Effects of Ca²⁺ concentration and Ca²⁺ channel blockers on noradrenaline release and purinergic neuroeffector transmission in rat tail artery

Abstract

1. The effects of Ca²⁺ concentration and Ca²⁺ channel blockers on noradrenaline (NA) and adenosine 5′-triphosphate (ATP) release from postganglionic sympathetic nerves have been investigated in rat tail arteries in vitro. Intracellularly recorded excitatory junction potentials (e.j.ps) were used as a measure of ATP release and continuous amperometry was used to measure NA release.

2. Varying the extracellular Ca²⁺ concentration similarly affected the amplitudes of e.j.ps and NA-induced oxidation currents evoked by trains of ten stimuli at 1 Hz.

3. The N-type Ca²⁺ blocker, ω-conotoxin GVIA (ω-CTX GVIA, 0.1 μM) reduced the amplitudes of both e.j.ps (evoked by trains of ten stimuli at 1 Hz) and NA-induced oxidation currents (evoked by trains of ten stimuli at 1 Hz and 50 stimuli at 10 Hz) by about 90%.

4. The ω-CTX GVIA resistant e.j.ps and NA-induced oxidation currents evoked by trains of 50 stimuli at 10 Hz were abolished by the non-selective Ca²⁺ channel blocker, Cd²⁺ (0.1 mM), and were reduced by ω-conotoxin MVIIC (0.5 μM) and ω-agatoxin IVA (40 nM).

5. Nifedipine (10 μM) had no inhibitory effect on ω-CTX GVIA resistant e.j.ps and NA-induced oxidation currents.

6. Thus both varying Ca²⁺ concentration and applying Ca²⁺ channel blockers results in similar effects on NA and ATP release from postganglionic sympathetic nerves. These findings are consistent with the hypothesis that NA and ATP are co-released together from the sympathetic nerve terminals.

Introduction

Depolarization-evoked entry of Ca²⁺ into nerve terminals plays a key role in triggering neurotransmitter release. In postganglionic sympathetic nerve terminals, this Ca²⁺ enters mainly through N-type Ca²⁺ channels (Maggi et al., 1988; Brock et al., 1989; De Luca et al., 1990; Smith & Cunnane, 1996; 1997; Wright & Angus, 1996), although neurotransmitter release can also be evoked by Ca²⁺ entering through other Ca²⁺ channel types. In particular, it has been demonstrated in a number of sympathetically innervated tissues that following the addition of a maximally effective concentration of the N-type Ca²⁺ channel blocker, ω-conotoxin GVIA (ω-CTX GVIA), postjunctional responses to sympathetic nerve stimulation remain (Smith & Cunnane, 1996; 1997; Wright & Angus, 1996; Waterman, 1997). This ω-conotoxin GVIA resistant response is best revealed when the frequency of nerve stimulation is raised. For example, in the guinea-pig vas deferens, ω-CTX GVIA abolished the excitatory junction potentials (e.j.ps) evoked by trains of stimuli at 1 Hz but trains of higher frequency stimuli (⩾3 Hz) evoked e.j.ps of small amplitude (Smith & Cunnane, 1996). These ω-CTX GVIA resistant e.j.ps were blocked by the non-selective Ca²⁺ channel blocker, Cd²⁺, indicating that they were due to Ca²⁺-dependent release of neurotransmitter.

The ω-CTX GVIA resistant e.j.ps in the guinea-pig vas deferens were unaffected by ω-agatoxin IVA (ω-AgTx IVA) and ω-conotoxin MVIIC (ω-CTX MVIIC) which block P- and Q-type Ca²⁺ channels (Randall & Tsien, 1995) or by nifedipine which blocks L-type Ca²⁺ channels. These findings indicate that ω-CTX GVIA resistant neurotransmitter release in guinea-pig vas deferens is due to the opening of Ca²⁺ channels not blocked by known selective Ca²⁺ channel antagonists. In contrast, in rat and mouse vas deferens (Wright & Angus, 1996; Waterman, 1997) and rat anococcygeus (Smith & Cunnane, 1997) the ω-CTX GVIA resistant responses are dependent on Ca²⁺ entering through P-/Q-type Ca²⁺ channels.

