Author's email address C. E. Hill: email@example.com
Pathway-specific effects of calcitonin gene-related peptide on irideal arterioles of the rat
Article first published online: 29 SEP 2004
The Journal of Physiology
Volume 505, Issue 3, pages 797–809, December 1997
How to Cite
Hill, C. E. and Gould, D. J. (1997), Pathway-specific effects of calcitonin gene-related peptide on irideal arterioles of the rat. The Journal of Physiology, 505: 797–809. doi: 10.1111/j.1469-7793.1997.797ba.x
- Issue published online: 29 SEP 2004
- Article first published online: 29 SEP 2004
- Received 19 May 1997; accepted 11 August 1997.
- 1Arteriolar diameter and membrane voltage have been measured to investigate the actions of calcitonin gene-related peptide (CGRP) in rat irideal arterioles.
- 2Activation of sensory nerves inhibited sympathetic vasoconstriction, reduced the accompanying 40–50 mV depolarization by 90% and caused a 4 mV hyperpolarization.
- 3The inhibition of vasoconstriction was prevented by either preincubation in l-NAME (10 μm), to inhibit nitric oxide production, by preincubation in the cell-permeant adenylate cyclase inhibitor dideoxyadenosine (1 mm) or by preincubation in the ATP-sensitive potassium channel blocker glibenclamide (10 μM). The subsequent addition of a nitric oxide donor to the glibenclamide solution inhibited nerve-mediated vasoconstriction, suggesting that the potassium channel involvement preceded the production of nitric oxide. The small hyperpolarization was not affected by l-NAME.
- 4Nerve-mediated vasodilatation persisted in the presence of l-NAME (10 μm) but was abolished with the CGRP1 receptor antagonist CGRP8–37.
- 5In arterioles preconstricted with the α2-adrenoceptor agonist UK-14304 (100 nm), exogenous CGRP caused a hyperpolarization and a dose-dependent vasodilatation, neither of which was affected by l-NAME (10 μm).
- 6In arterioles preconstricted with 30 mm KCl, CGRP (10 nm) caused vasodilatation but not hyperpolarization, suggesting that the hyperpolarization was not causal to the vasodilatation.
- 7Forskolin (30 nm), in the presence of l-NAME to prevent effects due to nitric oxide, caused vasodilatation.
- 8These results suggest that CGRP inhibits sympathetic nerve-mediated vasoconstriction through sequential increases in cyclic AMP and nitric oxide, while vasodilatation results from increases in cyclic AMP alone. The production of nitric oxide, but not its mechanism of action, appears to be dependent on the activation of ATP-sensitive potassium channels. The possible sites of action of these two pathways are discussed.
Calcitonin gene-related peptide (CGRP) is a potent vasodilator released from sensory motor nerves (Brain, Williams, Tippins, Morris & Macintyre, 1985; Kawasaki, Takasaki, Saito & Goto, 1988). The pathway by which CGRP produces vasodilatation appears to vary between vascular beds. In the aorta (Fiscus et al. 1991; Gray & Marshall, 1992a), superior mesenteric artery (Bratveit, Haugan & Helle, 1991), gastric mucosa (Holzer, Wachter, Jocic & Heinemann, 1994) and venules, but not arterioles, of rat striated muscle (Kim, Roberts & Joshua, 1995) and in mouse pial arterioles (Rosenblum, Shimizu & Nelson, 1993), vasodilatory effects of CGRP are mediated via the endothelium and are due to the production of nitric oxide. In the majority of vessels studied to date, however, CGRP causes vasodilatation through an endothelium-independent process, not involving nitric oxide. These vessels include the rat mesenteric, gastric, splenic and renal arteries (Bratveit et al. 1991; Li & Duckies, 1992; Amerini, Mantelli & Ledda, 1993; Holzer, Lippe, Jocic, Wachter, Erb & Heinemann, 1993; Gao, Nishimura, Suzuki & Yoshida, 1994), rabbit jejunal (La & Rand, 1993), rat and rabbit hepatic artery (Bratveit & Helle, 1991; Brizzolara & Burnstock, 1991), skin and skeletal muscle microcirculations (Persson, Hedqvist & Gustafsson, 1991; Ralevic, Khalil, Dusting & Helme, 1992; Brain, Hughes, Cambridge & O'Driscoll, 1993) dog basilar (Oyama et al. 1993), cat cerebral (Saito, Masaki, Uchiyama, Lee & Goto, 1989), dog lingual (Kobayashi, Todoki, Ozono & Okabe, 1995) and human uterine arteries (Bodelsson & Stjernquist, 1992).
In cat cerebral, rabbit mesenteric and ophthalmic arteries, CGRP has been shown to produce hyperpolarization (Saito et al. 1989; Nelson, Huang, Brayden, Hescheler & Standen, 1990; Zschauer, Uusitalo & Brayden, 1992) due to the activation of ATP-dependent potassium (KATP) channels (Nelson et al. 1990; Zschauer et al. 1992). This is secondary to receptor-activated increases in cyclic AMP and protein kinase A (Quayle, Bonev, Brayden & Nelson, 1994; Miyoshi & Nakaya, 1995). The role of this hyperpolarization in the vasodilatory action of CGRP, however, varies between vascular beds, being causal to part or all of the dilatation in rabbit mesenteric arteries, rat basilar and dog lingual artery (Nelson et al. 1990; Kitazono, Heistad & Faraci, 1993; Kobayashi et al. 1995) and of no consequence to the dilatation in other vessels, such as the small mesenteric and renal arteries of the rat (Gao et al. 1994; Lei, Mulvany & Nyborg, 1994; Gao, Nishimura, Suzuki & Nakai, 1995) and the rabbit ophthalmic artery (Zschauer et al. 1992).
