Healthy Blood Vessel Blog

the Division of Cardiology, Emory University School of Medicine (S.R., D.G.H.), and Veterans Administration Hospital (D.G.H.), Atlanta, Ga.

Correspondence to David G. Harrison, Professor of Medicine, Cardiology Division, Emory University School of Medicine, Atlanta, GA 30322. (Circulation. 1996,94:240-243.)

Key Words: Editorials • endothelium-derived factors • vasoconstriction • vasodilation

	Introduction

Top Introduction References

A clinically important aspect of this important role of the endothelium is that it is impaired in a variety of diseases and conditions. These include hypertension, hypercholesterolemia, diabetes, transplant atherosclerosis, congestive heart failure, and cigarette smoking. NO· has a variety of beneficial, potentially antiatherogenic effects in the vessel wall. These include inhibition of platelet aggregation, leukocyte adhesion, and inhibition of adhesion molecule expression. It is conceivable, therefore, that loss of an NO· effect in various disease conditions could contribute to the atherosclerotic process. On the basis of these considerations, there is considerable interest in developing treatment strategies to prevent the loss of endothelial NO· in these various disease states. If this could be accomplished, it has been reasoned that many of the early events in atherosclerosis could be prevented.

In this issue of Circulation, Mancini et al⁸ present a paper that examines the effectiveness of one such treatment strategy. These investigators examined a subgroup of patients with several risk factors that have been associated with loss of endothelial NO·. These patients all had established coronary atherosclerosis, and some had mild hypercholesterolemia. Some were active smokers, and more than half had a history of hypertension. The authors examined responses to intracoronary injections of acetylcholine and nitroglycerin at baseline and after 6 months of treatment with the ACE inhibitor Quinapril or placebo. Acetylcholine was chosen as a prototypical endothelium-dependent vasoactive agent, and nitroglycerin was used as an endothelium-independent vasodilator. In point of fact, in most human studies (as in the present study by Mancini et al), acetylcholine predominantly produces vasoconstriction via its direct action of muscarinic receptors on the vascular smooth muscle. This degree of constriction is blunted by the concomitant release of NO·. Therefore, the lack of constriction or presence of mild vasodilatation of a coronary artery during injection of acetylcholine can be used as a readout for the preserved release of NO·. In contrast, the development of vasoconstriction would indicate a deficiency in NO· release and a direct effect of acetylcholine on the vascular smooth muscle. In the present study, Mancini et al observed a 9% to 14% vasoconstriction in response to the highest dose (10^-4 mol/L) of acetylcholine at the time of initial study. At the 6-month follow-up, the response of the placebo group had not changed (9.4% to 10.5% constriction at baseline and follow-up, respectively), whereas the group treated with Quinapril exhibited a dramatic decrease in the vasoconstriction caused by acetylcholine (from 14.3% to 2.3% constriction at baseline and follow-up, respectively). About twice as many patients in the Quinapril group as in the placebo group exhibited dilation to acetylcholine at the time of follow-up. Responses to nitroglycerin were not changed in either the Quinapril or placebo groups from baseline to follow-up study. The authors concluded that ACE inhibition has a beneficial effect on endothelium-dependent regulation of vasomotion.

This study⁸ has several strengths. An impressively large number of patients was included, particularly for a study as time- and labor-intensive as this, and it is unlikely that the result could be attributed to any artifact resulting from an insufficient number of subjects. Studies such as this are difficult to perform because they involve a repeat catheterization and the protocols are time-consuming for the investigators and the subjects. The analyses of coronary artery diameter involved state-of-the-art quantitative techniques, and the investigators examined both a mild stenosis and other segments of the coronary arteries. Rather than using only the baseline values as a control, the investigators included a placebo-treated group. The effect of Quinapril treatment on the acetylcholine response was substantial, as great as has been reported for lipid lowering.⁹ A potential criticism of the study is that it is not clear whether another vasodilator-like drug could have the same effect. The dose of Quinapril chosen, however, had no effect on blood pressures in these subjects (who primarily were normotensive at the outset), and therefore it is unlikely that this result was simply due to vasodilation. This suggests that ACE inhibition likely has an interaction with the endothelial/NO· system that is independent of a vasodilator or hemodynamic effect.

