Studies on the Mode of Action of Octopamine, 5-Hydroxytryptamine And Proctolin on A Myogenic Rhythm in the Locust

The myogenic rhythm within the extensor-tibiae muscle of the locust hindleg was first described by Voskresenskaya (1959). It is confined to a bundle of muscle fibres at the proximal end of the leg (Evans & O’Shea, 1978; Hoyle, 1978). Within the myogenic bundle the rhythm is thought to be initiated by a small number of pacemaker fibres, and then to spread electrotonically to other muscle fibres in the bundle (Burns & Usherwood, 1978). The frequency of the rhythm can be modulated by a variety of pharmacologically active substances including glutamate and y-aminobutyric acid, the presumed neurotransmitters of the excitatory and inhibitory neurones respectively to this muscle (Burns & Usherwood, 1978; Evans & O’Shea, 1978). The rhythm is also reduced in frequency by octopamine, a biogenic amine released from a modulatory neurone that innervates this muscle (Hoyle, 1975 ; Evans & O’Shea, 1977, 1978).

Multiple classes of octopamine receptors exist in the locust extensor-tibiae muscle (Evans, 1980, 1981). The OCTOPAMINE₁ class receptors act to reduce the frequency of the myogenic rhythm and are assumed to be located postsynaptically (and/or extrajunctionally) on the muscle fibres of the myogenic bundle. The OCTOPAMINE₂ class receptors modulate neuromuscular transmission and muscle contraction in the extensor muscle (Evans & O’Shea, 1977; O’Shea & Evans, 1979; Evans & Siegler, 1982). They have been subdivided into a subclass 2A, located presynaptically on terminals of the slow motoneurone and subclass 2B located post-synaptically (and/or extrajunctionally) on all types of muscle fibres (Evans, 1981). The 2A receptors modulate both spontaneous and neurally evoked release of transmitter and the 2B receptors modulate the rate of relaxation of muscle tension. Multiple classes of receptors have also been described for a variety of other biogenic amines including catecholamines, serotonin and histamine. The classes can be distinguished pharmacologically and in many cases mediate their effects through different mechanisms (see Berridge, 1980; Snyder & Goodman, 1980; McGrath, 1983). Recently both subclasses of OCTOPAMINE2 receptor have been shown to mediate their actions through an increase in cyclic AMP levels brought about by an increase in the activity of adenylate cyclase (Evans, 1984a,b). The mode of action of the OCTOPAMINE₁ receptors has not been described.

In addition to its modulation by octopamine, the frequency of the locust myogenic rhythm is increased by 5-hydroxytryptamine (S-HT) (Evans & O’Shea, 1978) and by the pentapeptide, proctolin (May, Brown & Clements, 1979). However, no neurones containing 5-HT or proctolin have been reported to innervate the extensor muscle of the locust. Proctolin has been shown to induce increases in basal and maintained tension in this muscle over a prolonged time course (May et al. 1979), and the degree of the effect depends on the frequency of stimulation of the slow motoneurone to the muscle (Evans, 1982). Thus the actions of these modulators are assumed to be neuro-hormonally mediated via changes in their levels in the haemolymph.

The present study investigates the role of cyclic AMP and calcium in the mediation of the actions of octopamine, 5-HT and proctolin on the myogenic rhythm. Since the pacemaker fibres that initiate the rhythm are only a small fraction of the myogenic bundle, it has not been possible to relate levels of cyclic AMP in the bundle to the observed modulatory effects. Thus the actions of the various modulators have been compared with the effects of artificially raising cyclic AMP levels in the muscle using the phosphodiesterase inhibitor, isobutylmethylxanthine (IBMX), and the diterpene activator of adenylate cyclase, forskolin. The effects of cyclic AMP analogues, and of changed calcium levels, has also been studied. The results are discussed in terms of the possible modes of action of the different modulatory substances.

Experiments were performed at room temperature (21 °C) on adult Schistocerca americana gregaria (formerly S. gregaria) of either sex. The locusts were obtained from crowded laboratory cultures fed on wheat seedlings.

