The mechanism by which nitroglycerin (NTG) lowers blood pressure involves metabolism in the vascular smooth muscle cells to nitric oxide by a process thought to involve thiol donors such as glutathione (GSH) or L-cysteine (L-cys) [1-3]. NO then stimulates the soluble guanylyl cyclase within smooth muscle cells causing an increase in cyclic guanosine monophosphate (cGMP) [4,5]. It appears that depressed formation of cGMP is involved in the mechanism of tolerance to NTG, since the attenuated relaxation of tolerant blood vessels is associated with a reduced increase in tissue cGMP and patients tolerant to NTG show decreased formation of cGMP in their platelets [6-8]. A major hypothesis for tolerance is that continued administration of NTG causes a reduction in vascular thiol containing compounds that are essential for its action. The hypothesis was introduced by Needleman and Johnson in 1973  based on experiments showing that chronic exposure of vascular tissue to NTG caused a reduction in the total thiol content. The finding that L-cys is essential for stimulation of soluble guanylyl cyclase by NTG, possibly through the formation of nitrosothiols, and the recent identification of the involvement of glutathione-S-transferase and GSH in the metabolism of NTG by smooth muscle cells, suggests that L-cys and GSH may be the key thiol donors depleted by chronic NTG administration [2,3,7,10]. This idea is principally supported by experiments showing that thiol donors such as N-acetyl-L-cysteine (NAC) can ameliorate NTG tolerance [11-13], although this is not always the case [14,15]. Only recently has this hypothesis been directly tested by measuring the content of these two thiol donors in vascular tissue. In isolated tolerant pig and cow coronary arteries and in blood vessels removed from NTG tolerant rats, the levels of L-cys and GSH were found to be essentially unchanged [16-18]. As pointed out by Boesgaard et al. , one of the limitations of measurement of thiols in vascular tissue is that such measurements do not represent levels in a specific cell, e.g., the vascular smooth muscle cell. Previously, we showed that chronic treatment of cultured vascular smooth muscle cells with NTG results in tolerance in the form of reduced formation of cGMP when the cells were rechallenged with acute administration of NTG [19,20]. This cultured cell model provided the opportunity to test the thiol depletion hypothesis directly in the target smooth muscle cell.
cGMP, GSH, L-cys, NAC, L-buthionine-(S,R)-sulfoximine (BSO), 3-isobutyl-1-methylxanthine (IBMX), HEPES, and all other chemicals were purchased from Sigma Chemical Co. (St. Louis, MO). NTG (0.4 mg Nitrostat Registered Trademark tablets) was from Parke-Davis, Division of Warner-Lambert Co. (Morris Plains, NJ). Cell culture plastic ware was from Costar (Cambridge, MA). Culture media and antibiotics were from ICN Biomedicals, Inc. (Costa Mesa, CA).
Smooth muscle cells were cultured from pig coronary arteries by a procedure previously described [19,20]. Briefly, coronary arteries were dissected free and stored in Dulbecco's modified Eagle's medium (DMEM) plus penicillin 100 IU/mL, streptomycin 100 micro gram/mL, and fungizone 0.25 micro gram/mL at 4 degrees C overnight. Explants (2 times 2 mm) were placed in a culture dish and covered with DMEM plus 20% fetal bovine serum with antibiotics. After 5-7 days the tissue was removed and cells that had exited were grown to confluence, passaged using 0.25% trypsin, and grown in DMEM plus 10% fetal bovine serum containing antibiotics in an incubator with 5% CO2 in room air at 37 degrees C. Each isolation was stained with a monoclonal antibody to smooth muscle alpha-actin with a kit (SIH 903-A) obtained from Sigma Chemical Co. Experiments were performed on confluent cells in 48-well culture dishes at approximately 30-60 micro gram of protein per well at passages 2 through 6.
