Sevoflurane (SEV) has been reported to generate oxygen-derived free radicals, which can be one of the major causes for the impairment of nitric oxide (NO)-induced relaxation of vascular smooth muscle (1). It has also been reported that SEV reduced endothelium-dependent vasorelaxation by generating superoxide anion (2). In addition, there are reports that superoxide anion is involved in the breakdown pathway of NO (3,4), whereas other reports indicated that oxygen radicals affected soluble guanylate cyclase (sGC) (5,6).
Nitroglycerin (TNG) tolerance, a phenomenon of an impaired vasodilation response to TNG developing during long-term administration, has been reported to be associated with increased vascular superoxide anion production (7,8). This may be one of the mechanisms that mediate TNG tolerance. However, data concerning the mechanisms responsible for the development of TNG tolerance are conflicting, and it has been attributed to sulfhydryl (SH) depletion (9,10), guanylate cyclase desensitization (11,12), or both (13,14). It remains unknown, however, whether and how superoxide anion could affect TNG-induced vasorelaxation pathways.
If both TNG and SEV can generate oxygen radicals, SEV might enhance the development of TNG tolerance through the production of oxygen-derived free radicals during their simultaneous exposure. Isoflurane (ISO), another volatile anesthetic frequently used in clinical anesthesia, has not been reported to generate reactive oxygen species, despite its inhibitory effects on NO-mediated vasorelaxation. The first goal of the current study was to elucidate the mechanisms that mediate TNG tolerance in isolated rat aorta. The second was to investigate and compare the effects of two volatile anesthetics, SEV, a possible reactive oxygen species generating drug, and ISO, with no evidence of reactive oxygen species production, on the development of TNG tolerance.
All experimental procedures and protocols were approved by the Animal Use Committee of Wakayama Medical University.
Preparation of Rat Aortic Rings
Thirty male Wistar rats weighting 250–350 g were anesthetized with intraperitoneal injection of sodium pentobarbital (50 mg/kg) and killed by bleeding from carotid arteries. The chest was rapidly opened, and then the descending thoracic aorta was dissected and immersed in Krebs bicarbonate buffer solution of the following composition: NaCl 118.2; KCl 4.6; CaCl2 2.5; KH2PO4 1.2; MgSO4 1.2; NaHCO3 24.8; and glucose 10.0 (mM). The preparations were cleaned of excessive connective tissue and cut into ring segments 3 mm in length. The endothelium was denuded by inserting the closed tips of a metal hemostat into the ring segment and by rolling it gently on moistened filter paper. Removal of intact endothelium was later verified by the absence of the relaxation response to acetylcholine (10−6 M).
Isometric Tension Experiments
The ring segments were mounted vertically between stainless hooks in organ chambers (10 mL capacity) filled with Krebs bicarbonate solution which was maintained at 37°C ± 0.5°C and aerated continuously with a gas mixture of 5% CO2 and 95% O2. The pH of the solution was kept at 7.40 ± 0.05. One of the hooks was anchored and the other was connected to a force transducer (Nihon Kohden San-ei Instrument CO, Tokyo, Japan). Changes in isometric force were recorded on a chart recorder (Nihon Kohden San-ei Instrument CO). Optimal resting tension was adjusted to 3.0 g, which was the amount of stretch necessary to achieve the maximum contractile response to 30 mM KCl, as determined in preliminary experiments. All rings were pretreated with indomethacin (10−5 M), an inhibitor of cyclooxygenase, to prevent the generation of vasoactive prostanoid metabolites. Aortic rings were first contracted with 30 mM KCl to validate the vascular smooth muscle contractile function and were then washed 3 times with fresh bathing media every 15 min. SEV and ISO were introduced into the gas mixture through an agent-specific vaporizer (Sevotec 3, Fortec 3; Ohmeda, Steeton, UK) to provide the appropriate concentrations in the gas mixture. Their concentrations in the aerating gas mixture were monitored and adjusted by using an anesthetic monitor (model 303; Atom CO, Tokyo, Japan), which was calibrated using a standard calibrating gas mixture (Atom CO). The concentrations of anesthetics in the bathing solution were measured by gas chromatography as described previously (15). The concentrations of SEV at 1.7%, 3.4%, and 5.1% in bathing solution were 0.17 ± 0.03 mM, 0.28 ± 0.02 mM, and 0.39 ± 0.03 mM, respectively (n = 3). Those of ISO at 1.1%, 2.2%, and 3.3% were 0.17 ± 0.01 mM, 0.25 ± 0.03 mM, and 0.33 ± 0.01 mM, respectively (n = 3).
