Introduction

Top Abstract Introduction Methods Results Discussion References

Endothelium-dependent vasodilation is generally attributed to the release of NO (Furchgott and Zawadski, 1980; Ignarro et al., 1987; Palmer et al., 1987). However, depending on the species, vascular tissue and agonist, endothelium-dependent, but NO-independent, vasodilation can be observed, an effect attributed to the release of a hyperpolarizing factor (Chen et al., 1988; Komori et al., 1988; Pacicca et al., 1992), the identity of which remains to be elucidated. Our studies using the perfused heart and kidney of the rat provide evidence for a P450-derived product of AA as the mediator of BK-induced vasodilation that is independent of NO and prostaglandins (Fulton et al., 1992, 1995). Thus, responses to BK, which is a recognized stimulus for phospholipases, are attenuated by inhibitors of P450 and abolished by inhibitors of both phospholipase A₂ and phospholipase C (Fulton et al., 1996). Of the P450 metabolites, an epoxide (EET) is considered the most likely candidate for a hyperpolarizing factor, an idea supported by several recent reports (Bauersachs et al., 1994; Campbell et al., 1996; Hecker et al., 1994). Epoxides are vasodilators that are synthesized by the endothelium and act via Ca⁺⁺-activated K⁺ channels (Hu and Kim, 1993; Rosolowsky et al., 1991; Zou et al., 1994). Furthermore, in the coronary perfusate, we detected EETs by gas chromatography-mass spectometry, whereas products of the other pathway of P450-dependent AA metabolism, the HETEs, could not be detected under basal or stimulated conditions (Fulton et al., 1995). Nevertheless, the vasodilator potency of EETs in the kidney and heart is less than would be anticipated for a putative mediator of the vasodilator effect of BK; microgram quantities are required (Fulton et al., 1996), whereas nanogram quantities of BK elicit maximal dilation (Fulton et al., 1992). Therefore, we also considered the possibility that a by-product of P450-dependent metabolism of AA (i.e., a free radical) could contribute to the vascular action of BK because it exhibits absolute dependency on phospholipases and also involves mono-oxygenase activity (Fulton et al., 1995,1996). BK has been shown to stimulate the release of superoxide from endothelial cells and feline and murine cerebral arterioles (Holland et al., 1990; Kontos et al., 1990; Rosenblum, 1987; Shimizu et al., 1994). Several sources of superoxide generation within endothelial cells have been identified and encompass NAD(P)H-dependent electron transport chains, mitochondria, xanthine oxidase and AA oxygenases, including cyclo-oxygenase and lipoxygenase (Wolin, 1996). Similarly, reactions involving P450 generate superoxide (Bondy and Naderi, 1994), although this has not been established for isozymes that metabolize AA. The promiscuous reactivity of superoxide results in the genesis of other reactive oxygen species (Kukreja and Hess, 1992). Products of superoxide (i.e., hydrogen peroxide and hydroxyl radicals) induce vasodilation in a variety of vessels, and hydrogen peroxide hyperpolarizes endothelium-denuded porcine coronary arteries (Beny and von der Weid, 1991). Consequently, we investigated whether free radicals generated via P450-dependent metabolism of AA could fulfill the requirements as a mediator of endothelium-dependent, but NO-independent, vasodilation to BK in the rat heart. Thus, in this preparation, vasodilator responses to BK are independent of NO but susceptible to inhibitors of P450 and K⁺ channels (Fulton et al., 1995, 1994). We first established that P450-dependent metabolism of AA could generate superoxide and then addressed the release of superoxide from the perfused heart in response to BK. Finally, we determined the effects of intracellular and extracellular free radical scavengers on vasodilator responses to BK in the perfused rat heart in which NO and prostaglandin synthesis was inhibited to isolate the P450-dependent component of the response. The results provide evidence against an oxygen-derived free radical as the mediator of the coronary vasodilator effect of BK.

	Methods

Top Abstract Introduction Methods Results Discussion References

Renal cortical microsomes were used as a source of P450 to examine whether P450-dependent metabolism of AA generates superoxide. Male Wistar rats (300-400 g) were anesthetized with pentobarbital (65 mg/kg), and after a midline laparotomy, the aorta was cannulated below the renal arteries and ligated above the renal arteries. The kidneys were flushed free of blood with 0.9% saline and excised, and the cortex was separated and homogenized in Tris/sucrose buffer. The suspension was centrifuged at 10,000 × g for 15 min, and the resultant supernate was centrifuged at 100,000 × g for 60 min. The resulting pellet was washed twice and resuspended in potassium phosphate buffer (0.1 M), and the protein concentration was determined using the BioRad (Hercules, CA) method.