In some arteries, such as those of the mesenteric bed, vasoconstriction following nerve stimulation occurs as a result of membrane depolarization and calcium entry through voltage-dependent calcium channels (Hill, Hirst & van Helden, 1983; see Hirst & Edwards, 1989). In these vessels, CGRP released from sensory nerves has been shown to inhibit vasoconstriction (Han, Naes & Westfall, 1990; Li & Duckies, 1992). Since sensory nerves also cause a hyperpolarization and a reduction in the size of excitatory junctions potentials in this vessel (Meehan, Hottenstein & Kreulen, 1991), it is possible that the antagonism of sympathetic vasoconstriction could result from decreases in the influx of calcium through voltage-dependent calcium channels.
Nerve-mediated vasoconstrictions in irideal arterioles result from the activation by noradrenaline of α1B-adrenoceptors and release of intracellular calcium (Gould & Hill, 1994). These constrictions are independent of membrane depolarization, since they are not affected by blockade of voltage-dependent calcium channels. We have previously shown that CGRP released from sensory nerves causes an inhibition of sympathetic nerve-mediated vasoconstriction in irideal arterioles following the activation of CGRP1 receptors and release of nitric oxide (Hill & Gould, 1995). In the rat aorta, CGRP causes sequential increases in cyclic AMP and cyclic GMP (Gray & Marshall, 1992b), the downstream effector of nitric oxide. We were therefore interested to determine whether a similar pathway is activated in irideal arterioles by sensory nerves and whether hyperpolarization plays any role in the effect. During the course of our previous studies, we noted that CGRP could produce a dilatation of iris arterioles. In this study, we have also investigated the mechanism of this vasodilatation, in particular, whether it occurred via nitric oxide formation and/or membrane hyperpolarization or via cyclic AMP. A preliminary report of these findings has appeared (Hill & Gould, 1997).
All experiments were performed on Wistar rats aged 17–21 days postnatal. Animals were killed with an overdose of ether anaesthetic and both eyes were removed. The iris was dissected from the eye, cut in half and pinned flat in a recording chamber that was perfused with Krebs solution containing (mm): NaCl, 119.8; KCl, 5.0; NaHCO3, 25; NaH2PO4, 1.0; CaCl2, 2.5; MgCl2, 2.0; glucose, 22.0; gassed with 95% O2, 5% CO2 and warmed to 33 °C. Scopolamine hydrochloride (1 μM) was added to the superfusion solution to prevent the action of cholinergic nerves. Due to their small size, arterioles within the iris were not perfused. Preparations were equilibrated for 30 min before commencing an experiment.
Nerves in the iris preparations were stimulated at 10 Hz (pulses of 0.1 ms, 100 mA), via platinum electrodes positioned on either side of the preparation. Voltage was monitored during nerve stimulation to ensure that supramaximal voltage was used and hence changes in nerve-mediated responses could not be attributed to changes in voltage. Arteriolar diameter was measured using video microscopy and the computer program DIAMTRAK (T. Neild, Flinders University, Adelaide, Australia; Neild, 1989), which uses averaging techniques to locate the edges of the arterioles and record their difference as arteriolar diameter. Thus both nerve-evoked changes in diameter, as well as those produced by the addition of drugs, could be measured. Intracellular recordings from smooth muscle cells in the arteriolar walls were made using fine borosilicate glass microelectrodes, filled with 0.5 M KCl, and having resistances of 140–220 MΩ (Flaming Brown micropipette puller, Sutter Instrument Co.). Membrane potential was measured with an Axoclamp-2B (Axon Instruments). All membrane potential records were low-pass filtered, with a cut-off frequency of 1 kHz. Changes in vessel diameter in the region where the cell was impaled were recorded and stored on computer disk for analysis.
The following drugs were purchased commercially: benextramine tetrachloride, (–)-hyoscine (scopolamine) hydrochloride, forskolin (Sigma Chemical Co.), rat CGRP1–37, rat CGRP8–37, l-NAME, d-NAME hydrochloride, ±-S-nitroso-N-acetylpenicillamine (SNAP), 2’,5’-dideoxyadenosine (Sapphire Bioscience Pty Ltd, Alexandria, Australia), glibenclamide (ICN Biomedicals Inc.) and 5-bromo-6-(2-imidazolin-2-ylamino) quinoxaline (UK-14304, Pfizer Central Research, Sandwich, UK). Stock solutions were made at 1000 times or 10000 times working dilutions in distilled water, except for forskolin and glibenclamide, which were made up in absolute alcohol and DMSO, respectively. Control experiments using Krebs solution containing appropriate dilutions of absolute alcohol or DMSO showed that there were no effects of the diluent on either the contraction or the depolarization.
Drugs were superfused for 10–20 min. l-NAME or d-NAME were superfused for 20 min or, in the case of l-NAME, until the response to nerve stimulation every 15s did not diminish with repeated stimuli, before adding any other drug. Throughout the experiments, the responses to nerve stimulation were measured every 3 min (10 Hz, 1 s). These parameters have previously been found to produce consistent contractions in control solutions (Hill & Gould, 1995). In experiments designed to test the effect of sensory nerves, the preparations were stimulated at 10 Hz for 1 s every 15 s, since we have shown previously that, under these stimulation conditions, the inhibitory effects of sensory nerves can be best demonstrated (Hill & Gould, 1995).