In isometric tension recording experiments, the rings were randomly assigned into the following protocols and only one protocol was performed in each aortic ring.
Study 1: Tolerance to TNG in Rat Aorta.
The aortic rings were partially contracted with phenylephrine (PE, 3 × 10−7 M); the resulting contractions were ranged between 50% and 70% of those induced by 30 mM KCl. After a steady-state tension had been reached, concentration-response curves for TNG (10−8 –10−5 M) were obtained. The rings were then incubated in bathing media containing TNG (10−5 M) for 30 min to induce TNG tolerance. The rings were then washed with fresh bathing media at least 3 times every 10 min before the second response to TNG was determined. After complete washout of TNG and PE from the organ bath, the rings were again precontracted with PE. The second relaxant response to TNG was obtained and compared with the first response. TNG tolerance was confirmed by the reduced second response to TNG after 30 min exposure to TNG. To eliminate the possibility that the development of TNG tolerance may have been a result of the mere change in reactivity of the preparations over time, some rings were incubated for 30 min in the absence of TNG as a time control study.
Study 2: Modification of the Tolerance to TNG by Anesthetics.
In the next series of experiments, we investigated the effect of SEV and ISO on the development of TNG tolerance. After the first responses to TNG, the aortic rings were exposed to TNG for 30 min in the presence of SEV (1.7%, 3.4%, and 5.1%) or ISO (1.1%, 2.2%, and 3.3%). The second response to TNG was investigated after repeated washout of TNG and anesthetics from the organ chamber. As a result, we found SEV, but not ISO, had a significant effect on the development of TNG tolerance. Therefore, we used only SEV in the following experiments.
Study 3: The Effect of -SH Supplements on SEV-Induced Enhancement of the Tolerance to TNG.
To assess whether SH supplements affect TNG tolerance, the rings were pretreated with N-acetyl-l-cysteine (10−5 M) or reduced glutathione (10−5 M) 10 min before the first response to TNG. These drugs remained in the bathing media during the simultaneous exposure to TNG and SEV for 30 min.
Study 4: The Effect of Antioxidants on SEV-Induced Enhancement of the Tolerance to TNG.
To characterize the contribution of reactive oxygen intermediates to the development of TNG tolerance, the rings were pretreated with one of the following antioxidants; deferoxamine (10−5 M), catalase (1,200 U/mL), conventional superoxide dismutase (SOD: 150 U/mL), or lecithinized SOD (PC-SOD: 120 U/mL). PC-SOD has a higher lipophilicity and greater cell membrane permeability compared with conventional SOD.
Study 5: The Contribution of Oxygen Tension of the Bathing Solution to the SEV-Induced Enhancement of the Tolerance to TNG.
We examined the influence of oxygen tension in the bathing media on the development of TNG tolerance in the presence of SEV. Either hyperoxic gas mixture (95% O2 and 5% CO2) or normoxic gas mixture (25% O2, 5% CO2, and 70% N2) was introduced to the bathing solution at the beginning of the experiment and continuously aerated throughout the experimental protocol.
Study 6: Electron Spin Resonance Spectrometry.
Finally, to confirm where SEV generates reactive oxygen intermediates–inside or outside of the cell membrane of vascular smooth muscle–electron spin resonance spectrometry (ESR spectrometer, JWS-FE2XG;JOEL Ltd., Nihon Denshi, Tokyo, Japan) experiment was performed. Sixty min after adding spin trapping agents (DMPO:5,5-dimethyl-1-pyrroline N-oxide; 30 μL) to the bathing solution equilibrated with 95% O2 and 5% SEV, spectrum of radicals was measured and compared with MnO standard spectrum.