The intrinsic rhythm of contraction and relaxation in the extensor-tibiae muscle of the locust metathoracic leg was monitored by recording muscle tension almost isometrically with a force transducer attached to the distal end of the muscle apodeme (Evans & O’Shea, 1978). Known concentrations of drugs, dissolved in locust saline, were applied to the myogenic bundle at a constant rate of 1 ml min⁻¹, as described Evans & O’Shea (1978). Locust isotonic physiological saline (pH 6·8) contained 140mm-NaCl, 10mm-KCl, 4mm-CaC1₂, 4mm-NaHCO₃, 6 mm-NaH₂PO₄ (Usherwood & Grundfest, 1965) and 90 mm-sucrose. Changes in the potassium and calcium levels of the saline were balanced by corresponding changes in sodium levels, and in sodium-free solutions sodium was replaced by Tris. Stock solutions of forskolin were made by dissolving 1 mg in 100µl of ethanol. Appropriate ethanol controls were run for each experiment. (−)-N⁶-(R-phenyl-isopropyl)-adenosine was initially dissolved in dimethylsulphoxide (DMSO). It was applied to preparations in solutions containing 1 % DMSO, which by themselves had no effect on the myogenic rhythm.

It is possible to observe rhythms of different frequencies and amplitudes co-existing in the same myogenic muscle bundle (Hoyle, 1978; P. D. Evans, unpublished). This could occur if some of the pacemaker fibres become uncoupled, but still can sustain their own contractile rhythm. Such preparations were not included for study in the present investigation.

Drugs were obtained from the following sources: 3-isobutyl-l-methylxanthine (IBMX) (Aldrich); 8,(4-chlorophenylthio) adenosine 3′: 5′ monophosphate, cyclic; (−)-N⁶-(R-phenyl-isopropyl)-adenosine (Boeringer Mannheim); A23187 and forskolin (Calbiochem) ; 2′,5′ dideoxyadenosine (P-L Biochemical Inc.). I would like to acknowledge the gifts of samples of the following drugs from pharmaceutical companies: phentolamine mesylate (CIBA); SQ20009 (Squibb); D600 (Knoll AG); Amrinone (Sterling-Winthrop); amiloride (Merck, Sharp & Dohme); N-ethyl carboxamide adenosine (Byk Gulden). All other drugs were obtained from the Sigma Chemical Co.

Octopamine reduces the frequency of the myogenic rhythm (Fig. 1 A) whilst 5-HT and proctolin increase its frequency (Fig. 1B, C). Dose-response curves (not shown) reveal that the threshold for the stimulatory action of proctolin occurs between 10⁻¹² and 10⁻¹¹ M. The threshold for the stimulatory effect of 5-HT occurs at around 10⁻⁹M and that for the inhibitory action of octopamine between 10⁻¹⁰ and 10⁻⁹M (Evans & O’Shea, 1978).

Proctolin and 5-HT act on separate receptors to bring about their stimulatory effects. Fig. 2 shows an example of an experiment in which 10⁻⁵ M gramine does not alter the response of the myogenic rhythm to 10⁻¹⁰M proctolin, but completely inhibits the actions of 10⁻⁷M 5-HT. Gramine, at this concentration, itself produces a slight increase in the frequency of the rhythm. Bromo LSD (concentration not given) has also been reported to inhibit the actions of 5-HT, but not proctolin, on the myogenic rhythm (May et al. 1979).

Preliminary pharmacological data on the 5-HT activated receptor (P. D. Evans, unpublished) indicate that it has a very narrow specificity range. In this respect its responses appear to be more similar to the activation by 5-HT of the receptors for the peptide diuretic hormone on Rhodnius prolixus Malpighian tubules (Maddrell, Pilcher & Gardiner, 1971) than to the activation of specific 5-HT receptors, such as those of blowfly salivary gland (Berridge, 1972). This raises the possibility that the endogenous activator of this receptor on the locust myogenic bundle may be a second peptide hormone, rather than 5-HT itself.

To investigate the role of cyclic nucleotides in the modulation of the myogenic rhythm, cyclic nucleotide levels have been artificially elevated by applying phosphodiesterase inhibitors, forskolin and cyclic nucleotide analogues.

