Introduction

Top Abstract Introduction Methods Results Discussion References

Sensory nerves are involved in the control of vascular tone and appear to play an important role in certain types of vascular inflammation. The proinflammatory nature of neuropeptides released from sensory C-fibers has been demonstrated in several models of inflammation, and neurogenic components have been identified in inflammatory disease states, including rheumatoid arthritis (Levine et al., 1985) and asthma (Barnes, 1986).

Activation of sensory C-fibers may be achieved in several ways, including application of the agent capsaicin, the active ingredient of `chili pepper' (Jancso et al., 1967), which selectively activates sensory C-fibers (for reviews, see Holzer, 1991; Dray, 1992), or direct electrical stimulation of nerve fibers. C-fibers are located within the arteriolar side of the microcirculation (e.g., Fleming et al., 1989; Baluk et al., 1992), and activation of these nerves results in the release of the neuropeptides CGRP and SP (Holzer, 1991). It is generally assumed that C-fibers do not innervate the venous side of the microcirculation (e.g., McDonald, 1988; Baluk et al., 1992). Indeed, although there are studies demonstrating the innervation of larger veins, such as human saphenous vein (Herbst et al., 1992), there is no immunohistochemical evidence demonstrating SP or CGRP immunoreactivity in the walls of small veins. When arteriolar C-fibers are activated, CGRP is thought to act at the arteriolar level to produce vasodilatation, whereas SP travels downstream to increase venular permeability. Together, these effects produce greater net edema than does either alone. This synergism has been attributed to an increase in the hydrostatic pressure (due to selective arteriolar dilatation), coupled with increased movement of fluid out of the permeable venules.

Synergism between CGRP and SP has been demonstrated in the skin of rabbits and rats (Brain and Williams, 1985; Gamse and Saria, 1985). However, this has not been a universal finding, and other studies have demonstrated that CGRP has an inhibitory effect (Raud et al., 1991) or no effect (Ekblom et al., 1993) on the response to agents that increase venular permeability. These studies have all relied upon exploring the effects of the neuropeptides in vivo or in intact perfused vascular beds in vitro. In these integrated experimental systems, it is difficult to distinguish between direct effects of neurogenic stimulation on the component parts of the vasculature. To overcome this problem, we have undertaken studies of isolated microvessels in vitro. We previously demonstrated that C-fiber activation in small arteries relaxes constricted vessels and inhibits the contractile responses to sympathetic stimulation (Ahluwalia and Vallance, 1996a). We have now investigated the effects of neuropeptides and capsaicin on the reactivity of isolated small veins of the microcirculation of the rat mesentery. These results have been presented in preliminary form to The British Pharmacological Society (Ahluwalia and Vallance, 1996b).

	Methods

Top Abstract Introduction Methods Results Discussion References

Preparation of vessels. Male Wistar rats (240-280 g) were stunned and the neck dislocated. The mesentery was removed and placed in oxygenated cold (4°C) Krebs solution (composition, in mM: NaCl, 118; KCl, 4.69; CaCl₂, 2.5; KH₂PO₄, 1.18; MgSO₄, 1.18; NaHCO₃, 25; glucose, 11). Small second- or third-order veins were dissected out and mounted in cold Krebs' solution as ring preparations suspended between two stainless steel wires of 40-µm diameter, in an automated myograph (JP Trading, Aarhus, Denmark), for the measurement of isometric tension (Mulvany and Halpern, 1977). After vessels were mounted, the temperature of the bath was raised to 37°C and the Krebs solution was bubbled with 5% CO₂ in O₂ (pH after bubbling, 7.4). After a 45-min equilibration period, vessels were stretched in a stepwise manner to determine the relationship between the passive tension and the internal circumference, according to Laplace's equation. They were then stretched to 90% of the diameter achieved when the vessel was under a transmural pressure of 20 mm Hg (i.e., the pressure approximating conditions in vivo). Vessels were repeatedly constricted with the thromboxane A₂-mimetic U-46619 (0.1 µM), until the response was constant, before experimentation. Integrity of the endothelium was determined by assessing the relaxation response to bradykinin (1 µM) in vessels submaximally precontracted with U-46619 (10-100 nM). In experiments where removal of the endothelium was required, a hair was passed through the lumen of the vessels before measurement of the bradykinin response. Endothelium was considered removed when responses to bradykinin were <10% relaxation.