In experiments using the irreversible α-adrenoceptor antagonist benextramine, the drug was superfused for 10 min before being replaced with control Krebs solution in order to prevent non-specific effects. In experiments where contractions were induced with Krebs solution containing 30 mm KCl, the NaCl concentration was reduced by 25 mm and the preparations were preincubated in benextramine to prevent the effects of noradrenaline released from nerve terminals by the high potassium solution.
To test the effects of CGRP as a vasodilator, arterioles were pre-constricted with either the α2-adrenoceptor agonist UK-14304 (100 nm) or 30 mm KCl for 5 min before the addition of the CGRP. This was necessary as the majority of arterioles did not develop any spontaneous tone, thus making vasodilatory effects too small to study significantly. The effect of the vasoconstrictors, in particular UK-14304, diminished slightly with time and consequently time control experiments were performed in the presence of the vasoconstrictor alone. Control values may thus appear as small vasodilatations since they represent the constricted vessel diameter at two progressively longer times after the addition of the vasoconstrictor agent. These times corresponded to the time when the maximal effect of the vasodilator would have been recorded, expressed relative to the time when it would have first been added to the bath.
Only one experiment was conducted on each preparation and each experiment was repeated at least 3 times in preparations from different animals. Thus in all cases, n equals the number of animals. Results are expressed as the mean ±s.e.m. Statistical significance was tested using Student's paired and unpaired t tests. If multiple comparisons were made, an analysis of variance with 95% confidence limits was performed to detect a difference between treatment groups, followed by Student's t test with Bonferroni correction for multiple groups. A probability, P, of less than 0.05 was taken as statistically significant.
General observations of nerve-mediated responses
Transmural nerve stimulation (10 Hz, 1 s) every 3 min produced a consistent constriction of irideal arterioles and this response was entirely blocked by the α-adrenoceptor antagonist benextramine, as reported previously (Gould & Hill, 1994). In the majority of preparations, the arterioles did not develop any spontaneous tone and so nerve-mediated vasodilatation was not observed.
Some preparations, however, did develop some tone, possibly because of a larger volume of blood caught inside them during the dissection and possibly therefore a higher internal pressure. In these preparations, repetitive nerve stimulation (10Hz, Is) at intervals of 15s for 90s, produced a long, slow vasodilatation after the initial vasoconstriction. This vasodilatation resulted from the nerve-stimulated release of CGRP since it was completely abolished following the addition of the CGRP1 receptor antagonist CGRP8–37(1 μm, Fig.1).
Nerve-mediated vasodilatation was more obvious in preparations that had been pretreated with l-NAME (10 μM) to prevent effects due to nitric oxide. As reported previously (Hill & Gould, 1995), incubation in l-NAME caused a vasoconstriction such that the arterial diameter decreased by some 20%. These observations suggest that the near-maximal relaxation of irideal vessels in the present experiments results from the tonic release of nitric oxide from the endothelium. In preparations pretreated with l-NAME, both single and repetitive nerve stimulation produced vasodilatation (Fig. 2). These results demonstrate that the nerve-mediated vasodilatation itself does not result from the release of nitric oxide. Subsequent addition of the CGRP1 receptor antagonist CGRP8–37 (1 μM) to the l-NAME, however, did completely abolish the nerve-mediated vasodilatation (n= 4, Fig. 2) further confirming that the nerve-mediated effects were due to the release of CGRP.
Intracellular recordings from irideal arterioles showed that the resting membrane potential (RMP) was –65 ± 1.3 mV (n= 10) when preparations were superfused with control Krebs solution.
CGRP and inhibition of sympathetic nerve-mediated vasoconstriction
When perivascular nerves were repetitively stimulated (10 Hz, 1 s) at intervals of 15 s for 90 s, the sympathetic nerve-mediated constriction of iris arterioles was rapidly attenuated (Fig. 3; see also Hill & Gould, 1995). We have previously shown that this effect was due to activation of sensory nerves and the release of CGRP since it was prevented by pretreatment with capsaicin or CGRP8–37 (Hill & Gould, 1995).
Simultaneous intracellular recordings demonstrated that the nerve-mediated contraction was accompanied by a 40–50 mV depolarization of the cell membrane and that the reduction in contractile responses following repetitive nerve stimulation was accompanied by a 92 % decrease in the size of the depolarizations (Figs 3 and 4). The large decrease in amplitude of the intracellular potentials was accompanied by a 4 mV hyperpolarization of the cell membrane (Fig. 3; 3.8 ± 0.7 mV, n= 4).
Effect of nitric oxide.
The loss of the contraction and depolarization was completely prevented by preincubation in l-NAME (10 μM), but not d-NAME (10 μm, Fig. 4), as described previously (Hill & Gould, 1995), suggesting that the inhibition of sympathetic vasoconstriction was due to the release of nitric oxide. l-NAME did not, however, prevent the small hyperpolarization of the cell membrane seen after repetitive nerve stimulation (4.6 ± 0.4 mV, n= 5, for l-NAME and 4.1 ± 0.4 mV, n= 4, for d-NAME). These results suggest that the hyperpolarization is not due to nitric oxide and is not responsible for the loss of vasoconstriction during repetitive nerve stimulation.
l-NAME had no effect on RMP (–67 ± 2.4 mV, n= 7 for l-NAME and –65 ± 1.3 mV, n= 1 for d-NAME) suggesting that the amount of nitric oxide tonically released in these preparations does not produce any hyperpolarization.
Effect of glibenclamide.
