INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

THE HEART, together with other organs, has the capacity to develop a delayed self adaptation to ischemia-reperfusion (I/R) injury. This phenomenon, termed delayed preconditioning (PC), is usually characterized by a marked limitation of infarct size (9, 16) and a prevention of myocardial stunning (26). Moreover, our group showed that the protective effect of delayed PC extends to coronary endothelial cells (8, 10, 11). Given the well-known role of endothelial cells in the regulation of vascular tone, but also of platelet and leukocyte function, it is likely that such an endothelial protection may have many important consequences, for example in terms of prevention of vasospasm, thrombosis, or atherosclerosis. However, little is known about the mechanisms by which PC confers delayed coronary endothelial protection.

Among the possible mediators of delayed PC, nitric oxide (NO) has been evidenced to play a major role. Indeed, NO produced by the inducible form of NO synthase (iNOS) is a mediator of delayed PC at the level of the cardiomyocytes (2, 5, 6). In contrast, our previous results showed that iNOS was not involved at the level of the coronary endothelium (10), suggesting distinct mechanisms between endothelial and myocardial PC. However, the lack of role of iNOS does not exclude the implication of the other isoforms of NOS, especially endothelial NOS (eNOS). Indeed, in vivo activation of eNOS (by acetylcholine) protects coronary endothelial cells against reperfusion injury (22).

Investigation of the mechanisms of delayed PC has to be separated in two phases: the triggers, corresponding to the essential mechanisms occurring during or immediately after the PC stimulus, and the mediators, corresponding to mechanisms involved during prolonged ischemia and reperfusion, and ultimately responsible for protection. In this regard, we (11) showed recently that NO produced from eNOS (but not iNOS) during PC is an essential trigger of delayed endothelial protection, probably through its interactions with free radicals. However, it is still unknown whether eNOS also plays a role as a mediator of delayed endothelial PC (i.e., when produced during prolonged I/R).

Reperfusion after myocardial ischemia is associated with the release of free radicals, such as superoxide anions. These species react with NO to form peroxynitrite (1). This latter compound, although generally considered as deleterious, has, however, been shown to inhibit leukocyte rolling and adherence to endothelial cells and to exert cytoprotective effects against myocardial I/R injury (13, 20). Moreover, peroxynitrite administration before prolonged ischemia has beneficial effects on coronary endothelium in cats (20) and plays a central role as a trigger of delayed endothelial PC (11). In this context, it is possible that peroxynitrite derived from NO is also involved as a mediator of delayed endothelial PC. Thus we also examined the effects of the peroxynitrite scavenger seleno-L-methionine (25, 27, 28) on delayed endothelial PC.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Experimental preparation. Experiments were performed in male Wistar rats (Charles River; Saint-Aubin les Elbeuf, France), weighing between 350 and 400 g. Animals used in the in vitro vascular studies were assigned to nine experimental groups (Fig. 1). Experimentation was conducted in conformity with the "Guiding Principles for Research Involving Animals."

View larger version (26K): [in this window] [in a new window]   Fig. 1.   Experimental groups and treatment protocols. Solid boxes indicate periods of myocardial ischemia, and open boxes indicate periods of reperfusion. NG-nitro-L-arginine methyl ester (L-NAME), N-[3(aminomethyl)benzyl]acetamidine (1400W), 7-nitroindazole (7-NI) and seleno-L-methionine were given 10 min before coronary occlusion on day 2. I/R, ischemia-reperfusion; PC, preconditioning; iv, intravenous.

5

View larger version (15K): [in this window] [in a new window]   Fig. 2.   Effect of in vitro incubation with 7-NI (105 M) on the relaxations to acetylcholine in coronary arteries isolated from normal rats (n = 5).

