Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Activation of vascular smooth muscle ATP-sensitive K⁺ channels (K_ATP) either by endogenous factors, such as vasoactive intestinal peptide (Standen et al., 1989) and adenosine (Kleppisch and Nelson, 1995), or K⁺ channel-activating agents, such as levcromakalim (the active enantiomer of cromakalim) and pinacidil, causes hyperpolarization and relaxation of vascular smooth muscle. We have recently provided evidence that there may be important interactions between vasodilator pathways such as K_ATP channels and, for example, the cyclic nucleotide systems. Activation of K_ATP channels clearly does not contribute to the relaxant effects of either nitric oxide, cyclic GMP (White and Hiley, 1998a), or cyclic AMP (Omar et al., 2000) in rat mesenteric arteries. Hence, although previous studies have shown that these agents cause membrane hyperpolarization through activation of K_ATP in this artery (Garland and McPherson, 1992; Prieto et al., 1997), this does not contribute directly to their effects, but rather is revealed as a "silent" modulatory effect on the actions of K_ATP channel-activating agents such as levcromakalim (White and Hiley, 1998a; Omar et al., 2000).

We demonstrated previously that activators of the cyclic AMP system potentiated K⁺ channel activator-induced relaxation through activation of protein kinase A (Omar et al., 2000), consistent with previous findings (Linde and Quast, 1995; Kessler et al., 1997). The aim of the present study was to investigate whether these actions are shared by -adrenoceptor agonists, which also stimulate cyclic AMP synthesis through G_s protein-coupled receptors (both ₁ and ₂ subtypes; Zwaveling et al., 1996). Indeed, isoproterenol causes relaxation of mesenteric arteries (Heesen and De Mey, 1990; Graves and Poston, 1993) and hyperpolarizes rat mesenteric arteries through activation of K_ATP (Fujii et al., 1999). Randall and McCulloch (1995) also showed that activation of K_ATP may play a minor role in -adrenoceptor-induced relaxation of the rat perfused mesenteric bed.

The interactions between isoproterenol and levcromakalim were examined using protocols described previously (White and Hiley, 1998a,b). We show that isoproterenol potentiates levcromakalim-induced relaxation by ₁-adrenoceptor-mediated activation of protein kinase A; however, K_ATP channels do not appear to contribute to the relaxant effects of isoproterenol. Surprisingly, the relaxant effects of isoproterenol were not affected by protein kinase A inhibition by either Rp-cAMPS or H-89. However, an inhibitory effect of Rp-cAMPS was observed in the presence of BK_Ca inhibition by iberiotoxin. We conclude that a major mechanism for isoproterenol relaxation is protein kinase A-independent activation of BK_Ca. However, protein kinase A-dependent mechanisms are unmasked under conditions of BK_Ca inhibition.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Myograph Mounting of Arteries. Male Wistar rats (250-350 g; Tucks, Rayleigh, Essex, UK) were killed with an overdose of sodium pentobarbital (120 mg kg¹ i.p.; Sagatal, Rhône Mérieux, Harlow, Essex, UK). The mesentery was removed and placed in ice-cold, gassed (95% O₂, 5% CO₂) Krebs-Henseleit solution of the following composition: 118 mM NaCl, 4.7 mM KCl, 1.2 mM MgSO₄, 1.2 mM KH₂PO₄, 25 mM NaHCO₃, 2.5 mM CaCl₂, 10 mM D-glucose. Segments (2 mm in length) of third order branches of the superior mesenteric artery were removed and mounted in a Mulvany-Halpern myograph (model 500A; Danish Myo-technology, Aarhus, Denmark) as described in White and Hiley (1997a). Vessels were maintained at 37°C in Krebs-Henseleit solution, containing indomethacin (10 µM), and bubbled with 95% O₂, 5% CO₂. After equilibration, vessels were normalized to a tension equivalent to that generated at 90% of the diameter of the vessel at 100 mm Hg (Mulvany and Halpern, 1977). The mean vessel diameter under these conditions was 338 ± 4 µm and the mean resting tension was 3.7 ± 0.1 mN (both n = 151).