Addition of the KATP channel blocker glibenclamide (10 μM) caused a depolarization of about 6 mV in RMP (–59 ± 1.3 mV, n= 5 compared with –65 ± 1.3 mV in control Krebs solution, n= 10) but had no effect on the resting vessel diameter (98 ± 2.6% of control, n= 7). The size of the nerve-mediated depolarization was decreased in the drug solution (Fig. 3, 41 ± 2 mV, compared with 49 ± 1.9 mV, n= 4), although the size of the nerve-mediated contraction was unchanged. These results suggest that, in the arteriolar smooth muscle cells, KATP channels are open at rest, since their closure decreases RMP. This drug-induced depolarization was, however, not sufficient by itself to open voltage-dependent calcium channels and produce a contraction.
Glibenclamide prevented the inhibition of sympathetic nerve-mediated vasoconstriction following repetitive nerve stimulation (Fig. 3). The size of the contraction and that of the depolarization remained constant with successive stimuli (Fig. 3). However, in the presence of glibenclamide, the addition of the nitric oxide donor SNAP (10 μM) abolished the nerve-mediated vasoconstriction and significantly reduced the depolarization (Fig. 5; 21 ± 5.9 mV, compared with 41 ± 2 mV, n= 4). Addition of SNAP also produced a hyperpolarization of 6 mV (5.6 ± 0.6 mV, n= 4). Together, these results suggest that glibenclamide does not prevent the actions of nitric oxide and hence nitric oxide effects are not mediated by KATP channels. The hyperpolarization produced by SNAP, but not observed during release of endogenous nitric oxide, may suggest a higher dependency of nitric oxide on membrane channels than of nitric oxide in inhibiting vasoconstriction.
Effect of dideoxyadenosine.
The role of cyclic AMP in the inhibition of sympathetic vasoconstriction by sensory nerves was assessed using the membrane-permeant adenylate cyclase inhibitor dideoxyadenosine (1 mm). As with incubation in l-NAME and glibenclamide, incubation in dideoxyadenosine prevented the inhibition of sympathetic nerve-mediated vasoconstriction during repetitive nerve stimulation (Fig. 6). The inhibition of vasoconstriction reappeared on washout into control Krebs solution. These results suggest that cyclic AMP, nitric oxide and KATP channels are all involved in the pathway leading to the inhibition of vasoconstriction following activation of sensory nerves and release of CGRP. While effects of dideoxyadenosine on nerve-mediated vasodilatation were difficult to assess due to the near absence of tone in these preparations, they were suggestive of a role for cyclic AMP. Repetitive nerve stimulation produced a vasodilatation of 3 ± 1.4% (n= 4) in control and 0 ± 0.5% (n= 4) in dideoxyadenosine.
CGRP and vasodilatation
Due to the near absence of intrinsic tone in the majority of preparations and hence the difficulty in recording nerve-mediated dilatations, it was necessary to preconstrict arterioles in order to study further the mechanism of action of CGRP as a vasodilator. We chose to use the α2-adrenoceptor agonist UK-14304 (100 nm), as it produced a long-lasting, submaximal vasoconstriction, which was 30% of that produced by 50 mm KCl (11 ± 1.7%, n= 5 compared with 37 ± 2%, n= 21, when expressed as a percentage of resting vessel diameter). Since α2-adrenoceptor agonists also have inhibitory presynaptic effects on neurotransmitter release from both sympathetic and sensory nerves, nerve-mediated responses were no longer measured and vasodilator effects of CGRP were measured following its application to the super fusion solution.
The onset of the contraction produced by UK-14304 was co-incident with the onset of an 8 mV depolarization (8.0 ± 1.9 mV, n= 5). This depolarization was of the same order of magnitude as that produced by glibenclamide; however, glibenclamide did not produce a contraction.
When added to arterioles preincubated in glibenclamide, UK-14304 produced an additional depolarization of 7 mV (7.0 ± 1.3 mV, n= 3), although the size of the contraction produced by UK-14304 in the presence of glibenclamide was not significantly different to that seen in the absence of glibenclamide (12 ± 5.8%, n= 3 when expressed as a percentage of resting vessel diameter). These results suggest that the contraction due to UK-14304 was independent of the activation of voltage-dependent calcium channels and that the membrane channels affected by UK-14304 were unlikely to be KATP channels, since UK-14304 produced a depolarization after their closure.
In arterioles preconstricted with UK-14304, CGRP caused a concentration-dependent vasodilatation (Fig. 7). The vasodilator effects of CGRP were long lasting and did not show any signs of desensitization. Although UK-14304 produced a long-lasting vasoconstriction, there was some small loss of tone with time of exposure to the agonist. This is represented by the apparent vasodilatation in the absence of CGRP (Fig. 7).
Lack of a role for nitric oxide in the vasodilator action of CGRP (10 nm).
As stated earlier, l-NAME, but not d-NAME, prevented the inhibition of sympathetic nerve-mediated vasoconstriction and the decrease in amplitude of the nerve-mediated depolarizations seen with repetitive nerve stimulation (Fig. 4). In the present and subsequent sections on the role of nitric oxide in the vasodilatory action of CGRP, preparations were tested in this way with repetitive nerve stimulation after 15–20 min in l-NAME, to confirm that nitric oxide was no longer being released before further drugs were added.
In preliminary experiments, CGRP relaxed vessels that had been preconstricted with l-NAME. Unlike l-NAME, which increased the tone of arterioles, d-NAME had no effect on tone. It was therefore necessary to preconstrict all vessels with UK-14304 (100 nm) before addition of CGRP, in order to be able to assess its vasodilatory effects in control, d-NAME-treated preparations. Due to the vasoconstrictor actions of l-NAME, the baseline level of constriction was greater in the presence of l-NAME and UK-14304 than in the presence of d-NAME and UK-14304, although the difference was not significant (P > 0.05).
