The Modulatory Actions of FMRFamide and Related Peptides on Locust Skeletal Muscle

Immunocytochemical evidence has recently been presented for the presence of peptides related to the neuropeptide FMRFamide (Phe-Met-Arg-Phe-NH₂) in specific subsets of neurones within the nervous systems of insects including the Colorado potato beetle (Veenstra & Schooneveld, 1984), the locust (Myers & Evans, 1985a,b) and the cockroach (Verhaert, Grimmelikhuijzen & DeLoof, 1985). FMRFamide was originally isolated and identified from the nervous system of the clam Macrocallista nimbosa (Price & Greenberg, 1977). Since that time FMRFamide, and a family of structurally related peptides, have been shown to be present in nervous tissue from a wide range of invertebrate species, including the snail Helix (Price et al. 1985) and the horseshoe crab Limulus (Watson et al. 1984). Structurally related peptides have also been reported to be present in vertebrate nervous tissue (Dockray & Williams, 1983; Williams & Dockray, 1983; Dockray et al. 1983; O’Donohue et al. 1984).

In molluscs, FMRFamide is cardioexcitatory (see Price & Greenberg, 1980). In addition, it produces a variety of effects upon a range of non-cardiac muscles. For instance, it induces sustained contractions in molluscan smooth muscles such as the radula protractor muscle of the whelk Buscycon contrarium and the anterior byssus retractor muscle (ABRM) of the mussel Mytilus edulis (Price & Greenberg, 1977; Greenberg & Price, 1979; Painter, 1982). Further, in the tentacle retractor muscle of Helix aspersa, it produces a prolonged contracture on which phasic, rhythmical contractions may be superimposed (Cottrell, Greenberg & Price, 1983a; Cottrell, Schot & Dockray, 1983b). FMRFamide also has a variety of actions on gastropod neurones (Cottrell, 1983; Murphy, Lukowiak & Stell, 1985).

In the locust, the observations that FMRFamide-like immunoreactive neurones could be traced to identified skeletal muscles, and also that the immunoreactivity appeared concentrated in neurohaemal areas (Myers & Evans, 1985a,b) suggested that locust skeletal muscle would be an interesting target site in which to investigate the effects of FMRFamide and related peptides. Although we do not yet know the precise structures of the immunoreactive FMRFamide-like molecules in the locust nervous system, studies on the pharmacological profile of insect skeletal muscle responses will provide much useful information upon the nature and mode of action of the receptors mediating these effects. In addition such information will be useful in the development of appropriate bioassays for use during the isolation and purification of such peptides from the locust nervous system. The extensor-tibiae muscle of the hindleg of the locust was chosen for investigation since it is one of the most intensively studied insect skeletal muscles with a well-defined neuronal input from a limited number of neurones, all of which are physiologically identifiable (for reviews see Hoyle, 1983; Evans, 1985a; Evans & Myers, 1986). The present paper confirms and extends the results presented in a short communication (Walther, Schiebe & Voigt, 1984) published during the course of this work. It describes the actions of FMRFamide-like peptides on twitch, tetanic and maintained tension in this muscle and emphasizes the dependence of the effects upon the frequency of motor neurone stimulation. Regional differences in the responsiveness of the extensor-tibiae muscle to FMRFamide are described, and structure—activity studies give additional information on the structural requirements for a bioactive FMRFamide-like peptide in the locust.

Adult locusts (Schistocerca gregaria) of either sex were obtained from crowded laboratory colonies fed on wheat seedlings. Small batches of animals were left for 1–2h before use, after removal from the main culture, to minimize any initial potentiation effects due to elevated levels of octopamine in the haemolymph (see Evans, 1981; Davenport & Evans, 1984a,b).

Tension in the intact muscle and in the various regions of the extensor-tibiae muscle of the locust hindleg was measured almost isometrically with a tension transducer attached distally to the apodeme. Muscle blocks were isolated for tension recording by cutting the central apodeme on the proximal side of the bundle from which recordings were to be made, and by cutting the extensor nerve on the distal side of the block (Evans, 1985b). The slow extensor-tibiae (SETi) and the fast extensor-tibiae (FETi) motor neurones were excited by stimulating nerves 3b and 5, respectively, with a pair of silver hook electrodes. Patterned pulse trains were computer-generated and used to trigger the output of a stimulator (Evans & Siegler, 1982). An operational amplifier signal differentiator was used to measure continuously the rates of contraction and relaxation of neurally evoked tension in the different muscle regions (Buchan & Evans, 1980). Miniature end-plate potentials were recorded intracellularly from extensor-tibiae muscle fibres innervated by SETi, but not FETi, using microelectrodes, filled with 2 mol I⁻¹ potassium acetate, which had d.c. resistances in saline of 15–30 MΩ.

