Br. J. Pharmacol. (1991), 104, 817-822
X-1 Macmillan Press Ltd, 1991
Evidence for inhibition of sympathetic neurotransmission by endogenously released acetylcholine in the guinea-pig trachea 1Yvonne D. Pendry & 2Jennifer Maclagan Academic Department of Pharmacology, Royal Free Hospital School of Medicine, Rowland Hill Street, London NW3 2PF 1 Interactions between pulmonary cholinergic and noradrenergic nerves were studied in the innervated tracheal tube preparation isolated from guinea-pigs anaesthetized with urethane. Relaxations of the trachealis smooth muscle in response to postganglionic stimulation of the sympathetic nerve were recorded as decreases in the intraluminal pressure of the tracheal tube after the pressure had been raised with the stable thromboxane-mimetic, U46619. In contrast, contractions following preganglionic stimulation of the vagal nerve trunk were recorded as increases in intraluminal pressure. 2 In approximately half of the preparations studied, concurrent stimulation of the vagal nerve trunk inhibited relaxation responses elicited by stimulation of the sympathetic nerves. The vagi were stimulated at parameters which caused no change in intraluminal pressure, excluding the involvement of postjunctional mechanisms. 3 The effect of simultaneous stimulation of the sympathetic nerve trunk was studied on contractile responses evoked by preganglionic stimulation of the vagus nerve. In 80% of the preparations tested the vagal responses were inhibited. This inhibitory effect of sympathetic nerve stimulation was antagonized by propranolol. 4 The potassium channel agonist, cromakalim, endothelins 1 and 3 and the neuropeptides, vasoactive intestinal peptide, neurokinin A and substance P, did not significantly modulate sympathetic nerveinduced relaxations. 5 The anticholinesterase drug, physostigmine, induced a concentration-dependent increase in the intraluminal pressure of the tracheal tube and potentiated the postjunctional action of exogenously applied acetylcholine to contract the guinea-pig trachealis muscle. In the presence of higher concentrations of physostigmine both vagally-induced contractions and sympathetic nerve-induced relaxations were reduced. Atropine blocked both the inhibitory effect of physostigmine on sympathetic relaxations and its postjunctional contractile action on the trachealis smooth muscle. 6 It is concluded that, in the guinea-pig trachea, acetylcholine released endogenously from pulmonary parasympathetic nerves, either by anticholinesterase drugs or in response to nerve stimulation, can inhibit transmission in the adjacent sympathetic nerves via activation of prejunctional muscarinic heteroreceptors, probably of the M3 subtype. Keywords: Inhibitory muscarinic heteroreceptors; endogenous acetylcholine; sympathetic neurotransmission; guinea-pig trachea
Introduction The airways receive a dual innervation from both the parasympathetic and sympathetic nervous systems. In all species the dominant control of airway smooth muscle tone is exerted via the parasympathetic nerves, whereas the sympathetic innervation to airway smooth muscle is very speciesdependent and may be sparse or absent (Mann, 1971). However, even in species which lack a direct sympathetic supply to the airway smooth muscle, cholinergic and noradrenergic varicosities have been shown to lie in close apposition (Jones et al., 1980; Daniel et al., 1986). The close proximity of the two branches of the autonomic nervous systems suggests that they may interact at the prejunctional level and that either nerve may be able to modulate neurotransmission in the other type of nerve and thus modulate airway smooth muscle tone. This suggestion is supported by the observation that in the guinea-pig, stimulation of the sympathetic nerve trunk inhibits vagally-induced contractions of the innervated tracheal tube preparation via activation of f6adrenoceptors on the cholinergic nerve endings (McCaig, 1987). Similarly, in the dog (Cabezas et al., 1971) and cat (Baker & Don, 1987) sympathetic nerve stimulation inhibited the bronchoconstrictor effect of vagal stimulation in vivo. The reverse relationship, that is, modulation of sympathetic neurotransmission by acetylcholine released from the adjacent 1 Present address: Glaxo Group Research Limited, Ware, Hertfordshire SG12 ODP. 2 Author for correspondence.
parasympathetic nerves, has not yet been fully investigated in the airways. However, in other tissues which receive a dual innervation, such as the gut (Manber & Gershon, 1979) and heart (Loffelholz & Muscholl, 1970; Levy & Blattberg, 1976; Boyle & Pollock, 1988), there is evidence that endogenously released acetylcholine can inhibit noradrenaline release via activation of prejunctional muscarinic heteroreceptors on the sympathetic nerve terminals. We have previously demonstrated, using the guinea-pig isolated innervated tracheal tube preparation, that exogenously applied muscarinic agonists can inhibit noradrenaline release via activation of muscarinic M3 cholinoceptors on pulmonary sympathetic nerve terminals (Pendry & Maclagan, 1989; 1991). The aim of the present study was to determine whether these inhibitory muscarinic heteroreceptors could be activated by endogenous acetylcholine released from pulmonary parasympathetic nerves and whether stimulation of the parasympathetic nerves might inhibit the release of noradrenaline from the adjacent sympathetic nerves via this mechanism. A preliminary account of these findings has previously been reported to the British Pharmacological Society (Pendry & Maclagan, 1990).
Methods Male guinea-pigs (200400g) of the Dunkin-Hartley strain were killed with an anaesthetic overdose (urethane 1.5gkg- ' i.p.). The trachea with the sympathetic nerve trunk on the right hand side and both vagal nerves and recurrent laryngeal
818
Y.D. PENDRY & J. MACLAGAN
branches attached was removed. During the dissection, the pulmonary nerves were kept moist with Krebs-Henseleit solution (composition mM: NaCl2 118.4, KCl 4.7, NaHCO3 25.0, glucose 11.1, KH2PO4 1.16, MgSO47H20 1.19 and CaCI2 2.6) gassed with 95% 02 and 5% CO2. The trachea was cannulated at both ends and mounted horizontally, at its in vivo length, in an organ bath (25ml) containing gassed KrebsHenseleit solution maintained at 370C. The lumen of the trachea was filled with Krebs-Henseleit solution; one end was closed and the other end was attached to a Statham (P23AC) transducer to measure changes in the intraluminal pressure (ILP). Increases or decreases in ILP reflected smooth muscle contraction or relaxation, respectively. All ILPs were expressed as values in mmH20 above the minimum ILP in the presence of excess isoprenaline (10-3M) which was established at the end of each experiment. At the beginning of each experiment, tissues were left to equilibrate for 1 h in the presence of indomethacin (5 x 10-6M) to remove any prostaglandin-induced tone. Indomethacin was present in the Krebs-Henseleit solution throughout all experiments.
Effect of concurrent vagal and sympathetic stimulation The vagal nerve trunks and sympathetic stellate ganglion were stimulated through two separate bipolar platinum electrodes with trains of rectangular pulses delivered from two Grass S44 stimulators, via stimulator isolation units to avoid any 'leakage' of current. In the first part of this study, the tone of the guinea-pig trachealis muscle was raised with U46619 in order to observe relaxation responses following supramaximal postganglionic stimulation of the sympathetic nerve trunk (40 Hz max voltage, 0.2ms for 5s at 90s intervals) (Pendry & Maclagan, 1991). When the tone had reached a plateau, sympathetic nerve-induced relaxations were measured in the absence and presence of concurrent submaximal vagal stimulation (20Hz, 0.2ms for 180 or 270s). The stimulation voltage was chosen such that stimulation of the vagus alone caused no significant change in ILP in order to exclude involvement of physiological antagonism and of tone changes. In the second part of this study contractile responses following supramaximal preganglionic stimulation of the vagal bundles (30Hz, max voltage, 0.2ms for 5s at 45s intervals) were measured. When a stable response had been obtained, the effect of concurrent submaximal sympathetic stimulation (40 Hz, 0.2 ms, 90 s) on these vagally-induced contractions was investigated. Sympathetic parameters were chosen so that stimulation of the sympathetic nerve trunk alone did not alter the ILP, to exclude physiological antagonism.