Microsomes (150 µg), suspended in Krebs-Henseleit buffer, pH 7.4, containing 10 mM HEPES, 2.8 µM indomethacin, 250 µM lucigenin and 250 µM NADPH, were incubated with and without AA (10 µg), clotrimazole (10 µM) as an inhibitor of P450 or its vehicle (ethanol, 1%), and scavengers of superoxide, Tiron (10 mM) and SOD (400 U/ml). The concentration of clotrimazole was 10-fold that previously shown to attenuate renal and coronary vasodilator responses to BK (Fulton et al., 1992, 1995) and was used in excess to ensure inhibition of both epoxygenase and -hydroxylase (Harder et al., 1995). At 2 min before the reaction, microsomes were resuspended in 0.885 ml of Krebs' buffer, and inhibitors or respective vehicles (10 µl) were added, followed by NADPH (5 µl) and lucigenin (100 µl). The reaction was initiated by the addition of AA, and superoxide production was estimated by quantification of lucigenin chemiluminescence (Paky et al., 1993), measured as cpm (0.1-min interval) with a Packard Tricarb 1900TR liquid scintillation analyzer at 2 min after initiation of the reaction. A positive control was also used to address any nonspecific actions of the inhibitors. Thus, 0.1 U of xanthine oxidase, suspended in Krebs' buffer, pH 7.4, containing 10 mM HEPES and 250 µM lucigenin, was combined with its substrate, xanthine (100 µM), and tested against the inhibitors.

Cardiac slices were used initially to examine the effects of BK and AA on superoxide production, detected by chemiluminescence as described above. After anesthesia with pentobarbital (65 mg/kg) and intravenous heparin (1000 U/kg), a thoracotomy was performed, and the heart with attached aorta was excised, flushed free of blood and immersed in cold saline. The apex, aorta and the top of the atria were removed, and the heart was bisected. Transverse sections of cardiac tissue (~150 mg wet weight and ~2 mm thick) were obtained using a tissue slicer. Slices were placed in 7-ml glass scintillation vials containing 1 ml of Krebs' buffer with 50 µM nitroarginine, 50 µM NADPH, 10 mM HEPES and 250 µM lucigenin. To increase superoxide release, slices were preincubated for 30 min with diethyldithiocarbamate (1 mM) to inhibit endogenous Cu⁺⁺/Zn⁺⁺ superoxide dismutase. Thus, in preliminary studies, slices incubated in the absence of NADPH and diethyldithiocarbamate did not produce a detectable increase in superoxide after stimulation with BK or AA. Lucigenin-enhanced chemiluminescence, an index of superoxide production, was monitored as described 2 min after the stimulation of cardiac slices with BK (1 µg) or AA (5 µg).

Perfused heart. Male Wistar rats (weight, 350-400 g) were anesthetized with pentobarbital (65 mg/kg intraperitoneally) and given heparin (1000 U/kg). After thoracotomy, the heart was excised and perfused, via an aortic cannula, at constant flow with oxygenated (95% O₂/5% CO₂) Krebs' buffer at 37°C according to the method of Langendorff as modified by Broadley (1979). Flow was adjusted to 9-10 ml/min, which resulted in a basal perfusion pressure of 25-40 mm Hg, which increased to ~130-140 mm Hg as a result of nitroarginine (50 µM) in the buffer to which indomethacin (2.8 µM) had also been added. Nitroarginine and indomethacin were included in the perfusate to inhibit NO and prostanoid synthesis, respectively, and isolate the P450-dependent vasodilator response to BK. The contribution of free radicals to the coronary vasodilator effect of BK was examined using various inhibitors of intracellular and extracellular free radicals. Thus, vasodilator responses to BK (10-1000 ng) were assessed in the absence and presence of SOD (100 U/ml) to scavenge extracellular superoxide and Tiron (3 mM) or TEMPO (0.3 mM), superoxide scavengers that gain access to the intracellular milieu. The concentrations of Tiron and TEMPO that were chosen for these experiments were determined in preliminary studies to be devoid of nonspecific effects on vasodilator responses (see below). However, to verify that the concentrations of Tiron and TEMPO as well as SOD were effective in scavenging superoxide, we used cardiac slices as described. Basal superoxide production, determined by lucigenin-enhanced chemiluminescence, was abolished by 3 mM Tiron and reduced by 95% and 72% by 0.3 mM TEMPO and 100 U/ml SOD, respectively (n = 2). In experiments using SOD, catalase (400 U/ml) was also included to scavenge hydrogen peroxide. Vasodilator responses to cromakalim (5 µg), an ATP-sensitive K⁺ channel opener, or SCA40 (1 µg) or NS1619 (15 µg), agents reported to stimulate Ca⁺⁺-activated K⁺ channels, and in some cases hydrogen peroxide (60 µg), were also examined to assess any effects of the inhibitors unrelated to inhibition of free radicals as well as to assess the effectiveness of the various interventions.