Effects of exogenous neuropeptides. The effects of CGRP (0.1-10 nM) and SP (1-10,000 nM) were determined under conditions of basal tone and in preparations in which stable tone had been induced with U-46619 (10-100 nM). In unprecontracted vessels, contractile concentration-response curves for SP were repeated until constant, with a 45-min washing period between each curve. When the curves were constant, tissues were incubated with the NK₁ receptor antagonist RP 67,580 (0.1-1 µM, 30-min pretreatment), the inactive enantiomer RP 68,651 (1 µM) or the NK₂ receptor blocker SR 48968 (1 µM, 2-hr pretreatment). At the end of each experiment the maximal contraction in response to U-46619 (1 µM) was determined.

Responses to capsaicin. Effects of capsaicin were determined under conditions of basal tone and in preparations in which tone had been induced with U-46619 (10-100 nM). In precontracted preparations, a single relaxation concentration-response curve for capsaicin (1-30 µM) was constructed.

Materials. Capsaicin, ruthenium red, SP, indomethacin, L-NAME and U-46619 were purchased from Sigma Chemical Co. (Poole, UK), and CGRP and CGRP_8-37 were purchased from Bachem (Saffron Walden, UK). L-NMMA and ODQ were kind gifts of Dr. Salvador Moncada of Wellcome Research Laboratories (Kent, UK). RP 67,580 and RP 68,651 were kind gifts of Dr. C. Garret of Rhône Poulenc Rorer (Vitry sur Seine, France) and SR 48968 of Dr. X. Emonds-Alt of Sanofi Recherche (Montpellier, France). Solutions of ruthenium red and L-NMMA were prepared in distilled water on the day of use. Indomethacin was dissolved in 1% sodium carbonate in water on the day of use. Stock solutions were made as follows: U-46619 in ethanol, CGRP, CGRP_8-37 and SP in distilled water, capsaicin in ethanol and ODQ in dimethylsulfoxide. The final concentration of ethanol in the myograph bathing solution was <0.5%. All stock solutions were kept at 20°C until required. Dilutions were made in Krebs solution.

Data and statistics. For measurement of the effects of antagonists on the concentration-response curve for SP, the EC₅₀ for SP was determined for each experiment in control and antagonist-treated vessels and the shift was expressed as a concentration ratio of the EC₅₀ for SP (control) and the EC₅₀ for SP in the presence of antagonist. For capsaicin concentration-response curves, a maximum plateau response was not reached. To assess shifts in concentration-response curves, the concentration of capsaicin required to produce an equivalent relaxation (usually 50% of the maximum response achieved with 30 µM capsaicin in control studies) was calculated as a concentration ratio. Statistical significance was calculated using analysis of variance for multiple comparisons, followed by the Bonferroni test. When multiple comparisons were not required, an unpaired Students t test was used. P < .05 was considered statistically significant.

	Results

Top Abstract Introduction Methods Results Discussion References

The mean diameter of veins used in the present study was 230 ± 13.2 µm (n = 48).

Effects of exogenous neuropeptides. SP (1-1000 nM) caused a concentration-dependent contraction of unprecontracted veins (see fig. 1 for typical contractile response curve). The concentration-response curve was shifted to the right, in a concentration-dependent manner, by the NK₁ receptor antagonist RP 67,580 (0.1, 0.3 and 1 µM, n = 5 for each concentration), to give concentration ratios of 1.6 ± 0.3 and 3.0 ± 0.7 with 0.1 and 0.3 µM, respectively; with 1 µM a maximum was not reached in all experiments, but a range of 5.0 to >100 µM was achieved. The EC₅₀ for SP in these studies was 1.1 ± 0.2 µM (n = 15). The inactive enantiomer RP 68,651 (1 µM) had no effect on the response to SP (n = 4, not significant, data not shown), nor did NK₂ receptor blockade with SR 48,968 (1 µM, n = 4, data not shown). CGRP had no effect in unprecontracted veins, and neither CGRP (0.1-10 nM, n = 3) nor SP (0.1-1 µM, n = 3) had any relaxant effects on vessels submaximally precontracted with U-46619 (10-100 nM). Pretreatment with CGRP (10 nM) did not alter the response to SP (1-1000 nM, n = 3).

View larger version (9K): [in this window] [in a new window]   Fig. 1.   Typical contractile concentration-response curve for SP in isolated rat small mesenteric vein.