Extensor-tibiae muscles were prepared for cyclic AMP measurements by incubation as described previously (Evans, 1984a). At the end of the incubation period, muscles were rapidly frozen, dissected away from their surrounding cuticle and divided up into appropriate regions for analysis. The muscle samples were homogenized in an ice-cold concentrated hydrochloric acid : absolute ethanol mixture (l:60v/v) (Horn & McAfee, 1977). The cyclic AMP levels were assayed in the muscle extracts by a protein binding method (Brown, Ekins & Albano, 1972) using a commercial cyclic AMP assay kit (Amersham). Protein determinations were carried out according to Lowry, Roseburg, Farr & Randall (1951) using bovine serum albumin as standard.

FMRFamide and related peptides were superfused directly onto the surface of the muscle dissolved in a physiologically isotonic saline (pH 6·8) containing (in mmol I⁻¹) NaCl, 140; KC1, 10; CaCl₂, 4; NaHCO₃, 4; NaH₂PO₄, 6; and sucrose, 90. YGGFMRFamide, FMRFamide, LPLRFamide and RFamide were obtained from Peninsula Laboratories. All other peptides and drugs were obtained from Sigma except for phentolamine mesylate which was a gift from Ciba.

The contractions evoked by stimulation of the slow motor neurone (SETi) to the extensor-tibiae muscle of the locust hindleg are potentiated in amplitude by the presence of FMRFamide-related peptides in the muscle superfusate (Walther et al. 1984). The present study confirms this finding and extends it to show that the peptides of this family also increase the rates of both contraction and relaxation of slow motor neurone-evoked twitch tension in this muscle. Fig. 1 shows that a 30-s pulse of 10⁻⁶mol l⁻¹ FMRFamide increases the amplitude, and rates of contraction and relaxation, of twitch tension when SETi is stimulated at 1 Hz. These effects outlast the presence of the peptide in the superfusate by several minutes. In the presence of a prolonged pulse of 10⁻⁶ mol I⁻¹ FMRFamide, maximal effects on each of these parameters are reached after an exposure of 4–5 min (Fig. 1). FMRFamide, at concentrations up to 10⁻⁶mol I⁻¹, produces no observable effect on twitch tension generated by stimulating the fast motor neurone (FETi) to this muscle at 1 Hz. In addition, it has no observable effect on the amplitude or frequency of the myogenic rhythm of contraction and relaxation observed in this muscle.

The FMRFamide-related molluscan heptapeptide, YGGFMRFamide, which was identified in the octopus (see Greenberg, Price & Lehman, 1985) also produces the same qualitative effects as FMRFamide (Fig. 2A,B). However, a prolonged pulse of YGGFMRFamide at 10⁻⁶moll⁻¹ produces maximal effects in a much shorter time (2–3 min) than an equivalent pulse of FMRFamide. The effects of both FMRFamide and YGGFMRFamide are dose-dependent (Fig. 3). FMRFamide is almost an order of magnitude more potent than YGGFMRFamide, but both substances have the same maximal effects on the amplitude of slow twitch tension, its rate of relaxation (Fig. 3) and its rate of contraction (not shown). The threshold for an observable action of YGGFMRFamide on slow motor neurone-evoked tension occurs between 10⁻⁹ and 10⁻¹⁰moir⁻¹.

FMRFamide-related peptides have been reported to produce their effects on slow motor neurone-induced twitch tension by both presynaptic and postsynaptic effects (Walther et al. 1984). In the latter study, however, the presynaptic actions were only inferred from the fact that FMRFamide changed excitatory junctional potential amplitudes but not the time integrals of depolarizations caused by glutamate, the presumed rapid-acting transmitter of SETi. In the present study a direct measurement of the presynaptic action of FMRFamide has been obtained by studying its effects on the spontaneous release of transmitter from the slow motor neurone. Fig. 4 shows that a 3-min pulse of 10⁻⁶moll⁻¹ FMRFamide increases the spontaneous frequency of miniature end-plate potentials (Fig. 4A) but does not change their amplitude distribution (Fig. 4B) in a muscle fibre innervated by SETi but not FETi. This provides direct evidence for a presynaptic action of FMRFamide on the terminals of the slow motor neurone. However, FMRFamide must also have post-synaptic actions in this preparation since it is capable of modulating the rate of relaxation of twitch tension.