The following drugs were used: acetylcholine chloride (Dai-ichi Pharmaceutical CO, Tokyo, Japan), indomethacin (Nacalai Tesque, Inc., Kyoto, Japan), TNG (Nihon Kayaku, Tokyo), SEV (Maruishi Pharmaceutical CO, Osaka, Japan), and ISO (Dainabott, Osaka). PC-SOD were obtained from Asahi Glass CO, Ltd., Tokyo, Japan. PE, N-acetyl-l-cysteine, catalase, deferoxamine, and conventional SOD were obtained from Sigma Chemical (St. Louis, MO). All drugs, except indomethacin, SEV, and ISO, were prepared and diluted in distilled water. Indomethacin was first dissolved in 99.5% ethanol (Katayama Chemical CO, Osaka Drug, Japan), and the final concentration of ethanol added to the organ bath was <0.1% vol/vol, which did not affect the isometric tension of the aortic rings. Drug concentrations are expressed as the final concentrations in the organ bath.
All results shown in the text and figures are expressed as mean value ± sem. Relaxations were expressed as the percentages relative to the precontraction induced by PE (3 × 10−7 M). The logarithm of the TNG concentration producing 50% of the maximum relaxant response (EC50) was calculated by linear regression analysis, and the maximum relaxant response (Rmax) was determined. Rmax = 100% indicates complete reversal of PE precontraction. Statistical analyses were performed by Student’s t-test and Scheffé’s F-test after one-way analysis of variance, where appropriate. Differences at P < 0.05 were considered statistically significant. In isometric tension experiments, sample size (n values) equals the number of rats from which aortic rings were obtained.
Tolerance to TNG in Rat Aorta
Concentration-response curves of TNG-induced vasorelaxation are shown in Figure 1. TNG elicited a concentration-dependent relaxation of rat aortic rings. The maximum relaxation was obtained at a concentration of 10−5M TNG (first response). After exposure to TNG (10−5M) for 30 min, the relaxant response to TNG was markedly attenuated (second response), although the contractile response to PE was not affected. Rmax was significantly reduced from 97.9% ± 1.1% to 82.3% ± 1.5% (P < 0.01, n = 7). Exposure to TNG also caused a rightward shift of the concentration-relaxation curve for TNG. Logarithm of EC50 values in the first and the second responses were -7.38 ± 0.02 and -6.92 ± 0.07, respectively (P < 0.01, n = 7 each). The second response to TNG in the absence of a continuous application of TNG for 30 min was identical to the first response. Logarithm of EC50 values in the first and second responses were –7.40 ± 0.02 and -7.30 ± 0.05, respectively, and Rmax values in the first and second responses were 98.9% ± 1.1% and 97.5% ± 1.5%, respectively (n = 7).
Modification of the Tolerance to TNG by Anesthetics
Either SEV or ISO was administered at 1, 2, and 3 MAC (SEV: 1.7%, 3.4%, and 5.1%, ISO: 1.1%, 2.2%, and 3.3%, respectively) during the incubation with TNG (10−5 M) for 30 min between the first and second responses. In control experiments, aortic rings were not exposed to anesthetics. The second response to TNG of aortic rings previously exposed to SEV at a concentration of 3 MAC was significantly attenuated when compared with that of control group, indicating that SEV enhanced the development of the TNG tolerance (Fig. 2a). Although exposure of SEV at concentrations of 1 MAC (1.7%) and 2 MAC (3.4%) appeared to enhance TNG tolerance in a concentration-dependent manner, this effect did not reach statistical significance (n = 6, data were not shown). By contrast, exposure of ISO up to 3 MAC did not affect the relaxant response to TNG (n = 6) (Fig. 2b).
To test whether SEV alone induces TNG tolerance, the arterial rings were exposed to SEV at a concentration of 3 MAC for 30 min in the absence of TNG. SEV alone did not affect the second relaxation response to TNG, indicating that SEV alone does not have an ability to induce TNG tolerance (data not shown, n = 4).
The Effect of -SH Supplements on SEV-Induced Enhancement of the Tolerance to TNG
Neither N-acetyl-l-cysteine (10−5 M) nor reduced glutathione (10−5 M) pretreatment affected the responsiveness of aortic rings to a second application of TNG after simultaneous exposure to TNG and SEV (Fig. 3).