ω-CTX GVIA has a similar blocking action on postjunctional responses mediated by the release of the sympathetic co-transmitters noradrenaline (NA) and adenosine 5′-triphosphate (ATP) in the mouse, rat and guinea-pig vas deferens (Maggi et al., 1988; Brock et al., 1989; De Luca et al., 1990; Waterman, 1997) and the rat anococcygeus (Smith & Cunnane, 1997). These findings suggest that the release of NA and ATP are similarly affected by blockade of N-type Ca²⁺ channels. However, it has recently been reported that ω-CTX GVIA inhibited the nerve stimulation evoked overflow of endogenous NA from the guinea-pig vas deferens to a much greater extent than that of endogenous ATP (Westfall et al., 1996). Conversely, ω-AgTx IVA reduced the stimulus evoked release of ATP to a much greater extent than that of NA. These findings suggest that the release of NA and ATP depend primarily on the activation of N- and P-/Q-type Ca²⁺ channels respectively.

The main aim of the present study was to compare the action of Ca²⁺ channel blockers on action potential evoked release of NA and ATP from postganglionic sympathetic nerve innervating the rat tail artery. In addition, the Ca²⁺ concentration dependence of NA and ATP release in this tissue was determined. The release of endogenous NA release was measured directly by continuous amperometry (Gonon et al., 1993; Brock et al., 1997). This technique provides a ‘real time' measurement of NA release. In separate experiments conventional intracellular recording techniques were used to monitor electrically evoked e.j.ps, which provide a measure of ATP release (Sneddon & Burnstock, 1984; McLaren et al., 1995).

Methods

Female outbred Wistar rats (150–200 g) were killed by an overdose of pentobarbitone sodium (100 mg kg⁻¹, i.p.). Sections of proximal tail artery approximately 10 mm in length were removed and individual preparations pinned to the Sylgard (Dow-Corning) covered base of a 1 ml (electrophysiology) or 2 ml (amperometry) recording chamber. The chambers were perfused continuously at 3–5 ml min⁻¹ with physiological saline of the following composition (mM): NaCl 133.4, KCl 4.7, CaCl₂ 2.4, MgCl₂ 1.2, NaH₂PO₄ 1.3, NaHCO₃ 16.3 and glucose 9.8. The physiological saline was gassed with a mixture of 95% O₂ and 5% CO₂ (to pH 7.4) and maintained at 35–36°C. In all experiments, the physiological saline contained the α₁-adrenoceptor antagonist, prazosin (0.1 μM), to inhibit neurally evoked contraction due to NA release, idazoxan (1 μM) to inhibit α₂-adrenoceptor mediated autoinhibition of transmitter release and desipramine (1 μM) to inhibit neuronal uptake of released NA (Uptake₁). In the experiments where the effects of varying the Ca²⁺ concentration were investigated, only the CaCl₂ content of the physiological saline was varied. The proximal end of the artery was drawn into a suction stimulating electrode and the perivascular nerves were excited by electrical field stimulation (1 ms pulse width, 3–20 V for the electrophysiology experiments and 0.3 ms, 20 V for the amperometry experiments).

Electrochemical recording

The release of endogenous NA was monitored by continuous amperometry using a technique similar to that described in Brock et al. (1997). This electrochemical method measures the stimulus-evoked increase in NA concentration at the adventitial surface of the artery as an oxidation current, the amplitude of which is linearly related to the change in NA concentration (Gonon et al., 1993; Brock et al., 1997). The only difference between the present experiments and those described in Brock et al. (1997), was that all the carbon fibre electrodes were pretreated by dipping the electrode tips for 15–30 min in a 2.5% (w/v) solution of Nafion (Aldrich Chemical Company, Castle Hill, NSW, Australia). After dipping, the electrodes were allowed to dry for at least 1 h before use. This treatment reduced the decline in electrode sensitivity observed during prolonged periods of recording (see Brock et al., 1997).