Effect of an anticholinesterase drug The effect of the anticholinesterase, physostigmine, was studied (1) on the ILP, (2) on contractile responses of the guinea-pig trachea following addition of acetylcholine, and (3) on contractions evoked by stimulation of the preganglionic cholinergic nerve fibres (30 Hz, max voltage, 0.2 ms, 5 s). Sympathetic nerve-induced relaxations of the guinea-pig trachea (40 Hz, max voltage, 0.2 ms, 5 s) were also measured following at least 15min preincubation with physostigmine (10'- or 10-6M); in these experiments the intraluminal pressure was raised with U46619.
Drugs
Sigma/ Semat), physostigmine sulphate, urethane, indomethacin (Sigma), acetylcholine bromide, atropine sulphate (BDH), propranolol hydrochloride (ICI) and hexamethonium bromide (Koch Light), cromakalim (Beecham), neurokinin A, substance P, vasoactive intestinal polypeptide (VIP), endothelin 1 and 3 (Peninsula) and isoprenaline sulphate (Wellcome) were used. U46619 (1 la, 9a epoxymethano-prostaglandin H2;
Indomethacin (1mgmlP ) was dissolved daily in buffer, (KH2PO4 19.76mM and Na2HPO4 118.34mM; pH adjusted to 7.8 with NaOH by warming and sonication before addition to the Krebs-Henseleit solution) final concentration 5 x 10-6M. Stock solutions of U46619 were prepared by dissolving 1 mg in 1 ml ethanol and adding 0.9% saline to give a 1 mM solution. Cromakalim was dissolved in saline and sonicated for at least 10 min. Stock solutions of neurokinin A, substance P and VIP were prepared in distilled water. Endothelin 1 and 3 were dissolved in 0.001 M acetic acid. Aliquots of these stock solutions were stored at - 70°C and diluted daily in 0.9% saline. All resultant solutions were stored on ice throughout the experiment. All other drugs were dissolved and diluted in 0.9% saline. Stock solutions of isoprenaline were never kept for longer than 5 min and therefore no antioxidant was used.
Statistical analysis of data Sympathetic nerve-induced relaxations in the absence and presence of physostigmine, cromakalim and the peptides were compared by unpaired Student's t tests and were deemed significantly different when P < 0.05.
Results Effect of concurrent vagal and sympathetic stimulation Figure 1 shows an experiment in which the tone of the guineapig trachea was raised with U46619 (10-7M). As previously described, sympathetic nerve-induced relaxations increased as the ILP was increased (Pendry & Maclagan, 1991). When the tone had reached a plateau, repetitive stimulation at supramaximal voltage of the postganglionic sympathetic nerve trunk resulted in reproducible relaxations. When the vagal nerves were stimulated simultaneously with the sympathetic nerve trunk at submaximal voltages which caused no increase in ILP, inhibition of succeeding sympathetic relaxation responses was seen. This inhibitory effect of the parasympathetic nervous system on responses to sympathetic nerve stimulation was observed in approximately 50% of the preparations and was not dependent upon the reverse relationship (that is, inhibition of vagal responses by the sympathetic nervous system); nor was it related to stimulation of either the right or the left vagal nerve trunk. The inhibitory effect of the parasympathetic nervous system on the sympathetic neurotransmission was: (i) very variable between tissues; (ii) not reproducible in the same tissue; and (iii) declined with successive periods of stimulation. It was therefore not possible to investigate the effect of muscarinic antagonists on this inhibitory effect or to determine whether attenuation was greater Stimulate vagus 20 Hz 0.2 ms
-300 0 -200 MI
E E
-100 L _
,
-0 40 Hz 0.2 ms 5s
0
U46619 (100 lM
9go 80
s
~~~~~~~~~~~
Was
Wash
V max
Figure 1 Effect of concurrent stimulation of the vagal nerve trunk (shown by the brackets; preganglionic 20Hz, 0.2ms) on relaxation responses of the guinea-pig trachea induced by stimulation of the preganglionic sympathetic nerve (indicated by the dots; 40Hz, max voltage, 0.2ms, 5 s). Concurrent stimulation of the vagal nerve trunk did not alter the intraluminal pressure (ILP) but inhibited sympathetic nerve-induced relaxations.
PULMONARY VAGAL AND SYMPATHETIC NERVE INTERACTION Table I Effect of physostigmine on the intraluminal pressure (ILP) and postjunctional contractile action of acetylcholine to increase the tone of the guinea-pig trachealis smooth muscle
* 37% inhibition
a,
819
[120
80
Physostigmine concentration 40
-
I
E
120
E ..0)
1-
80
- 40
a,)
Propranolol present (10-5 6% inhibition
c
O
v0)
'0. Q C
4-.0
120
-80
-
*
30 Hz, 0.2 5 s, Vmax
* ms
I
40
*,
Stimulate sympathetic 40 Hz, 0.2 ms
0
45
90 S
Figure 2 Effect of concurrent stimulation of the sympathetic nerve trunk (shown by the brackets; 40 Hz, 0.2 ms), on contractile responses of the guinea-pig trachealis muscle induced by preganglionic stimulation of the vagal nerve trunk (shown by the dots; 30 Hz, max voltage, 0.2 ms, 5 s). (a) and (b) show two separate periods of concurrent sympathetic stimulation in the same tissue; (c) shows a period of concurrent sympathetic stimulation after the tissue had been preincubated with propranolol (10-1 M). Propranolol (10- M) antagonized the inhibitory effect of the sympathetic nervous system on contractions induced by vagal stimulation.
° 120 E E
C
100-
0 x
Z3
80-
a)
C.)-
CL
2 0-
100
(-logloM)
(mmH2O)
8.0 7.5 7.0 65 6.0
2.9 + 1.4 2.1 + 1.8 11.3 + 4.2 24.8 + 5.9 86.6 + 9.6
Increase (%) in contractile response to acetylcholine (10- M)
69.8 + 32 122.9 + 40
All values are means + s.e.mean of at least five experiments.
'co
.E I u
ILP increase
200
Intraluminal pressure (mmH2O)
Figure 3 Relaxations of the guinea-pig trachea induced by stimulation of the sympathetic nerve trunk (40Hz, max voltage, 0.2 ms, 5 s) in the absence (A) and presence of cromakalim 10- m (0) and 10- S M (-). The intraluminal pressure was raised with U46619.
when both vagi were stimulated simultaneously compared to the effect of stimulation of a single nerve. When vagal stimulation ceased, sympathetic nerve-induced relaxations returned immediately to control values (Figure 1). The reverse relationship is shown in Figure 2, where sympathetic nerve stimulation was superimposed on the effect of vagal nerve stimulation. Stimulation of the preganglionic vagal nerve fibres with supramaximal voltages resulted in reproducible contractile responses, which could be abolished with atropine. These contractile responses increased with increasing frequency, a maximum being observed at 30 Hz and were rapid in onset (0.5 s). In almost 80% of the preparations concurrent stimulations of the sympathetic nerve trunk resulted in a reduction in the size of the vagal contractile response. The degree of inhibition was reproducible in any one tissue. (Figure 2a, b). This inhibitory effect of the sympathetic nervous system on the parasympathetic nervous system was reduced by propranolol (10 5 M; Figure 2c). The neuropeptides, VIP (10-9-10-8 M; n = 4), neurokinin A (10-9-10-6M; n = 5) and substance P (108-10-6M; n = 3) and endothelins 1 and 3 (10-9-10-8M; n = 4) did not inhibit sympathetic nerve-induced relaxations; nor were such relaxations significantly different in the absence and presence of the potassium channel opener, cromakalim (10-6, 10-SM; Figure 3).
