EXCITATORY ACTIONS OF Antho-RFamide, AN ANTHOZOAN NEUROPEPTIDE, ON MUSCLES AND CONDUCTING SYSTEMS IN THE SEA ANEMONE CALLIACTIS PARASITICA

Ultrastructural studies have shown that transmission at most synapses in the nerve net of sea anemones is chemical. Dense-cored vesicles, associated with synaptic membrane specializations, have been found at interneuronal and neuromuscular junctions (Westfall, 1973). Physiological studies, demonstrating conduction failure following addition of excess magnesium ions, also suggest that neurotransmission in sea anemones is chemical (McFarlane, 1973). So far, however, the nature of neurotransmitter substances in sea anemones, or in coelenterates in general, has not been established (see Martin & Spencer, 1983, for a review). It has been shown that adrenaline excites slow muscle contractions in Calliactis parasitica, but the ineffectiveness of known adrenaline agonists and antagonists indicated that adrenaline was not a transmitter (Ross, 1960a,b).

Recently, using immunocytochemistry and radioimmunoassays, we have shown that a substance resembling the molluscan neuropeptide Phe-Met-Arg-Phe-amide (FMRFamide; Price & Greenburg, 1977) is abundant in the nervous systems of sea anemones and other coelenterates (Grimmelikhuijzen, 1983; Grimmelikhuijzen & Graff, 1985; Grimmelikhuijzen, Graff & Spencer, 1987). Of antisera against different fragments of FMRFamide, those against RFamide were especially potent in demonstrating the peptide. By using a radioimmunoassay with an RFamide antiserum, we have purified the RFamide-like peptide from the sea anemone Anthopleura elegantissima (Grimmelikhuijzen & Graff, 1985, 1986). This peptide, which was named Antho-RFamide, has the sequence <Glu-Gly-Arg-Phe-amide (Grimmelikhuijzen & Graff, 1986). It was found to be abundant in sea anemones and other anthozoans, and may be a neuropeptide generally produced by anthozoan species (Grimmelikhuijzen & Groeger, 1987).

The aim of the present study is to investigate the action of Antho-RFamide on muscles and conducting systems in the sea anemone Calliactis parasitica. There are at least three conducting systems in sea anemones: the through-conducting nerve net (TCNN), and two slow conduction systems (SSI and SS2). The cellular identities of SSI and SS2 remain enigmatic (McFarlane, 1982), although evidence is accumulating that they have a neural basis (McFarlane, 1984). The three conducting systems interact and control many aspects of behaviour, such as simple reflexes, longterm behavioural phases and complex behaviour patterns (McFarlane, 1982). We show here that Antho-RFamide has diverse excitatory actions in C. parasitica; it excites both the SSI and the SS2 and causes prolonged, repeated contractions of several slow muscle groups. It is possible, therefore, that Antho-RFamide is an excitatory transmitter in parts of the sea anemone nerve net.

Synthetic Antho-RFamide was used in the experiments described below. This material was very pure, as was shown by HPLC using a variety of reversed-phase columns (Grimmelikhuijzen & Graff, 1986). Calliactis parasitica were obtained from the Marine Biological Laboratory, Plymouth, and kept at room temperature (16–19°C). The study used isolated preparations, half-animal preparations (McFarlane, 1973) and whole animals.

The recording, stimulating and display apparatus were as previously described (McFarlane, 1984). All recordings were made from two polyethylene suction electrodes attached to widely separated tentacles to facilitate recognition of pulses. Pulses in the three known conducting systems (TCNN, SSI, SS2) have a distinctive appearance and although small (<10μV) are usually easy to detect. Pulses were identified before each recording sequence by applying a single shock to a stimulating electrode on the column base. Antho-RFamide was applied internally by injection through a cannula inserted into one of the cinclides, and was applied externally directly into the bath to give a known final concentration. To apply Antho-RFamide to a restricted area, solutions of the peptide in sea water were sucked into a 1 mm internal tip diameter suction electrode that was then attached to the column. This technique has been used previously to apply food extracts in studies of pre-feeding responses in Urticina eques (McFarlane & Lawn, 1972; Lawn, 1975).