How could an ACE inhibitor or angiotensin II modulate the ability of the endothelium to release NO·? One well-documented phenomenon relates to the fact that the angiotensin I–converting enzyme and the endothelial cell kininase are one and the same enzyme.¹ As a kininase, this enzyme is responsible for degradation of bradykinin. ACE inhibitors such as Quinapril, therefore, are capable of prolonging the half-life of any bradykinin that is in the proximity of the endothelium. Bradykinin is a potent stimulator of NO· and prostacyclin release. Therefore, an ACE inhibitor theoretically can promote release of these vasodilator substances by augmenting the effect of any bradykinin that is present.¹⁰ It is unclear how important this is in the intact human or animal. It is questionable whether or not there is sufficient bradykinin in proximity to the endothelium under normal circumstances to permit this effect. Furthermore, it is unlikely that prolonging the half-life of bradykinin (even if sufficient quantities of bradykinin were present) could enhance acetylcholine stimulation of NO· release. Thus, although this role of ACE inhibitors has been the focus of substantial investigation, it probably was not the mechanism underlying the beneficial effect observed in the study by Mancini and coworkers.⁸

Angiotensin II is one of the most potent vasoconstrictors produced in vivo. In addition, angiotensin II has been shown to stimulate transcription of preproendothelin and to promote the release of endothelin-1 from vascular cells.¹¹ In renovascular hypertension, angiotensin II has been implicated in stimulating production of the vasoconstrictor prostanoid PGH₂ by the endothelium.¹² An ACE inhibitor such as Quinapril, therefore, could improve vasomotor function by reducing these vasoconstrictor influences. However, it is difficult to understand how this would specifically improve responses to endogenously released NO· and not affect responses to nitroglycerin. Baseline coronary diameters were not different before and after Quinapril treatment, which suggests it was unlikely that Quinapril worked by simply reducing vasoconstrictor influences.

These considerations suggest that ACE inhibition had a unique effect on NO· synthesis and release or on its ultimate effect on the vascular smooth muscle. One potential mechanism to explain this might relate to an increase in expression of the NO synthase enzyme. Although this enzyme has been considered to be expressed constitutively, it is now clear that its expression is subject to modest degrees of regulation. Recent studies¹³ have shown that NO synthase expression can be modulated by shear stress, oxidized LDL, lysophosphatidyl choline, hypoxia, and manipulation of activity of protein kinase C. The latter may have relevance to the effect of an ACE inhibitor, because angiotensin II is a potent activator of protein kinase C.

A mechanism by which an ACE inhibitor very likely could improve the endothelial NO· effect is through modulation of vascular superoxide (·O₂^-) production. Even before the chemical nature of EDRF was identified, it was known that its half-life was prolonged by SOD and shortened by chemical generation of superoxide anions (·O₂^-).¹⁴ This interaction between superoxide and NO· was demonstrated initially in systems in which the ·O₂^- was produced artificially. Several conditions in experimental animals have been found in which vascular ·O₂^- production is increased and that seem to impair endothelium-dependent vascular relaxation via inactivation of NO·.³

One of the most important observations in the past few years regarding vascular sources of ·O₂^- is that both the endothelium and vascular smooth muscle contain membrane-bound oxidase(s) that use NADH and NADPH as substrates for electron transfer to molecular oxygen.¹⁵ These are likely multicomponent enzyme systems that are membrane bound and extramitochondrial. They have similarities to the neutrophil NADPH oxidase in that they possess flavin- and heme-binding regions that probably are important in the transfer of electrons. Recently, Fukui and coworkers¹⁶ cloned a component of the vascular smooth muscle oxidase, p22phox, which has 90% similarity to the neutrophil oxidase system. Thus, at least one component of the vascular oxidase system is similar to the neutrophil system. There are, however, very important differences between the vascular oxidases and the neutrophil oxidase. First, the output of the vascular oxidase is much lower than that of the neutrophil oxidase (nanomoles versus micromoles of ·O₂^- per minute per milligram of protein). Second, the vascular oxidase does not exhibit bursts of activity, as does the neutrophil system, but releases ·O₂^- constantly. This does not detract from the importance of the vascular oxidase system. The neutrophil oxidase system serves a bactericidal role, whereas the vascular oxidase may have other roles, such as modulation of the vasodilator effect of NO·.

