INTRODUCTION

Top Abstract Introduction Methods Results Discussion References

IN A VARIETY OF ARTERIES, a component of endothelium-dependent relaxation to ACh and bradykinin (BK) persists after inhibition of nitric oxide synthase (NOS) and cyclooxygenase. This NOS- and prostacyclin-independent relaxation is accompanied by an endothelium-dependent hyperpolarization of the vascular smooth muscle cell membrane potential and has been suggested to be mediated by an endothelium-derived hyperpolarizing factor (EDHF) (3, 4, 8, 15, 21, 24). Although the endothelium-dependent hyperpolarization of the vascular smooth muscle cell membrane potential has been widely reported, the identity of EDHF and its mechanism of action have remained elusive.

EDHF-mediated relaxation has been shown to be depressed by either apamin or charbydotoxin (ChTX) alone in some tissues, but more usually a complete block of EDHF-mediated relaxation is achieved only with a combination of apamin and ChTX (4, 14, 22). Initially this was thought to implicate both small-conductance (SK_Ca) and large-conductance Ca²⁺-activated potassium channels (BK_Ca) in the relaxation (4), but iberiotoxin, another peptide blocker of BK_Ca channels, cannot substitute for ChTX (14, 24), and there are only a handful of reports of an apamin-sensitive K⁺ current in arterial tissue (9, 14).

To date, no study has addressed the question of whether ChTX and apamin act on the endothelium or on the smooth muscle to block EDHF-mediated relaxations. The aim of this study was to determine the site of action of these toxins by comparing the effects of applying ChTX and apamin, either to the smooth muscle or to the endothelium, on EDHF-mediated relaxation of myogenic and phenylephrine (PE)-induced tone in isolated pressurized mesenteric arteries.

	METHODS

Top Abstract Introduction Methods Results Discussion References

Male Wistar rats (wt 200-300 g) were killed by an intraperitoneal injection of pentobarbital sodium (500 mg/kg) or by stunning and cervical dislocation. Third-order superior mesenteric arteries were dissected in physiological saline solution (PSS) containing (in mM) 119 NaCl, 4.7 KCl, 25 NaHCO₃, 1.18 KH₂PO₄, 1.8 or 2.5 CaCl₂, 1.2 MgSO₄, 11 glucose, and 0.027 EDTA. The pH was 7.4 when gassed with 95% O₂-5% CO₂.

Wire myography. Arteries (length 2 mm) were mounted in a Mulvany myograph at 37°C under normalized tension for measurement of isometric force, as previously described (23). The tissues were maintained in a static bath in PSS gassed with 95% O₂-5% CO₂, and containing 2.8 µM indomethacin and 100 µM N-nitro-L-arginine methyl ester (L-NAME). Concentration-effect curves for ACh were constructed by cumulative addition of ACh to arterial segments preconstricted with PE (1-3 µM).

Pressure myography. Leak-free segments of artery (length at least 1 mm) were mounted between two glass cannulas in an arteriograph (Living Systems Instrumentation, Burlington, VT) at room temperature (18-21°C) and pressurized to 80 mmHg, under conditions of no luminal flow. A set constant pressure was maintained via a pressure servo-control system (PS200, Living Systems Instrumentation). Pressure transducers at both ends of the artery allowed continual monitoring of intraluminal pressure. Arteries were viewed through a Nikon TMS inverted microscope, and a measurement of the internal diameter was made from a video image using a video dimension analyzer (V91, Living Systems Instrumentation). The arteriograph was continually superfused with the standard PSS (see METHODS) at a rate of 25 ml/min. The superfusing PSS was warmed to 37°C by passing it through a heat-exchange coil before entering the arteriograph. If a sustained reduction in arterial diameter was seen, a myogenic response was considered present. If the myogenic response was absent, the artery was constricted with 0.3-0.5 µM PE applied in the superfusate to give a level of tone similar to that of the myogenic response in other vessels. Pressure and diameter measurements were recorded to computer via a Digidata 1200B interface using AxoScope software (Axon Instruments, Foster City, CA).