Intracellular recording

Intracellular recordings were made using the technique described in Brock et al. (1997). To avoid the variable amplitude e.j.ps with an early fast component recorded in superficial cells (see Cassell et al., 1988), recordings were made preferentially from cells two or more cells deep in the media in which the e.j.ps decayed monoexponentially. To facilitate comparisons between different preparations and to reduce the effects of non-linear summation on e.j.p. amplitude, the stimulus intensity was adjusted at the start of each experiment to ensure that e.j.ps evoked by single stimuli were approximately 10 mV in amplitude.

The effects of varying Ca²⁺ concentration and of Ca²⁺ channel blockers on e.j.p. amplitude were determined in single cell experiments in which both control and test recordings were made during the same impalement.

Ca²⁺ concentration dependence experiments

In experiments where the effects of varying the Ca²⁺ concentration on oxidation currents and e.j.ps were investigated, the tissues were stimulated at 10 min intervals with three trains of ten pulses at 1 Hz, each train being separated by 90 s. Initially the tissues were superfused with physiological saline containing 2.4 mM Ca²⁺ for two periods of stimulation and then, immediately following the second and subsequent periods of stimulation, the Ca²⁺ concentration of the physiological saline was varied to 0.6, 0.9, 1.2, 1.8, 3.6 or 4.8 mM. The sequence of the changes in Ca²⁺ concentration in each experiment was randomized. In the electrochemistry experiments, only three changes in Ca²⁺ concentration were made before returning to 2.4 mM Ca²⁺ for the final period of stimulation. This procedure was followed because during control experiments (n=6), in which the Ca²⁺ was maintained throughout at 2.4 mM, the amplitudes of oxidation currents did not change detectably during six cycles of stimulation but, during experiments of longer duration, the amplitudes of the oxidation currents decreased significantly (Wilcoxon signed rank tests). No change in e.j.p. amplitude was detected during ten cycles of stimulation in the presence of 2.4 mM Ca²⁺ (n=6). Therefore in most electrophysiology experiments the full range of Ca²⁺ concentrations was tested before returning the Ca²⁺ concentration to 2.4 mM for the last period of stimulation. However, in some experiments, the impalements were lost before all the Ca²⁺ concentrations could be tested. All the data presented are normalized with respect to the initial values obtained under control conditions (2.4 mM Ca²⁺) at the start of each experiment.

Ca²⁺ channel blocker experiments

In all experiments investigating the effects of the Ca²⁺ channel blockers on oxidation currents and e.j.ps, the effects of adding ω-CTX GVIA were first determined. In this part of the experiments, the tissues were stimulated at 10 min intervals with three trains of ten pulses at 1 Hz or three trains of 50 pulses at 10 Hz (electrochemistry experiments only), each train being separated by 90 s. Immediately following the second period of stimulation, ω-CTX GVIA (0.1 μM) was applied to the tissues and the effects of this agent followed for two or three periods of stimulation. In the second part of each experiment, which investigated the effects of adding nifedipine (10 μM), ω-AgTx IVA (40 nM) or ω-CTX MVIIC (0.5 μM) on the ‘ω-CTX GVIA resistant' responses, all tissues were stimulated at 10 min intervals with three trains of 50 pulses at 10 Hz. The Ca²⁺ channel blocker was applied to the tissues immediately following the second period of stimulation and the effects of the agent followed for three periods of stimulation. At the end of all experiments, a further period of stimulation was recorded 20 mins after applying Cd²⁺ (0.1 mM) to block fully neurotransmitter transmitter release. In control experiments (n=6) in which the superfusing solution contained ω-CTX GVIA (0.1 μM), the amplitudes of the oxidation currents and the depolarizations evoked by trains of 50 pulses at 10 Hz declined significantly during a recording period of 30 mins (Wilcoxon signed rank test, P<0.05). For this reason the results obtained in the second part of the experiment were compared statistically with those for time matched ω-CTX GVIA-treated controls.