An important observation regarding the vascular NADH/NADPH oxidase is that its activity can be increased by angiotensin II. In studies of cultured vascular smooth muscle cells, Griendling and coworkers¹⁷ have shown that subnanomolar concentrations of angiotensin II increase NADH oxidase activity by severalfold in as little as 4 hours. The signaling processes involved seem to be related to activation of phospholipase A₂ and could be mimicked by treatment of cellular homogenates with arachidonic acid.

Recently, we¹⁸ have performed studies to determine whether angiotensin II can activate NADH-driven oxidases in vivo. Osmotic minipumps were used to infuse either angiotensin II (0.7 mg·kg^-1·d^-1) or norepinephrine (2.75 mg·kg^-1·d^-1). Both angiotensin II and norepinephrine increased blood pressure to a similar extent (190 mm Hg). Interestingly, vascular ·O₂^- production was doubled in rats made hypertensive by angiotensin II but was not changed in the rats made hypertensive by norepinephrine infusion. Studies of homogenates of vessels from these animals showed that the activity of the NADH oxidase was doubled in concert with the increase in vascular ·O₂^- production. Endothelium-dependent relaxations to acetylcholine and the calcium ionophore A23187 were abnormal in vessels from the angiotensin II–treated rats but were unaltered in the norepinephrine-treated rats. Finally, the endothelium-dependent vascular relaxations were restored toward normal by treatment of the vessels with a form of SOD (liposome-encapsulated SOD) that allows delivery of the SOD intracellularly. In summary, these studies demonstrated that hypertension caused by angiotensin II has a completely different effect on vascular ·O₂^- production and vascular reactivity from other forms of hypertension not associated with increased angiotensin II levels. In additional studies,¹⁸ we showed that even lower concentrations of angiotensin II, which had only minimal effects on blood pressure, also doubled NADH oxidase activity.

In these studies, we found that the angiotensin type-1 receptor antagonist losartan prevented the effect of angiotensin II and actually lowered vascular ·O₂^- production below the level observed in normal animals. These findings suggest that activation of the angiotensin type-1 receptor, even by the ambient levels of angiotensin II that are present in normal, physiological circumstances, may modulate vascular ·O₂^- production. Thus, it is conceivable that an ACE inhibitor could lower ·O₂^- rates of production even in situations in which angiotensin II is not elevated.

Activation of membrane oxidases and increases in vascular ·O₂^- production may have important roles in other conditions. Recently, we found that prolonged (3-day) nitroglycerin treatment is associated with an increase in rates of vascular ·O₂^- production and that this was at least partly responsible for the tolerance to nitroglycerin and cross-tolerance to endogenously released nitroglycerin.¹⁹

These new observations may provide insight into the mechanisms whereby both normal and elevated levels of angiotensin II can contribute to vascular disease. Increases in vascular ·O₂^- production may not only shorten the half-life of NO· but may contribute to oxidation of lipoproteins, damage to membrane lipids, and genesis of other radicals with a myriad of physiological and pathophysiological effects.²⁰ In contrast, measures that lower angiotensin II levels, such as treatment with an ACE inhibitor, may have beneficial effects on outcome by lowering vascular ·O₂^- production. This phenomenon may contribute to the beneficial effect of ACE inhibitors on outcome after myocardial infarction or in the setting of heart failure.²¹

In summary, there is mounting evidence that there are direct links between the renin/angiotensin system and the NO/L-arginine pathway in the endothelium and vessel wall. The underlying processes may involve modulation of NO· production, alterations of NO· degradation, or other, poorly understood phenomena. Of particular interest is the effect of angiotensin II on activation of membrane oxidases, leading to excessive degradation of NO· via ·O₂^-. Whatever the mechanism, the work of Mancini et al⁸ demonstrates the importance of this interaction and points to new mechanisms whereby ACE inhibitors and perhaps angiotensin II–receptor antagonists might benefit vascular function. Given that NO· may have a number of beneficial antiatherogenic effects, it will be important to determine whether ACE inhibitors can influence other properties of the endothelium and vascular smooth muscle.