In vitro vascular studies. Coronary endothelial function was assessed as described previously (7, 8, 10, 11, 22, 23). At the end of the experiments, the heart was removed and immediately placed in cold oxygenated Krebs buffer. A 1.5- to 2-mm-long segment of the coronary artery distal to the site of occlusion was carefully dissected and mounted in a small vessel myograph (JP Trading; Aarhus, Denmark). Normalization procedure was performed after an equilibration period, as described previously (19). Segments with an internal diameter <170 µm were excluded to avoid mechanical endothelial injury and unspecific dysfunction. Concentration-response curves to acetylcholine (10⁸ to 3.10⁵ M) were performed in serotonin-precontracted segments (10⁵ M). Serotonin was used because the epicardial coronary arteries in the rat do not express serotonin receptors on endothelial cells, and thus serotonin only induces smooth muscle cell contraction and no endothelium-dependent relaxations in these preparations (21). The changes in the relaxing responses to acetylcholine were used as a marker of endothelial dysfunction (after I/R) or endothelial protection (after PC). Changes in the responses to acetylcholine appear to reflect the extent of structural injury to endothelial cells in this model (7). In some cases, relaxing responses were assessed in the presence of 10⁴ M L-arginine (see RESULTS). Endothelium-independent relaxation to increasing concentrations of sodium nitroprusside (SNP) was also obtained in serotonin-precontracted arteries.

Statistical analysis. All results are expressed as means ± SE. In all experiments, n refers to the number of animals from which the arteries were taken. The significant differences in mean values between and within multiple groups were examined by repeated measures, two-way ANOVA, taking into account the nine groups (except for the responses to acetylcholine in seleno-L-methionine vs. untreated preconditioned rats, which were analyzed separately because they were performed in the absence of L-arginine; see RESULTS). A P value 0.05 was considered statistically significant.

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Hemodynamics. Compared with sham-operated animals, neither I/R nor PC affected blood pressure or heart rate. L-NAME induced a significant increase in blood pressure. Indeed, at the onset of ischemia (i.e., 10 min after administration), mean arterial pressure was 104 ± 7 and 146 ± 10 mmHg in untreated and L-NAME-treated rats, respectively (Table 1). However, this increase was transient, and no differences in blood pressure were observed at the end of ischemia (i.e., 30 min after administration). 7-NI, 1400W, or seleno-L-methionine did not affect heart rate or blood pressure at any time (Table 1).

Vascular diameters and relaxing responses to SNP. No differences in normalized vessel diameters were observed between the groups (data not shown). No significant differences in the responses to serotonin or to the SNP were observed between the groups, suggesting that the responsiveness of smooth muscle cells was intact after prolonged I/R of PC (data not shown).

Effects of in vivo L-NAME administration on in vitro responses to acetylcholine. In the present experimental conditions, in vivo administration of L-NAME in sham rats (i.e., in the absence of ischemia) markedly reduced the in vitro relaxing responses to acetylcholine. A similar inhibition was shown in rats subjected to I/R (Fig. 3) and to PC (maximal relaxations: untreated, 74 ± 5%; L-NAME, 17 ± 6%). However, this impaired response could be fully reversed by in vitro incubation of the arteries with 10⁴ M L-arginine. Most importantly, we also found that in vitro incubation with L-arginine did not affect the impaired relaxation to acetylcholine after I/R in the absence of L-NAME (Fig. 3B, comparison between untreated and untreated in vivo + L-arginine in vitro). Thus because in vitro L-arginine did not interfere with the assessment of I/R injury to the endothelium, all responses to acetylcholine were later assessed in the presence of 10⁴ M L-arginine. To compare the effects of L-NAME to those of the other NOS inhibitors tested, the responses to acetylcholine in animals treated with 7-NI and 1400W were also assessed in the presence of L-arginine.

View larger version (20K): [in this window] [in a new window]   Fig. 3.   Effect of L-NAME in vivo on the relaxation of the coronary arteries to acetylcholine in vitro. A: sham rats (n = 7). B: rats subjected to I/R without PC (n = 7). The responses to acetylcholine were assessed in the absence or in the presence of L-arginine (L-Arg) 104 M. *P < 0.05.

Effects of I/R and PC on endothelium-dependent relaxation. Compared with sham animals, the response to acetylcholine was markedly reduced after I/R alone (maximal relaxation: sham, 77 ± 3%; I/R, 44 ± 6%; P < 0.01). However, we prevented this impairment by PC (74 ± 5%; P < 0.01 vs. I/R) (Fig. 4).