Tolerance to NTG was induced by exposing the cells to 10 or 100 micro Meter NTG added to an incubation buffer for 60 min. The medium was removed and replaced with 0.5 mL/well of HEPES-buffered Earle's salts containing 0.05% bovine serum albumin (in mM, HEPES 25, NaCl 116, KCl 5.4, NaH2 PO4 0.89, MgSO4 0.81, CaCl (2) 1.8, and glucose 5.5), pH 7.3 at 37 degrees C. Solutions of NTG were prepared by dissolving Nitrostat Registered Trademark tablets in HEPES-buffered Earle's salts.
In order to determine recovery from tolerance, after the 60-min treatment with NTG to induce tolerance, the drug was washed out. The response of intracellular cGMP to an acute 2-min rechallenge with 200 micro Meter NTG was determined immediately after washout (0 time), and then again at 6, 12, and 24 h. The 2-min time point for the acute rechallenge with NTG was selected based on previous studies of the time-course of the cGMP response to NTG . Two minutes was the time of maximum response.
Cells were treated with thiol donors as follows: 1 mM GSH was added to the HEPES-buffered Earle's salts in 24 wells of a 48-well culture dish while the remaining 24 wells served as a time-matched untreated control. Supplemental GSH was present during induction of NTG tolerance and during the recovery period. Similar experiments were conducted with the addition of 1 mM L-cys and 1 mM NAC.
GSH synthesis was inhibited as follows: In 24 wells of a 48-well culture dish, BSO, a specific inhibitor of gamma-glutamylcysteine synthetase , was added in concentrations ranging from 0.01 to 0.5 mM for 18 h. The remaining 24 wells served as controls.
Intracellular cGMP was determined as follows: For basal levels, the culture medium was removed and intracellular cGMP was extracted by adding 200 micro Liter/well of 0.1 N HCl for 20 min at 4 degrees C. To determine the acute response of intracellular cGMP, cells were washed three times and then preincubated for 10 min at 37 degrees C with 0.5 mL/well HEPES-buffered Earle's salts plus 0.5 mM 3-isobutyl-l-methylxanthine. The cells were washed three times with buffer and incubated for an additional 2 min with 200 micro Meter of NTG. The reaction was terminated by removal of the buffer and the addition of 300 micro Liter/well of ice-cold 0.1 N HCl for 20 min at 4 degrees C to extract cGMP. cGMP was measured by radioimmunoassay as described previously [19,20].
GSH and L-cys were extracted and measured in a 20-micro Liter aliquot of HCl by high-pressure liquid chromatography  with a Waters Nova Pak C18 column and a Bioanalytical Systems LC-4 ampereometric detector using a Au/Hg amalgam electrode. The mobile phase consisted of a buffer containing 50 mM potassium phosphate and 5 mM 1-heptanesulfonic acid adjusted to pH 2.4 with phosphoric acid before the addition of 10% methanol. Standard curves were constructed with authentic GSH and L-cys for determination of the content of sample aliquots.
Cell protein was measured using the method of Bradford  by a Bio-Rad Protein Assay Kit (Bio-Rad Labs, Richmond, CA).
Time-course for the recovery from NTG tolerance was analyzed by two-way analysis of variance with repeated measurements. Student's t-test was used for comparisons of basal levels of cGMP, GSH, and L-cys between treated and untreated cells. All data are expressed as mean +/- SEM. A level of P < 0.05 was considered to be significant.
Confluent cells grew in a "hill-and-valley" pattern and displayed dark filaments when stained with the monoclonal antibody to smooth muscle alpha-actin. These results indicate that the cells were of smooth muscle origin. A photomicrograph has been published previously .
Initially we determined whether tolerance was associated with a reduction in intracellular thiol donors. Tolerance was induced by exposing cells to 100 micro Meter NTG for 60 min. This treatment caused an 83% reduction in the response of cGMP to rechallenge with 200 micro Meter NTG applied for 2 min. In control cells, basal intracellular cGMP was 2.6 +/- 0.2 pmol/mg cell protein and increased to 33.1 +/- 3.6 in response to 200 micro Meter NTG for 2 min as compared to an increase to 5.6 +/- 0.4 from a basal level of 2.3 +/- 0.1 in cells made tolerant to NTG (P < 0.01 vs control cells).