Effect of an anticholinesterase drug In the absence of nerve stimulation, the anticholinesterase drug, physostigmine (in the range 10-8 to 10-6M) caused a concentration-dependent increase in the ILP of the fluid-filled trachea which was slow to develop and reached a maximum after about 15min exposure, after which time the contraction was maintained or slowly declined. Physostigmine was also found to potentiate the postjunctional contractile action of exogenously applied acetylcholine in a concentrationdependent manner (Table 1). Figure 4 shows that physostignine at the lowest concentrations used (10'8, 3 x 10-8M) had very little effect on vagallyinduced contractions; at intermediate concentrations (10-', 3 x 10-7M) it caused an overall potentiation of vagallyinduced contractions, whereas in the presence of the highest concentration of physostigmine (10- 6M), vagally-induced contractions were reduced. Physostigmine, particularly at the higher concentrations, increased the duration of vagallyinduced contractions. The large error bars on Figure 5 are, however, an indication of the variability between tissues. The effect of physostigmine (10- 7M and 10-6 M) on sympathetic nerve-induced relaxations is shown in Figure 5. Sympa-
thetic relaxations were inhibited in the presence of physostigmine (10 6 M). In five experiments, atropine (10-7M) completely blocked both the inhibitory effect of physostigmine (10 -M) on sympathetic nerve-induced relaxations and its postjunctional action to contract the guinea-pig trachealis smooth muscle indicating the involvement of muscarinic receptors in both of these responses.
820
Y.D. PENDRY & J. MACLAGAN
20 -
t:> CU .C
10 CU
0
141t 0
~0 CU
020 -30
-40-... -8
-6
-7
[Physostigminel-log1o
M
Figure 4 Effect of physostigmine on contractions of the guinea-pig trachea induced by stimulation of the vagal nerve trunk (30Hz, max voltage, 0.2 ms, 5 s). Points are mean (n 5) with s.e.mean shown by vertical bars. >
N
E 150
E E 0
X.
Co
1001
4-
CL
0
Intraluminal pressure (mmH20) Figure 5 Sympathetic nerve-induced relaxation responses of the guinea-pig trachea (40 Hz, max voltage, 0.2 ms, 5 s), in the absence (A) and presence of physostigmine 10- 7m (A) and I 0- 'm (-). The intraluminal pressure was raised with U46619. Sympathetic nerve-induced relaxations were inhibited in the presence of the higher concentration of physostigmine (10-6M) (*p < 0.05; ***P < 0.005; mean of n > 5; s.e.mean shown by vertical bars).
Discussion
Interrelationships have been described between the two branches of the autonomic nervous system for tissues, such as the heart (Lofelholz & Muscholl, 1970; Boyle & Pollock, 1988) and ileum (Manber & Gershon, 1979), which receive a
dual innervation from both branches of the autonomic nervous system. Although in the airways a prejunctional on the effect of the sympathetic nervous the system inhibitory autonom-ic dualminetrvatedion fomr lborthoranches, very little parasympathetic nervous system has been reported, is known about the reverse relationship. It has, however, been trachea, exogenously applied muscarinic agonists can inhibit noradrenaline release via activation of muscarinic M3 cholino-
ceptors present on the sympathetic nerve terminals (Pendry & Maclagan, 1991). It was important to establish whether the inhibitory muscarinic receptors on pulmonary sympathetic nerve endings could be activated by acetylcholine released under physiological conditions from the adjacent cholinergic nerves. Using discrete stimulation of the pulmonary parasympathetic and sympathetic nerve trunks, we were able to confirm this interaction. Concurrent stimulation of the vagal nerve trunk with submaximal voltages inhibited sympathetic nerve-induced relaxations in 50% of the preparations. The involvement of postjunctional functional antagonism was excluded by stimulating the vagal nerve trunk at parameters which did not contract the trachealis smooth muscle. It proved to be very difficult to judge these parameters, and the lack of inhibition observed in half of the preparations may be due to variability between tissues in the metabolism of the transmitter by cholinesterase enzymes before reaching the muscarinic receptors on the airway smooth muscle and the sympathetic nerve endings. Stimulation of the preganglionic vagal nerve trunk not only stimulates the cholinergic fibres but also non-adrenergic, non-cholinergic (NANC) fibres which arise in the vagal bundle. The possibility that the inhibitory effect of concurrent vagal stimulation was due to release of neuropeptides from NANC nerves is, however, unlikely because when VIP, neurokinin A or substance P were administered exogenously, they did not inhibit sympathetic nerveinduced relaxations. In addition McCaig (1986) was unable to elicit vagally mediated NANC responses in the guinea-pig trachea in vitro during cholinergic blockade using similar stimulus parameters to those used in the present study. In McCaig's experiments, sustained stimulation with pulses of 2-4 ms duration and frequencies as great as 160 Hz were required to observe NANC responses. In contrast, in the experiments described in this paper the lower parameters of 0.2 ms and 20 Hz were sufficient to elicit an inhibitory effect of concurrent vagal stimulation upon sympathetic nerve-induced relaxations. Sympathetic nerve-induced relaxations were also unaffected by the presence of either endothelin 1 or endothelin 3, which are known to be formed by airway epithelial cells (Black et al., 1989). Thus it appears from the above results that, in the airways, acetylcholine released following preganglionic stimulation of the cholinergic nerves can activate inhibitory muscarinic receptors present on the adjacent sympathetic nerve endings. It is known that certain subtypes of muscarinic receptors which inhibit cell function are linked to the opening of potassium channals such as those studied in the heart (Sakmann et al., 1983). However, in the present experiments sympathetic nerve-induced relaxation responses were not altered in the presence of concentrations of the potassium channel opening drug, cromakalim which had previously been shown to inhibit vagally induced contraction by at least 50% (Hall & Maclagan, 1988). This suggests that the inhibitory muscarinic receptors on the pulmonary sympathetic nerves are not linked to potassium channels of the type opened by cromakalim, and a different second messenger system may be involved. Other workers have reported a prejunctional inhibitory effect of noradrenaline from sympathetic nerves on cholinergic transmission in the airways of various species (Vermiere & Vanhoutte, 1979; Jones et al., 1980; McCaig, 1987; Rhoden et al., 1988). We also recorded inhibition of vagally-induced contractile responses by noradrenaline released endogenously following concurrent stimulation of the sympathetic nerve trunk. The adrenoceptors involved, which are present on the cholinergic nerve endings, appear to be predominantly of the fl subtype. These findings are consistent with the observations that, in the airways, the two branches of the autonomic nervous system lie in close proximity and may interconnect (Jones et al., 1980; Daniel et al., 1986; Daniel, 1988). The present results confirm that functional interactions between these two nervous systems may also occur.