The inhibition of phosphodiesterase, the enzyme which breaks down cyclic nucleotides, leads to the elevation of cyclic nucleotide levels within tissues. In the locust extensor-tibiae muscle the phosphodiesterase inhibitor isobutylmethylxanthine (IBMX) increases the levels of both cyclic AMP and cyclic GMP (Evans, 1984a). When applied to the myogenic bundle, IBMX increases the frequency of the rhythm (Fig. 3A). The response reaches a maximum after a 2-min exposure and then declines despite the continued presence of IBMX in the superfusate. A dose-response curve for the maximum effect of IBMX is shown in Fig. 3B. The threshold lies between 3 × 10⁻⁷ M and 10⁻⁷ M and the maximum effect occurs at 10⁻⁴ M. At higher concentrations the degree of stimulation by IBMX declines. This biphasic effect of IBMX may be related to the fact that it increases the levels of both cyclic AMP and cyclic GMP in locust muscle (Evans, 1984a). In a small minority of preparations (in a single batch of 8 locusts out of 120 tested) IBMX was observed to induce an inhibition of the myogenic rhythm at doses of 10⁻⁴ M and above. This could reflect a difference in the endogenous levels of cyclic nucleotides in different preparations.

The effectiveness of several phosphodiesterase inhibitors at 10⁻⁴M are compared in Table 1. IBMX is the most effective compound, whilst papaverine is without effect on the myogenic rhythm up to a concentration of 10⁻⁴M. At concentrations above 5 × 10⁻⁴M both theophylline and caffeine induce a maintained contraction in the extensor-tibiae muscle.

If octopamine, proctolin or 5-HT act on the myogenic rhythm by increasing cyclic nucleotide levels, then their effects should be potentiated in the presence of a phosphodiesterase inhibitor. Fig. 4 shows that IBMX (10⁻⁵M) potentiates the stimulatory actions of proctolin (Fig. 4A) and 5-HT (Fig. 4B) on the myogenic rhythm. The effects can also be potentiated by theophylline (10⁻⁴M) (not shown). In similar experiments on the inhibitory action of octopamine on the myogenic rhythm, no consistent effects have been observed with either IBMX or theophylline at concentrations up to 10⁻⁴M.

Although IBMX increases the levels of cyclic AMP and cyclic GMP in locust muscle (Evans, 1984a), it and other methylxanthines are also potent antagonists of some classes of adenosine receptor (Wolff, Londos & Cooper, 1981). Thus it is essential to show that the elevation of specific cyclic nucleotide levels by other means has the same effect on the frequency of the myogenic rhythm.

The diterpene compound, forskolin, increases the activity of adenylate cyclase in a number of intact cellular preparations. It acts on the catalytic subunit, bypassing the receptor activation stage (Seamon & Daly, 1981). Forskolin also specifically increases the cyclic AMP levels in the locust extensor-tibiae muscle (Evans, 1984a). Fig. 5A shows that during a 5-min pulse of 10⁻⁵ M forskolin the frequency of the myogenic rhythm gradually increases to 230% of its rate in normal saline. This action of forskolin is potentiated by the presence of IBMX (10⁻⁶M) in the superfusate (Fig. 5B). The above evidence suggests that an increased level of endogenously generated cyclic AMP in the myogenic pacemaker fibres leads to an increase in the frequency of the myogenic rhythm.

Since increases in endogenously generated cyclic AMP increase the frequency of the myogenic rhythm, it would be expected that the exogenous application of cyclic AMP and its derivatives would induce a similar response. However, the application of cyclic AMP or dibutyryl cyclic AMP at high concentrations (up to 10⁻²M) produces a slight slowing of the rhythm. Fig. 6A shows that in the presence of 10⁻⁴M IBMX, dibutyryl cyclic AMP (10⁻²M) produces a rapid slowing of the rhythm. Further experiments showed that cyclic GMP and the 8-bromo-derivatives of cyclic AMP and cyclic GMP at concentrations of 10⁻³M reduced the frequency of the myogenic rhythm.