Intracellular GSH and L-cys levels in the control cells were 24.5 +/- 4.3 and 0.7 +/- 0.2 nmol/mg cell protein, respectively. In tolerant cells, GSH and L-cys levels were 27.4 +/- 4.8 and 1.4 +/- 0.4 nmol/mg cell protein, respectively, and were not significantly different from control cells. (Note: We cannot be certain of the precise cellular location of GSH and L-cys measured in these experiments. However, extensive washing of the cells did not further alter the values reported. We therefore assume that the values reported represent intracellular levels.)
(Figure 1) shows the effect of supplemental GSH on the development of and recovery from NTG tolerance. The left panel shows basal levels of cGMP and the response of cGMP to the acute administration of 200 micro Meter NTG in the presence and absence of 1 mM GSH. The right panel shows GSH in these cells. Basal levels of cGMP were 2.6 +/- 0.2 pmol/mg cell protein in untreated cells and 2.3 +/- 0.1 in cells supplemented with GSH. GSH did not affect the response of nontolerant cells to the acute administration of 200 micro Meter NTG. Exposure of the cells to 100 micro Meter NTG for 60 min caused tolerance in the response of cGMP when rechallenged with an acute 2 min application of 200 micro Meter NTG. Immediately after washout of NTG treatment (0 h in Figure 1) the response of intracellular cGMP was only twofold in tolerant cells in the presence and in the absence of supplemental GSH. In tolerant cells, cGMP increased from 2.6 +/- 0.2 to 5.6 +/- 0.4 pmol/mg cell protein without supplemental GSH (P < 0.01) and from 2.3 +/- 0.1 to 4.9 +/- 0.5 pmol/mg cell protein with 1 mM GSH supplementation (P < 0.01). This increase in cGMP was not statistically different between control cells and those supplemented with GSH. Furthermore, supplementation with GSH throughout the recovery period did not alter the time-course of recovery. For instance, at 24 h after washout of NTG the response of intracellular cGMP to rechallenge was sevenfold with or without GSH (left panel, Figure 1).
In cells supplemented with 1 mM GSH, levels of GSH were increased threefold to 68.8 +/- 12.4 nmol/mg cell protein as compared to 24.5 +/- 4.3 in nonsupplemented cells (P < 0.01, right panel, Figure 1). As can also be seen in Figure 1, increased intracellular levels of GSH were maintained during the induction of NTG tolerance and throughout the recovery period. Not shown in Figure 1, GSH supplementation also increased intracellular L-cys from 0.7 +/- 0.2 to 6.9 +/- 0.2 nmol/mg cell protein (P < 0.01).