PULMONARY VAGAL AND SYMPATHETIC NERVE INTERACTION
As we were able to show that endogenous acetylcholine can inhibit sympathetic neurotransmission, anticholinesterase drugs would be expected to potentiate this interaction. It is well known that anticholinesterase drugs, such as physostigmine and neostigmine, cause contraction of the trachealis muscle. It has been suggested that this contractile effect may not depend totally on their ability to prevent breakdown of acetylcholine but may also be due to potentiation of acetylcholine release from cholinergic nerve endings (Douglas, 1951; Carlyle, 1963; Kirkpatrick & Rooney, 1979). Although potentiation of vagally-induced contractions of the guinea-pig trachea by the anticholinesterases, neostigmine and physostigmine, has been reported (McCaig, 1986; Widmark & Waldeck, 1986), other workers have described inhibition of potassium-evoked release of [3H]-acetylcholine from the rat bronchi by the anticholinesterase, soman (Aas et al., 1986; 1987). In our experiments, the anticholinesterase, physostigmine, potentiated contractions at high concentrations and caused inhibition at low concentrations. Previous conflicting reports in the literature may be due to the concentration of the anticholinesterase used and to the contact time with the tissue. Aas (1988) found that exposure for 40h to the anticholinesterase, soman, reduces the acetylcholine-induced contraction of the rat bronchi, probably via a reduction in the number of muscarinic receptors. However, the inhibition of vagallyinduced contractions following 15 min preincubation with 10-6M physostigmine, in our experiments, does not appear to be due to desensitization of postjunctional muscarinic receptors, as 1O- 6M physostigmine caused a greater increase in the postjunctional contractile action of acetylcholine than was obtained with the lower concentration of 10-7M physostigmine. The inhibitory effect is therefore probably a result of
821
acetylcholine accumulating in the synapse and exerting a negative feedback via muscarinic autoreceptors on the cholinergic nerve endings so inhibiting its own release. This complication has previously been observed in gut experiments involving measurement of acetylcholine release in the presence of anticholinesterases to prevent its breakdown (Kilbinger, 1984). The importance of our finding is that the concentration of physostigmine which inhibits sympathetic nerve-induced relaxations is in the same concentration range as that which inhibits vagally-induced contractile responses. This inhibitory effect of physostigmine was blocked by atropine (10-7M). Similarly, McCaig (1986) reported an inhibitory effect of the anticholinesterase, neostigmine, on sympathetic relaxations, which was shown to be reduced by atropine. The inhibitory effect of the anticholinesterase drugs on sympathetic responses is likely to be due to accumulation of acetylcholine which activates inhibitory muscarinic heteroreceptors on the nearby sympathetic nerve terminals. This accumulation of transmitter is probably due to a combination of the anticholinesterase effects of physostigmine on spontaneously released ACh and a direct stimulatory effect of the drug on the cholinergic nerve terminals. This sympathoinhibitory effect may contribute to the bronchospasm observed with anticholinesterase drugs. The results described in this paper reinforce the hypothesis that in the airways, parasympathetic and sympathetic nerves can interact at the prejunctional level. Thus, either branch of the autonomic nervous system can modulate neurotransmission in the other branch, increasing the complexity of the neural control of airway smooth muscle tone. We wish to acknowledge the financial support of Pfizer Central Research, Sandwich, Kent.
References AAS, P. (1988). The toxic effect of an acetylcholinesterase-inhibitor on the cholinergic nervous system in airway smooth muscle. Toxicol., 49, 91-97. AAS, P., MALMEI, T. & FONNUM, F. (1987). The effect of soman on potassium evoked 3H-acetylcholine release in the isolated rat bronchi. Pharmacol. Toxicol., 60, 206-209. AAS, P., VEITEBERG, T. & FONNUM, F. (1986). In vitro effects of soman on bronchial smooth muscle. Biochem. Pharmacol., 35, 1793-1799. BAKER, D.G. & DON, H. (1987). Catecholamines abolish vagal but not acetylcholine tone in the cat trachea. J. Appl. Physiol., 63, 24902498. BLACK, P.N., GHATEI, M.A., TAKAHASHI, K., BRETHERTON-WATT, D., KRAUSZ, T., DOLLERY, C.T. & BLOOM, S.A. (1989). Formation of endothelin by cultured airway epithelial cells. FEBS, 255, 129132. BOYLE, S.J. & POLLOCK, D. (1988). Interactions between cholinergic and noradrenaline nerves in the rat isolated atria. Br. J. Pharmacol., 93, 25P. CABEZAS, G.A., GRAF, P.D. & NADEL, J.A. (1971). Sympathetic versus parasympathetic nervous regulation of airways in dogs. J. Appl. Physiol., 31, 651-655. CARLYLE, R.F. (1963). The mode of action of neostigmine and physiostigmine on the guinea-pig trachealis muscle. Br. J. Pharmacol. Chemother., 21, 137-149. DANIEL, E.E. (1988). Ultrastructure of airway smooth muscle. In Mechanisms in Asthma: Pharmacology, Physiology and Management. ed. Armour, C.L. & Black, J.L. pp. 179-203. New York: Alan R. Liss, Inc. DANIEL, E.E., KANNAN, M., DAVIES, C. & POSEY-DANIEL, V. (1986). Ultrastructural studies on the neuromuscular control of human tracheal and bronchial muscle. Respir. Physiol., 63, 109-128. DOUGLAS, W.W. (1951). The effect of some anticholinesterase drugs on isolated tracheal muscle of the guinea-pig. J. Physiol., 112, 20P. HALL, A.K. & MACLAGAN, J. (1988). Effect of cromakalin on cholinergic neurotransmission in the guinea-pig trachea. Br. J. Pharmacol., 95, 792P. JONES, T.R., KANNAN, M.S. & DANIEL, E.E. (1980). Ultrastructural study of guinea-pig tracheal smooth muscle and its innervation. Can. J. Physiol. Pharmacol., 8, 974-983.
KILBINGER, H. (1984). Presynaptic muscarinic receptors modulating acetylcholine release. Trends Biochem. Sci., 5, 103-105. KIRKPATRICK, C.T. & ROONEY, PJ. (1979). The contracture produced in tracheal smooth muscle by anticholinesterase. Br. J. Pharmacol., 67, 646P. LEVY, M.N. & BLATTBERG, B. (1976). Effect of vagal stimulation on the overflow of norepinephrine into the coronary sinus during cardiac sympathetic nerve stimulation in the dog. Circ. Res., 38, 81-85. LOFFELHOLZ, K. & MUSCHOLL, E. (1970). Inhibition by parasympathetic nerve stimulation of the release of the adrenergic transmitter. Naunyn-Schmiedebergs Arch. Pharmacol., 267, 181-184. MANBER, L. & GERSHON, M.D. (1979). A reciprocal adrenergiccholinergic axoaxonic synapse in the mammalian gut. Am. J. Physiol., 236, E738-E745. MANN, S.P. (1971). The innervation of mammalian bronchial smooth muscle, the localization of catecholamines and cholinesterases. Histochem. J., 3, 319-331. McCAIG, D.J. (1986). Autonomic responses of the isolated, innervated trachea of the guinea-pig: interaction with autonomic drugs, histamine and 5-hydroxytryptamine. Br. J. Pharmacol., 88, 239-248. McCAIG, D.J. (1987). Effect of sympathetic stimulation and applied catecholamines on mechanical and electrical responses to stimulation of the vagus nerve in guinea-pig isolated trachea. Br. J: Pharmacol., 91, 385-394. PENDRY, Y.D. & MACLAGAN, J. (1989). Evidence for prejunctional muscarinic receptors on pulmonary sympathetic nerves in the guinea-pig. Br. J. Pharmacol., 98, 777P. PENDRY, Y.D. & MACLAGAN, J. (1990). Vagally-released acetylcholine inhibits sympathetic neurotransmission in the guinea-pig trachea. Br. J. Pharmacol., 101, 590P. PENDRY, Y.D. & MACLAGAN, J. (1991). Evidence for prejunctional inhibitory muscarinic receptors on sympathetic nerves innervating guinea-pig trachealis muscle. Br. J. Pharmacol., 103, 1165-1171. RHODEN, K.J., MELDRUM, L.A. & BARNES, PJ. (1988). Inhibition of cholinergic neurotransmission in human airways by 162-adrenoceptors. J. Appl. Biol., 65, 700-705. SAKMANN, B., NOMA, A. & TRAUTWEIN, W. (1983). Acetylcholine activation of single muscarinic K+ channels in isolated pacemaker cells of the mammalian heart. Nature, 303, 250-253.
822
Y.D. PENDRY & J. MACLAGAN
VERMIERE, P.A. & VANHOUTTE, P.M. (1979). Inhibitory effects of catecholamines in isolated canine bronchial smooth muscle. J.
Appl. Phys., 46, 787-791.