In a previous study on the mode of action of OCTOPAMINE2 class receptors in the modulation of neuromuscular transmission in the locust extensor-tibiae muscle, the only exogenously applied cyclic AMP analogue that mimicked the effects of an endogenous increase in cyclic AMP was 8,(4-chlorophenylthio)-cyclic AMP (CPT cyclic AMP) (Evans, 1984b). This derivative is reported to be a hundred times more effective than dibutyryl cyclic AMP in the activation of cyclic AMP dependent protein kinase in rat liver, and to be more resistant to phosphodiesterase activity (Miller, Beck, Simon & Meyer, 1975). When a 5-min pulse of CPT cyclic AMP (10^{− 3}M) is applied to the myogenic bundle it produces a biphasic response (Fig. 6B). The rhythm slows initially, but then speeds up gradually so that at the end of the pulse exceeds the initial frequency before the CPT cyclic AMP application.

The failure of exogenously added cyclic AMP analogues to mimic the effects of increases in endogenous cyclic AMP could be explained if the exogenously added cyclic AMP analogues have sites of action that are not accessible to the endogenously generated cyclic AMP. The occupation of such extra sites, within the muscle fibres, could lead to the prolonged opening of calcium channels as occurs in cultured heart cells (cf. Cachelin, de Peyer, Kokubun & Reuter, 1983) or to the stimulation of adenosine receptors by the endogenous production of adenosine from the cyclic AMP derivatives (see Wolff et al. 1981). To test the latter possibility a range of adenosine analogues were applied at a concentration of 100µM (Table 2). The analogues fell into three categories: Class A derivatives, which are effective only in reducing the frequency of the rhythm; Class B derivatives, 2-deoxyadenosine and 2-Cl-adenosine which reduce the frequency of the rhythm at 10⁻⁴M, but increase its frequency at 10⁻³M; and a Class C derivative 2′ 5′ dideoxyadenosine, which increases the frequency of the rhythm at all concentrations up to 10⁻³M. Dipyridamole (10⁻⁴M), a blocker of adenosine uptake, preferentially blocks the increase in frequency mediated by 10⁻⁴M 2′ 5′ dideoxyadenosine, but has no effect on the decrease in frequency mediated by 10⁻⁴M adenosine (not shown). In addition the effects of adenosine (10⁻³M) are not blocked by IBMX (10⁻⁵M) or phentolamine (10⁻⁵M). Thus it appears that there are two sites of action for adenosine derivatives, one that accelerates the rhythm, and another that reduces its frequency.

The response of the myogenic rhythm to variations in the level of calcium in the superfusate is variable. In the majority of cases lowering the external calcium concentration from the control level of 4 mm to a level of 1 mm produces an initial speed up in the frequency of the rhythm of 68·5 ± 13·3% (±s.e., N = 20). This increased frequency lasts for a minute or two before declining to the control level (cf. Burns & Usherwood, 1978; Evans & O’Shea, 1978). However, in a minority of preparations (out of 25) the frequency of the rhythm declines upon exposure to saline containing 1 mm calcium. Increasing the external calcium concentration from 4mm to 10mm initially reduces the frequency of the rhythm in the majority of preparations by 1·7 ± 5·1 % (±s.e., N = 20). Again in a minority of preparations (3 out of 23) the inverse effect was observed. The reason for the variation in the responses to changes external calcium is unknown.

The internal levels of calcium in the muscle fibres of the myogenic bundle can also increased by exposure to a calcium ionophore. Fig. 7 shows the effect of a 5-min pulse of the calcium ionophore A23187 (10⁻⁵M) introduced into the muscle superfusate. This treatment reduces the frequency of the rhythm by 79·2 ± 3·8 % (s.e., = 4). This effect of the ionophore is not caused by a direct interaction with octopamine receptors, or by the release of endogenous octopamine from the terminals the DUMETi neurone, since the actions of the ionophore are not inhibited in the presence of phentolamine (10^{− 6}M), a drug that blocks octopamine receptors (Evans, 1981).