	Selected Abbreviations and Acronyms

	Footnotes

	References

Top Introduction References

1. Mombouli JV, Vanhoutte PM. Kinins and endothelial control of vascular smooth muscle. Annu Rev Pharmacol Toxicol. 1995;35:679-705.[Medline] [Order article via Infotrieve]
   
2. Lerman A, Burnett JC Jr. Intact and altered endothelium in regulation of vasomotion. Circulation. 1992;86(suppl III):III-12-III-19.
   
3. Harrison DG. Endothelial control of vasomotion and nitric oxide production. Cardiol Clin. 1996;14:1-15.[Medline] [Order article via Infotrieve]
   
4. Palmer RMJ, Ferrige AG, Moncada S. Nitric oxide release accounts for the biological activity of endothelium-derived relaxing factor. Nature. 1987;327:524-526.[Medline] [Order article via Infotrieve]
   
5. Nathan C, Xie QW. Nitric oxide synthases: roles, tolls, and controls. Cell. 1994;78:915-918.[Medline] [Order article via Infotrieve]
   
6. Myers PR, Minor RL Jr, Guerra R Jr, Bates JN, Harrison DG. The vasorelaxant properties of the endothelium-derived relaxing factor more closely resemble S-nitrosocysteine than nitric oxide. Nature. 1990;345:161-163.[Medline] [Order article via Infotrieve]
   
7. Stamler JS, Singel DJ, Loscalzo J. Biochemistry of nitric oxide and its redox-activated forms. Science. 1992;258:1898-1902.[Abstract/Free Full Text]
   
8. Mancini GBJ, Henry GC, Macaya C, O'Neill BJ, Pucillo AL, Carere RG, Wargovich TJ, Mudra H, Luscher TF, Klibaner MI, Haber HE, Uprichard ACG, Pepine CJ, Pitt B. Angiotensin-converting enzyme inhibition with Quinapril improves endothelial vasomotor dysfunction in patients with coronary artery disease: the TREND study (Trial on Reversing ENdothelial Dysfunction). Circulation. 1996;94:258-265.[Abstract/Free Full Text]
   
9. Treasure CB, Klein JL, Weintraub WS, Talley JD, Stillabower ME, Kosinski AS, Zhang J, Boccuzzi SJ, Cedarholm JC, Alexander RW. Beneficial effects of cholesterol-lowering therapy on the coronary endothelium in patients with coronary artery disease. N Engl J Med. 1995;332:481-487.[Abstract/Free Full Text]
   
10. Cachofeiro V, Sakakibara T, Nasjletti A. Kinins, nitric oxide, and the hypotensive effect of captopril and ramiprilat in hypertension. Hypertension. 1992;19:138-145.[Abstract/Free Full Text]
    
11. Hahn AW, Resink TJ, Scott-Burden T, Powell J, Dohi Y, Buhler FR. Stimulation of endothelin mRNA and secretion in rat vascular smooth muscle cells: a novel autocrine function. Cell Regul. 1990;1:649-659.[Medline] [Order article via Infotrieve]
    
12. Lin L, Mistry M, Stier CT Jr, Nasjletti A. Role of prostanoids in renin-dependent and renin-independent hypertension. Hypertension. 1991;17:517-525.[Abstract/Free Full Text]
    
13. Harrison DG, Venema RC, Arnal JF, Inoue N, Ohara Y, Sayegh H, Murphy TJ. The endothelial cell nitric oxide synthase: is it really constitutively expressed? Agents Actions. 1995;45:107-117.
    