Data analysis

All data were digitized (sampling frequencies of 0.04–0.2 kHz) and collected with a MacLab recording system and the program Scope (ADInstruments Pty Ltd, Castle Hill, N.S.W., Australia). Subsequent analysis was made with the computer program Igor Pro (Wavemetrics, Lake Oswego, OR, U.S.A.). To assess the effect of changing the Ca²⁺ concentration or of drug treatments, the electrophysiological or electrochemical records for individual tissues under each condition were averaged before measurements were made. The amplitudes of the electrophysiological and electrochemical signals evoked during the trains of stimuli at 1 Hz were determined as previously described in Brock et al. (1997). The amplitudes of the depolarizations evoked by 50 stimuli at 10 Hz were determined during last the 10 stimuli in the train. The amplitudes of the oxidation currents evoked by 50 pulses at 10 Hz were measured by averaging the period 100–200 ms following the last stimulus in the train.

Data are presented as means±s.e.mean. Statistical comparisons were made using non-parametric tests as indicated in the text. For comparisons between the effects of the various manipulations on oxidation currents and e.j.ps it was assumed that both provide measures of neurotransmitter release. Prior to analysing the Ca²⁺ concentration data, the number of measures for each Ca²⁺ concentration was standardized by assigning each measure a random number and using the nine highest values. Comparison between the selected data sets and the complete data sets using Mann-Whitney U-tests revealed no significant difference. In all tests P values <0.05 were considered to be significant.

Drugs

Prazosin, idazoxan, desipramine and nifedipine were supplied by Sigma Chemical Company (Castle Hill, N.S.W., Australia) and ω-CTX GVIA and ω-CTX MVIIC were supplied by Auspep Pty Ltd (Parkville, Vic, Australia). ω-AgTx IVA was a kind gift from Pfizer (Groton, CT, U.S.A.). Nifedipine was made up as a stock solution in ethanol. The peptides were prepared as stock solutions in water and stored at −20°C (the ω-AgTx IVA solution contained 1 mg ml⁻¹ cytochrome C). Drugs were added to the solution superfusing the tissue. During peptide application, a recirculation system of 20 ml volume was used to minimize the amounts of the toxins used.

Results

General observations

Under control conditions (2.4 mM Ca²⁺) the mean increase in oxidation current evoked by each stimulus during a train of 10 stimuli at 1 Hz was 0.5±0.1 pA (n=39) and the amplitude of the oxidation current measured at the end of a train of 50 stimuli at 10 Hz was 7.8±0.9 pA (n=25). Also under these conditions the mean amplitude of e.j.ps evoked during trains of 10 stimuli at 1 Hz was 9.7±0.4 mV and the membrane potential was −63.9±0.3 mV (n=30 tissues).

Effects of Ca²⁺ concentration

Both the amplitudes of the oxidations currents (Figure 1a) and the e.j.ps (Figure 1b) evoked during trains of 10 pulses at 1 Hz increased as the concentration of Ca²⁺ was raised from 0.6 to 4.8 mM. Figure 1c shows graphically the effects of Ca²⁺ concentration on the amplitudes of the oxidation currents and e.j.ps, relative to the values measured under control conditions (2.4 mM Ca²⁺) at the start of the experiment. The relationship between the Ca²⁺ concentration and the amplitudes of oxidation currents and e.j.ps was similar and statistical comparison revealed no significant difference between effects of Ca²⁺ concentration on the two data sets (Mann-Whitney U-test, P=0.53).

Figure 1.