Experimental Protocol. Concentration-response curves were established by precontracting vessels with methoxamine (10 µM; the mean tension generated was 14.1 ± 0.4 mN, n = 151) and then cumulatively adding increasing concentrations of the vasodilator agent under investigation. The vessels were then washed thoroughly. After generating a concentration-response curve to -adrenoceptor agonists, vessels were subsequently discarded to avoid possible effects due to incomplete washout. When used, iberiotoxin (50 or 100 nM), glibenclamide (10 µM), Rp-cAMPS (50 or 100 µM), and H-89 (5 µM) were incubated with vessels for 30 min before construction of a concentration-response curve to the vasorelaxant under investigation.

Vasodilator Interaction Studies. The effects of isoproterenol on responses to levcromakalim were evaluated according to the "standard tone" protocol described previously (White and Hiley, 1998a,b). The vasorelaxant effect of levcromakalim was tested by precontracting vessels with methoxamine (10 µM) and then cumulatively adding levcromakalim. Previous studies have shown that consistent concentration-response curves to levcromakalim can be observed in a single preparation (White and Hiley, 1997b).

Data and Statistical Analysis. Relaxation responses in myograph experiments are expressed as the percentage of relaxation of the tone induced by methoxamine. Data are given as the mean ± S.E.M. EC₅₀ values for cumulative responses were obtained from individual concentration-response curves by fitting the data to the following logistic equation:

(1)

where E is the effect (reduction in tone), A the concentration of the agonist, E_max the maximum effect, n_H the slope function, and EC₅₀ the concentration of relaxant giving half the maximal relaxation. The iterative curve-fitting procedure was carried out using KaleidaGraph (Synergy Software, Reading, PA) running on a Macintosh computer. Raw data for individual concentration-response curves were compared by two-way ANOVA. The Bonferroni/Dunn post hoc test was used for determining significant differences between factors. Student's unpaired t test was used for statistical comparison where responses to a single concentration of drug only were obtained. P values less than 0.05 were considered statistically significant.

Drugs. Methoxamine, carbachol, isoproterenol, dobutamine, iberiotoxin, and atenolol (all from Sigma, Poole, Dorset, UK), and Rp-cAMPS (Sigma/RBI, Natick, MA) were dissolved in distilled water. Indomethacin (Sigma) was dissolved in 5% (w/v) NaHCO₃ solution. Levcromakalim (SmithKline Beecham, Betchworth, Surrey, UK) was dissolved in 100% ethanol. Forskolin (Sigma), H-89 (Calbiochem, Nottingham, UK), and glibenclamide (Aldrich, Gillingham, Dorset, UK) were dissolved in 100% dimethyl sulfoxide.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Effect of Isoproterenol on Levcromakalim-Induced Relaxation. An original recording from an interaction experiment is shown in Fig. 1A. Levcromakalim caused concentration-dependent relaxations of methoxamine-precontracted arteries (EC₅₀ = 0.45 ± 0.04 µM, E_max = 91 ± 4%, n = 10; Fig. 1B). The presence of a near EC₃₀ concentration of the nonselective -adrenoceptor agonist isoproterenol potentiated (P < 0.05) the relaxant effect of levcromakalim (EC₅₀ = 0.17 ± 0.03 µM, E_max = 88 ± 5%, n = 10; Fig. 1B). However, no significant further potentiation was observed when levcromakalim relaxations were established in the presence of a near EC₅₀ concentration of isoproterenol (levcromakalim EC₅₀ = 0.20 ± 0.03 µM, E_max = 101 ± 5%, n = 8; Fig. 1B). Parameters for the interaction protocol are given in Table 1.

View larger version (23K): [in this window] [in a new window]   Fig. 1.   A, original recording showing an interaction experiment examining the effect of isoproterenol (near EC50 concentration) on levcromakalim in methoxamine-precontracted rat mesenteric arteries in the presence of endothelium. Under control conditions levcromakalim induced concentration-dependent relaxations (top) that were potentiated in the presence of isoproterenol (bottom). Parameters for the protocol used are given in Table 1. Vertical lines denote addition of drugs at the concentrations indicated. B, concentration-response curves for levcromakalim-induced relaxation of methoxamine-induced tone in the absence or presence of either low (near EC30 concentration) or high concentrations (near EC50 concentration) of isoproterenol. Responses to levcromakalim in the presence of both lower (P < 0.05) and higher (P < 0.01) concentrations of isoproterenol were significantly different from the control curve (two-way ANOVA with the Bonferroni/Dunn post hoc test). Logistic curve fit parameters are given in the text. Control, n = 10; low dose isoproterenol, n = 10; high dose isoproterenol, n = 8.