The potentiating actions of FMRFamide show some similarities with the actions of the biogenic amine, octopamine, on slow motor neurone twitch tension in the locust extensor-tibiae muscle (Evans & O’Shea, 1977; O’Shea & Evans, 1979; Evans, 1981). Therefore, the muscle’s response to FMRFamide was examined in the presence of 10^−smoll⁻¹ phentolamine, an effective blocking agent of octopamine receptors in this preparation (Evans, 1981). The responses to a 5-min pulse of 10⁻⁶moll⁻¹ FMRFamide were not reduced in the presence of 10⁻⁵ moll⁻¹ phentolamine (not shown) indicating that they are mediated via a receptor system which is independent of the octopaminergic system in this preparation. Further evidence for this conclusion is provided by the observation that the responses to submaximal doses of FMRFamide (e.g. 10⁻⁷moll⁻¹), unlike those to submaximal doses of octopamine (see Evans, 1984a,b), are not potentiated in the presence of the phosphodiesterase inhibitor isobutylmethylxanthine (IBMX) at a concentration of 10⁻⁷moll⁻¹ (not shown). This suggests that FMRFamide and related peptides do not bring about their effects in this preparation via a mechanism that increases cyclic nucleotide levels. This suggestion is confirmed by direct measurements of cyclic AMP levels in the extensor-tibiae muscle which do not increase after 10-min exposures to FMRFamide in the presence of 10⁻⁴moll⁻¹ IBMX at concentrations of FMRFamide between 10⁻¹¹ to 10⁻⁴moll⁻¹ (data not shown).

A preliminary screening of the actions of FMRFamide-related peptides on slow motor neurone twitch tension in the locust extensor muscle revealed that besides FMRFamide and YGGFMRFamide, the analogous peptides in which methionine was substituted by leucine, i.e. FLRFamide and YGGFLRFamide, were just as potent (Walther et al. 1984). The latter study concluded that the C-terminal sequence -Arg-Phe-NH₂ is essential for activity in the locust and that the preceding neutral amino acid (Met or Leu) may also be important, because γ₁-MSH, where the neutral amino acid is replaced by aspartate, is inactive. The results of specificity studies performed in the present study (see Table 1) confirm and extend these observations. The amidated dipeptide Arg-Phe-NH2 is inactive at concentrations up to 10⁻⁶moll⁻¹ and does not block the actions of FMRFamide. In addition, the related peptide Leu-Pro-Leu-Arg-Phe-NH2 (LPLRFamide in single letter notation) is almost an order of magnitude less potent than FMRFamide. This suggests that the presence of the phenylalanine residue at the beginning of the FMRFamide sequence and at position 4 of YGGFMRFamide may also be important for the action of these peptides on locust muscle.

Recent immunocytochemical studies have revealed the presence in the locust nervous system of cells immunoreactive to an antiserum raised against bovine pancreatic polypeptide (BPP), some of which are also positive for an antiserum raised against FMRFamide (Myers & Evans, 1985a,b). Thus the effect of a variety of pancreatic polypeptides was tested on slow motor neurone-induced twitch tension (Table 1). At concentrations up to 10⁻⁶moll⁻¹, however, 5-min pulses of bovine, human and avian pancreatic polypeptide had no effect on twitch tension produced by stimulating SETi at 1 Hz. In addition, they did not block the actions of FMRFamide on this preparation.