The response of the myogenic rhythm to octopamine has previously been shown to in-dependent upon the presence of calcium ions in the bathing medium (Evans & Shea, 1978). In saline containing no added calcium the rhythm speeded up, and a pulse of 10⁻⁶M DL-octopamine failed to abolish the rhythm, which became highly reegular. This could be a result of the uncoupling of the muscle fibres in the myogenic bundle. In an attempt to quantify this effect, muscles were exposed.to 5-min pulses 10⁻⁸M DL-octopamine in the presence of different levels of calcium (1,2,4,6 and 10 mm). At the control level of calcium (4 HIM) this procedure reduces the frequency the rhythm by 43·8 ± 10·2% (±s.e., N = 6). However, at the other levels of calcium the effect of octopamine is not significantly different. Similar experiments with the 5-min pulses of 5-HT (10⁻⁷M) and proctolin (10^{− 10}M) increase the frequency the rhythm by 391·0 ± 91·1 % (±s.e., N = 6) and 516-8 ± 185-5 % (±s.e., N = 6) respectively in the presence of 4mm calcium. Again no significant differences are observed with 1 mm or 10 mm calcium in the superfusate. It is possible that small changes in the responses to the modulators may not have been detected due to the large variability of the responses.

If octopamine, 5-HT or proctolin induce any changes in calcium permeability in the muscle fibres of the myogenic bundle, their effects should be antagonized by the presence of calcium channel blocking agents. In the present study neither the blocker amrinone (10⁻⁴M) nor D-600 (10⁻⁴M) antagonized the effects of octopamine (10⁻⁷M), 5-HT (10⁻⁶M) or proctolin (10⁻¹⁰M) on the myogenic rhythm. Both drugs did however reduce the frequency of the myogenic rhythm when applied alone at 10⁻⁴M, amrinone by 62·5 ± 7·9 % (±s.e., N = 3) and D-600 by 39·6 ± 6·3 % (±s.e., N=3).

Octopamine has been suggested to increase the respiratory rate of isolated cockroach nerve cords by increasing the sodium permeability of perineurial glial cells, which in turn stimulates the activity of the sodium pump (Steele & Chan, 1980). In addition, some of the effects of catecholamines on excitable cells in vertebrates have been suggested to be due to the activation of the sodium pump (see Phillis & Wu, 1981). Accordingly the effects of alterations in sodium and potassium permeability, and of changes in the activity of the sodium pump, have been investigated.

Total replacement of sodium ions in the muscle superfusate with Tris reduces the frequency of the myogenic rhythm by 69·9 ± 10·8% (±s.e., N=5). Under these conditions a 5-min pulse of 10⁻⁶M DL-octopamine still reduces the frequency of the myogenic rhythm by 40·3 ± 4·3 % (±s.e., N = 3). This suggests that the action of octopamine on the myogenic rhythm is not absolutely dependent on the presence of external sodium ions. The passive sodium permeability of membranes can be increased by the actions of drugs such as veratridine and amphotericin B (Narahashi, 1974, 1977). The application of 5-min pulses of these drugs at concentrations above 5 × 10⁻⁶M increases the frequency of the rhythm, and also causes a maintained contraction (see Fig. 8Ai). The sustained contraction is presumably brought about by the depolarizing actions of these drugs. In contrast, a 5-min pulse of amiloride (10⁻³M), a drug which decreases sodium permeability in a variety of epithelial preparations (see Lindeman, 1980), also increased the frequency of the myogenic rhythm by 424·0 ± 50·2% (±s.e., N =3) but did not produce an increase in basal tension (Fig. 8Aii). The reason for this discrepancy is not clear, but amiloride may also bind to potassium channels in some tissues such as mouse neuroblastoma cells (see Hugues et al. 1982).

The effect of changing the potassium concentration of the muscle superfusate on the frequency of the myogenic rhythm was found to be variable, as previously reported by Burns & Usherwood (1978). In the majority of preparations tested, changing from a control saline with a potassium concentration of 10 mm to a saline with no added potassium, initially speeded up the rhythm by 78·8 ± 28·3 % (±s.e., N = 6) during a 5-min exposure. However, in a minority of cases (N = 2) the rhythm slowed down. In all cases tested, increasing the potassium concentration of the bathing medium from the control level of 10 mm to a level of 20 mm, again induced a slight increase, 34·7 ± 7·9 % (s.e., N = 6), in the frequency of the rhythm, also as observed by Burns & Usherwood (1978). No significant differences in the slowing of the myogenic rhythm to a 5-min pulse of 10⁻⁹M DL-octopamine could be detected at either the elevated or lowered levels of potassium used.