14. Rubanyi GM, Vanhoutte PM. Oxygen-derived free radicals, endothelium and responsiveness of vascular smooth muscle. Am J Physiol. 1986;250:H815-H821.
    
15. Mohazzab KM, Kaminski PM, Wolin MS. NADH oxidoreductase is a major source of superoxide anion in bovine coronary artery endothelium. Am J Physiol. 1994;266:H2568-H2572.[Abstract/Free Full Text]
    
16. Fukui T, Lassegue B, Kai H, Alexander RW, Griendling KK. Cytochrome b558 -subunit cloning and expression in rat aortic smooth muscle cells. Biochim Biophys Acta. 1995;1231:215-219.[Medline] [Order article via Infotrieve]
    
17. Griendling K, Ollerenshaw JD, Minieri CA, Alexander RW. Angiotensin II stimulates NADH and NADPH activity in cultured vascular smooth muscle cells. Circ Res. 1994;74:1141-1148.[Abstract/Free Full Text]
    
18. Rajagopalan S, Kurz S, Munzel T, Tarpey M, Freeman B, Griendling K, Harrison D. Angiotensin II mediated hypertension in the rat increases vascular superoxide production via membrane NADH/NADPH oxidase activation: contribution to alterations of vasomotor tone. J Clin Invest. 1996;97:1916-1923.[Medline] [Order article via Infotrieve]
    
19. Munzel T, Sayegh H, Freeman BA, Tarpey MM, Harrison DG. Evidence for enhanced vascular superoxide anion production in nitrate tolerance: a novel mechanism underlying tolerance and cross-tolerance. J Clin Invest. 1995;95:187-194.
    
20. Berliner JA, Navab M, Fogelman AM, Frank JS, Demer LL, Edwards PA, Watson AD, Lusis AJ. Atherosclerosis: basic mechanisms—oxidation, inflammation, and genetics. Circulation. 1995;91:2488-2496.[Abstract/Free Full Text]
    
21. Latini R, Maggioni AP, Flather M, Sleight P, Tognoni G. ACE inhibitor use in patients with myocardial infarction: summary of evidence from clinical trials. Circulation. 1995;92:3132-3137.[Free Full Text
21. 
    

¹ Univ of Tokyo, Tokyo, Japan
² Kobe Univ, Kobe, Japan
³ Univ of Tokyo, Tokyo, Japan

Background: Epidemiological studies have shown that relative hypogonadism is associated with the higher incidence of cardiovascular disease, the mechanism of which remains unknown. We investigated whether relative hypogonadism would be related to endothelial dysfunction in men.

Methods: Consecutive 188 men (mean age ± SD = 47±15 years), who examined flow-mediated vasodilation (FMD) of the brachial artery using ultrasonography, were enrolled. The subjects with cardiovascular disease, malignancy, overt endocrine disease or having steroid hormones were excluded. Plasma hormone levels were determined after 12-hour fast in the morning, and the relationship between hormone levels and FMD was analyzed.

Results: Total and free testosterone and dehydroepiandrosterone-sulfate (DHEA-S) were significantly correlated with %FMD (r=0.262, 0.355 and 0.296, respectively; p<0.001),> was not. %FMD in the highest quartile of free testosterone was 1.7-fold higher than that in the lowest quartile (5.6±2.9 vs. 3.3±2.6 [mean±SD], p<0.05).> analysis revealed that total and free testosterone were related to %FMD independent of age, body mass index, hypertension, hypercholesterolemia, diabetes mellitus and smoking (ß=0.197 and 0.240, respectively; p<0.01),> index, systolic blood pressure, total cholesterol, HDL cholesterol, fasting plasma glucose, smoking and carotid intima-media thickness (ß=0.211 and 0.259, respectively; p<0.01).> was not significantly related to %FMD on multivariate analysis.

Conclusions: Low plasma testosterone level was associated with endothelial dysfunction in men independent of other risk factors, suggesting a protective effect of testosterone on the endothelium.