View larger version (17K): [in this window] [in a new window]   Fig. 4.   Effect of I/R and PC on endothelium-dependent relaxation of the coronary arteries to acetylcholine in the presence of L-arginine (104 M). *P < 0.05 and **P < 0.01 vs. sham; P < 0.05, P < 0.01 vs. PC.

Effects of L-NAME, 7-NI, or 1400W on coronary endothelial protection by PC. Responses to acetylcholine (performed in the presence of 10⁴ M L-arginine) are shown in Fig. 5. L-NAME abolished the protective effect of PC (maximal relaxation: PC 74 ± 5%, PC + L-NAME 41 ± 7%, P < 0.01; Fig. 5A). In contrast, in the presence of L-arginine, in vivo L-NAME administration was not associated with changes in the responses to acetylcholine in sham rats or in rats subjected to I/R without PC (Fig. 3, comparison between untreated and L-NAME in vivo + L-arginine in vitro).

View larger version (14K): [in this window] [in a new window]   Fig. 5.   Effect of nitric oxide synthase (NOS) inhibition in vivo on the protective effect of delayed PC. The endothelium-dependent relaxation of coronary arteries to acetylcholine was assessed in the presence of L-arginine 104 M in vitro, after in vivo administration of the nonselective NOS inhibitor L-NAME (5 mg/kg iv) (A), the selective inducible NOS (iNOS) inhibitor 1400W (1 mg/kg iv) (C), and the selective neuronal NOS (nNOS) inhibitor 7-NI (30 mg/kg iv) (B). The relaxing responses to acetylcholine were performed in the presence of L-arginine 104 M. *P < 0.05, **P < 0.01 vs. untreated.

Effect of peroxynitrite scavenging on coronary endothelial protection by PC. Figure 6 shows the effect of the peroxynitrite scavenger seleno-L-methionine given 10 min before prolonged I/R in arteries from preconditioned animals (in the absence of L-arginine). Although there was a trend toward inhibition, this effect did not reach statistical significance (group asterisk concentration interaction: P = 0.13).

View larger version (14K): [in this window] [in a new window]   Fig. 6.   Effect of the peroxynitrite scavenger seleno-L-methionine (3 mg/kg) administered in vivo 10 min before prolonged ischemia (day 2) on the protective effect of delayed PC. The relaxing responses to acetylcholine were performed in the absence of L-arginine.

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

The major finding of the present study is that in vivo administration of L-NAME before prolonged myocardial I/R abolished the coronary endothelial protective effect of PC, showing for the first time that NO is a mediator of the endothelial effects of delayed PC. Although L-NAME inhibits all isoforms of NOS, we also observed no effect of the selective inhibitors of iNOS or nNOS. This demonstrates that NO produced by eNOS, in addition to being a trigger of PC (11), also is a mediator of the coronary endothelial protective effect of delayed PC in rats.

In the present study, we used L-NAME at a dose of 5 mg/kg. In pilot experiments, we found that the administration of higher doses (10-15 mg/kg) was associated with a high mortality during prolonged ischemia, whereas lower doses only moderately increase blood pressure. However, the efficacy of L-NAME at the dose of 5 mg/kg was verified by the significant increase in blood pressure observed immediately after administration. The efficacy of eNOS blockade is also supported by the marked inhibition of the in vitro relaxations to acetylcholine (see below).