WIDMARK, E. & WALDECK, B. (1986). Physiological and pharmacological characterization of an in vitro vagus nerve-trachea preparation from guinea-pig. J. Auton. Pharmacol., 6, 187-194. (Received March 19, 1991 Revised July 17, 1991 Accepted August 5, 1991)
X-1 Macmillan Press Ltd, 1991
Evidence for inhibition of sympathetic neurotransmission by endogenously released acetylcholine in the guinea-pig trachea 1Yvonne D. Pendry & 2Jennifer Maclagan Academic Department of Pharmacology, Royal Free Hospital School of Medicine, Rowland Hill Street, London NW3 2PF 1 Interactions between pulmonary cholinergic and noradrenergic nerves were studied in the innervated tracheal tube preparation isolated from guinea-pigs anaesthetized with urethane. Relaxations of the trachealis smooth muscle in response to postganglionic stimulation of the sympathetic nerve were recorded as decreases in the intraluminal pressure of the tracheal tube after the pressure had been raised with the stable thromboxane-mimetic, U46619. In contrast, contractions following preganglionic stimulation of the vagal nerve trunk were recorded as increases in intraluminal pressure. 2 In approximately half of the preparations studied, concurrent stimulation of the vagal nerve trunk inhibited relaxation responses elicited by stimulation of the sympathetic nerves. The vagi were stimulated at parameters which caused no change in intraluminal pressure, excluding the involvement of postjunctional mechanisms. 3 The effect of simultaneous stimulation of the sympathetic nerve trunk was studied on contractile responses evoked by preganglionic stimulation of the vagus nerve. In 80% of the preparations tested the vagal responses were inhibited. This inhibitory effect of sympathetic nerve stimulation was antagonized by propranolol. 4 The potassium channel agonist, cromakalim, endothelins 1 and 3 and the neuropeptides, vasoactive intestinal peptide, neurokinin A and substance P, did not significantly modulate sympathetic nerveinduced relaxations. 5 The anticholinesterase drug, physostigmine, induced a concentration-dependent increase in the intraluminal pressure of the tracheal tube and potentiated the postjunctional action of exogenously applied acetylcholine to contract the guinea-pig trachealis muscle. In the presence of higher concentrations of physostigmine both vagally-induced contractions and sympathetic nerve-induced relaxations were reduced. Atropine blocked both the inhibitory effect of physostigmine on sympathetic relaxations and its postjunctional contractile action on the trachealis smooth muscle. 6 It is concluded that, in the guinea-pig trachea, acetylcholine released endogenously from pulmonary parasympathetic nerves, either by anticholinesterase drugs or in response to nerve stimulation, can inhibit transmission in the adjacent sympathetic nerves via activation of prejunctional muscarinic heteroreceptors, probably of the M3 subtype. Keywords: Inhibitory muscarinic heteroreceptors; endogenous acetylcholine; sympathetic neurotransmission; guinea-pig trachea
Introduction The airways receive a dual innervation from both the parasympathetic and sympathetic nervous systems. In all species the dominant control of airway smooth muscle tone is exerted via the parasympathetic nerves, whereas the sympathetic innervation to airway smooth muscle is very speciesdependent and may be sparse or absent (Mann, 1971). However, even in species which lack a direct sympathetic supply to the airway smooth muscle, cholinergic and noradrenergic varicosities have been shown to lie in close apposition (Jones et al., 1980; Daniel et al., 1986). The close proximity of the two branches of the autonomic nervous systems suggests that they may interact at the prejunctional level and that either nerve may be able to modulate neurotransmission in the other type of nerve and thus modulate airway smooth muscle tone. This suggestion is supported by the observation that in the guinea-pig, stimulation of the sympathetic nerve trunk inhibits vagally-induced contractions of the innervated tracheal tube preparation via activation of f6adrenoceptors on the cholinergic nerve endings (McCaig, 1987). Similarly, in the dog (Cabezas et al., 1971) and cat (Baker & Don, 1987) sympathetic nerve stimulation inhibited the bronchoconstrictor effect of vagal stimulation in vivo. The reverse relationship, that is, modulation of sympathetic neurotransmission by acetylcholine released from the adjacent 1 Present address: Glaxo Group Research Limited, Ware, Hertfordshire SG12 ODP. 2 Author for correspondence.
parasympathetic nerves, has not yet been fully investigated in the airways. However, in other tissues which receive a dual innervation, such as the gut (Manber & Gershon, 1979) and heart (Loffelholz & Muscholl, 1970; Levy & Blattberg, 1976; Boyle & Pollock, 1988), there is evidence that endogenously released acetylcholine can inhibit noradrenaline release via activation of prejunctional muscarinic heteroreceptors on the sympathetic nerve terminals. We have previously demonstrated, using the guinea-pig isolated innervated tracheal tube preparation, that exogenously applied muscarinic agonists can inhibit noradrenaline release via activation of muscarinic M3 cholinoceptors on pulmonary sympathetic nerve terminals (Pendry & Maclagan, 1989; 1991). The aim of the present study was to determine whether these inhibitory muscarinic heteroreceptors could be activated by endogenous acetylcholine released from pulmonary parasympathetic nerves and whether stimulation of the parasympathetic nerves might inhibit the release of noradrenaline from the adjacent sympathetic nerves via this mechanism. A preliminary account of these findings has previously been reported to the British Pharmacological Society (Pendry & Maclagan, 1990).
Methods Male guinea-pigs (200400g) of the Dunkin-Hartley strain were killed with an anaesthetic overdose (urethane 1.5gkg- ' i.p.). The trachea with the sympathetic nerve trunk on the right hand side and both vagal nerves and recurrent laryngeal
818
Y.D. PENDRY & J. MACLAGAN
branches attached was removed. During the dissection, the pulmonary nerves were kept moist with Krebs-Henseleit solution (composition mM: NaCl2 118.4, KCl 4.7, NaHCO3 25.0, glucose 11.1, KH2PO4 1.16, MgSO47H20 1.19 and CaCI2 2.6) gassed with 95% 02 and 5% CO2. The trachea was cannulated at both ends and mounted horizontally, at its in vivo length, in an organ bath (25ml) containing gassed KrebsHenseleit solution maintained at 370C. The lumen of the trachea was filled with Krebs-Henseleit solution; one end was closed and the other end was attached to a Statham (P23AC) transducer to measure changes in the intraluminal pressure (ILP). Increases or decreases in ILP reflected smooth muscle contraction or relaxation, respectively. All ILPs were expressed as values in mmH20 above the minimum ILP in the presence of excess isoprenaline (10-3M) which was established at the end of each experiment. At the beginning of each experiment, tissues were left to equilibrate for 1 h in the presence of indomethacin (5 x 10-6M) to remove any prostaglandin-induced tone. Indomethacin was present in the Krebs-Henseleit solution throughout all experiments.
Effect of concurrent vagal and sympathetic stimulation The vagal nerve trunks and sympathetic stellate ganglion were stimulated through two separate bipolar platinum electrodes with trains of rectangular pulses delivered from two Grass S44 stimulators, via stimulator isolation units to avoid any 'leakage' of current. In the first part of this study, the tone of the guinea-pig trachealis muscle was raised with U46619 in order to observe relaxation responses following supramaximal postganglionic stimulation of the sympathetic nerve trunk (40 Hz max voltage, 0.2ms for 5s at 90s intervals) (Pendry & Maclagan, 1991). When the tone had reached a plateau, sympathetic nerve-induced relaxations were measured in the absence and presence of concurrent submaximal vagal stimulation (20Hz, 0.2ms for 180 or 270s). The stimulation voltage was chosen such that stimulation of the vagus alone caused no significant change in ILP in order to exclude involvement of physiological antagonism and of tone changes. In the second part of this study contractile responses following supramaximal preganglionic stimulation of the vagal bundles (30Hz, max voltage, 0.2ms for 5s at 45s intervals) were measured. When a stable response had been obtained, the effect of concurrent submaximal sympathetic stimulation (40 Hz, 0.2 ms, 90 s) on these vagally-induced contractions was investigated. Sympathetic parameters were chosen so that stimulation of the sympathetic nerve trunk alone did not alter the ILP, to exclude physiological antagonism.