In the next series of experiments, NTG tolerance and recovery was examined in the presence and absence of supplementation with 1 mM L-cys. In these experiments, 10 micro Meter NTG for 60 min was used to produce a lesser degree of tolerance than that associated with the 100 micro Meter NTG used in the above-mentioned experiments with supplemental GSH. This was done to determine whether supplementation with a thiol donor could affect less pronounced tolerance than that produced by 100 micro Meter NTG. Seen in the left panel of Figure 2, such treatment caused a 54% reduction in the intracellular cGMP response to rechallenge with 200 micro Meter NTG for 2 min (compare bars labeled CON with bars labeled 0 h after NTG pretreatment) as compared to the 83% reduction with 100 micro Meter NTG treatment (see Figure 1). Also seen in the left panel of Figure 2, supplementation with 1 mM L-cys did not alter the cGMP response in tolerant cells either immediately after washout of NTG or during the recovery period even though the level of intracellular L-cys increased from 1.3 +/- 0.2 to 30.3 +/- 4.9 nmol/mg cell protein (P < 0.01) and remained increased through the recovery period (right panel). Not shown in Figure 2, supplementation with L-cys also increased intracellular GSH from 25.3 +/- 4.0 to 41.7 +/- 3.0 (P < 0.05) which remained at this level throughout recovery. Additionally, experiments conducted with 1 mM NAC added to the incubation buffer did not alter the cGMP response in control cells or alter the induction of or recovery from NTG tolerance. In control cells, 200 micro Meter NTG for 2 min increased cGMP from a basal level of 2.4 +/- 0.2 pmol/mg protein to 24.6 +/- 1.1 (P < 0.01). In cells supplemented with 1 mM NAC, cGMP increased to 27.6 +/- 0.9 pmol/mg protein from a basal level of 2.8 +/- 0.2 (P < 0.01) and was not significantly different from control cells. After 60 min treatment with 10 micro Meter NTG to induce tolerance, 200 micro Meter NTG for 2 min increased cGMP to 11.2 +/- 0.7 pmol/mg protein in control cells and to 12.6 +/- 0.2 in cells supplemented with NAC. These values were not significantly different. At 24 h after the induction of tolerance, there was no difference in the response of cGMP to 200 micro Meter NTG between control cells and those supplemented with NAC (17.4 +/- 0.3 pmol/mg protein vs 19.1 +/- 1.3, respectively).
(Figure 3) shows the effect of lowering intracellular GSH on the response of cGMP to NTG. Exposing cells to increasing concentrations of BSO for 18 h decreased intracellular GSH but had no effect on the response of cGMP to NTG. For instance in cells treated with 0.5 mM BSO, intracellular GSH content was reduced by 77% from 32.5 +/- 1.8 nmol/mg cell protein in untreated cells to 7.4 +/- 0.4 in treated cells (P < 0.01). This reduction in intracellular GSH was not associated with a reduced response of cGMP to NTG. In control cells 200 micro Meter NTG increased cGMP from 0.8 +/- 0.1 pmol/mg cell protein basal level to 21.8 +/- 0.6 (P < 0.01) and to 22.3 +/- 1.0 in cells treated with 0.5 mM BSO (P < 0.01 vs basal).
To our knowledge, this is the first report of a direct test in vascular smooth muscle cells of the thiol donor depletion hypothesis for NTG tolerance. The major findings are: 1) treatment with NTG that produces tolerance in the form of reduced response of cGMP was not associated with a reduction in intracellular levels GSH or L-cys; 2) increasing intracellular levels of GSH threefold and L-cys by 23-fold did not prevent or reduce the degree of NTG tolerance nor alter the recovery from tolerance; 3) supplementation of the incubation buffer with 1 mM NAC did not alter tolerance or recovery, and; 4) lowering intracellular levels of GSH by 77% did not alter the response of cGMP to NTG. The model that we used here--cGMP responsiveness in cultured pig coronary smooth muscle cells--was characterized in previous studies [24,25]. These studies showed that tolerance was dependent on the time of exposure as well as the concentration of NTG, and that the concentration-response curve of cGMP in tolerant cells was almost flat even when rechallenged with 1 mM NTG. Tolerance was selective for nitrates, in that the response of intracellular cGMP to atrial natriuretic peptide was not reduced and the response of intracellular cyclic adenosine monophosphate to isoproterenol and forskolin was not affected. Additionally, complete recovery from NTG tolerance occurred in 72 h. Thus, these cultured smooth muscle cells exhibit many of the characteristics of NTG tolerance seen in patients, intact animals, and isolated vascular tissue and appear to provide a reasonable model for studying NTG tolerance at the cellular level.