Effects of Ca²⁺ concentration on oxidation currents and e.j.ps evoked by trains of 10 stimuli at 1 Hz. (a and b) Overlaid traces showing oxidation currents (a) and e.j.ps (b) recorded in a range of Ca²⁺ concentrations (0.6–4.8 mM). (c) Graph showing the effects of Ca²⁺ concentration on the amplitudes of oxidation currents (n=9) and e.j.ps (n=9) evoked during the trains of stimuli. The data presented are normalized with respect to values obtained under control conditions (2.4 mM Ca²⁺) at the start of each experiment.

In the electrophysiological experiments the effect of Ca²⁺ concentration on membrane potential was not determined. However, the time constant of decay of the e.j.ps (Table 1), which provides a measure of the membrane time constant (see Cassell et al., 1988), did not change significantly over the range of Ca²⁺ concentrations used (Kruskal-Wallis test, P=0.31). In addition, the smooth muscle membrane potential measured at the start (−63.1±0.7 mV) and end (−62.6±0.7 mV) of each experiment did not differ significantly (Wilcoxon signed rank test, P=0.16, n=11 tissues).

Table 1.

Effects of Ca²⁺ concentration on the e.j.p. time constant of decay (τ e.j.p.; n=9 for all values)

Effects of ω-CTX GVIA

ω-CTX GVIA (0.1 μM) markedly reduced the amplitudes of the oxidation currents and e.j.ps evoked during trains of ten pulses at 1 Hz (Figure 2a,Figure 2c, Wilcoxon signed rank test; oxidation current P<0.05, n=8; e.j.p P<0.001, n=22) and the amplitudes of the oxidation currents evoked by 50 pulses at 10 Hz (Figure 2b, P<0.001, n=26). This inhibitory effect was clearly apparent within 10 min of application and reached a plateau level after 20 min (Figure 2d). At 20 min, the inhibitory effect of ω-CTX GVIA on oxidation currents and on e.j.ps did not differ significantly (Kruskal-Wallis test, P=0.52).

Figure 2.

Effects of ω-CTX GVIA (0.1 μM) on oxidation currents and e.j.ps. (a and b) Oxidation currents evoked by trains of ten stimuli at 1 Hz (a) and 50 stimuli at 10 Hz (b) before (upper traces) and 20 min after application of ω-CTX GVIA (lower traces). (c) E.j.ps evoked by trains of ten stimuli at 1 Hz before (upper trace) and 20 min after application of ω-CTX GVIA (lower trace). (d) Graph showing the mean inhibitory effect of ω-CTX GVIA on the amplitude of oxidations currents evoked by ten stimuli at 1 Hz (n=8) and 50 stimuli at 10 Hz (n=26) and e.j.ps evoked by ten stimuli at 1 Hz (n=22).

In a small number of experiments the concentration of ω-CTX GVIA was raised to 0.3 μM, 20 min after the initial application of the 0.1 μM solution. The amplitude of the oxidation currents evoked by 50 pulses at 10 Hz and the mean amplitude of e.j.ps evoked by 10 pulses at 1 Hz, measured 20 min after applying 0.3 μM ω-CTX GVIA, decreased by 14±4% (n=3) and 6±8% (n=5) respectively compared to the values measured just prior raising the concentration of ω-CTX GVIA. In tissues exposed only to 0.1 μM ω-CTX GVIA, the amplitude of the oxidation currents and e.j.ps decreased respectively by 16±8% (n=6) and 9±5% (n=14) over the same time interval. Thus in the rat tail artery, as in other sympathetically innervated tissues (see e.g. Smith & Cunnane, 1996; 1997), 0.1 μM ω-CTX GVIA is a maximally effective concentration for blocking neurotransmitter release.

Effects of nifedipine, ω-AgTx IVA and ω-CTX MVIIC on ω-CTX GVIA resistant responses

In tissues treated with the ω-CTX GVIA (0.1 μM), stimulation with 50 pulses at 10 Hz produced clearly discernible oxidation currents and membrane depolarizations due to summation of e.j.ps (Figure 3a,b). The amplitude of the oxidation currents measured at the end of the train of stimuli was 1.0±0.2 pA (n=28) and the amplitude of membrane depolarization was 7.9±0.8 mV (n=26). Application of the inorganic Ca²⁺ channel blocker, Cd²⁺ (0.1 mM), abolished these ω-CTX GVIA resistant responses (Figure 3a(i) and Figure 3b(i)), confirming that they were due to Ca²⁺-dependent release of neurotransmitter.