----------------------

Discussion

The results of the present study demonstrate that testosterone (100 pM – 10 M) causes potent and rapid vasorelaxations in the rat mesenteric arterial bed. The acute vasorelaxant effects of testosterone are partly mediated via endothelium- and NO-dependent, but not via receptor-dependent mechanisms. However, a substantial proportion of vasorelaxation to testosterone is mediated by increasing potassium efflux through BK_Ca channels, but not via K_ATP or K_V channels.

In the present study, it was shown that testosterone caused acute vasorelaxations in the rat mesenteric arterial bed pre-contracted with methoxamine. These results are consistent with previous findings which show that testosterone induces vasorelaxation in contracted blood vessels, such as rat thoracic aorta (Costarella et al., 1996; Perusquia et al., 1996; Honda et al., 1999) and porcine (Chou et al., 1996), and human coronary arteries (Webb et al., 1999). Several studies also demonstrated that acute vasorelaxant effects of testosterone were observed at pharmacological concentrations (100 nM – 300 M) (Yue et al., 1995; Chou et al., 1996; Costarella et al., 1996; Crews & Khalil, 1999; Honda et al., 1999). Interestingly, the present findings showed that testosterone, at low concentrations (100 pM – 1 nM), induced acute vasorelaxation, suggesting, therefore, that physiological (ca 20 nM; Johnson & Everitt, 1980) levels of testosterone may influence vascular tone.

Crews & Khalil (1999) demonstrated, in rat aortic strips, that vasorelaxation to relatively high concentrations of testosterone (100 nM – 10 M) had a rapid time of onset between 30 s and 2 min. Moreover, acute vascular effects of testosterone were also previously reported in an in vivo study by Chou et al. (1996) showing that testosterone (0.1 – 1 M) caused coronary relaxation within 90 to 120 s. Our results confirm that relaxant responses of the rat mesenteric arterial bed to testosterone were rapid in onset, occurring within 30 s to 20 min. Therefore, it is probable that testosterone-induced vasorelaxation is due to non-genomic mechanisms.

In contrast to the present study, others have demonstrated that testosterone enhanced vasoconstrictor responses to various agents, such as PGF₂, potassium chloride, endothelin-1 and 5-HT (Farhat et al., 1995; Teoh et al., 2000), and attenuated the effects of the endothelium-dependent vasorelaxants, bradykinin and the calcium ionophore, A23187 (Teoh et al., 2000). Clearly, in the present study testosterone caused relaxation rather than augmentation of methoxamine-induced tone. Although when K⁺ channels were blocked vasorelaxations to testosterone were abolished in the presence of 60 mM KCl or TBA, and that, under these conditions, contractile responses to testosterone were uncovered. Therefore, it appears that testosterone can cause mesenteric vasoconstriction, but only when K⁺ channels are inhibited.

Previous findings have demonstrated that vasorelaxation induced by testosterone was inhibited in the presence of NOS inhibitors in rat thoracic aorta (Costarella et al., 1996) and canine coronary artery (Chou et al., 1996). In the present investigation, we similarly found that testosterone-induced vasorelaxation was partly sensitive to L-NAME. Furthermore, we also found that removal of the endothelium opposed vasorelaxation to testosterone. These observations point to testosterone acting partly via endothelium- and NO-dependent pathways. Indeed, the effects of NOS inhibitor and removal of the endothelium had comparable effects, implying that NO is the principal endothelium-derived autacoid mediating these responses. The present findings are in agreement with the results of Chou et al. (1996), which showed that testosterone-induced vasorelaxation was inhibited in de-endothelialized canine coronary arteries. However, in our study there was a substantial endothelium-independent component. Indeed, other studies have demonstrated that neither removal of the endothelium nor the inclusion of NOS inhibitors affected vasorelaxation induced by testosterone in aortae from rat and rabbit, and coronary arteries from pig and rabbit (Yue et al., 1995; Perusquia et al., 1996; Crews & Khalil, 1999). Yue et al. (1995) also showed that methylene blue, an inhibitor of NO-activated guanylyl cyclase had no effects on testosterone-induced vasorelaxation in rabbit coronary artery and aorta. Furthermore, testosterone did not affect eNOS activity in the cultured bovine aortic endothelial cells (Goetz et al., 1999).