Our results at the cellular level are in agreement with those of three other recent studies in which NTG tolerance was induced and levels of thiol donors in whole vascular tissue were actually measured. Gruetter and Lemke  induced tolerance to NTG in bovine coronary rings. In nontolerant tissue, 1 micro Meter NTG caused a 83% relaxation in K+ precontracted rings and a 10-fold increase in the tissue level of cGMP. These effects were almost completely abolished in tolerant tissue. In tolerant tissue, levels of L-cys were not reduced; however, GSH decreased significantly by 30%. Supplementation with 1 mM L-cys increased the level of L-cys in the coronary rings approximately 20-fold, a value in close agreement with our result, but had no effect on tolerance measured either as relaxation or increase in tissue cGMP. Similar findings were reported in isolated pig coronary segments where tissue levels of GSH were not reduced in segments made tolerant to relaxation by NTG . Additionally, treatment of these pig coronary arteries with diethyl maleate reduced tissue GSH levels by 95% but had no effect on concentration-relaxation curves to NTG . Recently, Boesgaard et al.  conducted the first in vivo study in rats made tolerant to NTG. In these tolerant animals, the blood pressure response to a bolus injection of NTG was reduced but levels of L-cys and GSH in the aorta and the vena cava were not reduced from control. Also, animals tolerant to NTG were treated with an infusion of NAC. This infusion potentiated the blood pressure response but did not return the response to normal, even though tissue levels of L-cys in aorta and vena cava increased approximately threefold. Taken collectively with these previous studies in isolated vascular tissue and intact animals, our findings do not support the hypothesis implicating reduced levels of intracellular GSH or L-cys as a mechanism for NTG tolerance. Further, the results from our study seem to eliminate the argument that, although levels of GSH and L-cys in vascular tissue are unaffected by NTG tolerance, these thiol donors may still be reduced in the critical pool of the vascular smooth muscle cell. It should be emphasized that our results are specific for GSH and L-cys and do not eliminate the possibility that some other unknown thiol donor plays a crucial role. On the other hand, intracellular levels of thiol donors may not be involved at all. NTG tolerance could result from a direct desensitization of the soluble guanylyl cyclase as has been reported recently . Additionally, alterations in circulatory compensatory mechanisms associated with NTG infusion, i.e., increased plasma volume, increased plasma renin activity and reduced plasma atrial natriuretic peptide, have been suggested to account for tolerance . Finally, it is possible that the explanation for the efficacy of NAC administration reported in some cases of NTG tolerance is due to an extracellular pathway for the conversion of NTG to an active agent in the presence of high concentrations of thiols  and has nothing to do with replenishing depleted stores of intracellular thiol donors.
The authors wish to acknowledge Mr. and Mrs. Z. Fordham for their cooperation and generous donation of porcine hearts. We appreciate the technical assistance of Jennifer Schulte.
1. Fung H-L, Chung S-J, Bauer JA, et al. Biochemical mechanism of organic nitrate action. Am J Cardiol 1992;70:4B-10B.
2. Hill KE, Hunt RW Jr, Jones R, et al. Metabolism of nitroglycerin by smooth muscle cells: involvement of glutathione and glutathione S-transferase. Biochem Pharmacol 1992;43:561-6.
3. Ignarro LJ, Lippton H, Edwards JC, et al. Mechanism of vascular smooth muscle relaxation by organic nitrates, nitrites, nitroprusside, and nitric oxide: evidence for the involvement of S-nitrosothiols as active intermediates. J Pharmacol Exp Ther 1981;218:739-49.
4. Wong SK-F, Garbers DL. Receptor guanylyl cyclases. J Clin Invest 1992;90:299-305.
5. Yuen PST, Garbers DL. Guanylyl cyclase linked receptors. Annu Rev Neurosci 1992;15:193-225.
6. Kieth RA, Burkman AM, Sokoloski TD, Fertel RH. Vascular tolerance to nitroglycerin and cyclic GMP generation in rat aortic smooth muscle. J Pharmacol Exp Ther 1982;221:525-31.