Effect of an anticholinesterase drug The effect of the anticholinesterase, physostigmine, was studied (1) on the ILP, (2) on contractile responses of the guinea-pig trachea following addition of acetylcholine, and (3) on contractions evoked by stimulation of the preganglionic cholinergic nerve fibres (30 Hz, max voltage, 0.2 ms, 5 s). Sympathetic nerve-induced relaxations of the guinea-pig trachea (40 Hz, max voltage, 0.2 ms, 5 s) were also measured following at least 15min preincubation with physostigmine (10'- or 10-6M); in these experiments the intraluminal pressure was raised with U46619.
Drugs
Sigma/ Semat), physostigmine sulphate, urethane, indomethacin (Sigma), acetylcholine bromide, atropine sulphate (BDH), propranolol hydrochloride (ICI) and hexamethonium bromide (Koch Light), cromakalim (Beecham), neurokinin A, substance P, vasoactive intestinal polypeptide (VIP), endothelin 1 and 3 (Peninsula) and isoprenaline sulphate (Wellcome) were used. U46619 (1 la, 9a epoxymethano-prostaglandin H2;
Indomethacin (1mgmlP ) was dissolved daily in buffer, (KH2PO4 19.76mM and Na2HPO4 118.34mM; pH adjusted to 7.8 with NaOH by warming and sonication before addition to the Krebs-Henseleit solution) final concentration 5 x 10-6M. Stock solutions of U46619 were prepared by dissolving 1 mg in 1 ml ethanol and adding 0.9% saline to give a 1 mM solution. Cromakalim was dissolved in saline and sonicated for at least 10 min. Stock solutions of neurokinin A, substance P and VIP were prepared in distilled water. Endothelin 1 and 3 were dissolved in 0.001 M acetic acid. Aliquots of these stock solutions were stored at - 70°C and diluted daily in 0.9% saline. All resultant solutions were stored on ice throughout the experiment. All other drugs were dissolved and diluted in 0.9% saline. Stock solutions of isoprenaline were never kept for longer than 5 min and therefore no antioxidant was used.
Statistical analysis of data Sympathetic nerve-induced relaxations in the absence and presence of physostigmine, cromakalim and the peptides were compared by unpaired Student's t tests and were deemed significantly different when P < 0.05.
Results Effect of concurrent vagal and sympathetic stimulation Figure 1 shows an experiment in which the tone of the guineapig trachea was raised with U46619 (10-7M). As previously described, sympathetic nerve-induced relaxations increased as the ILP was increased (Pendry & Maclagan, 1991). When the tone had reached a plateau, repetitive stimulation at supramaximal voltage of the postganglionic sympathetic nerve trunk resulted in reproducible relaxations. When the vagal nerves were stimulated simultaneously with the sympathetic nerve trunk at submaximal voltages which caused no increase in ILP, inhibition of succeeding sympathetic relaxation responses was seen. This inhibitory effect of the parasympathetic nervous system on responses to sympathetic nerve stimulation was observed in approximately 50% of the preparations and was not dependent upon the reverse relationship (that is, inhibition of vagal responses by the sympathetic nervous system); nor was it related to stimulation of either the right or the left vagal nerve trunk. The inhibitory effect of the parasympathetic nervous system on the sympathetic neurotransmission was: (i) very variable between tissues; (ii) not reproducible in the same tissue; and (iii) declined with successive periods of stimulation. It was therefore not possible to investigate the effect of muscarinic antagonists on this inhibitory effect or to determine whether attenuation was greater Stimulate vagus 20 Hz 0.2 ms
-300 0 -200 MI
E E
-100 L _
,
-0 40 Hz 0.2 ms 5s
0
U46619 (100 lM
9go 80
s
~~~~~~~~~~~
Was
Wash
V max
Figure 1 Effect of concurrent stimulation of the vagal nerve trunk (shown by the brackets; preganglionic 20Hz, 0.2ms) on relaxation responses of the guinea-pig trachea induced by stimulation of the preganglionic sympathetic nerve (indicated by the dots; 40Hz, max voltage, 0.2ms, 5 s). Concurrent stimulation of the vagal nerve trunk did not alter the intraluminal pressure (ILP) but inhibited sympathetic nerve-induced relaxations.
PULMONARY VAGAL AND SYMPATHETIC NERVE INTERACTION Table I Effect of physostigmine on the intraluminal pressure (ILP) and postjunctional contractile action of acetylcholine to increase the tone of the guinea-pig trachealis smooth muscle
* 37% inhibition
a,
819
[120
80
Physostigmine concentration 40
-
I
E
120
E ..0)
1-
80
- 40
a,)
Propranolol present (10-5 6% inhibition
c
O
v0)
'0. Q C
4-.0
120
-80
-
*
30 Hz, 0.2 5 s, Vmax
* ms
I
40
*,
Stimulate sympathetic 40 Hz, 0.2 ms
0
45
90 S
Figure 2 Effect of concurrent stimulation of the sympathetic nerve trunk (shown by the brackets; 40 Hz, 0.2 ms), on contractile responses of the guinea-pig trachealis muscle induced by preganglionic stimulation of the vagal nerve trunk (shown by the dots; 30 Hz, max voltage, 0.2 ms, 5 s). (a) and (b) show two separate periods of concurrent sympathetic stimulation in the same tissue; (c) shows a period of concurrent sympathetic stimulation after the tissue had been preincubated with propranolol (10-1 M). Propranolol (10- M) antagonized the inhibitory effect of the sympathetic nervous system on contractions induced by vagal stimulation.
° 120 E E
C
100-
0 x
Z3
80-
a)
C.)-
CL
2 0-
100
(-logloM)
(mmH2O)
8.0 7.5 7.0 65 6.0
2.9 + 1.4 2.1 + 1.8 11.3 + 4.2 24.8 + 5.9 86.6 + 9.6
Increase (%) in contractile response to acetylcholine (10- M)
69.8 + 32 122.9 + 40
All values are means + s.e.mean of at least five experiments.
'co
.E I u
ILP increase
200
Intraluminal pressure (mmH2O)
Figure 3 Relaxations of the guinea-pig trachea induced by stimulation of the sympathetic nerve trunk (40Hz, max voltage, 0.2 ms, 5 s) in the absence (A) and presence of cromakalim 10- m (0) and 10- S M (-). The intraluminal pressure was raised with U46619.
when both vagi were stimulated simultaneously compared to the effect of stimulation of a single nerve. When vagal stimulation ceased, sympathetic nerve-induced relaxations returned immediately to control values (Figure 1). The reverse relationship is shown in Figure 2, where sympathetic nerve stimulation was superimposed on the effect of vagal nerve stimulation. Stimulation of the preganglionic vagal nerve fibres with supramaximal voltages resulted in reproducible contractile responses, which could be abolished with atropine. These contractile responses increased with increasing frequency, a maximum being observed at 30 Hz and were rapid in onset (0.5 s). In almost 80% of the preparations concurrent stimulations of the sympathetic nerve trunk resulted in a reduction in the size of the vagal contractile response. The degree of inhibition was reproducible in any one tissue. (Figure 2a, b). This inhibitory effect of the sympathetic nervous system on the parasympathetic nervous system was reduced by propranolol (10 5 M; Figure 2c). The neuropeptides, VIP (10-9-10-8 M; n = 4), neurokinin A (10-9-10-6M; n = 5) and substance P (108-10-6M; n = 3) and endothelins 1 and 3 (10-9-10-8M; n = 4) did not inhibit sympathetic nerve-induced relaxations; nor were such relaxations significantly different in the absence and presence of the potassium channel opener, cromakalim (10-6, 10-SM; Figure 3).