CGRP (10 nm) reversed the vasoconstriction produced by UK-14304 in the presence of either l-NAME or d-NAME (Fig. 8). CGRP produced a hyperpolarization of 7.0 ± 0.8 mV (n= 4) in l-NAME and 7.9 ± 1.5 mV (n= 5) in the presence of d-NAME. When the dilatatation due to CGRP was expressed as a percentage of the UK-14304-constricted diameter, CGRP could be seen to produce a vasodilatation amounting to 12 ± 1.3% (n= 5) and 28 ± 8.4% (n= 4) in the presence of d-NAME and l-NAME, respectively, due to the larger constriction in the presence of l-NAME. These results provide evidence that the vasodilatation and the hyperpolarization produced by CGRP, like those following the nerve-stimulated release of CGRP (see earlier), do not depend on the production of nitric oxide.
Lack of a role for hyperpolarization in the vasodilator action of CGRP.
To determine whether the vasodilatation produced by CGRP was dependent on the membrane hyperpolarization that accompanied it, arterioles were exposed to 30 mm KCl in the superfusion solution. This increase in the external concentration of K+ produced a large depolarization of more than 30 mV (RMP in 30 mm KCl –30 ± 0.6 mV, n= 3). Under these conditions, CGRP did not produce a hyperpolarization.
In 30 mm KCl, the equilibrium potential for K+ ions (Ek), calculated from the Nernst equation and assuming that the intracellular K+ concentration is 130 mm, is more negative (–39 mV) than that recorded here. This suggests that other ions, such as sodium, calcium and chloride, which have more positive equilibrium potentials (ENa=+50 mV, ECa > +150 mV, Ecl=–20 mV, Hirst & Edwards, 1989) must be contributing to the RMP under these conditions.
When the smooth muscle membrane had depolarized by 31 ± 1.5 mV (n= 3), the arterioles began to constrict. The final constriction was 70 % of that produced by 50 mm KCl (26 ± 3.8%, n= 9, when expressed as a percentage of resting vessel diameter).
Some preparations were pretreated with l-NAME to prevent any possible nitric oxide-mediated vasodilator effects of CGRP uncovered when the arterioles were preconstricted with KCl. The vasoconstriction in the presence of l-NAME and KCl was greater than that in d-NAME and KCl due to the vasoconstrictor actions of l-NAME (Fig. 9). Assessment of the efficacy of the l-NAME treatment was again determined by repetitive nerve stimulation, as in Fig. 4.
In the presence of d-NAME or l-NAME and KCl, CGRP (10 nm) produced a significant vasodilatation (Fig. 9, P < 0.05, paired t test). This vasodilatation was therefore recorded under conditions where CGRP could not produce a hyperpolarization. When this dilatation was expressed as a percentage of the KCl-constricted diameter, the dilatation amounted to 14 ± 2.5% (n= 5) and 19 ± 5% (n= 3) in the presence of d-NAME and l-NAME, respectively. These results suggest that the vasodilatory actions of CGRP are not due to hyperpolarization of the cell membrane.
A role for cyclic AMP in vasodilatation.
The ability of increases in cyclic AMP to produce vasodilatation directly were tested using forskolin, which stimulates adenylate cyclase. Forskolin (30 nm) caused a hyperpolarization of about 11 mV (RMP, –76 ± 0.3 mV; n= 3) and abolished the nerve-mediated vasoconstriction and associated depolarization in the presence of d-NAME (10 μm; Figs 10 and 11A). To prevent possible effects of cyclic AMP being due to the sequential activation of nitric oxide as previously described (Gould, Vidovic & Hill, 1995), preparations were pretreated with l-NAME. Under these conditions, forskolin (30 nm) did not abolish the nerve-mediated contraction or depolarization (74 ± 7.7% of contraction in control Krebs solution, n= 4, and 85 ± 12.1% of depolarization in control Krebs solution, n= 3; Figs 10 and 11A) since these latter effects were indirect and due to the release of nitric oxide. Forskolin (30 nm) added to the l-NAME did, however, cause a significant vasodilatation (P < 0.05, paired t test; Fig. 11B and C, n= 4) and hyperpolarization (RMP, –72 ± 2.3; n= 3). The persistence of the vasodilatation when l-NAME was present with the forskolin demonstrates that cyclic AMP can cause vasodilatation independently of nitric oxide.
CGRP released from sensory nerves had two distinct effects on iris arterioles: it inhibited nerve-mediated vasoconstriction and it caused vasodilatation. These two effects occurred through different intracellular pathways. The inhibition by sensory nerves of nerve-mediated constriction of irideal arterioles was prevented by either inhibition of nitric oxide synthase or inhibition of adenylate cyclase, suggesting the sequential involvement of the two enzymes. In a previous study we showed that increases in intracellular cyclic AMP also inhibited nerve-mediated constrictions of irideal arterioles, but that this could be prevented when nitric oxide synthesis was inhibited (Gould, Vidovic & Hill, 1995). Taken together, these results suggest that, as in the rat aorta (Gray & Marshall, 1992b), CGRP activates receptors leading to sequential stimulation of adenylate cyclase, to increase cyclic AMP, and nitric oxide synthase, to increase nitric oxide. The absence of nitric oxide synthase in sympathetic nerves in the iris (C. E. Hill, unpublished observations) suggests that the inhibition of sympathetic vasoconstriction occurs postsynaptically and not pre-synaptically. The absence of constitutive nitric oxide synthase in smooth muscle cells points to the involvement of endothelial cells in these events.