Materials. The following reagents, which were obtained from the Sigma Chemical (St. Louis, MO), were dissolved in deionized water before use: Tiron, TEMPO, SOD, catalase, xanthine, xanthine oxidase, lucigenin (bis-N-methylacridinium nitrate), 1,3-dimethyl-2-thiourea, BK, sodium nitroprusside and nitroarginine. Clotrimazole, cromakalim and indomethacin were also obtained from Sigma and were dissolved in ethanol, acetone and 4.2% NaHCO₃, respectively. Clotrimazole was subsequently diluted with water. SCA 40, a gift from Dr. Cervoni (American Home Products, Pearl River, NY), was dissolved in 10% ethanol, and NS1619, obtained from Research Biochemicals International (Natick, MA), was dissolved in ethanol and diluted with water. Sodium arachidonate from NuChek (Elysian, MN) was dissolved in water, divided into aliquots, sealed under nitrogen, and stored at 70°C.

Data analysis. Results are presented as mean ± S.E.M. Data were compared by analysis of variance, and individual values were compared by Neuman-Keuls test. A value of P < .05 was considered statistically significant.

	Results

Top Abstract Introduction Methods Results Discussion References

P450. The addition of AA to indomethacin-pretreated renal cortical microsomes, a rich source of P450 mono-oxygenases, produced a 5-fold increase in lucigenin-enhanced chemiluminescence (an index of superoxide production) over the control value (vehicle-treated microsomes). The increase in chemiluminescence was abolished by the P450 inhibitor clotrimazole (fig. 1). The scavengers of superoxide, SOD (extracellular) and Tiron (intracellular and extracellular), significantly reduced lucigenin-enhanced chemiluminescence, with Tiron being the most effective (fig. 1). The effects of these agents were also tested on a known superoxide-generating system, xanthine and xanthine oxidase, which significantly increased chemiluminescence. The increase was inhibited by both SOD and Tiron but was not affected by clotrimazole (fig. 1).

View larger version (9K): [in this window] [in a new window]   Fig. 1.   Lucigenin-enhanced chemiluminescence as an index of superoxide production by (A) rat renal cortical microsomes in the presence of AA (10 µg; n = 15) and (B) xanthine oxidase (XO) in the presence of xanthine (X; 100 µM; n = 8). Effects of clotrimazole (CTZ; 10 µM, n = 6), SOD (400 U/ml; n = 7 for microsomes and n = 5 for XO) and Tiron (10 mM; n = 7 for microsomes and n = 5 for XO). Results are expressed as the increase in chemiluminescence from the respective controls without AA or xanthine. P < .01 relative to vehicle-treated microsomes (A) or vehicle-treated xanthine oxidase (B).

Cardiac slices. Superoxide production, estimated by lucigenin-enhanced chemiluminescence, was subsequently investigated in cardiac slices stimulated with BK and AA. The addition of either agent was without a significant effect in stimulating superoxide production. Thus, chemiluminescence was 31 ± 4 × 10⁴ cpm for the control vs. 25 ± 5 × 10⁴ cpm and 26 ± 4 × 10⁴ cpm for slices stimulated with AA and BK, respectively. In the absence of NADPH, chemiluminescence was 11 ± 2 × 10⁴ cpm for the control compared with 12 ± 2 × 10⁴ cpm for the corresponding slices treated with BK.