Effects of capsaicin. Capsaicin had no effect on unprecontracted vessels but caused a concentration-dependent relaxation of precontracted vessels (fig. 2). The responses to capsaicin were significantly (P < .001) attenuated in animals denervated of C-fibers by pretreatment with capsaicin (n = 5) (fig. 3), compared with the responses in animals that had been treated with vehicle only (n = 5). The responses to bradykinin (to check for intact endothelium) were not significantly different in the two groups.

View larger version (9K): [in this window] [in a new window]   Fig. 2.   Typical relaxation concentration-response curve for capsaicin in isolated rat small mesenteric vein.

View larger version (12K): [in this window] [in a new window]   Fig. 3.   Inhibition of capsaicin-induced relaxation after chemical destruction of sensory C-fibers in rats. Animals were treated with either capsaicin (50 mg/kg) () or vehicle () s.c. 4 days before use. Values shown are mean ± S.E.M. ***P < .001, for statistical differences.

View larger version (13K): [in this window] [in a new window]   Fig. 4.   Effect of ruthenium red on capsaicin-induced relaxation of rat small mesenteric veins. Veins were pretreated with ruthenium red (30 µM, n = 5) () or no pretreatment (n = 9) () for 30 min before construction of the concentration-response curve for capsaicin. Values shown are mean ± S.E.M. *P < .05; **P < .01.

View larger version (11K): [in this window] [in a new window]   Fig. 5.   Role of endothelial NO in capsaicin-induced relaxation of rat small veins. A, concentration-response curves for capsaicin in the absence (n = 11) () and presence of L-NMMA (100 µM, n = 5) (), L-NAME (100 µM, n = 6) (), L-NAME (300 µM, n = 3) () and ODQ (1 µM, n = 4) (). B, concentration-response curves for capsaicin constructed in preparations with intact endothelium (n = 9) () and endothelium-denuded preparations (n = 5) (). Values shown are mean ± S.E.M. *P < .05; ***P < .001.

	Discussion

Top Abstract Introduction Methods Results Discussion References

This study demonstrates that local activation of C-fibers with capsaicin relaxes small mesenteric veins. This effect was endothelium-dependent but was not mediated by SP or CGRP and did not involve generation of prostanoid but did appear to involve NO. These results are the first direct demonstration of functional responses to sensory stimulation in the venous microcirculation.

Capsaicin, in concentrations similar to those that affect arterioles and small arteries, caused a concentration-dependent relaxation of precontracted small veins. This effect was markedly attenuated in veins taken from animals in which C-fibers had been largely destroyed by systemic capsaicin treatment (Gamse, 1982; Maggi et al., 1988), suggesting that the relaxant effect seen in vitro was due to C-fiber activation. This suggestion is further supported by the finding that the concentration-response curve for capsaicin was shifted to the right by ruthenium red, an agent that blocks the cation channel associated with vanilloid receptors (Dray et al., 1990), the receptor/binding site for capsaicin on sensory nerves (Winter et al., 1990; Szallasi and Blumberg, 1990). The finding that small veins respond to capsaicin in the same manner as do small arteries (Ahluwalia and Vallance, 1996a) suggests that both vessel types possess functionally active sensory C-fibers that respond to the same stimuli.

CGRP is released upon activation of arterial C-fibers and appears in the venous effluent of perfused beds stimulated by capsaicin or electrical field stimulation (Manzini et al., 1991). However, in the vessels we studied, CGRP had no effect on venous tone and therefore is unlikely to contribute to the capsaicin response. Our finding in isolated veins is similar to results from studies in the perfused mesentery, demonstrating that CGRP has no effect on the reactivity of the venous side of the mesenteric microcirculation (Claing et al., 1992). Similarly, studies in humans have demonstrated that CGRP does not dilate capacitance vessels of the hand (Marshall et al., 1988) or human saphenous vein (McEwan et al., 1988). In contrast, a recent study using intravital microscopy of the rat cremaster muscle demonstrated that capsaicin, applied topically, dilates arterioles and venules (Kim et al., 1995) and that the response is blocked by the CGRP receptor antagonist CGRP_8-37. The differences between our study and this earlier study might be due to differences between vascular beds. However, it is possible that the capsaicin-induced and CGRP-dependent venodilatation seen in the in situ cremaster muscle preparation is due to an effect of a mediator released from the arterial side of the circulation in response to CGRP, rather than a direct effect of CGRP on the veins themselves.