The responses of the extensor-tibiae muscle to octopamine vary in different parts of the muscle in relation to the proportions of slow and intermediate muscle fibres present (Evans, 1985b). To investigate the effects of FMRFamide on different regions of this muscle, SETi-induced twitch tension was recorded from seven blocks of muscle fibres designated blocks a—f and 135c,d in a proximal to distal direction according to the nomenclature of Hoyle (1978) (Table 2). A 5-min pulse of 10⁻⁶moll⁻¹ FMRFamide produced increases in slow twitch tension amplitude which varied from 63·6% in region a to 105·2% in region f. The relaxation rate of slow twitch tension also increased by different amounts in each of the regions of the muscle innervated by SETi. Stimulation of SETi at 1 Hz also produced tension responses from blocks d and 135c,d, but these responses were small and the changes due to FMRFamide were difficult to quantify. Table 2 also includes data for regional variations in responsiveness to octopamine from Evans (1985b) for comparison.

The extensor-tibiae muscle of the locust is modulated by a range of other neuroeffectors including octopamine and the neuropeptide proctolin (Evans & O’Shea, 1977, 1978; May, Brown & Clements, 1979), so the ability of these substances to interact with the modulatory effects of FMRFamide-like peptides was assessed on this muscle. Fig. 5 shows that low doses of DL-octopamine (10⁻⁸mol I⁻¹), which by themselves produce very little effect on the amplitude of slow twitch tension and its rate of contraction, do not block or potentiate the actions of a 5-min pulse of 10⁻⁷moll⁻¹ FMRFamide. The effects of higher concentrations of DL-octopamine on these two parameters are additive to those of FMRFamide. At 10⁻⁸moll⁻¹, DL-octopamine causes an increase in the rate of relaxation of slow twitch tension to which the effect of the 5-min pulse of 10⁻⁷ mol l⁻¹ FMRFamide is additive.

Proctolin, at a concentration of 10⁻⁹moll⁻¹, has a variety of effects on the extensor-tibiae muscle including an increase in the frequency of the myogenic rhythm and an increase in basal tension (May et al. 1979; Evans, 1982, 1984C). In addition, in some preparations it is possible to demonstrate an increase in the amplitude of slow motor neurone-evoked tension, together with increases in its relaxation and contraction rates induced by proctolin. However, if a 5-min pulse of 10⁻⁶mol I⁻¹ FMRFamide is introduced into the muscle superfusate in the presence of 10⁻⁹ mol I⁻¹ proctolin, the effects of FMRFamide are not blocked and are additive to the effects of proctolin on twitch tension amplitude, and its rates of contraction and relaxation (data not shown).

A number of the modulatory effects of octopamine (Evans & Siegler, 1982) and proctolin (Evans, 1982) on basal and neurally evoked tension in the locust extensortibiae muscle are dependent on the frequency of motor neurone stimulation. To test the effects of FMRFamide, the SETi motor neurone was stimulated at several frequencies and at each a 5-min pulse of 10⁻⁶moll⁻¹ FMRFamide was introduced into the muscle superfusate (Fig. 6). The result depended upon the frequency at which the motor neurone was being stimulated. At 1 Hz there was a pronounced increase in twitch amplitude but very little increase in basal tension. At 3 Hz there was a longer proportional increase in twitch amplitude which was superimposed upon an increase in basal tension. The increase in basal tension was maximal between 5 and 7 Hz.

As a further test of the effects of FMRFamide on muscle tension, SETi was stimulated every 60s with 10-s trains of pulses that were increased in frequency stepwise from 3 to 50 Hz. The ability of the muscle to develop and maintain tension in normal saline was compared with that in the presence of 10⁻⁶ mol l⁻¹ FMRFamide (Fig. 7). In normal saline, tension gradually increased throughout each 10-s stimulus and at the higher frequencies tested produced a smooth tetanic tension plateau. In the presence of FMRFamide the height of the individual tension transients visible up to 7 Hz was increased and this was superimposed upon an increased tension plateau. At frequencies of 10 Hz and above the height of the tetanic plateau was increased. The rate of rise and fall of tetanic tension did not appear to be altered in the presence of FMRFamide. At 20 Hz the threshold for an effect of FMRFamide on the height of the tetanic plateau occurred between 10⁻⁸ and 10⁻⁷moll⁻¹ and the effects of 10⁻⁷ and 10⁻⁶moll⁻¹ FMRFamide reached a maximum within 3min of the start of a prolonged pulse (Fig. 8).