View this table: [in this window] [in a new window]   TABLE 1 Parameters for experiments investigating the effect of -adrenoceptor agonists on levcromakalim-induced relaxation EC30 and EC50 indicate the concentrations of drug added to give 30 or 50% relaxation of methoxamine-induced tone, respectively. The concentrations added were titrated in individual vessels and the range of concentrations used is shown. Restored level of tone indicates the tone established after addition of the vasorelaxant by further addition of methoxamine. n indicates the number of animals studied.

Effects of Isoproterenol, Salbutamol, Dobutamine, and Atenolol on Rat Mesenteric Arteries. Isoproterenol caused concentration-dependent relaxation of methoxamine-precontracted mesenteric arteries (EC₅₀ = 35 ± 3 nM, E_max = 101 ± 3%, n = 4; Fig. 2A). The isoproterenol concentration-response curve was shifted rightward in parallel manner (P < 0.001) by factors of 3- and 16-fold in the presence of 1 and 10 µM atenolol (a ₁-selective adrenoceptor antagonist), respectively (Fig. 2A).

View larger version (15K): [in this window] [in a new window]   Fig. 2.   A, concentration-response curves for relaxation by isoproterenol of methoxamine-induced tone in endothelium intact rat mesenteric arteries in the absence or presence of either 1 or 10 µM atenolol. Responses to isoproterenol in the presence of 1 or 10 µM atenolol were significantly different from the control curve (P < 0.001; two-way ANOVA with the Bonferroni/Dunn post hoc test). B, concentration-response curves for relaxation by dobutamine and salbutamol in the absence or presence of 10 µM atenolol. Salbutamol responses in the absence or presence of atenolol were not significantly different. Logistic curve fit parameters are given in the text. C, original recording showing relaxation by dobutamine of methoxamine-precontracted rat mesenteric arteries in the presence of endothelium. Vertical lines denote addition of drugs at the concentrations indicated. Note that the relaxations induced by submaximal concentrations of dobutamine are poorly sustained. All data are n = 4 except for B, salbutamol, in the presence of atenolol, where n = 6.

Effect of Isoproterenol in Presence of Atenolol on Levcromakalim-Induced Relaxation of Mesenteric Arteries. The mean concentration of isoproterenol (20 nM) established as causing around 30% relaxation of methoxamine-induced tone evoked a significantly (P < 0.05) smaller relaxation in the presence of 10 µM atenolol (Table 1). Moreover, the ability of this concentration of isoproterenol to potentiate levcromakalim-induced relaxations was abolished in the presence of atenolol (Fig. 3). Since the relaxations to dobutamine were not sustained, it was not possible to determine its effects on levcromakalim-induced relaxations (data not shown).

View larger version (19K): [in this window] [in a new window]   Fig. 3.   Concentration-response curves for relaxation by levcromakalim of methoxamine-precontracted arteries in the absence or presence of a near EC30 concentration of isoproterenol and 10 µM atenolol. There was no significant difference between the two curves (two-way ANOVA). Data for the tone in the protocols are given in Table 1. n = 8 for both curves.

Effect of Protein Kinase A Inhibition on Actions of Isoproterenol. The presence of the protein kinase A inhibitor Rp-cAMPS (50 µM) did not significantly reduce the relaxation induced by the mean near EC₃₀ concentration of isoproterenol used previously (Table 1). In contrast, isoproterenol was no longer able to potentiate subsequent relaxations to levcromakalim in the presence of Rp-cAMPS (Fig. 4A).