Under resting conditions, iris arterioles exhibited little or no tone. Arterioles in which nitric oxide synthesis was inhibited demonstrated a significant contraction implicating the tonic release of nitric oxide in the relaxed state of the vessels. Contraction was also seen in these vessels after exposure to 30 mm KCl. In this case the contraction did not commence until the membrane of the smooth muscle cells had been depolarized by some 30 mV. Contractions have been reported to be initiated by smaller depolarizations than this in other blood vessels (see Hirst & Edwards, 1989), suggesting that differences may exist between the voltage dependency of calcium channels in iris arterioles and other vascular beds. On the other hand, the α2-adrenoceptor agonist UK-14304 caused a contraction which was co-incident with a depolarization of only 8 mV. The KATP channel blocker glibenclamide caused a similarly sized depolarization but failed to cause a contraction. These results support the conclusions that a very large depolarization is required to open sufficient voltage-dependent calcium channels in iris arterioles to initiate a contraction, that contraction due to UK-14304 does not result from calcium entering through voltage-dependent calcium channels, that KATP channels contribute to RMP and that their closure per se does not produce contraction.
In addition to cyclic AMP and nitric oxide, KATP channels were involved in the sensory nerve-mediated inhibition of vasoconstriction since glibenclamide also prevented the phenomenon. While these results may suggest that the mechanism of nitric oxide action involves hyperpolarization, several pieces of data suggest that this is not the case. Inhibition of nitric oxide synthesis led to a pronounced vasoconstriction but this was not accompanied by any change in RMP. Furthermore, a small hyperpolarization of the cell membrane was observed to accompany the inhibition of vasoconstriction but this hyperpolarization persisted after inhibition of nitric oxide synthesis. Finally, addition of a nitric oxide donor, SNAP, in the presence of glibenclamide caused an immediate loss of nerve-mediated vasoconstrictions indicating that the effect of glibenclamide was upstream of the production of nitric oxide. Surprisingly, SNAP did produce a small hyperpolarization. It is possible, then, that nitric oxide can produce membrane hyperpolarization in smooth muscle cells of iris arterioles, but only when present at higher concentrations than those achieved following nerve stimulation or during the tonic release of nitric oxide. Nitric oxide may act to inhibit nerve-mediated vasoconstriction by increasing cyclic CMP, due to the stimulatory effect of nitric oxide on guanylate cyclase, and interfering with the release of calcium from internal stores as has been described in other systems (Chen & Rembold, 1992; McDaniel, Chen, Singer, Murphy & Rembold, 1992; Murthy, Seven, Grider & Makhlouf, 1993; Murthy & Makhlouf, 1995). This mechanism may well operate in iris arterioles since the calcium required for contraction is derived from intracellular stores (Gould & Hill, 1994).
Recent studies have shown that KATP channels are activated by sequential increases in cyclic AMP and protein kinase A and inhibited by increases in ATP (Quayle et al. 1994; Miyoshi & Nakaya, 1995). Activation of CGRP1 receptors could lead to sequential increases in cyclic AMP, opening of KATP channels in endothelial cells, hyperpolarization and increases in nitric oxide production. What, then, is the link between KATP channels and nitric oxide production? The availability and transport of the nitric oxide substrate l-arginine appears to modulate agonist-stimulated nitric oxide production, in spite of apparently saturating intracellular concentrations (Palmer, Bees, Ashton & Moncada, 1988; Aisaka, Gross, Griffith & Levi, 1989; see Bogle, Baydoun, Pearson & Mann, 1996 for discussion). In addition, there is a linear relationship between cell membrane potential and influx of l-arginine over the potential range of –23 to –90 mV (Bussolati et al. 1989), such that depolarization can decrease agonist-stimulated transport of l-arginine and cyclic GMP levels (Bogle et al. 1996). Taking these points into consideration, we might propose that the hyperpolarization resulting from the opening of the KATP channels leads to increased nitric oxide production through an increase in l-arginine transport. The failure to observe any significant hyperpolarization, which might correspond to the potassium channel opening, the evidence that KATP channels are already open at rest in the smooth muscle cells and the involvement of endothelial cells in the production of nitric oxide suggests that the smooth muscle cells and the endothelial cells are not electrically coupled in this vessel.
The present study has demonstrated that exogenous CGRP can cause a dose-dependent vasodilatation of preconstricted rat irideal arterioles. In some arterioles with higher resting tone, stimulation of sensory nerves also caused vasodilatation via the activation of CGRP1 receptors. Both of these vasodilatations were independent of nitric oxide production. These results provide further evidence that, in the majority of blood vessels, the vasodilatory effects of CGRP are not dependent on the endothelium and the production of nitric oxide.
Coincident with the vasodilatation, CGRP caused a small hyperpolarization of about 8 mV, which was still recorded when nitric oxide production was inhibited. In the present study, increases in intracellular cyclic AMP, using forskolin, also caused hyperpolarizations of about 10 mV. Since hyperpolarization has been shown to be causal to CGRP-induced vasodilatation in some vessels (Nelson et al. 1990; Kitazono et al. 1993; Kobayashi et al. 1995), we tested the effects of CGRP after preconstriction with 30 him KCl, in which solution CGRP did not produce any hyperpolarization. Under these conditions, CGRP still produced a vasodilatation, suggesting that in iris arterioles, the vasodilatation does not result from the hyperpolarization. Interestingly, in conscious rats, mean arterial pressure was attenuated by CGRP but this was not affected by glibenclamide (Abdelrahman, Wang, Chang & Pang, 1992), perhaps indicating that potassium channels do not contribute to the hypotensive effects of CGRP in the majority of small, resistance vessels.