7. Kukovetz WR, Holzmann S. Mechanisms of nitrate-induced vasodilation and tolerance. Eur J Clin Pharmacol 1990;38:S9-14.
8. Watanabe H, Kakihana M, Ohtsuka S, et al. Platelet cGMP: a potentially useful indicator to evaluate the effects of nitroglycerin and nitrate tolerance. Circulation 1993;88:29-36.
9. Needleman P, Johnson EM Jr. Mechanism of tolerance development to organic nitrates. J Pharmacol Exp Ther 1973;184:709-15.
10. Torresi J, Horowitz JD, Dusting GJ. Prevention and reversal of tolerance to nitroglycerine with N-acetylcysteine. J Cardiovasc Pharmacol 1985;7:777-83.
11. May DC, Popma JJ, Black WH, et al. In vivo induction and reversal of nitroglycerin tolerance in human coronary arteries. N Engl J Med 1987;317:805-9.
12. Newman CM, Warren JB, Taylor GW, et al. Rapid tolerance to the hypotensive effects of glyceryl trinitrate in the rat: prevention by N-acetyl-L-cysteine but not N-acetyl-D-cysteine. Br J Pharmacol 1990;99:825-9.
13. Boesgaard S, Petersen JS, Aldershvile J, et al. Nitrate tolerance: effect of thiol supplementation during prolonged nitroglycerin infusion in an in vivo rat model. J Pharmacol Exp Ther 1991;258:851-6.
14. Parker JO, Farrel B, Lahey KA, Rose BF. Nitrate tolerance: the lack of effect of N-acetylcysteine. Circulation 1987;76:572-82.
15. Dupuis J, Lalonde G, Lemieux R, Rouleau JL. Tolerance to intravenous nitroglycerin in patients with congestive heart failure: role of increased intravascular volume, neurohumoral activation and lack of prevention with N-acetylcysteine. J Am Coll Cardiol 1990;16:923-931.
16. Gruetter CA, Lemke SM. Dissociation of cysteine and glutathione levels from nitroglycerin-induced relaxation. Eur J Pharmacol 1985;111:85-95.
17. Sakanashi M, Matsuzaki T, Aniya Y. Nitroglycerin relaxes coronary artery of the pig with no change in glutathione content or glutathione S-transferase activity. Br J Pharmacol 1993;103:1905-8.
18. Boesgaard S, Aldershvile J, Poulsen HE, et al. Nitrate tolerance is not associated with the depletion of arterial or venous thiol levels. Circ Res 1994;74:115-20.
19. Zhang LM, Castresana MR, Stefansson S, Newman WH. Tolerance to sodium nitroprusside: studies in porcine vascular smooth muscle cells. Anesthesiology 1993;79:1094-1103.
20. Zhang LM, Castresana MR, Newman WH. Tolerance to nitroglycerin in vascular smooth muscle cells: recovery and cross-tolerance to sodium nitroprusside. Anesth Analg 1994;78:1053-9.
21. Griffith OW, Meister A. Potent and specific inhibition of glutathione synthesis by buthionine sulfoximine (S-N-butyl homocysteine sulfoximine). J Biol Chem 1979;254:7558-60.
22. Grossman SJ, Simson J, Jollow DJ. Dapsone-induced hemolytic anemia: effect of N-hydroxy dapsone on the sulfhydryl status and membrane proteins of rat erythrocytes. Toxicol Appl Pharmacol 1992;117:208-17.
23. Bradford MM. A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Anal Biochem 1976;72:248-54.
24. Romanin C, Kukovetz WR. Tolerance to nitroglycerin is caused by reduced guanylate cyclase activation. J Mol Cell Cardiol 1989;21:41-8.
25. Fung H-L, Chong S, Kowaluk E, et al. Mechanism for the pharmacologic interaction of organic nitrates with thiols. Existence of an extracellular pathway for the reversal of nitrate vascular tolerance by N-acetylcysteine. J Pharmacol Exp Ther 1988;245:524-30.