The effect of reducing potassium exit from the muscle fibres of the myogenic bundle can be assessed by using drugs or ions which block potassium channels. Tetraethylammonium (TEA) ions, which reduce passive potassium permeability (Narahashi, 1974, 1977), also reduce the frequency of the myogenic rhythm when introduced into the superfusate at concentrations above 10⁻⁴M (Fig. 8Bi). In addition, 4-aminopyridine (4-AP), a drug that reduces potassium currents (Narahashi, 1974, 1977), also reduces the frequency of the rhythm. A 5-min pulse of 10⁻⁶M 4-AP reduces the frequency of the rhythm by 29·2 ±9·5% (±s.e., N=3). The effects of TEA (up to 10⁻²M) and 4-AP (up to 10⁻⁶M) were not blocked in the presence of 10⁻⁵M phentolamine, suggesting that they were not produced by an activation of octopamine receptors. Further, the addition of barium ions, which all block potassium channels (Nagel, 1979; Frank & Rohani, 1982) to the muscle superfusate also causes a reduction in the frequency of the rhythm (Fig. 8Bii). At concentrations up to 15 mm barium ions fail to alter the inhibitory actions of a 5-min pulse of 10⁻⁷M DL-octopamine. Apamin, a drug that selectively blocks calcium-dependent potassium channels (Hugues et al. 1982), produced no change in the frequency of the myogenic rhythm at a concentration of 10⁻⁶ M, and did not significantly alter the effect on the rhythm of octopamine (10⁻⁶M), 5-HT (10⁻⁶M) or proctolin (10⁻¹⁰M). In addition, valinomycin, an antibiotic that selectively increases potassium permeability (see Narahashi, 1974), did not have any apparent effect on the myogenic rhythm during a 10-min exposure at concentrations up to 1⁻⁵ M.

Any effects of octopamine that are mediated by the stimulation of the sodium pump, as have been suggested for some actions of catecholamines (Phillis & Wu, 1981), should be inhibited by the sodium pump inhibitor, ouabain. Ouabain, at concentrations above 10⁻⁵M, speeds up the myogenic rhythm (Fig. 9A) but at this concentration does not block the actions of a 5-min pulse of 10⁻⁷M DL-octopamine (Fig. 9B). At concentrations higher than 5 × 10⁻⁵M, the stimulatory effects of ouabain cannot be overcome by pulses of 10⁻⁷M DL-octopamine. Ethacrynic acid, a supposed sodium transport inhibitor, had no effect on the myogenic rhythm up to a concentration of 10⁻⁴ M. It is not clear if the ouabain-induced increase in the frequency of the myogenic rhythm is the result of ionic changes in the pacemaker fibres or due to a reduced breakdown, and consequent accumulation, of ATP in the fibres.

Substitution of sodium ions by lithium ions can also inhibit the actions of the sodium pump. This occurs because lithium ions can replace sodium ions in the ionic mechanism responsible for the generation of action potentials, but cannot substitute for them in the activation of the sodium pump (Keynes & Swan, 1959). The complete substitution of sodium by lithium did not produce any significant change in the frequency of the myogenic rhythm, nor did it block the actions of 5-min pulses of 10⁻⁶M DL-octopamine in inhibiting the myogenic rhythm. This further suggests that the octopamine-induced inhibition of the myogenic rhythm is not mediated by an activation of the sodium pump.

The rhythmical contractions of the extensor-tibiae muscle of the metathoracic leg of the locust are myogenic in origin (Hoyle & O’Shea, 1974) and can be localized to a proximal bundle of muscle fibres (Evans & O’Shea, 1978; Hoyle, 1978). Intracellular recordings from these muscle fibres show rhythmic oscillations of the membrane potential, with slow depolarizations accompanying the contractions (Burns & Usherwood, 1978) and May et al. (1979) observed that each of the slow depolarizations terminated in an active membrane response. Burns & Usherwood (1978) further observed that the muscle fibres in the myogenic bundle were electrically coupled, but they were unable to identify a pacemaker site. The mechanism underlying these rhythmic contractions must await the identification of the ionic currents present and their pharmacological properties.

The application of either proctolin or 5-HT to the myogenic bundle of the locust results in a dose-dependent increase in the frequency of the myogenic rhythm. The results of the present study suggest the possibility that these compounds may mediate their effects by increasing the level of cyclic AMP in the tissue.