The Effect of Antioxidants on SEV-Induced Enhancement of the Tolerance to TNG
Pretreatment with PC-SOD (120 U/mL), deferoxamine (10−5 M), and catalase (1200 U/mL) did significantly restore the hyporesponsiveness to TNG in rings pre-exposed with TNG and SEV (Fig. 4). PC-SOD also completely restored the impaired relaxation response to TNG in rings pre-exposed with TNG alone (data not shown, n = 4). By contrast, conventional SOD (150 U/mL) had no effect on the impaired response to TNG.
The Contribution of Oxygen Tension of the Bathing Solution to the SEV-Induced Enhancement of the Tolerance to TNG
The Po2 of the bathing solution aerated with hyperoxic (95% O2 and 5% CO2) and normoxic (25% O2, 5% CO2, and 70% N2) gas mixture was 538.7 ± 37.2 mm Hg and 170.3 ± 2.4 mm Hg, respectively (n = 6 each). The pH, Pco2, and HCO3 values in the bathing solution aerated with normoxic gas mixture were identical to those with hyperoxic gas mixture. The effects of oxygen tension of the bathing solution on the second response to TNG are summarized in Figure 5. There were no differences in the contractile response to PE and the first relaxation response to TNG between the normoxic and hyperoxic groups. The second response to TNG of vascular rings in bathing solution aerated with normoxic gas mixture was not impaired even after 30-min exposure to TNG and SEV, suggesting that reduced oxygen tension prevents acquisition of the TNG tolerance of aortic rings.
No significant spin-trapped spectrum was detected in the solution aerated with hyperoxic gas mixture in the presence of both TNG (10−5M) and SEV (5%) for 60 min without vascular tissue.
The main findings of our study are as follows. Aortic rings exposed to a large concentration of TNG showed a decreased relaxation response to TNG. In the presence of a large concentration of SEV during TNG exposure, the hyporesponsiveness of the rings to TNG was enhanced. ISO had no effect on the reduced response to TNG. Pretreatment of the rings with SH supplements (-SH donor) failed to affect the TNG response. Antioxidants such as catalase, deferoxamine, and PC-SOD, but not conventional SOD, restored the reduced response to TNG. Furthermore, reduced oxygen tension in the bathing solution almost restored the impaired vasorelaxation response to TNG. ESR experiments demonstrated no significant spin-trapped spectrum in the bathing solution.
TNG tolerance is an impaired vasorelaxation response to TNG developing during long-term administration. TNG tolerance was experimentally induced by exposure to TNG (10−5 M) for 30 minutes in endothelium-denuded rat aortic rings (16), consistent with our findings. We found in a preliminary study that longer exposure to TNG suppressed the second contractile response to PE too strongly to allow for the assessment of TNG-induced relaxation response even after repeated washing with fresh bathing fluid and that smaller concentrations of TNG and/or <30 minutes of exposure did not induce TNG tolerance. Furthermore, time-control studies indicate that the reduced second response to TNG in rings pre-exposed with TNG cannot be attributed to the merely functional change in the aortic rings within the time course. Lawson et al. (16) reported that TNG tolerance developed in endothelium-intact rat aortic rings incubated with TNG (10−6 M) for 10 min. Other investigators have also demonstrated that the presence of intact endothelium markedly impaired sensitivity to TNG in TNG-pretreated vessels, including rabbit aortas and pig coronary arteries (7,17). In the present study, endothelium-denuded rat aortic rings were used in order to exclude the effects of volatile anesthetics on NO released from endothelium, as anesthetics can decrease the synthesis or the action of the endothelium-derived NO. We aimed to examine the effect of anesthetics on the TNG tolerance of the vascular smooth muscle devoid of the influences of endothelium and to elucidate its mechanisms.