Figure 3.

Effects of various Ca²⁺ channel blockers on ω-CTX GVIA resistant oxidation currents (a) and e.j.ps (b) evoked by trains of 50 stimuli at 10 Hz. (a and b) Overlaid traces recorded before and 30 min after (indicated by the arrowhead) application of Cd²⁺ (0.1 mM, a(i) and b(i)), nifedipine (10 μM, a(ii) and b(ii)), ω-AgTx IVA (40 nM, a(iii) and b(iii)) and ω-CTX MVIIC (0.5 μM, a(iv) and b(iv)).

Figure 3 and Figure 4 show the effects of applying the Ca²⁺ channel blockers nifedipine (10 μM, Figure 3a(ii) and Figure 3b(ii)), ω-AgTx IVA (40 nM, Figure 3a(iii) and Figure 3b(iii)) or ω-CTX MVIIC (0.5 μM, Figure 3a(iv) and Figure 3b(iv)) on the oxidation currents and membrane depolarizations. The effects of these agents were compared 30 min following their addition to the superfusing solution (Figure 4). At this time there was a significant difference between the effects of the Ca²⁺ channel blockers on the amplitudes of both the oxidation currents and membrane depolarizations (Kruskal-Wallis test, P<0.01 for both oxidation currents and membrane depolarizations).

Figure 4.

Graph showing the effects of 30 min applications of various Ca²⁺ channel blockers on the amplitude of ω-CTX GVIA resistant oxidation currents and e.j.ps evoked by trains of 50 stimuli at 10 Hz. The % change for tissues treated with 10 μM nifedipine (n=6), 40 nM ω-AgTx IVA (n=6) or 0.5 μM ω-CTX MVIIC (n=6) is plotted relative to that for time matched ω-CTX GVIA (0.1 μM) treated controls (n=6). The s.e.mean bars for the control data are plotted on 100% line. ^*P<0.05, ^**P<0.01 (one-tailed test).

In comparison with ω-CTX GVIA treated control tissues, ω-CTX MVIIC and ω-AgTx IVA significantly reduced the amplitude of both the oxidation currents and membrane depolarizations (see Figure 4). This statistical analysis was made by a nonparametric comparison of the ω-CTX GVIA control group data to that of the other test groups (see Zar, 1984). Nifedipine had no inhibitory effect on either the oxidation currents or the membrane depolarizations. Comparison between the effects of each Ca²⁺ blocker on the oxidation currents and membrane depolarizations revealed no significant difference (Mann-Whitney U-tests; nifedipine P=0.26; ω-AgTx IVA P=0.08; ω-CTX MVIIC P=0.42).

Table 2 shows the effects of the Ca²⁺ channel blockers on both the time constant of decay of the stimulation-evoked membrane depolarizations and the smooth muscle membrane potential measured at the end of the electrophysiological experiments, following the addition of Cd²⁺ (0.1 mM). Comparison of the differences in time constant and membrane potential measured at the start and end of the experiments revealed no significant difference between the effects of the Ca²⁺ blockers (Kruskall-Wallis test P=0.07 for the time constant and 0.91 for the membrane potential).

Table 2.

Effects of Ca²⁺ channel blockers on the smooth muscle resting membrane potential and the time constant of decay of the membrane depolarization

Discussion

The present study demonstrates that both the NA-induced oxidation currents and the e.j.ps evoked by electrical stimulation of the perivascular sympathetic nerves supplying the tail artery have a similar dependence on extracellular Ca²⁺ concentration and sensitivity to the blocking action of a range of Ca²⁺ channel antagonists. As e.j.ps in the rat tail artery are most probably caused by neuronally released adenosine 5′-triphosphate (ATP) acting at P2X-purinoceptors (Sneddon & Burnstock, 1984; McLaren et al., 1995), the present findings indicate that the release of both NA and ATP is modulated in parallel.