Contractions were recorded on kymographs. Sphincter muscle contractions of whole animals were monitored by attaching a hook to the upper column at right angles to the longitudinal axis of the body. The isolated preparations included sphincter muscle rings, mid-column circular muscle rings and longitudinal body wall strips (containing parietal muscles). The simplest sphincter preparation was a ring cut from the upper 0·5 cm of the body, complete with attached tentacles. Sphincter preparations were sometimes partly trimmed by removing the tentacles, and sometimes fully trimmed by removing ectodermal and endodermal tissue. As the ectoderm is pigmented, and the mesogloea white, it was obvious when the ectoderm had been totally removed. Light microscope examination of stained sections confirmed removal of the endoderm. The other preparations were as described before (e.g. Ross, 1960a,b,McFarlane, 1974a,b). No anaesthetic was used as excess magnesium delays recovery and leads to abnormal responses of sphincter preparations (Lawn, 1976). Preparations were used 24–48 h after operation.

When Antho-RFamide was applied externally to 15 intact anemones (final concentration 10⁻⁶ mol 1⁻¹), the only movements shown were oral disc expansion and slight mouth opening; two animals detached. Repeated contractions were not evident. However, when the peptide, at a concentration of 10⁻⁷-10⁻⁶mol 1⁻¹ was injected into the coelenteron of anemones, or applied to various isolated preparations, there was an increase in muscle tone, contraction amplitude and contraction frequency in several muscle groups (Fig. 1). When injected into an intact anemone attached to a shell, there was an almost immediate response (Fig. 1A). (In the example shown in Fig. 1A spontaneous sphincter muscle contractions were unusually large and frequent, but the peptide had a similar effect on quiescent anemones.) The increase in tone, amplitude and frequency rarely lasted more than 30 min.

The three types of isolated sphincter preparations, intact, partly trimmed and fully trimmed, all contracted after addition of Antho-RFamide (Fig. 1). Calliactis parasitica sphincter muscle shows both fast and slow contractions (Ross, 1960b). The contractions evoked here were always slow. Partly trimmed preparations showed a low threshold (10⁻⁷mol⁻¹), a long latency and repeated washes were required to stop the response (Fig. IB). Untrimmed sphincter preparations only responded at higher concentrations (10⁻⁶mol 1⁻¹). The latency, however, was shorter than in the partly trimmed preparations, and the effect was quickly terminated by washing. Interpretation of the response of untrimmed and partly trimmed sphincter rings is complicated because these preparations also include another muscle group, the endodermal circulars. Complete removal of the ectoderm and endoderm, however, leaves just the sphincter muscle, which is embedded in the mesogloea (Robson, 1965). Fully trimmed sphincter preparations were not spontaneously active, but Antho-RFamide again caused an increase in contraction tone, amplitude and frequency (Fig. 1C). Compared with the partly trimmed preparation, the response latency was short and the action comparatively easy to wash out, but the threshold was the same, about 10⁻⁷mol1⁻¹. The reduction in latency presumably arose through removal of some permeability barriers.

Circular muscle rings from the mid-column, below the sphincter, and muscle rings from the column base were always spontaneously active and also responded with an increase in tone, frequency and contraction amplitude. With the mid-column muscle ring the action was slow to start and difficult to wash out (Fig. ID). Mid-column rings did not respond to concentrations below 10⁻⁶ mol 1⁻¹. From the orientation of the rings it was clear that the contractions were only due to circular muscles and that no other muscle groups were involved.

Two other muscle preparations were tried: longitudinal column strips containing parietal muscles and circular rings cut from the pharynx wall. Both were spontaneously active and responded to Antho-RFamide (10⁻⁶moll⁻¹) with increased tone, contraction amplitude and contraction frequency.

Long-term monitoring of electrical activity from tentacles (McFarlane, 1973) has shown that the TCNN is spontaneously active, with short bursts of 3–12 pulses. Bursts are 20–60 min apart in unstimulated animals attached to shells. The SS2 is also spontaneously active and the pulses appear at a low frequency, with no obvious pattern of bursting. SSI activity, however, does not occur spontaneously.