Effect of an anticholinesterase drug In the absence of nerve stimulation, the anticholinesterase drug, physostigmine (in the range 10-8 to 10-6M) caused a concentration-dependent increase in the ILP of the fluid-filled trachea which was slow to develop and reached a maximum after about 15min exposure, after which time the contraction was maintained or slowly declined. Physostigmine was also found to potentiate the postjunctional contractile action of exogenously applied acetylcholine in a concentrationdependent manner (Table 1). Figure 4 shows that physostignine at the lowest concentrations used (10'8, 3 x 10-8M) had very little effect on vagallyinduced contractions; at intermediate concentrations (10-', 3 x 10-7M) it caused an overall potentiation of vagallyinduced contractions, whereas in the presence of the highest concentration of physostigmine (10- 6M), vagally-induced contractions were reduced. Physostigmine, particularly at the higher concentrations, increased the duration of vagallyinduced contractions. The large error bars on Figure 5 are, however, an indication of the variability between tissues. The effect of physostigmine (10- 7M and 10-6 M) on sympathetic nerve-induced relaxations is shown in Figure 5. Sympa-
thetic relaxations were inhibited in the presence of physostigmine (10 6 M). In five experiments, atropine (10-7M) completely blocked both the inhibitory effect of physostigmine (10 -M) on sympathetic nerve-induced relaxations and its postjunctional action to contract the guinea-pig trachealis smooth muscle indicating the involvement of muscarinic receptors in both of these responses.
820
Y.D. PENDRY & J. MACLAGAN
20 -
t:> CU .C
10 CU
0
141t 0
~0 CU
020 -30
-40-... -8
-6
-7
[Physostigminel-log1o
M
Figure 4 Effect of physostigmine on contractions of the guinea-pig trachea induced by stimulation of the vagal nerve trunk (30Hz, max voltage, 0.2 ms, 5 s). Points are mean (n 5) with s.e.mean shown by vertical bars. >
N
E 150
E E 0
X.
Co
1001
4-
CL
0
Intraluminal pressure (mmH20) Figure 5 Sympathetic nerve-induced relaxation responses of the guinea-pig trachea (40 Hz, max voltage, 0.2 ms, 5 s), in the absence (A) and presence of physostigmine 10- 7m (A) and I 0- 'm (-). The intraluminal pressure was raised with U46619. Sympathetic nerve-induced relaxations were inhibited in the presence of the higher concentration of physostigmine (10-6M) (*p < 0.05; ***P < 0.005; mean of n > 5; s.e.mean shown by vertical bars).
Discussion
Interrelationships have been described between the two branches of the autonomic nervous system for tissues, such as the heart (Lofelholz & Muscholl, 1970; Boyle & Pollock, 1988) and ileum (Manber & Gershon, 1979), which receive a
dual innervation from both branches of the autonomic nervous system. Although in the airways a prejunctional on the effect of the sympathetic nervous the system inhibitory autonom-ic dualminetrvatedion fomr lborthoranches, very little parasympathetic nervous system has been reported, is known about the reverse relationship. It has, however, been trachea, exogenously applied muscarinic agonists can inhibit noradrenaline release via activation of muscarinic M3 cholino-
ceptors present on the sympathetic nerve terminals (Pendry & Maclagan, 1991). It was important to establish whether the inhibitory muscarinic receptors on pulmonary sympathetic nerve endings could be activated by acetylcholine released under physiological conditions from the adjacent cholinergic nerves. Using discrete stimulation of the pulmonary parasympathetic and sympathetic nerve trunks, we were able to confirm this interaction. Concurrent stimulation of the vagal nerve trunk with submaximal voltages inhibited sympathetic nerve-induced relaxations in 50% of the preparations. The involvement of postjunctional functional antagonism was excluded by stimulating the vagal nerve trunk at parameters which did not contract the trachealis smooth muscle. It proved to be very difficult to judge these parameters, and the lack of inhibition observed in half of the preparations may be due to variability between tissues in the metabolism of the transmitter by cholinesterase enzymes before reaching the muscarinic receptors on the airway smooth muscle and the sympathetic nerve endings. Stimulation of the preganglionic vagal nerve trunk not only stimulates the cholinergic fibres but also non-adrenergic, non-cholinergic (NANC) fibres which arise in the vagal bundle. The possibility that the inhibitory effect of concurrent vagal stimulation was due to release of neuropeptides from NANC nerves is, however, unlikely because when VIP, neurokinin A or substance P were administered exogenously, they did not inhibit sympathetic nerveinduced relaxations. In addition McCaig (1986) was unable to elicit vagally mediated NANC responses in the guinea-pig trachea in vitro during cholinergic blockade using similar stimulus parameters to those used in the present study. In McCaig's experiments, sustained stimulation with pulses of 2-4 ms duration and frequencies as great as 160 Hz were required to observe NANC responses. In contrast, in the experiments described in this paper the lower parameters of 0.2 ms and 20 Hz were sufficient to elicit an inhibitory effect of concurrent vagal stimulation upon sympathetic nerve-induced relaxations. Sympathetic nerve-induced relaxations were also unaffected by the presence of either endothelin 1 or endothelin 3, which are known to be formed by airway epithelial cells (Black et al., 1989). Thus it appears from the above results that, in the airways, acetylcholine released following preganglionic stimulation of the cholinergic nerves can activate inhibitory muscarinic receptors present on the adjacent sympathetic nerve endings. It is known that certain subtypes of muscarinic receptors which inhibit cell function are linked to the opening of potassium channals such as those studied in the heart (Sakmann et al., 1983). However, in the present experiments sympathetic nerve-induced relaxation responses were not altered in the presence of concentrations of the potassium channel opening drug, cromakalim which had previously been shown to inhibit vagally induced contraction by at least 50% (Hall & Maclagan, 1988). This suggests that the inhibitory muscarinic receptors on the pulmonary sympathetic nerves are not linked to potassium channels of the type opened by cromakalim, and a different second messenger system may be involved. Other workers have reported a prejunctional inhibitory effect of noradrenaline from sympathetic nerves on cholinergic transmission in the airways of various species (Vermiere & Vanhoutte, 1979; Jones et al., 1980; McCaig, 1987; Rhoden et al., 1988). We also recorded inhibition of vagally-induced contractile responses by noradrenaline released endogenously following concurrent stimulation of the sympathetic nerve trunk. The adrenoceptors involved, which are present on the cholinergic nerve endings, appear to be predominantly of the fl subtype. These findings are consistent with the observations that, in the airways, the two branches of the autonomic nervous system lie in close proximity and may interconnect (Jones et al., 1980; Daniel et al., 1986; Daniel, 1988). The present results confirm that functional interactions between these two nervous systems may also occur.