A potential problem with the electrophysiology experiments investigating the effects of Ca²⁺ concentration on e.j.p amplitude, is that released ATP activates non-selective cation channels with a relatively high Ca²⁺ permeability (see Benham, 1989; Evans et al., 1996). However, it is unlikely that changing the Ca²⁺ concentration significantly affects the size of the ATP-gated unitary currents for the following reasons. (1) In vascular smooth muscle bathed in a physiological solution, Ca²⁺ carries only 5–10% of the inward current through ATP-activated channels (Benham, 1989). (2) The ATP-evoked current in cells homologously expressing the P2X₁-purinoceptor subunit, which is the predominant subunit present in vascular smooth muscle (Kennedy, 1996), is unaffected by changes in the extracellular Ca²⁺ concentration (0.3–10 mM, Evans et al., 1996). (3) The amplitudes of spontaneous excitatory junction currents recorded in guinea-pig vas deferens, which are believed to result from release of single quanta of ATP from the postganglionic sympathetic nerve terminals, are unaffected by changes in the extracellular Ca²⁺ concentration ( 0.3– 1.8 mM, Cunnane & Ziogas, 1990).

The Ca²⁺ channel blocker ω-CTX GVIA reduced the amplitude of both the oxidation currents and e.j.ps by about 90%, suggesting that neurotransmitter release from sympathetic nerves innervating the tail artery is triggered primarily by the opening of N-type Ca²⁺ channels. A similar finding has also been reported for rat tail artery labelled with [³H]NA, in which ω-CTX GVIA (0.1 μM) reduced the stimulus evoked release of tritium by about 90% (Clasbrummel et al., 1989).

In the presence of a maximally effective concentration of ω-CTX GVIA, electrical stimulation of the nerves still evoked oxidation currents and e.j.ps. These resistant responses were blocked by the non-selective Ca²⁺ channel blocker, Cd²⁺, demonstrating that they were due to Ca²⁺-dependent release of neurotransmitter. The resistant responses were also substantially reduced in amplitude by ω-CTX MVIIC (80–90% inhibited), which, under the conditions of the experiment, would be expected to block P- and Q-type Ca²⁺ channels (Randell & Tsien, 1995). A low concentration of ω-AgTx IVA (40 nM), which should fully block P-type Ca²⁺ channels (K_d ∼1 nM in rat cerebellar Purkinje neurons, Mintz & Bean, 1993), also reduced the amplitude of the ω-CTX GVIA resistant oxidation currents and membrane depolarizations. The L-type Ca²⁺ channel blocker, nifedipine, had no inhibitory effect on either the oxidation currents or membrane depolarizations. Comparison of the effects of ω-CTX MVIIC, ω-AgTx IVA and nifedipine on oxidation currents and e.j.ps revealed no significant difference. These findings indicate that ω-CTX GVIA resistant release of NA and ATP is due primarily to opening of P-/Q-type Ca²⁺ channels.

There was no significant difference in the magnitude of the inhibitory effect of ω-CTX GVIA on oxidation currents evoked by trains of ten pulses at 1 Hz and 50 pulses at 10 Hz, or on e.j.ps evoked by ten pulses at 1 Hz. The effects of ω-CTX GVIA on e.j.ps evoked by higher frequencies of stimulation were not investigated because, under control conditions, e.j.ps summate and trigger muscle action potentials. However, the findings with the NA-induced oxidation current suggest that the magnitude of the inhibitory effect of ω-CTX GVIA is not changed if the frequency of stimulation is altered. This finding contrasts with those of previous studies in which the postjunctional response has been used to monitor neurotransmitter release from postganglionic sympathetic nerves. For example, in mouse, rat and guinea-pig vas deferens, the ω-CTX GVIA resistant responses of the smooth muscle were only revealed when the frequency of stimulation was increased (Smith & Cunnane, 1996; Wright & Angus, 1996; Waterman, 1997). These findings have led to the conclusion that raising the frequency of stimulation recruits the ω-CTX GVIA resistant component of neurotransmitter release. This does not appear to be the case in rat tail artery.