View larger version (11K): [in this window] [in a new window]   Fig. 4.   A, concentration-response curves for relaxation by levcromakalim of methoxamine-precontracted arteries in the absence or presence of a low (near EC30) concentration of isoproterenol and 50 µM Rp-cAMPS. B, concentration-response curves for relaxation by isoproterenol in the absence or presence of the protein kinase A inhibitor Rp-cAMPS (50 or 100 µM). C, effect of Rp-cAMPS (50 µM) on forskolin-induced relaxation of methoxamine-precontracted arteries. Data for the tone in the protocols are given in Table 1. In A, n = 6 for both curves. There was no significant difference between the two curves (two-way ANOVA). In B, the control response is reproduced from Fig. 2A for clarity; with 50 µM Rp-cAMPS, n = 4; with 100 µM Rp-cAMPS, n = 3. Responses to isoproterenol in the presence of either concentration of Rp-cAMPS were not significantly different from the control curve. In C, n = 4 for both columns. *, indicates statistically significant difference (P < 0.05) compared with control values (Student's unpaired t test).

Effect of Glibenclamide on Isoproterenol-Induced Relaxation of Rat Mesenteric Arteries. Figure 5 shows that the relaxant effect of isoproterenol (control EC₅₀ = 39 ± 4 nM, E_max = 78 ± 2%, n = 8) was significantly (P < 0.01) enhanced by the presence of the K_ATP channel blocker glibenclamide (10 µM; isoproterenol EC₅₀ = 22 ± 4 nM, E_max = 95 ± 4%, n = 8). Glibenclamide caused no significant effect in the additional presence of either 50 or 100 µM Rp-cAMPS (Fig. 5).

View larger version (18K): [in this window] [in a new window]   Fig. 5.   Concentration-response curves for relaxation by isoproterenol in the absence or presence of the KATP channel inhibitor glibenclamide (10 µM) and Rp-cAMPS (50 or 100 µM). Control, n = 8; with glibenclamide, n = 8; in the presence of glibenclamide and Rp-cAMPS, n = 4 for both curves. Responses to isoproterenol in the presence of glibenclamide were significantly different (P < 0.001) from the control curve; responses in the presence of either concentration of Rp-cAMPS were not significantly different from the control curve (two-way ANOVA followed by Bonferroni/Dunn post hoc test). , control; , +10 µM glibenclamide; , +10 µM glibenclamide + 50 µM Rp-cAMPS; , +10 mM glibenclamide + 100 µM Rp-cAMPS.

Effect of Iberiotoxin, Alone or in Presence of Rp-cAMPS, H-89, or Glibenclamide, on Isoproterenol-Induced Relaxation. The presence of iberiotoxin (50 nM), an inhibitor of large conductance Ca²⁺-activated K⁺ channels (BK_Ca), significantly (P < 0.001) attenuated relaxation to isoproterenol (EC₅₀ = 71 ± 22 nM, E_max = 60 ± 5%, n = 4). This was a maximally effective concentration of iberiotoxin, since 100 nM gave no further inhibition (isoproterenol EC₅₀ = 72 ± 6 nM, E_max = 73 ± 2%, n = 4; data not shown).

View larger version (16K): [in this window] [in a new window]   Fig. 6.   Concentration-response curves for relaxation by isoproterenol in the presence of the BKCa inhibitor iberiotoxin (50 nM) and the protein kinase A inhibitors Rp-cAMPS (50 or 100 µM) (A) and H-89 (5 µM) (B). All curves are n = 4 except for isoproterenol in the presence of H-89 (n = 5) and isoproterenol in the presence of iberiotoxin and Rp-cAMPS (100 µM, n = 6). In A, all three experimental curves were significantly different from the control curve (P < 0.001). Moreover, responses to isoproterenol in the presence of iberiotoxin and 50 µM Rp-cAMPS (P < 0.01) or 100 µM Rp-cAMPS (P < 0.01) were significantly reduced compared with responses in the presence of iberiotoxin alone. In B, responses to isoproterenol in the presence of H-89 were not significantly different from the control curve. However, iberiotoxin significantly reduced responses (P < 0.001 compared with control curve), and the combination of iberiotoxin and H-89 caused significantly greater inhibition (P < 0.05) than iberiotoxin alone. All comparisons were carried out using a two-way ANOVA followed by the Bonferroni/Dunn post hoc test.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

This study shows that isoproterenol potentiates the relaxant effects of K_ATP channel activators such as levcromakalim in rat mesenteric arteries through a protein kinase A-dependent mechanism mediated by the activation of ₁-adrenoceptors. The lack of effect of glibenclamide and Rp-cAMPS/H-89 against isoproterenol-induced relaxations could logically suggest that neither activation of K_ATP channels nor protein kinase A is involved. Crucially, however, we have shown that this is not the case. Inhibition of BK_Ca by iberiotoxin attenuated the effects of isoproterenol, but also unmasked an inhibitory effect of Rp-cAMPS and H-89. This is evidence that there may be redundancy in vasodilator pathways, such that certain mechanisms may only become important when others are inhibited. These results also provide the first demonstration of the importance of direct G protein-mediated activation of BK_Ca in an intact blood vessel preparation.