Perfused heart. Figure 2 is a representative tracing from six experiments of the simultaneous recording of perfusion pressure and lucigenin-enhanced chemiluminescence. Administration of BK, AA and cromakalim produced dilation of the coronary circulation but failed to increase the amount of lucigenin chemiluminescence detected. Consequently, we tested the effects of hypoxia and reoxygenation on the production of superoxide to validate this method of detection using an isolated heart. Interruption of coronary flow and reduced O₂ tension resulted in a rapid decline in both the perfusion pressure and the chemiluminescence signal. Reperfusion and reoxygenation elicited a rapid rise in perfusion pressure and a dramatic increase in luminescence (fig. 2).

View larger version (28K): [in this window] [in a new window]   Fig. 2.   Representative tracing from six experiments of simultaneous recordings of perfusion pressure- and lucigenin-enhanced chemiluminescence from the isolated, perfused heart of the rat showing the effects of BK (1 µg), AA (5 µg), cromakalim (CK; 5 µg) and hypoxia (stop flow) followed by reperfusion (start flow). The perfusate contained nitroarginine (50 µM) to inhibit NO synthesis and elevate perfusion pressure from 30 to 40 mm Hg to 130-140 mm Hg.

Perfused heart and radical scavengers. Vasodilator responses to BK, SCA40 and hydrogen peroxide were determined in nitroarginine- and indomethacin-treated hearts, and the effects of free radical scavengers were assessed. In some experiments, responses to NS 1619, an opener of Ca⁺⁺-activated K⁺ channels, were also assessed. This agent was included because it, unlike SCA40, has been shown to directly activate these channels (Macmillan et al., 1995).

View larger version (23K): [in this window] [in a new window]   Fig. 3.   Effect of SOD (400 U/ml) plus catalase (400 U/ml), denoted by open bars (n = 6), and vehicle (solid bars; n = 6) on vasodilator responses to BK, SCA40 and hydrogen peroxide (H2O2) in isolated, perfused hearts treated with indomethacin (2.8 µM) and nitroarginine (50 µM) to elevate perfusion pressure (PP) from ~40 mm Hg to 140 mm Hg. ** P < .01 vs. control.

View larger version (20K): [in this window] [in a new window]   Fig. 4.   Effect of Tiron (3 mM; hatched bars; n = 4) and TEMPO (0.3 mM; open bars; n = 4) on (A) BK-induced vasodilation and (B) vasodilation in response to NS1619, SCA40 and cromakalim in the isolated, perfused rat heart treated with indomethacin (2.8 µM) and nitroarginine (50 µM) to elevate perfusion pressure from ~40 mm Hg to 140 mm Hg. * P < .05 compared with control (solid bars; n = 4).

	Discussion

Top Abstract Introduction Methods Results Discussion References

We have thus far shown that BK-induced vasodilation in the isolated perfused heart of the rat is independent of NO but susceptible to inhibitors of P450 (Fulton et al., 1995) and phospholipases (Fulton et al., 1996) and is mediated via activation of a charybdotoxin-sensitive K⁺ channel (Fulton et al., 1994), results that are consistent with a P450-dependent metabolite of AA as an endothelium-derived hyperpolarizing factor. Indeed, a series of recent reports support this interpretation, providing evidence for an EET as a hyperpolarizing factor (Bauersachs et al., 1994; Campbell et al., 1996; Hecker et al., 1994). EETs have been shown to be vasorelaxant and to increase the open probability of Ca⁺⁺-activated K⁺ channels (Rosolowsky et al., 1991; Hu and Kim, 1993; Zou et al., 1994). However, the relative lack of vasodilator potency of EETs in isolated preparations prompted us to examine other products of P450-dependent metabolism of AA as putative mediators of NO-independent vasodilation to BK.

It is well established that P450-dependent mechanisms generate free radicals as by-products and that free radicals or their derivatives exhibit vasoactivity (Bondy and Naderi, 1994; Rosenblum, 1987). For example, superoxide can result in the formation of hydrogen peroxide, which is a vasodilator and stimulates K⁺ channels (Beny and von der Weid, 1991). Furthermore, the dilator effect of BK in cat cerebral arterioles has been attributed to free radicals generated via cyclo-oxygenase metabolism of AA (Kontos et al., 1990), which is released consequent to phospholipase stimulation by the peptide. Therefore, in this study, we examined the potential contribution of free radicals, generated by P450-dependent conversion of AA, to the coronary vasodilator effect of BK in the rat heart. Our approach was to first establish that P450-dependent metabolism of AA resulted in the formation of free radicals and, subsequently, to determine whether BK stimulated the release of superoxide in the heart and whether free radical scavengers modified the P450-dependent vasodilator effect of the peptide.