The effects of ω-CTX GVIA on oxidation currents and e.j.ps in rat tail artery suggests that NA and ATP release are modulated in parallel. Similar conclusions have been made for the mouse, rat and guinea-pig vas deferens where ω-CTX GVIA had a similar inhibitory effect on the noradrenergic and purinergic components of the neurally evoked contraction (Maggi et al., 1988; Brock et al., 1989; De Luca et al., 1990; Waterman, 1997). Also in rat anococcygeus, both the slow (noradrenergic) and fast (purinergic) components of the e.j.p. evoked by sympathetic nerve stimulation were affected in a similar manner by a range of Ca²⁺ channel antagonists (Smith & Cunnane, 1997). It was therefore surprising that Westfall et al. (1996) reported that the release of endogenous NA in guinea-pig vas deferens was markedly inhibited by ω-CTX GVIA (0.1 μM) whereas that of endogenous ATP was only slightly inhibited. The release of ATP was substantially reduced by ω-AgTx IVA (0.1 μM). These findings for ATP release in the guinea-pig vas deferens are at odds with those for the purinergic component of the neurally evoked contraction and for e.j.ps which are very substantially reduced by ω-CTX GVIA (Maggi et al., 1988; Brock et al., 1989; Smith & Cunnane, 1996). In addition, in the guinea-pig vas deferens, e.j.ps are insensitive to ω-AgTx IVA (0.1 μM, Smith & Cunnane, 1996). Since the purinergic component of the neurally evoked contractions and the e.j.ps are believed to provide a measure of junctionally active ATP in guinea-pig vas deferens (see Stjärne, 1995), the ATP overflow detected in the study of Westfall et al. (1996) is unlikely to reflect that released at neuroeffector junctions.

In the present study it has been assumed that e.j.ps provide a measure of ATP release. However, if the ATP overflow detected in the study of Westfall et al. (1996) is shown to represent that released at neuroeffector junctions, then it must be concluded that the e.j.ps in the guinea-pig vas deferens are not due to released ATP. Both in muscular arteries and rodent vasa deferentia the e.j.ps evoked by electrical stimulation are resistant to α- and β-adrenoceptor antagonists but are blocked by a range of agents known to block or desensitize smooth muscle P2X-purinoceptors (see Brock and Cunnane, 1993). Furthermore, the non-cholinergic e.j.ps evoked in guinea-pig bladder detrusor smooth muscle when the parasympathetic nerves are excited are also blocked by agents that prevent activation of purinoceptors by ATP (Hashitani & Suzuki, 1995; Bramich & Brading, 1996). Taken together these findings provide strong supporting evidence for the view that e.j.ps evoked in muscular arteries and in rodent vasa deferentia are not due to released NA and are mediated through activation of purinoceptors. As ATP is predominant purine co-localized with NA in synaptic vesicles isolated from sympathetic nerve terminals (see Klein & Lagercrantz, 1981), it is reasonable to assume that the e.j.ps in these tissues are due to activation of purinoceptors by released ATP.

In conclusion, the results of the present study indicate that the release of NA and ATP from sympathetic nerves supplying the rat tail artery is similarly affected by changes that modify Ca²⁺ entry into the nerve terminals. Previous studies have also shown in tail artery that the amplitudes of oxidation currents and e.j.ps are increased and decreased in a parallel manner activating prejunctional β-adrenoceptors and α₂-adrenoceptors respectively (Msghina et al., 1992; Brock et al., 1997). Taken together these findings provide strong support for the hypothesis that NA and ATP are co-released together from the sympathetic nerve terminals.

Abbreviations

Excitatory junction potential

e.j.p.