Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

It is well documented that neurotransmitter release from a variety of central and peripheral nerve endings can be inhibited through the activation of presynaptic adenosine A₁- and histamine H₃-receptors [see reviews by Fredholm and Dunwiddie (1988) and Schlicker et al. (1994)]. These receptor types have also been described on cholinergic nerve terminals of the enteric nervous system where receptor stimulation leads to a decrease in acetylcholine (ACh) release [for example, Hew et al. (1990), Poli et al. (1993), Nitahara et al. (1995), and Barajas-Lopez et al. (1996)].

The coexistence of two or more populations of presynaptic receptors on the same neuron could provide the opportunity for interactions at the level of the signaling pathway(s) employed in the control of neurotransmitter release. Very recently, it has been demonstrated that A₁- and H₃-receptors interact in the modulation of neurogenic, cholinergic twitch contractions of the guinea pig duodenum (Poli et al., 1997). Thus, simultaneous activation of receptors with near maximal agonist concentrations produced an inhibitory response that was significantly less than the sum of effects produced when activated individually. This suggests that there is a common step in the transduction mechanism(s) coupled to A₁- and H₃-receptors that becomes saturated at high agonist concentrations.

At present, the intraneuronal events that occur in the inhibition of transmitter release after A₁- and H₃-receptor activation remain uncertain. A large amount of evidence suggests that these receptors are coupled to their effector(s) by the G_i family of guanosine nucleotide-binding proteins (G-proteins) based largely on the ability of pertussis toxin to either dramatically reduce or abolish the effects of receptor activation (Fredholm et al., 1990; Scholz and Miller, 1992; Endou et al., 1994; Schlicker et al., 1994; Celuch, 1995). However, the intraneuronal events that occur beyond the G-protein level have yet to be resolved completely. It is well accepted that the influx of Ca²⁺ through N-type Ca²⁺ channels into a neuron is vital for the process of neurotransmission (Miller, 1987). An obvious mechanism for inhibiting this process would be to reduce N-type channel conductance and diminish Ca²⁺ influx. An inverse relationship between intraneuronal Ca²⁺ accumulation and H₃- and A₁-receptor-mediated inhibition (Dowdle and Maske, 1980; Poli et al., 1994) supports this notion, whereas either potentiation or diminution of presynaptic neuromodulation with the N-type Ca²⁺ channel blocker, -conotoxin GVIA (CTX), confirms a prominent role for the N-type channel (Endou et al., 1994; Fossier et al., 1994; Mynlieff and Beam, 1994; Poli et al., 1994; Scholz and Miller, 1996). The involvement of other types of Ca²⁺ channel may also be important. For example, Ambrosio et al. (1996) found that A₁-receptor-mediated inhibition of glutamate release at rat striatal terminals was unmodified by CTX but was attenuated in the presence of the P/Q type blocker, -conotoxin MVIIC. Mechanisms independent of Ca²⁺ influx have also been proposed for the A₁-receptor (Scholz and Miller, 1992; Barajas-Lopez et al., 1996).

An unequivocal role for cyclic nucleotides has yet to be elucidated. For example, A₁-receptor-mediated suppression of cholinergic transmission in guinea pig myenteric ganglia has been associated with decreased intracellular cAMP levels (Borasio et al., 1995), whereas more recent work has shown the mechanism to be independent of the cAMP cascade (Barajas-Lopez et al., 1996). Similarly, elevating intracellular cyclic nucleotide levels has failed to modify the action of H₃-receptor agonists on central noradrenergic transmission (Celuch, 1995) or cholinergic transmission in the guinea pig duodenum (Poli et al., 1993). In contrast, activation of H₃-receptors on myenteric neurones of the guinea pig small intestine diminished forskolin-induced ACh release (Yau and Youther, 1994), indicating a role for cAMP.

The aim of the present investigation was to examine the nature of the signal-transduction mechanism(s) that couple H₃- and A₁-receptor activation to the modulation of cholinergic neuroeffector transmission in the guinea pig small intestine. This was achieved by evaluating the influence of selective Ca²⁺ channel blockade or altered intracellular cyclic nucleotide metabolism/activity on the action of adenosine and the H₃-selective agonist, R-()-methylhistamine (RAMH) (Arrang et al., 1987). Electrically evoked twitch contractions of the guinea pig ileum were used as a measure of ACh release.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Tissue Preparation and Stimulation. Dunkin-Hartley guinea pigs (600-800 g) of either sex that had previously been fasted overnight were stunned by a blow to the head and sacrificed by exsanguination. Two to four segments of 2-cm length were removed from the distal portion of the ileum and mounted in 25-ml organ baths containing Krebs' solution (composition in mM: NaCl, 118.3; KCl, 4.7; MgSO₄, 1.2; KH₂PO₄, 1.2; NaHCO₃, 25; glucose, 11.1; CaCl₂, 2.5). This was aerated with 95% O₂/5% CO₂ and maintained at 37°C. The tissues were allowed to equilibrate for 1 h under a resting tension of 1 g before commencing stimulation.