There have been numerous reports, with conflicting data, concerning the mechanism of TNG tolerance. SH depletion (9,10), sGC desensitization (11,12), or both (13,14) have been suggested as possible mechanisms. In our present study, SH supplements, such as N-acetyl-l-cysteine and reduced glutathion, had little effect on the development of TNG tolerance in rat aorta. It has been reported that TNG preferentially dilates the larger coronary arteries (17) and that N-acetyl-l-cysteine can improve the sensitivity to TNG, mainly in small coronary arteries (18). The mechanisms contributing to TNG tolerance may depend on the type and size of vessels. The lack of SH compounds is not involved in the SEV-induced enhancement of TNG tolerance, at least in rat aorta. Lawson et al. (16) showed that previous exposure of endothelium-intact vascular rings to acetylcholine (10−6 M) for 10 minutes resulted in the development of TNG tolerance. Waldman et al. (11) reported that sGC purified from TNG-pretreated rat aorta exhibited persistent desensitization to nitrate-induced activation, indicating the desensitization of sGC as the possible mechanism that mediates TNG tolerance.
It has been postulated that TNG tolerance is associated with increased vascular superoxide anion production (7,8), suggesting the possibility that oxygen radicals, such as superoxide anion, might affect the TNG relaxation pathway by influencing the biotransformation of TNG to NO or the NO – sGC pathway. Subsequently, oxygen radicals can impair sensitivity to TNG. In the present study, antioxidants including PC-SOD, deferoxamine, and catalase significantly restored the responsiveness to the second application of TNG in rings pre-exposed to TNG and SEV. Furthermore, reduced oxygen tension in the bathing solution improved the decrease in sensitivity to TNG. It seems likely that the enhancement of TNG tolerance is mediated by the production of oxygen-derived radicals. Unlike PC-SOD, conventional SOD failed to improve the reduced response to TNG. PC-SOD has a higher lipophilicity and greater membrane permeability (19,20), whereas conventional SOD is less permeable, suggesting that oxygen radicals were generated within vascular smooth muscle cells rather than in the bathing fluid. This is further supported by the findings that no significant spin-trapped spectrum was detected in the bathing solution aerated with hyperoxic gas mixture in the ESR spectrometry experiments.
Volatile anesthetics, including SEV and ISO, are known to affect the NO – sGC – cyclic GMP pathway, resulting in impairment of endothelium-dependent vasorelaxation in isolated vessels (1,18). However, its mechanism seems different. Yoshida and Okabe (1) have reported in isolated rat aorta that SEV possesses the ability to generate oxygen-derived free radicals, which can inactivate endothelium-derived NO and subsequently impair the endothelium-dependent vasorelaxation. By contrast, generation of reactive oxygen species by ISO has never been reported. This property of ISO may account for the lack of enhancement of TNG tolerance. Although the exact mechanisms by which SEV enhances the TNG tolerance remain unknown, it might be related to oxygen free radical generation. SEV has an ability to generate oxygen-derived free radicals (1). In the present study, oxygen radical scavengers restored the impaired relaxation response to TNG in rings pre-exposed to TNG alone and in combination with SEV. These findings suggest that reactive oxygen radicals may be involved in the development of TNG tolerance or its enhancement induced by SEV, although the relative contribution of the scavengers to reactive oxygen species formed in TNG treatment alone or in combination with SEV is indistinguishable.
Reactive oxygen radicals could affect the NO – sGC – cGMP pathway by the inactivation of the NO molecule as well as by the desensitization of sGC activity (Fig. 6). TNG acts as a NO-donor and released NO reacts rapidly with superoxide anion to generate peroxynitrite, which is also a highly reactive radical that might affect the intracellular signaling pathway that mediates TNG-induced vasorelaxation. Because SEV existed only during exposure to TNG for 30 minutes and was completely removed from the bathing fluid until the start of the second response to TNG in the present study, it is likely that reactive oxygen radicals released during TNG and SEV exposure impaired the response to TNG by inducing the desensitization of sGC, rather than by inhibiting the action of NO released from TNG. Exposure to SEV alone for 30 minutes failed to develop the TNG tolerance in rat aorta, suggesting SEV alone does not have an ability to generate oxygen radicals enough to induce TNG tolerance. Based on the findings that only large concentrations of SEV (3 MAC) significantly enhanced TNG tolerance, SEV appears to act as a booster but not an initiator in the pathway. Further study would be required to clarify the exact mechanism underlying SEV-induced enhancement of TNG tolerance. Moreover, because the enhancement of TNG tolerance induced by SEV is significant only at supra-clinical concentrations in endothelium-denuded vessels, the results from the present study may not be directly extrapolated to the clinical situation.