The presence of a near EC₃₀ concentration of isoproterenol caused a leftward shift in the concentration-response to levcromakalim, similar to the effect of other cyclic AMP-modulating agents (Omar et al., 2000). Interestingly, a higher concentration of isoproterenol (causing approximately 50% relaxation of precontracted tone) exerted no additional potentiating effect. This suggests that the mechanism by which isoproterenol modulates K_ATP is maximally activated at concentrations of the drug where the relaxant mechanism clearly is not, which may indicate that the two involve distinct intracellular pathways. This contrasts with our previous findings that near EC₃₀ and near EC₅₀ concentrations of cyclic AMP or cyclic GMP modulating agents caused different effects (White and Hiley, 1998a; Omar et al., 2000). Unfortunately, technical considerations rendered it impossible for us to evaluate the concentration dependence of these effects further; lower concentrations of isoproterenol (EC_10-20) generally produced relaxations that reversed rapidly, hence it was not possible to evaluate their effects on a subsequent concentration-response curve to levcromakalim. On the other hand, higher concentrations of isoproterenol (EC_70-80) were unsuitable because further addition of methoxamine did not reliably return the vessels to the stable, standard tone level.

Previous studies have shown that rat small mesenteric arteries express both ₁- and ₂-adrenoceptors, with the former being the more important in causing vasorelaxation (Graves and Poston, 1993; Zwaveling et al., 1996). In the present study, atenolol caused parallel rightward shifts in the concentration-response curves to isoproterenol, with 10 µM atenolol abolishing the relaxation to isoproterenol at the concentration used to induce a near 50% relaxation in the levcromakalim interaction experiments. The selective ₁-agonist dobutamine was also found to be a highly potent relaxant of precontracted vessels, whereas the selective ₂-agonist salbutamol caused similar effects only at 1000-fold higher concentrations. Our observation that salbutamol-induced relaxations were insensitive to atenolol indicates that activating ₂-adrenoceptors can evoke relaxation in mesenteric arteries; however, it is clear that isoproterenol-induced relaxation normally occurs predominantly through the activation of ₁-adrenoceptors.

Atenolol, at a concentration of 10 µM, abolished the relaxant effect of the near EC₃₀ concentration of isoproterenol, and also inhibited the ability of isoproterenol to potentiate levcromakalim-induced relaxation. These findings clearly show that the modulatory action of isoproterenol on levcromakalim-induced relaxation also occurs through the activation of ₁-adrenoceptors. It is unfortunate that the short-lived vasorelaxant action of dobutamine meant that it was not possible to use it to confirm this conclusion.

Previous studies have generally assumed that -adrenoceptor-induced vasorelaxation is primarily mediated through cyclic AMP (Heesen and De Mey, 1990), produced by G_s protein stimulation of adenylyl cyclase. Somewhat surprisingly, therefore, the protein kinase A inhibitor Rp-cAMPS did not inhibit isoproterenol-induced relaxation in the present study. The effectiveness of the Rp-cAMPS was confirmed by our observation that 50 µM Rp-cAMPS significantly inhibited relaxations induced by the adenylyl cyclase activator forskolin, an effect that has also previously been demonstrated using only 25 µM Rp-cAMPS (McKinnon et al., 1996). Interestingly, although the relaxant effect of isoproterenol was unchanged by 50 µM Rp-cAMPS, the presence of the protein kinase A inhibitor did inhibit the ability of isoproterenol to potentiate levcromakalim-induced relaxations.