Effect of Adenosine and RAMH. To confirm the identity of the receptors being examined, cumulative concentration-response curves for the inhibition of electrically evoked contractions were established for adenosine and RAMH in the absence and presence of the receptor selective antagonists, 1,3-dipropyl-8-cyclopentylxanthine (DPCPX) and thioperamide, respectively (Arrang et al., 1987; Lohse et al., 1987). Repeat concentration-response curves were constructed within one tissue, and tissues were incubated with or without antagonists for 30 min before establishing the second curve.

Role of Ca²⁺ Channels and Cyclic Nucleotides. Some of the following manipulations predictably modified the twitch magnitude in a concentration- and/or time-dependent manner. Concentrations of agents were chosen that elicited a detectable (to indicate the effectiveness of the procedure) although minimal effect on the twitch response (typically, a 15 to 20% change in the twitch height compared with controls; see Tables 1 and 3). Single, submaximal concentrations of agonist were tested to minimize the contribution of any time-dependent changes in twitch magnitude caused by the manipulation being examined. Agonist-induced responses were allowed to plateau before washing out.

Statistical Analysis. All drug concentrations presented are final organ bath concentrations and volume of drug added did not exceed 1% of the total organ bath volume. Drug effects are expressed as percentage change of twitch contraction magnitude and calculated as 1  (force exerted in presence of drug/corresponding force exerted in absence of drug) × 100.

Drugs and Solutions. The following drugs were used: adenosine, atropine, 8-Br-cGMP, 2-chloroadenosine (2-CA), carbachol, CTX, NMHA, nifedipine, and tetrodotoxin (Sigma-Aldrich, Poole, UK); MDL-12330A and (R_p)-cAMPS (Alexis Corp. Ltd., Nottingham, UK); DPCPX, ODQ, and thioperamide maleate (Tocris-Cookson, Bristol, UK); DMPPO (synthesized and kindly donated by Glaxo-Wellcome, Les Ulis, France); RAMH (a kind donation from Bioprojet, Paris, France); Ro-20-1724 (Calbiochem-Novachem Ltd., Nottingham, UK); SKF-95654 and SKF-96231 (synthesized and kindly donated by SmithKline Beecham Pharmaceuticals, Harlow, UK).

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Effects of Adenosine and RAMH. Electrical stimulation of whole ileal segments produced reproducible twitch responses, which were abolished by either tetrodotoxin (0.1 µM) and atropine (1 µM), confirming their neurogenic, cholinergic nature. Both adenosine (0.2-60 µM) and RAMH (4-600 nM) caused a concentration-dependent depression of electrically evoked twitch contractions with respective maximal inhibitory responses of around 90 and 60% of original twitch contraction magnitude, and EC₅₀ values of 2.1 ± 0.4 µM (n = 9) and 15.2 ± 3.5 nM (n = 6). The A₁-receptor antagonist, DPCPX (20 nM), and the H₃-receptor antagonist, thioperamide (60 nM), antagonized the effect of adenosine and RAMH, respectively, with dissociation constants (K_B) of 8.9 ± 0.7 and 22.4 ± 0.7 nM. The agonists had no effect on comparable carbachol-evoked contractions.

Effect of Ca²⁺ Channel Blockers. Initial studies showed that CTX caused a slow, ongoing depression of electrically evoked twitch contractions. Thus, 10 nM CTX caused ~15% relaxation after 5-min incubation during a train of stimuli, which reached a maximum of ~90% after 30 min. Therefore, a 5-min incubation period was chosen for the agonist studies. Comparable contractile responses to carbachol were unaffected (Table 1). Nifedipine (20 nM) inhibited both carbachol- and electrically evoked contractile responses (Table 1). The inhibitory effects of adenosine and RAMH were significantly greater in the presence of CTX but were unmodified in the presence of nifedipine (Fig. 1 and Table 2).

View larger version (53K): [in this window] [in a new window]   Fig. 1.   Original traces illustrating the potentiation of adenosine- and RAMH-mediated inhibition of electrically evoked twitch contractions of the guinea pig ileum by CTX

Role of cAMP. MDL-12330A caused a small reduction in the amplitude of electrically evoked twitch contractions (Table 3) and significantly potentiated the inhibitory responses to both adenosine and RAMH compared with controls (Table 4). (R_p)-cAMPS also inhibited electrically evoked twitch contractions but did not modify the contractile response to exogenously applied carbachol (Table 3). The inhibitory responses to adenosine and RAMH were unchanged in the presence of (R_p)-cAMPS (Table 4). Although SKF-95654 did not influence either carbachol- or electrically evoked contraction amplitude, Ro-20-1724 reduced both contractile responses, being more effective on the response to carbachol (Table 3). SKF-95654 and Ro-20-1724 had no effect on the inhibitory actions of adenosine or RAMH (Table 4).

