration-response curve of GTN (Fig. 5, Table 2). Interestingly, the time of subjection of the arteries to NO is most likely more important than the concentration of NO itself. Preincubation with a single application of 100 µM of DEA/NO, which liberated almost 10 times more NO compared with SPER/NO (Fig. 2), did not change the pD₂ value of GTN (6.51 ± 0.12, six experiments), whereas a single application of SPER/NO or a repetitive application of DEA/NO did (100 µM, every 8 min within 30 min) (Table 2). In contrast, pretreatment with GTN did not decrease the vasorelaxing potency of the NO donor SNAP (Table 2), indicating that under these conditions, the capacity to produce cGMP by NO or the efficacy of cGMP itself is not affected.

View this table: [in this window] [in a new window]   TABLE 2 Vasodilator potencies of GTN, SNAP, and DEA/NO as evaluated by cumulative application to ring segments of porcine coronary arteries precontracted with 50 µM PGF2 In some experiments, preincubation with GTN (100 µM), SNAP (100 µM), DEA/NO (four times 100 µM every 8 min), or SPER/NO (200 µM) preceded cumulative applications. Given are mean pD2- values (in log M) and standard error values for n individual experiments.

View larger version (26K): [in this window] [in a new window]   Fig. 5.   Inhibition of GTN-induced vasorelaxation of isolated porcine coronary arteries after pretreatment with the different spontaneous NO donors SNAP, DEA/NO, and SPER/NO compared with controls (preincubation with vehicle). The vasodilatory response is expressed as percentage of precontraction induced by 50 µM PGF2, and each dose-response curve was plotted by taking the respective mean values of one or two ring preparations from coronary arteries of 6-10 different animals [mean ± standard error (bars)]. The respective concentrations for half-maximal vasodilation are given in Table 2. It is evident that preincubation with any of the NO donors substantially decreased the vasorelaxing potency of GTN.

Activation of cGMP-dependent protein kinase in intact arteries. To study the sensitivity of vascular cGMP-dependent protein kinase, cumulative applications of 8-pCPT-cGMP, a congener of cGMP, were performed in porcine coronary arteries pretreated with vehicle, SPER/NO, or GTN. None of these pretreatments altered the vasodilator activity of 8-pCPT-cGMP indicating an unchanged activity of vascular cGMP-dependent protein kinase (Table 3). However, the very low vasorelaxing potency of 8-pCPT-cGMP in porcine coronary arteries might have masked any potential variation of the activity of cGMP-dependent protein kinase. Thus, the experiments were repeated using rat aorta. In this vessel type, 8-pCPT-cGMP is a much more potent vasodilator (Fig. 6). Nevertheless, preincubation of rat aorta with 200 µM SPER/NO for 30 min had no effect on relaxations in response to 8-pCPT-cGMP. These results indicate that prolonged subjection of vascular smooth muscle to either NO or GTN does not change the activity of cGMP-dependent protein kinase.

View this table: [in this window] [in a new window]   TABLE 3 Vasodilator potency of 8-pCPT-cGMP in isolated ring segments of porcine coronary artery precontracted with 50 µM PGF2 In these experiments, preincubation with vehicle (0.01 M NaOH and 0.9% NaCl), GTN (100 µM), or SPER/NO (200 µM) preceded cumulative application of 8-pCPT-cGMP. Given are mean pD2 values (in log M) and standard error value of n individual experiments. No significant differences were observed.

View larger version (20K): [in this window] [in a new window]   Fig. 6.   Lack of effect of pretreatment with the spontaneous NO donor SPER/NO compared with controls (preincubation with vehicle) on the vasodilator potency of 8-pCPT-cGMP, a cGMP analog that penetrates cells, does not undergo hydrolysis by phosphodiesterases, and specifically activates cGMP-dependent protein kinase. The vasodilatory response is expressed as percentage of precontraction induced by 3 µM phenylephrine, and each dose-response curve was plotted by taking the respective mean values of one or two ring preparations from rat aorta of nine different animals [mean ± standard error (bars)]. It is evident that preincubation with SPER/NO had no effect on activation of cGMP-dependent protein kinase.

	Discussion

Top Summary Introduction Materials & Methods Results Discussion References

We studied the effect of NO pretreatment on vascular bioactivation of GTN, the activity of soluble guanylate cyclase, the activity of cGMP-dependent protein kinase, and the vasorelaxing potency of GTN. Our main finding was that NO can reduce vascular formation of dinitrate metabolites from GTN and its vasorelaxing activity as well. For this effect, a continuous subjection of blood vessels to NO is more important than the concentration of NO itself. Our results suggest that the preferential venodilation, which is typical for organic nitrates such as GTN, is at least in part the result of the low endogenous production of NO by the vascular endothelium in veins.