Intraluminal perfusion. Changes of intraluminal solution in pressurized arteries were achieved using microfil syringe fillers (World Precision Instruments), which are fine tubes (0.164-mm OD, 0.1-mm ID) fabricated from fused silica with a Luer-Lok fitting bonded to one end. Three microfil syringe fillers were inserted through the Tygon tubing connecting the cannula distal to the pressure servo and sealed in place using flowable silicone adhesive (RS Components). When fully assembled, the microfil tubes entered the distal cannula (1.16-mm ID) and approached the vessel to within 3 mm. This is shown schematically in Fig. 1. Small volumes (typically 20-30 µl) of solution were introduced into the distal cannula (and subsequently, the lumen of a vessel) over ~5 s using micrometer syringes (200-µl capacity; Gilmont). The servo-controlled peristaltic pump then acted to back off the increase in pressure due to the volume of fluid injected. Small changes in intraluminal pressure were transient and were monitored by both proximal (P1) and distal (P2) pressure transducers. The volume per unit length of the glass used to fabricate the cannulas was 10 µl/cm. Given this value, the introduced volume (20-30 µl), and the placement of the microfil tubes within the cannulas, it is clear that injected solution completely replaced the solution within the artery. Drugs were washed out of the lumen by flushing. To flush the lumen, a three-way tap on the distal end of the artery was opened to an open-ended column of saline, the height of which was adjusted to provide a pressure of 75 mmHg (5 mmHg less than the servo-pressure). This caused the servo-controlled peristaltic pump to push fluid down the pressure gradient through the lumen of the artery, thus flushing to waste. For a typical 4-min flush, ~100 µl of PSS passed through the vessel, which was sufficient to displace the luminal contents ~60 mm through the cannula and associated tubing, and away from the artery.

View larger version (23K): [in this window] [in a new window]   Fig. 1.   Schematic representation of intraluminal perfusion of pressurized artery. P1 and P2, proximal and distal pressure transducers, respectively. See METHODS for details.

Intraluminal solutions. During all experiments, the PSS, both superfusing and within the lumen of the pressurized vessel, contained 100 µM L-NAME and 2.8 µM indomethacin. All intraluminal solutions were gassed before being loaded into the cannulas and the microfils.

Superfusion. Gassed PSS was circulated at 25 ml/min using a peristaltic pump (Gilson, Villiers-le-Bel, France). The volume contained in the arteriograph, heat exchanger, and associated tubing was 25 ml, and the total volume of PSS was either 500 ml (control) or 100 ml (toxins).

Drugs. All drugs were made up as stock solutions in Milli-Q water, unless otherwise stated, and diluted in the experimental solution. ChTX was made up in 150 mM NaCl, and indomethacin was made up in 2% Na₂CO₃. All drugs were supplied by Sigma or Calbiochem.

	RESULTS

Top Abstract Introduction Methods Results Discussion References

Wire myography. ACh (0.01-10 µM) applied to isometrically mounted arteries relaxed PE-stimulated tone in a concentration-dependent manner, with an EC₅₀ of 58 nM (n = 10). Neither ChTX (50 nM, n = 4) nor apamin (50 nM, n = 4) significantly affected the relaxation to ACh. However, a combination of apamin and ChTX completely and reversibly abolished the relaxation to ACh (n = 5) (Fig. 2).

View larger version (21K): [in this window] [in a new window]   Fig. 2.   ACh-induced relaxation of isometric rings of rat mesenteric artery. Concentration-dependent relaxation of phenylephrine (PE)-constricted (1-3 µM) arterial rings to ACh in presence of 100 µM N-nitro-L-arginine methyl ester (L-NAME) and 2.8 µM indomethacin (n = 10). Experiment was repeated with 50 nM apamin (n = 4), 50 nM charybdotoxin (ChTX) (n = 4), and 50 nM apamin plus 50 nM ChTX (n = 5). Plotted points are means ± SE; n = no. of experiments.

Pressure myography. On warming from room temperature to 37°C, some arteries developed myogenic tone (n = 3). ACh (10 µM), applied intraluminally, relaxed tone, dilating pressurized rat mesenteric arteries. In the event that myogenic tone failed to develop, arteries were constricted with PE (0.3-0.5 µM, n = 5). PE was applied cumulatively until the level of tone was similar to the level of tone observed in myogenically active arteries. ACh (10 µM) applied intraluminally also relaxed the PE-induced constriction. Examples of dilatation of a pressurized artery in response to intraluminal application of ACh (10 µM) are shown in Figs. 3 and 4. A transient increase in pressure on both P1 and P2 pressure transducers indicates the period during which the intraluminal solution is exchanged. The ACh dilatation was maintained over a time course of several minutes, after which some desensitization of the response occurred. The dilatation was fully reversible on flushing. Periods of flushing can just be seen in the pressure measurements as a 5-mmHg drop in pressure (see METHODS). ACh-induced dilatations were reproducible, which confirms that flushing of the lumen did not damage the endothelium. Intraluminal application of control PSS did not elicit a dilatation (see Fig. 3, A and B), demonstrating that ACh-induced relaxations could not be explained either by flow or the transient increase in pressure observed during microinjection.