Role of cGMP. ODQ caused a small but significant potentiation of electrically evoked twitch contractions, whereas similar treatment with NMHA had no effect (Table 3). Neither drug affected carbachol-evoked contractions. The inhibitory effect of adenosine was unmodified by either NMHA or ODQ. In contrast, NMHA attenuated and ODQ increased the inhibitory action of RAMH (Table 5). Both SKF-96231 and DMPPO produced a small but significant reduction of the electrically evoked twitch response (Table 3). Administration of SKF-96231 onto a carbachol-induced plateau response caused a similar loss in muscle tone (Table 3). The PDE V inhibitors also caused a significant potentiation of the inhibitory response to adenosine, whereas the response induced by RAMH remained unaffected (Table 5).

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

The present study examines the signal transduction mechanisms mediating presynaptic adenosine A₁- and histamine H₃-receptor-mediated inhibition of cholinergic neurotransmission in the isolated guinea pig ileum. We have shown that activation of A₁- and H₃-receptors inhibited neurogenic, cholinergic twitch contractions of the ileum elicited by electrical stimulation. This confirms previous findings that these receptors can inhibit neuronal ACh release in the small intestine (Hew et al., 1990; Poli et al., 1993; Yau and Youther, 1994; Nitahara et al., 1995; Barajas-Lopez et al., 1996; Lee and Parsons, 1998).

N-type Ca²⁺ channels have been implicated in governing transmitter release from a variety of peripheral neurones in numerous tissues (Miller, 1987). ACh release from cholinergic neurones of the ileum are no exception, as shown in the current study. Thus, neurogenic twitch contractions were almost completely abolished by CTX at a concentration that had no effect on postsynaptic carbachol-induced contractions. In contrast, the L-type Ca²⁺ channel blocker, nifedipine, caused comparable inhibition of both pre- and postsynaptically mediated contractions. CTX also enhanced the neuroinhibitory effect of adenosine and RAMH. This was not due to the decrease in the contractile response caused by the toxin, because nifedipine, which alone caused a similar depression, had no effect on presynaptic inhibition. Instead, the data suggest that A₁- and H₃-receptors are targeting the same site as the toxin, i.e., the N-type Ca²⁺ channel.

A similar potentiation of H₃-receptor activity by N-type Ca²⁺ channel blockade has been observed in the inhibition of noradrenaline release in guinea pig atria (Endou et al., 1994) and ACh release in guinea pig duodenum (Poli et al., 1994). In other cases, A₁- and H₃-receptor-mediated inhibition has become less marked or absent in the presence of CTX (Fossier et al., 1994; Mynlieff and Beam, 1994). Nevertheless, both positive and negative cooperativity between agonist and toxin would suggest an interaction through a shared signaling pathway.

There is considerable evidence to suggest that the AC/cAMP signaling cascade is important in neurotransmission and neuromodulation. Thus, pharmacological interventions designed to elevate intracellular cAMP have been shown to enhance neurotransmitter release (Zhang et al., 1996), and increasing cAMP production is regarded as a principle pathway coupled to facilitatory presynaptic receptors (Majewski et al., 1990). Because A₁- and H₃-receptors are known to decrease neuronal AC activity (Yau and Youther, 1994; Borasio et al., 1995), we investigated whether a negative coupling with AC/cAMP could be involved in their inhibitory actions on ACh release.

The inhibitory effects of adenosine and RAMH were significantly greater in tissues pretreated with the AC inhibitor, MDL-12330A. By analogy to the results obtained with CTX earlier, this would suggest that A₁- and H₃-receptor activation and MDL-12330A are interacting at a common effector, presumably AC. If A₁- and H₃-receptor-mediated inhibition of neurotransmission occurs through decreasing AC activity, then elevating endogenous cAMP should counteract their effect. However, inhibiting cAMP metabolism with either the PDE III or PDE IV inhibitors, SKF-95654 and Ro-20-1724, respectively, had no effect on adenosine or RAMH, indicating a cAMP-independent mechanism. Despite being used at a concentration known to inhibit PDE III in whole tissue (Murray et al., 1992), SKF-95654 also failed to modify contractions to neurally released ACh or exogenously applied carbachol, suggesting that PDE III is unimportant in both pre- and postjunctional events at the neuromuscular junction. Therefore, it is perhaps not surprising that A₁- and H₃-receptor activation was unaffected by SKF-95654. Inhibiting PDE IV, on the other hand, caused ~15% reduction in neurally evoked twitch contraction magnitude. An increase in ACh release would have been expected, although this is probably being tempered functionally by the postsynaptic actions of Ro-20-1724, evidenced by a slightly greater reduction in carbachol tone (~25%). So, Ro-20-1724 appeared to be successful in promoting intracellular accumulation of cAMP yet had no effect on A₁- and H₃-receptor-mediated neuromodulation.