The behavioural significance of the effects of octopamine on the extensor-tibiae muscle have been assessed (Evans & Siegler, 1982) in experiments where SETi was stimulated using the pattern of SETi activity reported for a free-walking locust by Burns & Usherwood (1979). Fig. 9A shows the tension profile produced by a 10-step sequence of stimuli to SETi in normal saline, whilst Fig. 9B shows that produced after exposure of the muscle to 10⁻⁶moll⁻¹ FMRFamide. In the presence of FMRFamide each of the steps produces more tension and the rates of increase and decrease of tension during the steps are increased. Up to the fifth step the tension maintained between the periods of stimulation is increased in the presence of FMRFamide but after this it is equal to that of the controls.

The effects of FMRFamide and related peptides on the amplitude (Walther et al. 1984), and the rates of contraction and relaxation of SETi-evoked twitch contractions in the locust extensor-tibiae muscle are very similar to those induced by octopamine (Evans & O’Shea, 1977; O’Shea & Evans, 1979; Buchan & Evans, 1980). Both the amine and peptides from this family increase the amplitude of the twitches and the rates of contraction and relaxation in a dose-dependent manner. In addition, the effects of both FMRFamide and octopamine outlast their presence in the superfusate by several minutes. There are, however, differences in the time courses of their actions. The effects of a prolonged pulse of 10⁻⁶moll⁻¹ FMRFamide on the amplitude and the rates of contraction and relaxation of twitch tension do not reach a maximum until after 4–5 min of exposure. With shorter pulses of FMRFamide maximal effects are not achieved. Octopamine, on the other hand, can achieve almost maximal potentiation of all three parameters in less than 1 min, although the time course of this effect is longer with longer exposure. A 30-s pulse (10⁻⁶moll⁻¹) of DL-octopamine takes almost 1 min to achieve maximum effects, whereas a 6-min pulse takes 2–3 min to reach a maximum. O’Shea & Evans (1979) suggest that a desensitization of octopamine receptors may be occurring during prolonged exposures to octopamine. It is possible that the slower time course of the FMRFamide potentiation may be due to the fact that FMRFamide is probably not the endogenous mediator of these effects. The related N-terminally extended peptide, YGGFMRFamide, produces maximum effects in a much shorter time (2-3min), which is more similar to the effects brought about by a prolonged pulse of octopamine (O’Shea & Evans, 1979). In addition, the threshold for an observable effect of YGGFMRFamide is 10–100 times lower than that for the FMRFamide itself. Another difference between the effects of FMRFamide-like peptides and octopamine on this preparation is that the dose-response curves for the effects of FMRFamide on twitch tension amplitude, relaxation rate and contraction rate can be superimposed, with roughly equal maximal effects of 200% increases occurring at 10⁻⁶moll⁻¹. In contrast, the effects of octopamine on relaxation rate are much bigger and have a lower threshold than its effects on twitch amplitude and contraction rate.

The effects of FMRFamide on slow twitch and tetanic tension in the extensortibiae muscle depend upon the frequency at which SETi is stimulated. At frequencies of stimulation above 1 Hz (3–10Hz), where the twitches summate but tetanus is incomplete, FMRFamide causes a marked increase in basal tension in the muscle. In addition, it also increases the height of the plateau of smooth tetanic contractions. These effects are the exact opposite to those induced by octopamine, which causes a large reduction in basal tension over this frequency range (Evans & Siegler, 1982). It is, however, similar to the effect of the pentapeptide, proctolin, on this preparation (May et al. 1979; Evans, 1982). Data on the effects of proctolin on locust skeletal muscle are rather sparse, but proctolin also increases basal tension in a frequency-dependent manner in this preparation (Evans, 1982), although the time courses of the actions of the two peptides are different. Following a 30-s pulse of proctolin while stimulating SETi at 1, 5 or 7 Hz, tension is increased in each case slowly up to a maximum, where it is maintained for 30–60 s. The muscle then relaxes abruptly, in a sharp step, during the recovery phase (Evans, 1982). The increase in basal tension brought about by FMRFamide continues to rise for as long as the peptide is present (5 min was the longest time tested). On removal of FMRFamide from the superfusate, after a further brief increase, the tension gradually declines. As well as the difference in the recovery rates, it is interesting to note that FMRFamide also increases the size of the individual twitch contractions on top of the increased basal tension. Proctolin does not usually produce such an effect, although in some preparations it has been observed to produce small increases in twitch tension amplitude, relaxation and contraction rates (P. D. Evans & C. R. Myers, unpublished observations). The mechanisms for the above actions of FMRFamide are unknown, but a possible explanation is that the effects of FMRFamide on the contraction and relaxation rates of the individual twitch contractions allow the muscle to continue to contract on top of an increased basal tension by decreasing the fusion of twitches which would otherwise result from the increased frequencies of stimulation. The effects of FMRFamide on SETi stepping patterns are also consistent with this explanation since FMRFamide increases the rates of rise and fall of tension in the individual steps, the amplitude of which increases. This is similar to the effect of octopamine on this pattern of stimulation (Evans & Siegler, 1982) but the basal tension maintained between the steps is not decreased by FMRFamide as it is by octopamine. A detailed behavioural analysis of the consequences of the actions of FMRFamide-like peptides will require information on the timing of their release and the degree of changes in their circulating haemolymph levels.