Treatments that elevate the endogenous level of cyclic AMP increase the frequency of the myogenic rhythm. The phosphodiesterase inhibitor, IBMX, which increases cyclic nucleotide levels (both cyclic AMP and cyclic GMP) in the bulk of the extensortibiae muscle (Evans, 1984a) increases the frequency of the myogenic rhythm, as does the diterpene compound, forskolin. Forskolin specifically activates adenylate cyclase in a variety of intact and broken cells preparations (Seamon & Daly, 1981) and specifically increases cyclic AMP levels in the locust extensor-tibiae muscle (Evans, 1984a). In addition IBMX potentiates the actions of 5-HT, proctolin and forskolin in speeding up the myogenic rhythm.

The results of applying cyclic nucleotide analogues to the myogenic bundle, differ however, from those obtained by increasing the endogenous levels of cyclic AMP. Dibutyryl cyclic AMP and 8-bromo cyclic AMP slowed the rhythm down, rather than speeding it up as expected, whilst CPT cyclic AMP produced biphasic effects. This contrasts with the actions of these derivatives on the potentiation of neuromuscular transmission and on muscular contraction in the same muscle. Here CPT cyclic AMP mimics the actions of forskolin and IBMX, but the other analogues are inactive (Evans, 1984a,b). Similarly, the oscillatory behaviour of the R15 neurone oiAplysia californica differs following extracellular or intracellular application of cyclic AMP analogues (Levitan, Harmer & Adams, 1979). Extracellular application of either 8-CPT cyclic AMP or 8-benzylthiocyclic AMP stimulates the oscillatory activity of R-15, whereas intracellular injection of these analogues initially hyperpolarizes the cell and abolishes oscillatory activity.

Attempts to demonstrate biochemically increases in cyclic AMP mediated by proctolin and 5-HT in the isolated myogenic bundle have not been successful although it has been found that octopamine increases cyclic AMP levels as it does in the rest of the muscle (P. D. Evans, in preparation). It is possible that the myogenic bundle only contains a small number of pacemaker fibres, which determine the frequency of the rhythm, plus a much larger number of follower fibres whose responses will dominate in biochemical measurements of the isolated myogenic bundle.

Proctolin is thought to act in different ways in different tissues. Cyclic AMP has been implicated in the responses of the cockroach hindgut to proctolin (Cook, Holman & Marks, 1975 ; Jennings, Steele & Starratt, 1983), but a simultaneous elevation of calcium may also occur. In addition in the locust, Locusta migratoria, proctolin acts on the hindgut by increasing intracellular calcium levels (Dunbar & Huddart, 1982). However, such a proctolin-mediated increase in intracellular calcium levels is unlikely to occur in the myogenic bundle of the locust since treatments that increase intracellular calcium, such as elevation of extracellular calcium and the application of the calcium ionophore A23187, reduce the frequency of the rhythm. Proctolin has also been reported to increase the activity of adenylate cyclase in locust nervous tissue Hiripi, Rózsa & Miller, 1979). Thus the mode of action of proctolin receptors needs to be examined for each preparation where it has physiological actions. This is to be particularly emphasized since Jennings et al. (1983) have reported apparent differences in the mode of action of proctolin in the hindgut of different species of cockroach.

One of the two classes of 5-HT receptor described in other preparations also brings about its actions by increasing cyclic AMP levels (see Berridge, 1980; Snyder & Goodman, 1980). In the salivary glands of the blowfly, Calliphora erythrocephala, 5-HT2 receptors mediate their actions through cyclic AMP, whilst 5-HTi receptors act through calcium (Berridge & Heslop, 1981). The receptor that mediates the actions of 5-HT on the locust myogenic rhythm has not been extensively examined pharmacologically, but the possibility exists that it is a receptor for a second peptide hormone rather than a true 5-HT receptor (see Results).

OCTOPAMINE₁ receptors, which decrease the frequency of the myogenic rhythm (Evans, 1981), are unlikely to act by increasing the levels of cyclic AMP. The phosphodiesterase inhibitor IBMX and the adenylate cyclase activator, forskolin, increase the frequency of the rhythm and the octopamine responses are not potentiated by IBMX. Thus increasing cyclic AMP levels alone cannot mimic the activation of OCTOPAMINE₁ receptors. Either some additional mechanism gates the effects of such cyclic AMP changes or an entirely different mechanism, not involving changes in cyclic AMP levels, is used. In the present study a number of other possible mechanisms have been examined.