Since the effects of CGRP on vasodilatation were not due to hyperpolarization, we tested the effects of increasing intra-cellular cyclic AMP using the adenylate cyclase activator forskolin. Since our previous studies had indicated that increases in cyclic AMP could lead to synthesis of nitric oxide (Gould et al. 1995), it was necessary to pretreat with l-NAME to eliminate such effects. Under these conditions, forskolin was able to cause a vasodilatation of irideal arterioles that could not be due to nitric oxide.
The present study has shown that activation of sensory nerves causes inhibition of nerve-mediated vasoconstriction through a series of steps comprising the activation of adenylate cyclase, the opening of KATP channels and the synthesis of nitric oxide. These results suggest that the receptors that mediate the inhibitory effects of CGRP on vasoconstriction are located on the endothelial cells. We further suggest that the nitric oxide diffuses back into the smooth muscle cells to inhibit intracellularly released calcium. In contrast, sensory nerve-mediated vasodilatation results from the action of CGRP at CGRP1 receptors on the smooth muscle cells. The vasodilatory effects of CGRP are due to increases in intracellular cyclic AMP and do not involve nitric oxide or changes in membrane potential, although hyperpolarization was recorded coincident with the dilatation. The lower sensitivity to CGRP of the inhibition of vasoconstriction compared with the vasodilatation could be due to the more restricted access of the neurally released CGRP to the endothelial, than the smooth muscle receptors. Since CGRP-induced vasodilatation was seen in arterioles preconstricted in a variety of ways, i.e. α2-adrenoceptor agonist, l-NAME or activation of voltage-dependent calcium channels, we suggest that the mechanism of its action is post-calcium release. This may involve a decrease in sensitivity of the contractile apparatus to calcium or a modification of the calcium extrusion system.
- 1992). Mechanism of the vasodilator action of calcitonin gene-related peptide in conscious rats. British Journal of Pharmacology 106, 45–48. , , & (
- L-Arginine availability determines the duration of acetylcholine-induced systemic vasodilatation in vivo. Biochemical and Biophysical Research Communications 163, 710–717. , , &
- 1993). Nitric oxide is not involved in the effects induced by non-adrenergic non-cholinergic stimulation and calcitonin gene-related peptide in the rat mesenteric vascular bed. Neuropeptides 25, 51–55.DOI: 10.1016/0143-4179(93)90068-L , & (
- 1992). Smooth muscle dilatation in the human uterine artery induced by substance P, vasoactive intestinal polypeptide, calcitonin gene-related peptide and atrial natriuretic peptide: relation to endothelium-derived relaxing substances. Human Reproduction 7, 1185–1188. & (
- 1996). Regulation of L-arginine transport and nitric oxide release in superfused porcine aortic endothelial cells. Journal of Physiology 490, 229–241. , , (
- 1993). The contribution of calcitonin gene-related peptide (CGRP) to neurogenic vasodilator responses. Agents and Actions 38, C19–21. , , & (
- 1985). Calcitonin gene-related peptide is a potent vasodilator. Nature 313, 54–56. , , , & (
- 1991). Effects of calcitonin gene-related peptide (CGRP) on regional haemodynamics and on selected hepato-splanchnic arteries from the rat: a comparison with VIP and atriopeptin II. Scandinavian Journal of Clinical and Laboratory Investigation 51, 167–174. , & (
- 1991). Vasodilation by calcitonin gene-related peptide (CGRP) and by transmural stimulation of the methoxamine-contracted rat hepatic artery after pretreatment with guanethidine. Scandinavian Journal of Clinical and Laboratory Investigation 51, 395–401. & (
- 1991). Endothelium-dependent and endothelium-independent vasodilatation of the hepatic artery of the rabbit. British Journal of Pharmacology 103, 1206–1212. & (
- 1989). Influx of l-arginine is an indicator of membrane potential in human fibroblasts. American Journal of Physiology 256, C930–935. , , , , , & (
- 1992). Cyclic nucleotide-dependent regulation of Mn2+ influx, [Ca2+]i, and arterial smooth muscle relaxation. American Journal of Physiology 263, C468–473. & (
- 1991). Calcitonin gene-related peptide (CGRP)-induced cyclic AMP, cyclic GMP and vasorelaxant responses in rat thoracic aorta are antagonized by blockers of endothelium-derived relaxant factor (EDRF). Neuropeptides 20, 133–143.DOI: 10.1016/0143-4179(91)90063-O , , , , , & (
- 1995). Calcitonin gene-related peptide-induced relaxation in isolated small superior mesenteric arteries from adult stroke-prone spontaneously hypertensive rats. Clinical and Experimental Pharmacology and Physiology 22, 8109–111. , , & (
- 1994). Relaxant effects of calcitonin gene-related peptide on isolated small renal arteries in stroke-prone spontaneously hypertensive rats. Journal of Smooth Muscle Research 30, 9–19. , , & (
- 1994). Alpha,B receptors and intracellular calcium mediate sympathetic nerve induced constriction of rat irideal blood vessels. Journal of the Autonomic Nervous System 50, 139–150.DOI: 10.1016/0165-1838(94)90004-3 & (
- 1995). Cross talk between receptors mediating contraction and relaxation in the arterioles but not the dilator muscle of the rat iris. British Journal of Pharmacology 115, 828–834. , & (