The actions of FMRFamide on locust skeletal muscle have some parallels and some differences with its observed actions on other invertebrate muscles. Thus in the locust, the ability of FMRFamide to increase maintained tension is similar to its effects on Helix tentacle retractor muscle (Cottrell et al. 1983a), on the anterior byssus retractor muscle of Mytilus edulis (Painter, 1982) on the isolated radula protractor muscle of Busycon contrarium (Greenberg & Price, 1979) and on leech heart (Kuhlman, Li & Calabrese, 1985). However, unlike its actions on the Helix tentacle retractor muscle and leech heart, FMRFamide does not induce or modulate the myogenic activity in the locust extensor-tibiae muscle, although it can do so in locust heart (C. M. Myers & P. D. Evans, in preparation).

The responses of the extensor-tibiae muscle to application of FMRFamide differ in different regions of the muscle, as has been found with the effects of octopamine on this muscle (Evans, 1985b). However, the variation in responsiveness to the peptide and the amine is different although their effects occur in the same blocks of muscle (i.e. blocks a, e and/) which contain the highest proportion of slow- and intermediate-type muscle fibres (Hoyle, 1978; Evans, 1985b). FMRFamide causes large increases in the amplitude of SETi-induced twitch contractions, with the largest increases occurring in region f. The effects on the relaxation rate are not as great as those on the amplitude, but again are most prominent in region f. The converse is true for octopamine. Octopamine-induced effects are greatest on the relaxation rates, with the largest increase again occurring in region f. Its effects on twitch amplitude are much smaller and are the same in all three regions measured. These results support the conclusion that the effects of FMRFamide are not mediated via an interaction with the octopaminergic receptor system.

FMRFamide has no effect on twitch tension induced by stimulation of the fast motor neurone (FETi) in the extensor-tibiae muscle, unlike octopamine which causes a large increase in the relaxation rate but has little effect on the rate of contraction or the amplitude of the twitches (Evans, 1981). This suggests that the actions of FMRFamide-like peptides in the extensor muscle may be confined to the slow muscle fibres which are innervated by SETi alone, receptors for these peptides being absent from the intermediate muscle fibres that are innervated by both SETi and FETi.

FMRFamide-related peptides produce their effects on slow motor neurone-induced twitch tension in the locust extensor-tibiae muscle by a combination of presynaptic and postsynaptic effects. Contractions, but not the depolarization, induced by the application of pulses of a high-[K⁺] saline to the muscle are potentiated by FMRFamide suggesting a postsynaptic effect on the excitation-contraction coupling mechanism (Walther et al. 1984). Further, the observation that FMRFamide also increases the rate of relaxation of twitch tension provides more evidence for a postsynaptic site of action. Evidence for a presynaptic site of action is provided by the observation that FMRFamide increases the average amplitude of excitatory junctional potentials but not the time integrals of depolarizations caused by superfusion of glutamate, the presumed rapid-acting transmitter of SETi (Walther et al. 1984). In addition, a direct measurement of the presynaptic action of FMRFamide has been obtained in the present study where it has been shown to increase the frequency, but not the amplitude distribution, of spontaneous miniature end-plate potentials in muscle fibres innervated by SETi but not FETi. The relative contributions of the presynaptic and postsynaptic components to the overall FMRFamide-induced increases in twitch tension amplitude and the rate of contraction are unknown at present.