A structurally unrelated protein kinase A inhibitor, H-89, also only slightly inhibited isoproterenol-induced relaxation, even when used at a concentration (5 µM) more than 100-fold greater than its K_i for inhibition of protein kinase A (48 nM; Chijiwa et al., 1990). Evidence that H-89 was active at this concentration is provided by our observation that methoxamine-induced contractions were attenuated by this agent. This is most likely to be due to inhibitory effects on other protein kinases such as protein kinase C and myosin light chain kinase that are involved in the contractile mechanism and that are also inhibited by H-89 (Chijiwa et al., 1990).

Relaxations to isoproterenol were slightly potentiated by the K_ATP inhibitor glibenclamide, despite the fact that isoproterenol hyperpolarizes rat mesenteric arteries through activation of glibenclamide-sensitive K_ATP channels (Fujii et al., 1999). Indeed, glibenclamide exerted no inhibitory effect in any of the experiments performed in the present study. The lack of inhibitory effect of glibenclamide on relaxation to isoproterenol is in agreement with previous work by Huang and Kwok (1997). Randall and McCulloch (1995) reported that activation of K_ATP might contribute to -adrenoceptor-mediated vasorelaxation of the rat perfused mesenteric bed. However, in that study, a maximally effective concentration of glibenclamide (10 µM) exerted only a modest inhibitory effect.

These findings would appear to indicate that activation of glibenclamide-sensitive K_ATP channels is not involved in isoproterenol-induced relaxation to any large extent. The effects of the -adrenoceptor agonist therefore provide another example of the ability of cyclic nucleotide-modulating agents to influence the effects of K_ATP-activating agents through a silent mechanism, that is, one that does not itself cause vasorelaxation. We have previously shown this to be the case for nitric oxide donors and cyclic GMP (White and Hiley, 1998a) as well as other modulators of cyclic AMP (Omar et al., 2000).

We then addressed the possibility that isoproterenol might cause vasorelaxation through protein kinase A-independent mechanisms. This has been postulated previously (Huang and Kwok, 1997) with the most likely mechanism being direct G_S protein-mediated activation of BK_Ca by the -adrenoceptors, which has been demonstrated by electrophysiological studies in vascular and airway smooth muscle (Scornik et al., 1993; Kume et al., 1994). Indeed, in the present study, inhibition of BK_Ca with iberiotoxin attenuated isoproterenol-induced relaxations. Crucially, however, both Rp-cAMPS and H-89 were found to inhibit isoproterenol-induced relaxations in the presence of iberiotoxin at concentrations that were inactive in the absence of the BK_Ca inhibitor. However, glibenclamide caused no inhibition in the presence of either Rp-cAMPS or iberiotoxin.

It is also reasonable to conclude that the lack of sensitivity of isoproterenol relaxations to Rp-cAMPS and H-89 under normal conditions is due to a major mechanism for relaxation being protein kinase A-independent activation of BK_Ca. Only when this mechanism is inhibited by iberiotoxin is an underlying protein kinase A-dependent (and hence Rp-cAMPS- and H-89-sensitive) mechanism unmasked. This is interesting because previous electrophysiological studies have shown that cyclic AMP and -adrenoceptor stimulation can activate BK_Ca (Sadoshima et al., 1988; Song and Simard, 1995). The present findings are of importance to studies investigating vasodilator mechanisms, since it is clear that, where vasorelaxation occurs through multiple pathways, there may be redundancy between them. Hence, one mechanism may be inhibited without affecting the global response due to other mechanisms being able to compensate fully. It may therefore be incorrect to assume from a negative result obtained with a certain inhibitor or antagonist (e.g., Rp-cAMPS or H-89 in the present study) that the respective mechanism is not activated by the vasorelaxant in question (isoproterenol), since this mechanism may in fact be active, but masked by other vasorelaxant pathways (e.g., protein kinase A-independent activation of BK_Ca).

In summary, the present study has demonstrated that isoproterenol potentiates levcromakalim-induced relaxation through ₁-adrenoceptor-mediated activation of protein kinase A. However, isoproterenol-induced relaxation seems largely to involve protein kinase A-independent activation of BK_Ca under normal conditions, with protein kinase A-dependent pathways only being unmasked under conditions of BK_Ca blockade with iberiotoxin. To our knowledge this is the first time that activation of BK_Ca channels, presumably directly mediated by a G protein, has been shown to be functionally important in an intact blood vessel preparation.