PULMONARY VAGAL AND SYMPATHETIC NERVE INTERACTION
As we were able to show that endogenous acetylcholine can inhibit sympathetic neurotransmission, anticholinesterase drugs would be expected to potentiate this interaction. It is well known that anticholinesterase drugs, such as physostigmine and neostigmine, cause contraction of the trachealis muscle. It has been suggested that this contractile effect may not depend totally on their ability to prevent breakdown of acetylcholine but may also be due to potentiation of acetylcholine release from cholinergic nerve endings (Douglas, 1951; Carlyle, 1963; Kirkpatrick & Rooney, 1979). Although potentiation of vagally-induced contractions of the guinea-pig trachea by the anticholinesterases, neostigmine and physostigmine, has been reported (McCaig, 1986; Widmark & Waldeck, 1986), other workers have described inhibition of potassium-evoked release of [3H]-acetylcholine from the rat bronchi by the anticholinesterase, soman (Aas et al., 1986; 1987). In our experiments, the anticholinesterase, physostigmine, potentiated contractions at high concentrations and caused inhibition at low concentrations. Previous conflicting reports in the literature may be due to the concentration of the anticholinesterase used and to the contact time with the tissue. Aas (1988) found that exposure for 40h to the anticholinesterase, soman, reduces the acetylcholine-induced contraction of the rat bronchi, probably via a reduction in the number of muscarinic receptors. However, the inhibition of vagallyinduced contractions following 15 min preincubation with 10-6M physostigmine, in our experiments, does not appear to be due to desensitization of postjunctional muscarinic receptors, as 1O- 6M physostigmine caused a greater increase in the postjunctional contractile action of acetylcholine than was obtained with the lower concentration of 10-7M physostigmine. The inhibitory effect is therefore probably a result of
821
acetylcholine accumulating in the synapse and exerting a negative feedback via muscarinic autoreceptors on the cholinergic nerve endings so inhibiting its own release. This complication has previously been observed in gut experiments involving measurement of acetylcholine release in the presence of anticholinesterases to prevent its breakdown (Kilbinger, 1984). The importance of our finding is that the concentration of physostigmine which inhibits sympathetic nerve-induced relaxations is in the same concentration range as that which inhibits vagally-induced contractile responses. This inhibitory effect of physostigmine was blocked by atropine (10-7M). Similarly, McCaig (1986) reported an inhibitory effect of the anticholinesterase, neostigmine, on sympathetic relaxations, which was shown to be reduced by atropine. The inhibitory effect of the anticholinesterase drugs on sympathetic responses is likely to be due to accumulation of acetylcholine which activates inhibitory muscarinic heteroreceptors on the nearby sympathetic nerve terminals. This accumulation of transmitter is probably due to a combination of the anticholinesterase effects of physostigmine on spontaneously released ACh and a direct stimulatory effect of the drug on the cholinergic nerve terminals. This sympathoinhibitory effect may contribute to the bronchospasm observed with anticholinesterase drugs. The results described in this paper reinforce the hypothesis that in the airways, parasympathetic and sympathetic nerves can interact at the prejunctional level. Thus, either branch of the autonomic nervous system can modulate neurotransmission in the other branch, increasing the complexity of the neural control of airway smooth muscle tone. We wish to acknowledge the financial support of Pfizer Central Research, Sandwich, Kent.
References AAS, P. (1988). The toxic effect of an acetylcholinesterase-inhibitor on the cholinergic nervous system in airway smooth muscle. Toxicol., 49, 91-97. AAS, P., MALMEI, T. & FONNUM, F. (1987). The effect of soman on potassium evoked 3H-acetylcholine release in the isolated rat bronchi. Pharmacol. Toxicol., 60, 206-209. AAS, P., VEITEBERG, T. & FONNUM, F. (1986). In vitro effects of soman on bronchial smooth muscle. Biochem. Pharmacol., 35, 1793-1799. BAKER, D.G. & DON, H. (1987). Catecholamines abolish vagal but not acetylcholine tone in the cat trachea. J. Appl. Physiol., 63, 24902498. BLACK, P.N., GHATEI, M.A., TAKAHASHI, K., BRETHERTON-WATT, D., KRAUSZ, T., DOLLERY, C.T. & BLOOM, S.A. (1989). Formation of endothelin by cultured airway epithelial cells. FEBS, 255, 129132. BOYLE, S.J. & POLLOCK, D. (1988). Interactions between cholinergic and noradrenaline nerves in the rat isolated atria. Br. J. Pharmacol., 93, 25P. CABEZAS, G.A., GRAF, P.D. & NADEL, J.A. (1971). Sympathetic versus parasympathetic nervous regulation of airways in dogs. J. Appl. Physiol., 31, 651-655. CARLYLE, R.F. (1963). The mode of action of neostigmine and physiostigmine on the guinea-pig trachealis muscle. Br. J. Pharmacol. Chemother., 21, 137-149. DANIEL, E.E. (1988). Ultrastructure of airway smooth muscle. In Mechanisms in Asthma: Pharmacology, Physiology and Management. ed. Armour, C.L. & Black, J.L. pp. 179-203. New York: Alan R. Liss, Inc. DANIEL, E.E., KANNAN, M., DAVIES, C. & POSEY-DANIEL, V. (1986). Ultrastructural studies on the neuromuscular control of human tracheal and bronchial muscle. Respir. Physiol., 63, 109-128. DOUGLAS, W.W. (1951). The effect of some anticholinesterase drugs on isolated tracheal muscle of the guinea-pig. J. Physiol., 112, 20P. HALL, A.K. & MACLAGAN, J. (1988). Effect of cromakalin on cholinergic neurotransmission in the guinea-pig trachea. Br. J. Pharmacol., 95, 792P. JONES, T.R., KANNAN, M.S. & DANIEL, E.E. (1980). Ultrastructural study of guinea-pig tracheal smooth muscle and its innervation. Can. J. Physiol. Pharmacol., 8, 974-983.
KILBINGER, H. (1984). Presynaptic muscarinic receptors modulating acetylcholine release. Trends Biochem. Sci., 5, 103-105. KIRKPATRICK, C.T. & ROONEY, PJ. (1979). The contracture produced in tracheal smooth muscle by anticholinesterase. Br. J. Pharmacol., 67, 646P. LEVY, M.N. & BLATTBERG, B. (1976). Effect of vagal stimulation on the overflow of norepinephrine into the coronary sinus during cardiac sympathetic nerve stimulation in the dog. Circ. Res., 38, 81-85. LOFFELHOLZ, K. & MUSCHOLL, E. (1970). Inhibition by parasympathetic nerve stimulation of the release of the adrenergic transmitter. Naunyn-Schmiedebergs Arch. Pharmacol., 267, 181-184. MANBER, L. & GERSHON, M.D. (1979). A reciprocal adrenergiccholinergic axoaxonic synapse in the mammalian gut. Am. J. Physiol., 236, E738-E745. MANN, S.P. (1971). The innervation of mammalian bronchial smooth muscle, the localization of catecholamines and cholinesterases. Histochem. J., 3, 319-331. McCAIG, D.J. (1986). Autonomic responses of the isolated, innervated trachea of the guinea-pig: interaction with autonomic drugs, histamine and 5-hydroxytryptamine. Br. J. Pharmacol., 88, 239-248. McCAIG, D.J. (1987). Effect of sympathetic stimulation and applied catecholamines on mechanical and electrical responses to stimulation of the vagus nerve in guinea-pig isolated trachea. Br. J: Pharmacol., 91, 385-394. PENDRY, Y.D. & MACLAGAN, J. (1989). Evidence for prejunctional muscarinic receptors on pulmonary sympathetic nerves in the guinea-pig. Br. J. Pharmacol., 98, 777P. PENDRY, Y.D. & MACLAGAN, J. (1990). Vagally-released acetylcholine inhibits sympathetic neurotransmission in the guinea-pig trachea. Br. J. Pharmacol., 101, 590P. PENDRY, Y.D. & MACLAGAN, J. (1991). Evidence for prejunctional inhibitory muscarinic receptors on sympathetic nerves innervating guinea-pig trachealis muscle. Br. J. Pharmacol., 103, 1165-1171. RHODEN, K.J., MELDRUM, L.A. & BARNES, PJ. (1988). Inhibition of cholinergic neurotransmission in human airways by 162-adrenoceptors. J. Appl. Biol., 65, 700-705. SAKMANN, B., NOMA, A. & TRAUTWEIN, W. (1983). Acetylcholine activation of single muscarinic K+ channels in isolated pacemaker cells of the mammalian heart. Nature, 303, 250-253.
822
Y.D. PENDRY & J. MACLAGAN
VERMIERE, P.A. & VANHOUTTE, P.M. (1979). Inhibitory effects of catecholamines in isolated canine bronchial smooth muscle. J.
Appl. Phys., 46, 787-791.
WIDMARK, E. & WALDECK, B. (1986). Physiological and pharmacological characterization of an in vitro vagus nerve-trachea preparation from guinea-pig. J. Auton. Pharmacol., 6, 187-194. (Received March 19, 1991 Revised July 17, 1991 Accepted August 5, 1991)