Inhibition of vascular bioactivation of GTN by NO. The inhibitory effect of NO on vascular bioactivation of GTN is a new observation (Fig. 4, Table 1). The NO-induced impairment of formation of GDN-metabolites occurs in parallel with a substantially decreased vasodilator activity of GTN (Fig. 5, Table 2). A decreased vasodilator activity of GTN also occurred after pretreatment with NO of bovine coronary arteries (Zhang et al., 1994) and porcine vena cordis magna (Kojda et al., 1994). Generation of NO from GTN in tissues is most likely an enzymatic process, but a nonenzymatic cleavage of organic nitrates in the presence of thiols such as cysteine also occurs (Feelisch and Noack, 1987; Chung and Fung, 1990). It has been shown earlier that GTN-induced vasorelaxation is preceded by vascular formation of 1,2-GDN and 1,3-GDN (Brien et al., 1986). In our study, the formation of the GDNs was almost completed after 2 min, and a 1:1 ratio of 1,2-GDN to 1,3-GDN of was observed. Similar results were obtained previously (Fung et al., 1984). Theoretically, formation of 1,2-GDN should be twice as great as formation of 1,3-GDN. Thus, our results and those of others suggest that enzymatic denitration of GTN in the vascular wall preferentially occurs at C2 of the molecule.

Effects of NO pretreatment on soluble guanylate cyclase and cGMP-dependent protein kinase. The reduction in the vasodilator potency of GTN induced by pretreatment with SNAP and GTN correlated with a reduction in vascular cGMP accumulation. Preincubation with GTN showed the strongest effect on both GTN-induced vasodilation and GTN-induced vascular cGMP accumulation (see Results). Previous studies have provided evidence that a desensitization of soluble guanylate cyclase occurs after pretreatment with GTN (Axelsson and Andersson, 1983) or as a consequence of endogenous NO production (Moncada et al., 1991). As shown in Fig. 3, our results not only confirm the results of previous studies but also suggest that desensitization of soluble guanylate cyclase might be the result of a direct interaction between NO and the enzyme (Schulz and Böhme, 1984). It seems conceivable to suggest nitrosation and transnitrosylation reactions as underlying mechanisms (Barnett et al., 1994). Soluble guanylate cyclase is known to contain free sulfhydryl groups that are essential for activation of the enzyme. This is consistent with the control of its activity in mammalian cells by redox mechanisms (Goldberg and Haddox, 1977) and suggests that nitrosation of free sulfhydryl groups of the enzyme might occur.

Effects on the hemodynamic profile of organic nitrates. It has been shown that the production of NO by the vascular endothelium also suppresses the vasodilator activity of GTN and that desensitization of soluble guanylate cyclase within the arterial smooth muscle probably is involved (Alheid et al., 1987; Moncada et al., 1991; Kojda et al., 1994). In these studies, inhibitors of NO synthase or endothelial denudation increased the vasodilator potencies of organic nitrates such as GTN both in vitro and in vivo. Recently, we were able to demonstrate that disruption of the endothelial NO synthase gene in mice increases the relaxant potency of GTN in mouse aorta (Kojda et al., 1997). Thus, endogenous production of NO by endothelial NO synthase most likely decreases the vasodilator potency of GTN.

View larger version (28K): [in this window] [in a new window]   Fig. 7.   Suggested mechanism underlying the preferential venodilation elicited by organic nitrates. Prolonged treatment with exogenous NO in concentrations of <3 µM can diminish both vascular bioactivation of GTN and activation of soluble guanylate cyclase. Because production of endogenous NO in the vascular endothelium is substantially lower in veins than in arteries, it is likely that NO-induced inhibition of soluble guanylate cyclase (see Fig. 3) and of bioactivation of organic nitrates (see Fig. 4) also is lower in venous blood vessels. The inhibitory effect of NO on the enzymatic bioactivation seems to be more important for the preferential venodilation induced by organic nitrates than NO-induced desensitization of guanylate cyclase. The nitrovasodilator sodium nitroprusside, which undergoes a different bioactivation process but also acts via NO, does not show a venoselective action.

Mechanism and kinetics of NO release by spontaneous NO donors. In this study, spontaneous NO donors were used instead of NO. NO rapidly reacts with superoxide anions present in all oxygenated physiological buffers to form peroxynitrite, which is pharmacologically active (Beckman and Crow, 1993). The rapid generation of peroxynitrite from NO and superoxide (Goldstein and Czapski, 1995) implies that a delayed release of NO from NO donors prevents a rapid increase in the peroxynitrite concentration. Furthermore, a considerable portion of the NO donor molecules diffuse next to target cells before NO is released, which reduces the probability of oxidation of NO before induction of pharmacological actions such as vasodilation.