- 1992a). Nitric oxide synthesis inhibitors attenuate calcitonin gene-related peptide endothelium-dependent vasorelaxation in rat aorta. European Journal of Pharmacology 212, 37–42.DOI: 10.1016/0014-2999(92)90069-G & (
- 1992b). Human a-calcitonin gene-related peptide stimulates adenylate cyclase and guanylate cyclase and relaxes rat thoracic aorta by releasing nitric oxide. British Journal of Pharmacology 107, 691–696. & (
- 1990). Calcitonin gene-related peptide is the endogenous mediator of nonadrenergic-noncholinergic vasodilatation in rat mesentery. Journal of Pharmacology and Experimental Therapeutics 255, 423–428. , & (
- 1995). Modulation of sympathetic vasoconstriction by sensory nerves and nitric oxide in rat irideal arterioles. Journal of Pharmacology and Experimental Therapeutics 273, 918–926. & (
- 1997). Pathway specific effects of calcitonin gene-related peptide on rat irideal arterioles. Proceedings of the Australian Neuroscience Society 8, 58. & (
- 1983). Development of sympathetic innervation to proximal and distal arteries of the rat mesentery. Journal of Physiology 338, 129–147. , & (
- 1989). Sympathetic neuroeffector transmission in arteries and arterioles. Physiological Reviews 69, 546–604. & (
- 1993). Nitric oxide-dependent and -independent hyperaemia due to calcitonin gene-related peptide in the rat stomach. British Journal of Pharmacology 110, 404–410. , , , , & (
- 1994). Vascular bed-dependent roles of the peptide CGRP and nitric oxide in acid-evoked hyperaemia of the rat stomach. Journal of Physiology 480, 575–585. , , & (
- 1988). Calcitonin gene-related peptide acts as a novel vasodilator neurotransmitter in mesenteric resistance vessels of the rat. Nature 335, 164–167.DOI: 10.1038/335164a0 , , & (
- 1995). Differences in the capsaicin-induced dilation of arterioles and venules in rat striated muscle. Journal of Pharmacology and Experimental Therapeutics 273, 605–610. , & (
- 1993). Role of ATP-sensitive K+ channels in CGRP-induced dilatation of basilar artery in vivo. American Journal of Physiology 265, H581–585. , & (
- 1995). Calcitonin gene-related peptide mediated neurogenic vasorelaxation in the isolated canine lingual artery. Japanese Journal of Pharmacology 67, 329–339. , , & (
- 1993). Neurogenic vasodilatation in isolated perfused segments of rabbit jejunal artery. Clinical and Experimental Pharmacology and Physiology 20, 355–358. & (
- 1994). Characterization of the CGRP receptor and mechanisms of action in rat mesenteric small arteries. Pharmacology and Toxicology 74, 130–135. , & (
- 1992). Effect of endothelium on the actions of sympathetic and sensory nerves in the perfused rat mesentery. European Journal of Pharmacology 210, 23–30.DOI: 10.1016/0014-2999(92)90647-M & (
- 1992). Nitrovasodilators relax arterial smooth muscle by decreasing [Ca2+]i and uncoupling stress from myosin phosphorylation. American Journal of Physiology 263, C461–467. , , , & (
- 1991). Capsaicin-sensitive nerves mediate inhibitory junction potentials and dilatation in guinea-pig mesenteric artery. Journal of Physiology 443, 161–174. , & (
- 1995). Calcitonin gene-related peptide activates the K+ channels of vascular smooth muscle cells via adenylate cyclase. Basic Research in Cardiology 90, 332–336. & (
- 1995). Interaction of cA-kinase and cG-kinase in mediating relaxation of dispersed smooth muscle cells. American Journal of Physiology 268, C171–180. & (
- 1993). Inhibition of IP3 and IP3-dependent Ca2+ mobilization by cyclic nucleotides in isolated gastric muscle cells. American Journal of Physiology 264, G967–974. , , & (
- 1989). Measurement of arteriole diameter changes by analysis of television images. Blood Vessels 26, 48–52. (
- 1990). Arterial dilations in response to calcitonin gene-related peptide involve activation of K+ channels. Nature 344, 770–773.DOI: 10.1038/344770a0 , , , & (
- 1993). Role of nitric oxide in the cerebral vasodilatory responses to vasopressin and oxytocin in dogs. Journal of Cerebral Blood Flow and Metabolism 13, 285–290. , , , , , & (
- 1988). L-Arginine is the physiological precursor for the formation of nitric oxide in endothelium-dependent relaxation. Biochemical and Biophysical Research Communications 153, 1251–1256. , , & (
- 1991). Nerve-induced tachykinin-mediated vasodilatation in skeletal muscle is dependent on nitric oxide formation. European Journal of Pharmacology 205, 295–301.DOI: 10.1016/0014-2999(91)90913-B , & (
- 1994). Calcitonin gene-related peptide activated ATP-sensitive K+ currents in rabbit arterial smooth muscle via protein kinase A. Journal of Physiology 475, 9–13. , , & (
- 1992). Nitric oxide and sensory nerves are involved in the vasodilator response to acetylcholine but not calcitonin gene-related peptide in rat skin microvasculature. British Journal of Pharmacology 106, 650–655. , , & (
- 1993). Endothelium-dependent effects of substance P and calcitonin gene-related peptide on mouse pial arterioles. Stroke 24, 1043–1047. , & (
- 1989). Calcitonin gene-related peptide and vasodilator nerves in large cerebral arteries of cats. Journal of Pharmacology and Experimental Therapeutics 248, 455–462. , , , & (
- 1992). Role of endothelium and hyperpolarization in CGRP-induced vasodilatation of rabbit ophthalmic artery. American Journal of Physiology 263, H359–365. , & (