The effects of external bath application of Antho-RFamide on the conducting systems are shown in Fig. 2A, a portion of the complete record. Frequency/time plots for the SSI and SS2 pulses in the complete record are shown in Fig. 2B,C, and the effects on SSI, SS2 and TCNN pulse frequency are summarized in Table 1. In the bath applications the peptide was normally added in the corner and several minutes elapsed before a clear response began. The only system that showed a marked increase in activity was the SSI (see Discussion), with the threshold being 10⁻⁷mol1⁻¹.

To investigate where Antho-RFamide acts on the ectoderm, a suction electrode containing a 5× 10⁻⁶mol 1⁻¹ solution of peptide was attached to various ectodermal surfaces, and evoked activity was monitored with recording electrodes on the tentacles. Attachment of the peptide-containing electrode to tentacles proved impossible as they always contracted. One successful contact was made on the oral disc and the result was a short burst of SSI pulses (6 pulses at a mean frequency of 1 pulse every 3 s) before electrode contact was lost. There was no problem with attachment to the column: the electrode stayed on for long periods. In most trials (7 out of 10) no response was evoked from the column. In the remainder, a discharge of SSI pulses occurred (Fig. 2D). The longest recorded sequence consisted of 45 SSI pulses with the interval between the pulses being 13–15 s for most of the response. Application of 5×10⁻⁶mol1⁻¹ peptide to the column never evoked SS2 pulses. In two cases a few TCNN pulses appeared immediately on application of the electrode but it was not clear if these arose from mechanical stimulation or a genuine action of peptide on the TCNN.

Responses to Antho-RFamide injected into the coelenteron are shown in Fig. 3. Fig. 3A is an extract from the record and Fig. 3B,C show SSI and SS2 frequencies during the complete record. Table 1 compares pulse frequencies before and after injection of the peptide. With a single injection of Antho-RFamide (10⁻⁶mol1⁻¹) there was an early, marked increase in SS2 activity that was complete within 2 min. It was also obvious that the SSI responded to injected Antho-RFamide. In the example shown in Fig. 3 the SSI discharge lasted about 14min. There was no marked change in TCNN activity: a short burst of TCNN pulses was seen (Fig. 3A) but as it came 3 min after injection, it may have been a spontaneous event. After the response had stopped, a second application of Antho-RFamide (2× 10⁻⁶ mol 1⁻¹) again resulted in an increase in SSI and SS2 activity (Table 1). In six other animals where injection was attempted, two anemones contracted within seconds and the recording electrodes became detached. In the other four animals SSI and SS2 pulse frequencies increased markedly after injection but the records were not complete. The records that could be obtained without loss of electrode contact are summarized in Fig. 4, which shows dose/response curves for the actions considered to be significant. As the SSI is normally silent, any increase in pulse frequency represents an action of Antho-RFamide, so the threshold (between 10⁻⁸and 10⁻⁷mol1⁻¹) is easily detected. The SS2, however, is spontaneously active, and the threshold is consequently more difficult to determine (below 10⁻⁶mol1⁻¹).

The slow conduction systems, SSI and SS2, may be nerve nets and, as conduction in both systems is quickly blocked by excess magnesium (McFarlane, 1973), transmission is probably via chemical synapses. The SSI occurs throughout the ectoderm but it may have transmesogloeal connections to the endoderm (McFarlane, 1976; Lawn, 1980). The SSI is not spontaneously active, so excitation by externally applied Antho-RFamide was easily detected in the present study (Fig. 2; Table 1). Externally applied Antho-RFamide might act either on the SSI chemoreceptors or on synapses within the SSI system. An effect on the SSI chemoreceptors is unlikely because when the peptide was applied locally to the column the SSI response did not show adaptation (Fig. 2D), whereas a marked decline in pulse frequency with time was seen when food extract was applied to the column of Urticina eques (McFarlane & Lawn, 1972). This suggests that Antho-RFamide is not simply mimicking the action of a pre-feeding excitant.