The mechanisms of the pre- and postsynaptic actions of FMRFamide are also unknown at present. Blocking experiments using phentolamine, an effective blocking agent of the actions of octopamine in this preparation (Evans, 1981), indicate that FMRFamide does not mediate any of its effects by the activation of octopamine receptors. A similar conclusion has been reached for the potentiating actions of FMRFamide on acetylcholine-evoked and electrically stimulated contractions in the anterior byssus retractor muscle of the mussel (Muneoka & Matsuura, 1985). In the locust muscle, the actions of FMRFamide do not appear to be mediated -via increases in cyclic nucleotide levels. This again suggests that FMRFamide and octopamine are working on separate receptor systems in this preparation. In all non-cardiac muscles from molluscs so far examined, except for the gill muscles of Aplysia (Weiss et al. 1984), FMRFamide does not appear to mediate its excitatory effects via mechanisms that use cyclic nucleotides. In contrast, in some molluscan neurones, FMRFamide can induce changes in potassium and calcium conductances, some of which are mediated by cyclic nucleotides in some cells but not in others (Colombaioni, Paupardin-Tritsch, Vidal & Gerschenfeld, 1985).

The nature of the endogenous FMRFamide-like peptide that modulates slow motor neurone twitch tension in the locust extensor-tibiae muscle has not yet been elucidated. However, the different potencies of the FMRFamide-related peptides described in this study and the preliminary communication of Walther et al. (1984) have given some clues as to the structural requirements for the endogenous peptide. The studies of Walther et al. (1984) have indicated that the carboxyl-terminal sequence -Arg-Phe-NH₂ is essential for activity. They also indicate that activity is maintained in this preparation if the methionine preceding these two residues is replaced by another neutral amino acid (such as leucine, as in FLRFamide) but not by an acidic amino acid (such as aspartate, as in γ₁-melanocyte stimulating hormone). The present work confirms these observations and in addition shows that the C-terminal dipeptide Arg-Phe-NH₂ is not active. This suggests that this is not the only structural requirement and that the N-terminal may also play an important role. This possibility is reinforced by the fact that YGGFMRFamide is so much more potent than FMRFamide itself on this preparation. The related peptide, LPLRFamide, first isolated from chicken brain (Dockray et al. 1983), is an order of magnitude less potent than FMRFamide. This lends support to the above argument, since the proline residue may alter the structure of the N-terminal quite substantially. Proline has a pronounced ‘kink’ in its structure which may make an important contribution to the secondary structure of a polypeptide chain; for example, in avian pancreatic polypeptide, the proline residues form a partial helix, such that the proline rings contribute towards hydrophobic interactions in the tertiary structure of the molecule (Blundell & Wood, 1982). In addition to revealing FMRFamide-like peptides in the locust nervous system, immunocytochemical studies indicate the presence of pancreatic polypeptide-like (PP-like) peptides in the brain, corpora cardiaca, abdominal ganglia and neurohaemal organs (Myers & Evans, 1985a,b). However, none of the closely related peptides, avian PP, bovine PP or human PP, had any effect on the extensor-tibiae muscle at 10⁻⁶ mol I⁻¹. These peptides all terminate in the sequence -Arg-Tyr-NH2, and thus it seems likely that the receptors in this muscle are quite specific for the C-terminal sequence -Arg-Phe-NH₂. It should be noted, however, that the residue which precedes arginine in the pancreatic polypeptides is proline, which as discussed above is important in the three-dimensional structure of the peptide and will thus affect its biological activity. Immunocytochemical evidence (see Evans & Myers, 1986) has also revealed that some of the neurones containing FMRFamide-like peptides in the locust are also immunoreactive with a monoclonal antibody raised against another peptide first isolated from the molluscan nervous system, namely small cardioactive peptide (SCP_B) (Morris et al. 1982). SCP_B antagonizes the actions of FMRFamide on various molluscan preparations (Murphy et al. 1985; P. E. Lloyd, personal communication) but does not alter any of the modulatory actions of FMRFamide described in the present paper. SCP_B does not modulate neuromuscular contraction directly in the locust extensor muscle but it does produce a dose-dependent acceleratory effect upon the myogenic rhythm (Evans & Myers, 1986).

Future studies on the biological role of FMRFamide-like peptides in locust skeletal muscle require the isolation and sequencing of the endogenous peptides from the locust. The extensor-tibiae muscle of the locust hindleg will undoubtedly be of much use as a bioassay during these isolation procedures.