Using rat renal cortical microsomes as a source of P450, we verified, using lucigenin-enhanced chemiluminescence, that metabolism of AA by P450, like other P450-dependent reactions, resulted in the formation of superoxide. The addition of AA to microsomes pretreated with indomethacin to eliminate cyclo-oxygenase activity stimulated an increase in chemiluminescence that was inhibited by clotrimazole and scavengers of superoxide, confirming both the source and identity of superoxide. Although both scavengers of superoxide, Tiron and SOD, were equally efficacious in reducing superoxide generated by xanthine oxidase, Tiron was much more effective against superoxide produced by renal microsomes. This presumably relates to the ability of Tiron, but not SOD, to penetrate lipid membranes and therefore scavenge intravesicular superoxide (Ledenev et al., 1986).

Cultured endothelial cells, which lose the ability to express P450, generate superoxide in response to BK as a result of cyclo-oxygenase-dependent metabolism of AA (Holland et al., 1990; Shimizu et al., 1994). To evaluate a potential role of superoxide generated by P450, we obtained measurements of superoxide from freshly isolated tissues stimulated with BK. However, we were unable to demonstrate increases in superoxide release in cardiac slices pretreated with indomethacin and nitroarginine to inhibit cyclo-oxygenase and NO synthase and mimic the conditions under which we have demonstrated P450-dependent coronary vasodilation to BK (Fulton et al., 1995). To improve the sensitivity of superoxide measurements and to associate changes in perfusion pressure with superoxide levels, hearts were placed in a light-shielded box in close apposition to a photon multiplier so that simultaneous changes in perfusion pressure and superoxide could be monitored (Mohazzab-H et al., 1996). However, the P450-dependent vasodilator action of BK was not associated with increases in superoxide. Thus, BK elicited a vasodilator response in the perfused heart that was not associated with a increase in lucigenin-enhanced chemiluminescence. Furthermore, metabolism of AA by cyclo-oxygenase, an established source of superoxide, also failed to increase the chemiluminescent signal. However, the method of detection does not appear to be a limitation in that reperfusion of the heart after a period of hypoxia, a known stimulus for superoxide formation, resulted in a dramatic increase in superoxide, 5-fold base-line, demonstrating that endogenously produced superoxide can be detected using this method. Consequently, the results suggest that BK does not stimulate the formation of superoxide or its derived metabolites from the heart in sufficient quantities to account for the vascular effects. Nevertheless, to exclude the possibility that undetected changes in superoxide or its derivatives contribute to the coronary vasodilator action of BK, we determined the effect of various scavengers of superoxide on the dilator response in the perfused heart. The combination of enzymes, SOD and catalase, scavengers of superoxide and hydrogen peroxide, respectively, were without effect on the vasodilator response to BK but abolished that to hydrogen peroxide, showing that effective concentrations were used. These results, which tend to exclude superoxide and hydrogen peroxide as potential mediators of BK-induced dilation, should be interpreted with the knowledge that SOD and catalase do not readily cross cell membranes and may not access sites of intracellular superoxide production (Beckman et al., 1988). However, the lack of effect of these agents on BK-induced vasodilation is good evidence against an endothelium-derived reactive oxygen species that is released to exert its effect on the underlying smooth muscle. Furthermore, free radical scavengers that do penetrate to intracellular sites, Tiron and TEMPO, were also without effect on the coronary vasodilator action of BK. These results indicate that although P450-dependent metabolism of AA generates superoxide, it does not contribute to the vasodilator action of BK and is in agreement with those of Beny and von der Weid (1991), who excluded hydrogen peroxide as the hyperpolarizing factor mediating the vasodilator effect of BK.

In conclusion, we have demonstrated that the P450-dependent coronary vasodilator response to BK that requires the activation of phospholipases and Ca⁺⁺-activated K⁺ channels is not associated with the release of superoxide and is not affected by scavengers of reactive O₂ species. By exclusion, these observations further support the role of a P450-derived metabolite of AA, most probably an EET, as the putative hyperpolarizing factor that mediates the vasodilator effect of BK in the rat heart.