Biogenic amines bring about their actions in a variety of vertebrate excitable tissues by the activation of the sodium pump (Phillis & Wu, 1981). Consistent with this is the observation that ouabain, a sodium pump inhibitor, speeds up the myogenic rhythm. However, the action of 10⁻⁷ M DL-octopamine is not blocked by 10⁻⁵ M ouabain or by complete substitution of sodium by lithium ions in the muscle superfusate.

Octopamine has been reported to increase sodium permeability in the perineurial cells of cockroach nerve cord (Steele & Chan, 1980). However, in the locust, octopamine inhibits the myogenic rhythm in sodium-free solutions. In addition, drugs such as veratridine and amphotericin B, which increase resting sodium permeability, increase rather than decrease the frequency of the rhythm.

A variety of biogenic amines either increase (Ascher, 1972; Swann & Carpenter, 1975; Drummond, Benson & Levitan, 1980; Ascher & Chesnoy-Marchais, 1982) or decrease (Klein & Kandel, 1980) potassium conductances. Since barium ions and drugs such as TEA and 4-AP, all of which block potassium channels, slow the myogenic rhythm, octopamine would have to decrease a potassium conductance to bring about a similar slowing of the myogenic rhythm. However, none of these drugs at the concentrations used interfered with the responses of the myogenic rhythm to octopamine. Nonetheless it is difficult to rule out an action of octopamine on potassium channels since the concentrations of drugs necessary to block the potassium channels completely would also abolish the myogenic rhythm.

The pharmacological similarities between α-adrenergic receptors and octopamine receptors (Evans, 1981) suggest that OCTOPAMINE₁ receptors could raisa intracellular calcium, either by increased calcium permeability or by a release from internal stores (cf. Van Breeman & Siegel, 1980; Deth & Lynch, 1981 ; Hertog, 1981 ; Taylor, Reinhart & Bygrave, 1983). Consistent with this hypothesis, increased internal calcium levels in the myogenic bundle, brought about by elevated external calcium concentrations or by the calcium ionophore A23187, reduced the frequency of the rhythm. Against it, octopamine action was not sensitive to external calcium ion concentration and was not blocked by the calcium channel antagonists amrinone or D-600. However, an octopamine-mediated increase in internal calcium levels due to a release from an internal calcium store is not ruled out by the present evidence. Indeed it has recently been suggested that in a variety of tissues various receptor agonists, including biogenic amines, use inositol-1,4,5-trisphosphate as a second messenger to induce the release of calcium from non-mitochondrial intracellular calcium stores (Berridge, 1983; Berridge et al. 1983; Streb, Irvine, Berridge & Schulz, 1983).

A second possibility, raised by the similarities between a-adrenergic receptors and octopamine receptors, is that octopamine actions could be mediated via a localized release of adenosine from the myogenic bundle (cf. Fredholm, 1976; Fredholm & Hedqvist, 1978; Fredholm & Sollevi, 1981). Consistent with this, adenosine inhibits the myogenic rhythm. An attempt to characterize pharmacologically the site of action of adenosine in this preparation did not allow its identification as either an extracellular ‘R’ receptor or an intracellular ‘P’ site as has been possible in a variety of vertebrate tissues (see Wolff et al. 1981). However, it did reveal two sites of action for adenosine analogues, one that speeds the rhythm up and one that slows it down. Further studies are required to determine if adenosine is involved in the inhibition of the myogenic rhythm by octopamine.

Thus the mode of action of the OCTOPAMINE₁ receptors in reducing the frequency of the myogenic rhythm in the locust extensor-tibiae muscle remains unclear and requires further experimentation. It does, however, appear to be different from that of the OCTOPAMINE₂ receptors which potentiate neuromuscular transmission and muscle contraction in the same muscle (Evans, 1984a,b). The data support the idea that different classes of octopamine receptor mediate their actions in different ways.

I would like to thank Drs M. J. Berridge, D. B. Morton and M. V. S. Siegler for their constructive criticism of the manuscript and for useful discussions during the course of the work.