Clearly endothelial-derived NO contributes modestly to testosterone-induced vasorelaxation, therefore, we then sought to investigate the role of K⁺ channels. We found that vasorelaxation to testosterone was completely inhibited by raising the extracellular K⁺ concentrations to 30 and 60 mM, indicating that testosterone acutely causes vasorelaxation mainly by increasing K⁺ efflux (Adeagbo & Triggle, 1993). Furthermore, we also showed that testosterone-induced vasorelaxation was reduced by TBA, a selective inhibitor of calcium-activated K⁺ channels and ChTx, a selective BK_Ca channel inhibitor, but not by glibenclamide, a K_ATP channel inhibitor or 4-AP, a voltage-sensitive K⁺ channel inhibitor. From these results, we suggest that, in the rat mesenteric arterial bed, testosterone causes vasorelaxation via selective activation of BK_Ca channels. In agreement with our observations, Honda et al. (1999) recently demonstrated that tetraethylammonium, an inhibitor of calcium-activated potassium channels, inhibited vasorelaxation to testosterone in aortic rings from spontaneous hypertensive rats. In addition, previous studies also showed that glibenclamide had no effects on testosterone-induced vasorelaxation in coronary artery from rabbit (Yue et al., 1995) and dog (Chou et al., 1996). In contrast, neither glibenclamide or 4-AP inhibited vasorelaxation induced by testosterone in aortic rings from both Wistar-Kyoto rats and spontaneous hypertensive rats (Honda et al., 1999).

One possibility is that the endothelial-derived NO may act via activation of BK_Ca channels (Tare et al., 1990). However, there was a substantial endothelium-dependent component of vasorelaxation to testosterone. This suggests that endothelial-derived NO cannot account for all of the BK_Ca channel activation, which must occur by some other mechanism.

The present findings showed that testosterone-induced vasorelaxation was not inhibited by a testosterone receptor antagonist, flutamide, but potentiated. From these results, it is indicated that vasorelaxation induced by testosterone is mediated via testosterone receptor-independent pathway. Our results are consistent with a previous study in rabbit coronary artery and aorta which showed that vasorelaxation to testosterone was unaffected by flutamide (Yue et al., 1995). However, Murphy et al. (1999) demonstrated that testosterone inhibited contractile responses to prostaglandin F₂ and KCl in pig coronary smooth muscle cells, and this was sensitive to flutamide. This may point to species and/or regional differences in the involvement of testosterone receptors. The potentiated relaxation in the presence of flutamide might actually point to testosterone receptors being coupled to vasoconstriction. Blocking these receptors with flutamide might leave the non-receptor mediated vasorelaxation unopposed, leading to enhanced relaxation.

Based on the possibility that testosterone could be metabolized to oestradiol by an aromatase enzyme, we investigated the possibility that testosterone was metabolized to a vasoactive mediator by using an aromatase inhibitor, aminoglutethimide. Our results showed that aminoglutethimide had no inhibitory effects on testosterone-induced vasorelaxation. The present findings are in agreement with a previous study by Yue et al. (1995) which demonstrated that, in rabbit coronary artery, an aromatase inhibitor, aminoglutethimide had no effect on vasorelaxation induced by testosterone. In addition, Chou et al. (1996) also showed that pre-treatment with an oestrogen receptor antagonist, ICI 182,780 had no effects on vasorelaxation of canine coronary artery to testosterone (1 M). Taken together, it is suggested that vasorelaxation to testosterone is unlikely to be due to vasorelaxant effects of oestrogen (Yue et al., 1995; Chou et al., 1996). The small potentiation due to aminoglutethimide might be explained by reduced local metabolism of testosterone.

In summary, our findings demonstrate that, in the rat mesenteric arterial bed, testosterone causes acute vasorelaxations at physiologically relevant concentrations, which are partly mediated by endothelial NO via receptor-independent pathways. The inhibitory effects of TBA and ChTx on testosterone-induced responses suggest that vasorelaxation to testosterone involves mainly BK_Ca channel activation.