L-NAME administration in vivo induced a marked inhibition of the in vitro endothelium-dependent relaxations of the coronary arteries to acetylcholine. Thus coronary NOS inhibition was not reversed after the heart was taken out and during in vitro incubation, despite several changes in the incubation buffer during isolation and preparation of the arteries. Because L-NAME is usually considered to be a reversible inhibition of eNOS, this suggests that the inhibitor has a long half-life in rats and that significant concentrations of the inhibitor persist at the time of death (i.e., 90 min after injection) and also that L-NAME strongly binds to eNOS even during prolonged in vitro incubation. Indeed, in most in vitro studies, it is usually observed that the inhibition of NO-mediated relaxation induced by incubation with L-NAME or nitro-L-arginine cannot be fully overcome when the arteries are incubated in inhibitor-free medium. Although some nonarginine-based molecules may also inhibit NOS and may also potentially not present the same persistent in vitro inhibition of the endothelial responses, most of these nonarginine NOS inhibitors are either poorly active as inhibitors of endothelial NOS (e.g., guanidine and thioureas) or show poor selectivity toward NOS. Thus none of those inhibitors were tested in the present experiments. Consequently, we chose to use only L-NAME, which has been used in numerous experiments as "selective" NOS inhibitors. However, some experiments suggested that L-NAME may in some conditions exert effects that are independent of NOS inhibition, including muscarinic antagonism (3), inhibition of catalase or of peroxide-mediated responses. However, those nonselective effects of L-NAME (including muscarinic antagonism) are not reversed by L-arginine, and were only demonstrated in few in vitro models; there is currently no evidence that they do occur in vivo and are relevant at the doses of L-NAME used.

We were able to reverse the effect of L-NAME blockade in vivo by the incubation of the coronary arteries in the presence of an excess L-arginine. Importantly, in vitro incubation with L-arginine (in the absence of L-NAME) did not improve the impaired responses to acetylcholine after I/R (Fig. 3B) nor did it affect it in sham rats (Fig. 3A). This lack of effect of L-arginine is also supported by comparison between the present data (in the presence of L-arginine) and our previous experiments performed in the absence of L-arginine (10, 11). This contrasts with other pathophysiological situations associated with endothelial dysfunction (e.g., hypercholesterolemia) and probably reflects the fact that the endothelial dysfunction observed in vitro after I/R in vivo is mostly the consequence of a reduced number of structurally intact cells, as observed with electron microscopy (7), and not that of a selective defect affecting NOS activity. Thus incubation of the arteries with excess of L-arginine after in vivo blockade of eNOS allowed us to test the role of NO produced in vivo during I/R as a mediator of the endothelial protective effect of PC.

As mentioned above, we observed a persistent inhibition of eNOS in vitro when arteries were taken out 90 min after injection of L-NAME, suggesting a prolonged inhibitory effect of eNOS. However, in the same conditions, the increase in blood pressure was transient and lasted <30 min (Table 1). To reconcile this apparent discrepancy, we performed a series of experiments in which we measured cardiac output by echocardiography in anesthetized closed-chest animals. In these conditions, we found that 30 min after L-NAME, cardiac output was decreased by 27 ± 2%, and total peripheral resistance was increased by 30 ± 4%, whereas blood pressure was back to pretreatment value (data not shown). Thus the marked increased in total peripheral resistance suggests that eNOS is indeed still inhibited and that the return of blood pressure is mostly due to the marked decrease in cardiac output.

Investigation of the physiological role of the eNOS is limited by the lack of selective inhibitors of this isoform. To overcome this limitation and allow assessment of the role of eNOS in endothelial PC, we compared the effects of L-NAME to those of selective inhibitors of iNOS or nNOS in the same experimental conditions.

The role of iNOS was assessed with the use of 1400W, which is one of the most potent and selective iNOS inhibitor (4). The dose of 1400W was chosen because it effectively blocks some of the iNOS-mediated effects of lipopolysaccharides, including increased vascular leakage (3) and hypotension (4, 10). We found that iNOS inhibition by 1400W had no effect on the endothelial protective effect of PC, in agreement with our previous results (10). We chose to repeat these experiments first to confirm our previous results and second because the present experimental protocol differed slightly from the previous one (especially regarding the use of L-arginine). Our present results confirm the lack of iNOS in delayed endothelial PC in rats, which also extends to the triggering phase (11), and further suggest that the mechanisms of endothelial PC differ from those observed at the level of the cardiomyocytes (2, 5).