View larger version (19K): [in this window] [in a new window]   Fig. 3.   ACh-induced relaxation of pressurized rat mesenteric artery. A: diameter measurements of artery pressurized at 80 mmHg. P1 and P2, pressure at proximal and distal ends of artery, respectively, throughout experiment. Tone developed after addition of 0.5 µM PE to superfusate (not shown). Intraluminal perfusion of control buffer did not mediate relaxation. Intraluminal perfusion of 10 µM ACh produced a dilatation. B: summary of paired data for intraluminal perfusion of control buffer (n = 7; 3 arteries). Arterial diameters: 1) initial, 246 ± 6.89 µm; 2) level of tone (pooled from PE-stimulated and myogenic vessels), 176 ± 10.7 µm; 3) intraluminal perfusion of control buffer, 174 ± 10.9 µm. Diameters 2 and 3 are not significantly different. In both B and C,  represents mean pressure under each condition. C: summary of paired data for intraluminal perfusion of 10 µM ACh (n = 8; 8 arteries). Arterial diameters: 1) initial arterial diameter, 254 ± 20.7 µm; 2) level of tone (pooled from PE-stimulated and myogenic vessels), 184 ± 17.3 µm; 3) intraluminal perfusion of 10 µM ACh, 253 ± 19.7 µm. * Significantly different from diameter 2.

View larger version (21K): [in this window] [in a new window]   Fig. 4.   Effects of toxins on ACh-induced relaxation. A: diameter measurements of artery pressurized at 80 mmHg. Tone developed after addition of 0.5 µM PE to the superfusate (shown). In presence of 50 nM ChTX and 50 nM apamin in lumen, ACh relaxation was abolished. Effect was reversible on flushing of lumen with control physiological saline solution. Application of 50 nM ChTX and 50 nM apamin to the superfusate did not depress ACh relaxation. B: effects of toxins on ACh-induced relaxation of 0.3-0.5 µM PE-stimulated tone (n = 5). Arterial diameters: 1) initial, 279 ± 20.7 µm; 2) PE-stimulated tone, 200 ± 22.4 µm; 3) ACh relaxation of PE-stimulated tone, 276 ± 21.6 µm; 4) 50 nM ChTX and 50 nM apamin in lumen, 197 ± 30.1 µm; 5) ACh in presence of 50 nM ChTX and 50 nM apamin in lumen, 210 ± 23.8 µm; 6) 50 nM ChTX and 50 nM apamin in superfusate, 183 ± 28.6 µm; and 7) ACh in presence of 50 nM ChTX and 50 nM apamin in superfusate, 272 ± 23.0 µm. C: effects of toxins on ACh-induced relaxation of myogenic tone (n = 3). Column definitions are as in B. Arterial diameters: 1, 211 ± 27.6 µm; 2, 157 ± 13.0 µm; 3, 213 ± 22.1 µm; 4, 159 ± 4.58 µm; 5, 161 ± 19.7 µm; 6, 150 ± 30.2 µm; and 7, 208 ± 22.2 µm. P values for comparison with level of tone (2) are shown above columns. * Significantly different from diameter 2. Values are means ± SE.

	DISCUSSION

Top Abstract Introduction Methods Results Discussion References

In several recent reports, relaxation that persists after inhibition of NOS and cyclooxygenase has been ascribed to an unidentified EDHF. In resistance arteries EDHF-mediated hyperpolarization correlates strongly with NO- and prostanoid-independent relaxation, implying a causal relationship (25) (for recent review, see Ref. 6). Whereas no single K⁺-channel blocker has been shown to be completely effective, a cocktail of apamin and ChTX abolishes EDHF-mediated hyperpolarization and relaxation (4, 15, 16, 22, 24). In previous studies the toxins have been applied to isolated vessels by superfusion, such that the endothelium and the smooth muscle are exposed simultaneously and so the site of action is not known. Our data clearly show that apamin and ChTX block EDHF by an action at the endothelial surface and not an action in the smooth muscle.

In the present study we have used L-NAME to inhibit NOS. Kemp and Cocks (10) reported that in human coronary arteries, oxyhemoglobin (HbO₂), which will scavenge NO, further depressed relaxations to BK in the presence of L-arginine analogs and indomethacin. This implies that NOS-independent NO may contribute to EDHF. It has also been shown in human omental arteries (15) that the relaxation to BK, which was resistant to a combination of indomethacin (10 µM) and either HbO₂ (10 µM) or L-NAME (300 µM), was similar, ruling out NO as a component of EDHF. In both studies the NO- and prostanoid-independent relaxation to BK was abolished by depolarizing concentrations of KCl, which is consistent with an EDHF.