In conclusion, SEV at a concentration of 3 MAC has an enhancing effect on the TNG tolerance under hyperoxic conditions. This effect of SEV may be mediated by an additive generation of oxygen-derived free radicals in vascular smooth muscle cells.
1. Yoshida K, Okabe E. Selective impairment of endothelium-dependent relaxation by sevoflurane. Anesthesiology 1992;76:440–7.
2. Maria MA, Luis MA, Antonio LF, et al. Sevoflurane reduces endothelium-dependent vasorelaxation: role of superocide anion and endothelin. Can J Anaesth 2002;49:471–6.
3. Gryglewski RJ, Palmer RMJ, Moncada S. Superoxide anion is involved in the breakdown of endothelium-derived vascular relaxing factor. Nature 1986;320:454–6.
4. Rubanyi GM, Vanhoutte PM. Superoxide anion and hyperoxia inactivate endothelium-derived relaxing factor. Am J Physiol 1986;250:H822–7.
5. Kontos HA, Wei EP. Hydroxyl radical-dependent inactivation of guanylate cyclase in cerebral arterioles by methylene blue and by LY83583. Stroke 1993;24:427–34.
6. Brune B, Schmidt KU, Ullrich V. Activation of soluble guanylate cyclase by carbon monoxide and inhibition by superoxide anion. Eur J Biochem 1990;192:683–8.
7. Münzel T, Sayegh H, Freeman BA, et al. Evidence for enhanced vascular superoxide anion production in nitrite tolerance. J Clin Invest 1995;95:187–94.
8. Münzel T, Kurz S, Rajagopalan S, et al. Hydralazine prevents nitroglycerin tolerance by inhibiting activation of membrane-bound NADH oxidase. J Clin Invest 1996;98:1465–70.
9. Brien JF, McLaughlin BE, Breedon TH, et al. Biotransformation of glyceryl trinitrate occurs concurrently with relaxation of rabbit aorta. J Pharmacol Exp Ther 1986;237:608–14.
10. Axelsson KL, Ahlner J. Nitrate tolerance from a biochemical point of view. Drugs 1987;33:63–8.
11. Waldman SA, Rapoport RM, Ginsburg R, Murad F. Desensitization to nitroglycerin in vascular smooth muscle from rat and human. Biochem Pharmacol 1986;35:3525–31.
12. Romanin CH, Kumovetz WR. Tolerance to nitroglycerin is caused by reduced guanylase cyclase activation. J Mol Cell Cardiol 1989;21:41–8.
13. Needleman P, Johnson EM. Mechanism of tolerance development to organic nitrates. J Pharmacol Exp Ther 1973;184:709–15.
14. Ignarro LJ, Kadowitz PJ. The pharmacological and physiological role of cyclic GMP in vascular smooth muscle relaxation. Pharmacol Toxicol 1985;25:171–91.
15. Unger, P, Berkenboom, G, Brékine D, Fontaine J. Nitrate tolerance and aging in isolated rat aorta. Eur J Pharmacol 1993;248:145–9.
16. Lawson DL, Haught WH, Mehta P, Mehta JL. Studies of vascular tolerance to nitroglycerin: Effects of N-acetylcysteine, NG-monomethyl l-arginine, and endothelin 1. J Cardiovasc Pharmacol 1996;28:418–24.
17. Murohara T, Kugiyama K, Yasue H. Interactions of nitrovasodilators, atrial natriuretic peptide and endothelium-derived nitric oxide. J Vasc Res 1996;33:78–85.
18. Nakamura K, Terasako K, Toda H, et al. Mechanisms of inhibition of endothelium-dependent relaxation by halothane, isoflurane, and sevoflurane. Can J Anaesth 1994;41:340–6.
19. Sellke FW, Myers PR, Bates JN, Harrison DG. Influence of vessel size on the sensitivity of porcine microvessels to nitroglycerin. Am J Physiol 1990;258:H515–20.
20. Sellke FW, Tomanek RJ, Harrison DG. L-cysteine selectively potentiates nitroglycerin-induced dilation of small coronary microvessels. J Pharmacol Exp Ther 1991;258:365–9.