The role of nNOS was assessed using 7-NI, administered intravenously at the dose of 30 mg/kg. This dose was chosen on the basis that it effectively blocks brain nNOS in mice without functional evidence of eNOS inhibition (18) and also limits cerebral infarct size in rats (30). The lack of functional evidence of eNOS inhibition by 7-NI is also supported by the fact that it did not affect blood pressure in our protocol, unlike L-NAME. It is also supported by our observation (Fig. 2) that 7-NI did not affect the coronary responses to acetylcholine in vitro. We found that this nNOS inhibitor had no effect on PC-induced coronary endothelial protection, which is consistent with the absence of this isoform in coronary endothelial cells. Indeed, cardiac nNOS has been found in nerve terminals (24), in cardiac conduction tissue (17), and in the sarcoplasmatic reticulum (29), but its presence has not been reported at the level of endothelial cells. In this context, it is not surprising to observe a lack of effect of nNOS inhibition in our experimental protocol.

Taken together, the lack of effect of iNOS or nNOS inhibition, together with the marked effect of the nonselective inhibitor, suggests that the NO-mediated endothelial protective effect of PC involves eNOS. It should be acknowledged, however, that the present conclusion is based only on the use of inhibitors of NOS. Given the limitations discussed above, i.e., the uncertainties regarding the selectivity of the inhibitors and the possible problems due to the use of L-arginine in our experiments, and the fact that we did not measure NOS activity or the production of NO, our results must be interpreted with caution regarding the exact role of endothelium-derived NO in delayed endothelial PC.

At present, little is known about the specific involvement of eNOS as a mediator of delayed PC. In contrast, there is large evidence that iNOS is a mediator of PC against stunning (2) and infarction (5) and also contributes to adenosine-induced myocardial protection (6). However, as mentioned before, iNOS is not involved in delayed endothelial PC. This difference is probably due to the different cellular targets and to the differential effects of NO on cardiac myocytes and coronary endothelial cells (10). However, although the isoforms of NOS involved appear to differ, NO seems to be a common mediator of all cardiac protective effects of delayed PC, whether at the level of the myocytes or of the coronary endothelial cells.

One possibility to explain the involvement of NO in delayed endothelial PC would be that it interacts with reactive oxygen species, mainly superoxide anions, to form peroxynitrite. Indeed, these species have been shown to inhibit leukocyte rolling and adherence to endothelial cells and to exert paradoxical cardioprotective effects when administered during prolonged myocardial I/R (13, 20). Thus the inhibition of endothelial PC by L-NAME might possibly be due to a prevention of the formation of peroxynitrites, secondary to the inhibition of NO production. To test this hypothesis, we administered seleno-L-methionine, which selectively scavenges peroxynitrites without affecting NO and the superoxide anions of hydrogen peroxide (28) and has been shown to inhibit neutrophil adhesion in some models (27). In these conditions, seleno-L-methionine did not affect PC. Indeed, although there was a trend toward inhibition, this did not reach statistical significance. We think it is unlikely that this lack of significant effect is due to an insufficient dosage, because we (11) showed previously that the same dose of seleno-L-methionine completely abolished endothelial protection when given at the same dose during PC.

It must be noted that the selectivity of seleno-L-methionine for peroxinitrite has been so far tested only in vitro (28). Thus our results with this scavenger have to be interpreted with caution because it may also scavenge other reactive oxygen species. The possibility that it scavenges NO in our experiments is, however, unlikely, because the compound did not increase blood pressure (Table 1). Moreover, the lack of apparent effect (whether beneficial or detrimental) of seleno-L-methionine probably rules out the hypothesis that it acts as a superoxide or hydroxyl radical scavenger in our experiments because such scavengers effectively prevent endothelial dysfunction in this model (8).

The fact that seleno-L-methionine had no effect does not support the hypothesis of a role for peroxynitrite as a mediator of delayed endothelial PC and argues for a direct effect of NO. Interestingly, we found that the peroxynitrite scavenger had no effect in the present study, i.e., when given during prolonged I/R, whereas it completely abolished endothelial protection when given at the same dose during PC (11). Thus although NO produced from eNOS plays a central role both as a trigger and a mediator of endothelial PC, the triggering and mediator phases differ with regard to the role of peroxynitrite.

In conclusion, our results suggest that the delayed PC-induced coronary endothelial protection against I/R injury is mediated by NO produced by eNOS.