SSI activity can also be excited by internal application of Antho-RFamide (Fig. 3A,B). Here, however, two types of response were detected. One, not seen with ectodermal application, consisted of bursts of SSI pulses (underlined in Fig. 3A). Similar bursts of SSI pulses occur in detached anemones during the behaviour termed the pre-settling phase (McFarlane, 1983), and have also been seen following single high-intensity shocks to the endoderm (Jackson & McFarlane, 1976). The other response, a low-frequency discharge with no obvious patterning, was similar to that seen with ectodermal application and may be due to the peptide leaking from the coelenteron and acting on the ectoderm.

The SS2 is an endodermal system which is spontaneously active (McFarlane, 1973). The SS2 is also chemosensitive: it responds to proline and reduced glutathione during the feeding response (McFarlane, 1975). The SS2 clearly responds to internally applied Antho-RFamide (Table 1): pulse frequencies of 6·8 and 10·5 pulses min⁻¹ are up to three times higher than normal levels. As the SS2 pulses appear within seconds of injection, they probably arise from a direct action of the peptide on SS2 receptors or SS2 conducting units. Externally applied Antho-RFamide does not appear to excite the SS2 (Table 1): only very small changes in pulse frequency occurred, and the levels reached were within the normal range of spontaneous activity.

There is no strong evidence for any action of Antho-RFamide on the TCNN. The TCNN shows bursts of spontaneous activity, and a burst can be triggered by a single touch or a single shock (McFarlane, 1974b). Consequently it is difficult to interpret the origin of any burst seen following Antho-RFamide application. For example, the burst of four TCNN pulses seen 3 min after injection in Fig. 3A could be a spontaneous event, and the few TCNN pulses which were sometimes seen with local application of peptide to the column (Fig. 2B) may have arisen through mechanical stimulation from the suction electrode containing the peptide. The effects of Antho-RFamide on the TCNN as presented in Table 1 are all very small (an increase of 0·5—1·5 pulses min⁻¹) and they are within the normal variation of TCNN activity.

Are the observed muscle contractions due to SSI and SS2 activity evoked by Antho-RFamide? Although the SSI can excite circular muscle contractions, it does not affect the sphincter muscle (McFarlane, 1976; Lawn, 1980). The SSI, therefore, cannot be responsible for the excitatory actions of Antho-RFamide on all the muscles. The same holds for the SS2 which also has no action on the sphincter muscle (McFarlane, 1974a).

Muscles are normally activated by the TCNN, with pulse frequency determining which muscles contract (Ross, 1957). The level of TCNN activity following addition of Antho-RFamide is insufficient to explain any of the observed muscle contractions so it is possible that the peptide may be acting presynaptically at TCNN neuromuscular junctions or postsynaptically on the muscles.

Antho-RFamide is a naturally occurring peptide in the sea anemone Anthopleura elegantissima (Grimmelikhuijzen & Graff, 1986) and in the pennatulid Renilla kôllikeri (Grimmelikhuijzen & Groeger, 1987). As two phylogenetically widely separated anthozoans produce the same peptide, it is likely that Antho-RFamide occurs in all anthozoans, including Calliactis parasitica. Immunocytochemistry has shown that RFamide immunoreactivity in sea anemones is always associated with neurones (Grimmelikhuijzen et al. 1987), suggesting that Antho-RFamide is exclusively a neuropeptide. In C. parasitica, RFamide immunoreactivity is present in neurones in all parts of the body (C. J. P. Grimmelikhuijzen, D. Graff & I. D. McFarlane, in preparation). This, together with the excitatory actions of Antho-RFamide on the SSI, SS2 and slow muscles, suggests that Antho-RFamide may be an excitatory neurotransmitter or neuromodulator.

We would like to thank Kate Boothby for assistance with the electrophysiological recordings, and the Deutsche Forschungsgemeinschaft for financial support (Gr762/4). CJPG is the recipient of a Heisenberg Fellowship (Gr762/1).