Experiments with (R_p)-cAMPS lend further support to a cAMP-independent mechanism. The PKA inhibitor caused a significant reduction in the magnitude of the electrically evoked twitch response but had no postsynaptic effects. Thus, the action of (R_p)-cAMPS was purely prejunctional and consistent with a blockade of cAMP, leading to diminished ACh release. Despite this, (R_p)-cAMPS had no effect on the actions of either adenosine or RAMH.

Taken together, the current data would suggest that, despite being used to define a role for AC in presynaptic modulation elsewhere (Silinsky, 1984; Correia-de-Sa and Ribeiro, 1994), the effect of MDL-12330A in this study appears to be unrelated to an action on the AC system. Moreover, A₁- and H₃-receptor activation is not linked to reducing cAMP during presynaptic inhibition. This is in good agreement with investigations on other neuronal systems (Fredholm et al., 1990; Poli et al., 1993; Celuch, 1995).

MDL-12330A has also been reported to inhibit cGMP PDE V (Hunt and Evans, 1980), which may explain the results reported earlier. This, coupled with evidence for the involvement of the GC/cGMP system in neurotransmitter release (for example, Greenberg et al., 1990), prompted us to examine the role of cGMP in mediating receptor-induced presynaptic inhibition.

Inhibiting cGMP metabolism with the PDE V inhibitors, SKF-96231 and DMPPO, resulted in a reduced twitch height caused by an action on the smooth muscle. The action of adenosine on ACh release was potentiated by both SKF-96231 and DMPPO, whereas that of RAMH was unaffected, suggesting a role for cGMP in A₁-receptor signal transduction. However, the PDE V inhibitors share a structural similarity to agents that block adenosine uptake, and uptake blockade has been demonstrated to potentiate the presynaptic effect of adenosine in the guinea pig ileum (Lee and Parsons, 1998). Because SKF-96231 attenuated the inhibitory effect of the A₁-agonist 2-CA, which is resistant to uptake (Daly, 1983), the effect of the PDE V inhibitors would appear to be due to uptake blockade. Moreover, mimicking a rise in intracellular cGMP with 8-Br-cGMP, or inhibiting GC with NMHA or ODQ had no effect on adenosine, suggesting that cGMP is not involved as a second messenger for A₁-receptors.

NMHA and ODQ did, however, exert opposing effects on H₃-receptor-mediated modulation, i.e., NMHA reduced, and ODQ potentiated, the inhibitory action of RAMH. This would suggest that the H₃-receptor is dependent on the GC/cGMP system, although the failure of either PDE V inhibition or 8-Br-cGMP to influence the inhibitory action of RAMH argues against this. The actions of NMHA and ODQ are difficult to explain. ODQ increased the magnitude of cholinergic twitch contractions without affecting postsynaptic responses to carbachol, suggesting that preventing endogenous cGMP production leads to increased ACh release. However, NMHA had no effect pre- or postsynaptically. This, coupled with their opposing influences on H₃-receptor function, would indicate two different mechanisms of action with at least one compound exerting an effect independent of GC. Garthwaite et al. (1995) initially characterized ODQ as a selective inhibitor of soluble GC, but in the same study ODQ also reduced atrial natriuretic factor-induced cGMP production, indicating an influence on the particulate isoform of the enzyme. Additionally, it has recently emerged that, in addition to soluble GC, ODQ can inhibit NO synthase and nitrovasodilator metabolism (Schmidt et al., 1998), although either of these actions would ultimately diminish soluble cyclase activity. NMHA has also been found to inhibit agonist-induced elevation of cGMP (Deguchi et al., 1978), which would also imply an inhibitory effect on particulate GC. It would, therefore, appear that NMHA and ODQ can influence the activity of both forms of the GC, but there could be marked differences in their relative selectivity. This and other so far uncharacterized effects may account for the disparate results recorded here.

In summary, the present study has confirmed the existence of presynaptic A₁- and H₃-receptors on parasympathetic nerve terminals in the guinea pig ileum, the activation of which inhibits electrically evoked neurogenic, cholinergic twitch contractions. Receptor activation probably limits the availability of intraneuronal Ca²⁺ for neurotransmitter release via restricted influx through N-type Ca²⁺ channels. This mechanism does not appear to occur through the intermediacy of AC/cAMP or GC/cGMP. The involvement of a second messenger system in presynaptic modulation may well be reserved for presynaptic facilitatory receptors.