The relaxant effects of capsaicin we observed are unlikely to be due to SP. SP had no effect on the precontracted preparations but caused reproducible concentration-dependent contractions in uncontracted veins. This response was mediated by the activation of NK₁ receptors, because the selective nonpeptide NK₁ receptor antagonist RP 67,580 antagonized the responses, whereas the inactive enantiomer RP 68,651 had no effect. The apparent efficacy of RP 67580 is somewhat lower than would be expected for the rat NK₁ receptor, normally with IC₅₀ values in the 1 nM range (for review, see Maggi et al., 1993). The differences in the present study may reflect the involvement of other tachykinin receptors in mediation of this response, but the lack of effect of SR48968 excludes a significant NK₂ component. The EC₅₀ for SP is also relatively high for a pure NK₁ receptor response, indicating that SP may not be the natural agonist for this system in this preparation. Interestingly, it was demonstrated in the perfused rat mesentery that the venoconstriction produced after electrical stimulation was attenuated by NK₁ receptor antagonism, despite the insensitivity of this preparation to SP (Claing et al., 1992). The use of selective NK₁ and NK₃ receptor agonists such as septide and senktide would help to determine the exact tachykinin receptor subtypes involved in mediation of the responses seen in this preparation. Furthermore, the recent development of NK₃ receptor antagonists may also help to clarify the receptor systems involved.

In contrast to the contractile effect of SP, capsaicin did not cause contraction in unpreconstricted vessels; when vessels were preconstricted to an approximately EC₅₀ level with U-46619, capsaicin relaxed the veins. These observations suggest that, under these experimental conditions, SP was not released in quantities sufficient to alter venous tone. Our functional study demonstrating that neither CGRP nor SP is likely to be involved in the relaxant capsaicin response is consistent with anatomic studies that failed to detect CGRP or SP-like immunoreactivity surrounding small veins. Whereas the anatomic studies have been interpreted as suggesting that there is no sensory innervation of the venous microcirculation, our study clearly demonstrates that the nerves do respond to local sensory stimulation.

Consistent with previous studies (Kim et al., 1995), removal of the endothelium abolished the relaxant response to caspsaicin, but indomethacin given in doses sufficient to inhibit cyclooxygenase had no effect. The use of NO synthase inhibitors in this system produced varied responses. L-NMMA (100 µM) had no significant effect on the responses to capsaicin, whereas L-NAME (100 and 300 µM) produced a concentration-dependent rightward shift of the capsaicin response, suggesting that NO is involved. The apparently increased efficacy of L-NAME over L-NMMA is not surprising, because it is known that L-NAME is more potent at inhibiting NO synthase activity than is L-NMMA in certain systems (e.g., Rees et al., 1990). Interestingly, it is also clear that L-NAME shows approximately 100-fold greater potency against neuronal NO synthase activity than does L-NMMA (e.g., Salter et al., 1995). Confirmation of involvement of some NO was obtained with the use of the specific guanylate cyclase inhibitor ODQ (Garthwaite et al., 1995). The concentration of ODQ used in these studies was previously shown to inhibit guanylate cyclase; in this study, confirmation of this effect was obtained because the relaxation response to the NO donor glyceryl trinitrate was completely blocked in the presence of ODQ. Our findings are most consistent with the release of a factor from sensory C-fibers that traverses the thin wall of the vein to stimulate the endothelium to release NO or an NO-containing molecule and cause relaxation of vascular smooth muscle by activating guanylate cyclase. An alternative explanation would be that the endothelium responds to noxious stimuli and that capsaicin might activate the endothelium through a mechanism similar to that by which it activates sensory nerves. Although there is evidence for the presence of sensory neuropeptides within the endothelium (Milner et al., 1989), the vanilloid capsaicin receptor has not been found on cells other than neuronal cells, and studies exploring the effects of physical or chemical denervation show complete loss of capsaicin binding in peripheral tissues of rats (Szallasi et al., 1993), suggesting that neurons are the sole target for capsaicin.

These studies demonstrate for the first time that capsaicin, an agent known to selectively activate sensory C-fibers, relaxes isolated small veins of rats. The response is endothelium-dependent and appears to be mediated by NO. It is independent of prostanoid generation and is not mediated by CGRP or SP. The existence of a dilator sensory system on the venous side of the microcirculation that responds to the same stimuli as the nerves present on the arterial side suggests that in some situations hydrostatic pressure might not be altered in response to sensory nerve activation. These studies demonstrate the importance of examining the reactivity of both sides of the microcirculation individually, and they provide insight into the mechanisms of the diverse (pro- or anti-inflammatory) effects of activation of C-fibers seen in vivo.