It is clear from our methods that introducing ACh, toxins, or control PSS into the lumen of the arteries involves a pressure change and luminal flow within the artery. This has been shown previously to induce vasodilatation and might be mediated by opening of BK_Ca channels (18). Figure 3 shows clearly that injection of control PSS resulted in a transient pressure increase that would be associated with flow, but this did not induce dilatation. Therefore, we conclude that neither pressure per se nor flow contributes to the dilatation induced by injection of ACh.

In our experiments, 1-3 µM PE-induced isometric force was abolished by 1-10 µM ACh. ChTX and apamin (each 50 nM) blocked the effect of ACh only when applied in combination. Isobaric myograph experiments were performed using 10 µM ACh (supramaximal in isometric recordings) to dilate both 0.3-0.5 µM PE-constricted and myogenic arteries to their passive diameter. It has been reported that agonist sensitivity is higher in isobaric myography compared with isometric myography (2, 5, 7). PE was used in isobaric experiments at concentrations that gave a level of tone similar to that observed in myogenic arteries, and wall forces were equivalent in the two situations. Falloon et al. (7) have shown that the concentration-effect relationship for vasorelaxation to ACh was not different in rat mesenteric arteries using isometric and isobaric myography. Therefore, it is justifiable to use a single supramaximal concentration of ACh (10 µM). ACh-induced vasorelaxation can be completely abolished by 50 nM ChTX and apamin applied intraluminally either to pressurized arteries or to the PSS superfusing isometric rings.

Several mechanisms have been proposed to explain the effect of apamin and ChTX on EDHF-mediated relaxations. It is unlikely that ChTX (in combination with apamin) blocks EDHF by an action on BK_Ca channels because neither iberiotoxin, which is selective for BK_Ca (12), nor tetraethylammonium ions are effective. However, it is well established that ChTX can block voltage-gated K⁺ (K_v) channels (20), and these channels may be the target of ChTX. Channels that bind and are blocked by apamin are K⁺ selective, voltage independent, and calcium sensitive, with a small unitary conductance and are referred to as SK channels. Cloned SK channels share little homology with other K_v channels (11), and there have been few reports of an apamin-sensitive, calcium-activated K⁺ current in either endothelial (19) or arterial muscle cells (9). Given that K⁺ channels activated by EDHF must pass enough current to substantially hyperpolarize the membrane and that SK current is difficult to find in patch-clamp studies, it seems unlikely that SK channel density is sufficient to account for relaxations to EDHF.

The requirement for apamin and ChTX to block relaxation could be explained if EDHF activates at least two channels, a K_v channel blocked by ChTX and an SK channel blocked by apamin. Alternatively, EDHF-mediated hyperpolarization may be dependent on activation of a channel that shares characteristics of BK_Ca and K_v (24). Interestingly, Zygmunt and co-workers (24, 25) reported that apamin can significantly enhance ChTX binding. Thus it might be possible for apamin to block EDHF via an allosteric effect that increases ChTX binding, rather than by acting independently to block an apamin-sensitive channel.

In previous reports, it has been assumed that the toxins act on the smooth muscle and do not affect EDHF synthesis or EDHF release from the endothelium. Our data from experiments on pressurized arteries clearly show that apamin and ChTX block EDHF-mediated relaxation by an action at the endothelial surface. It is unlikely for two reasons that toxins applied intraluminally were required to diffuse to the muscle cells to block EDHF-mediated dilatations. First, simultaneous application of the toxins with ACh (i.e., no preaddition of toxins) resulted in complete inhibition of EDHF (data not shown). Second, external application of toxins tended to further constrict pressurized myogenic and PE-stimulated vessels, consistent with the report of Brayden and Nelson (see Ref. 1), whereas intraluminal application did not alter vessel diameter.

Three mechanisms could explain an endothelial target for toxins: 1) EDHF is an as yet unidentified factor whose synthesis or release by the endothelium is blocked in the presence of apamin and ChTX; 2) EDHF is not a diffusible factor but an endothelium-derived hyperpolarizing current transmitted to smooth muscle through gap junctions after hyperpolarization of the endothelium (13); or 3) EDHF is potassium (6). Loss of K⁺ through endothelial K⁺ channels might raise the concentration of K⁺ within the media of the vessel to promote hyperpolarization of smooth muscle by two independent mechanisms. First, increased extracellular K⁺ concentration would increase current passed by inward rectifier K⁺ channels, which appear to be expressed preferentially in small arteries (17). Second, elevated extracellular K⁺ concentration would tend to increase the activity of Na⁺-K⁺-ATPase, which is consistent with evidence that EDHF is partially ouabain sensitive (8, 16).

In conclusion, this is the first report to establish that the combination of apamin and ChTX block EDHF-mediated relaxation by an action on the endothelium, contradicting the fundamental assumption that these toxins block K⁺ conductance(s) in smooth muscle.