Single shocks to the column sometimes evoke tentacle contractions, ranging from slight movement of a few scattered tentacles to rapid bending or shortening of all the tentacles. Some individuals are more responsive than others. Complex bursts of electrical activity follow single shocks, but only in tentacles that contract.
These single shocks excite pulses in two conducting systems - the through-conducting nerve net (TCNN) and the ectodermal slow conduction system (SS1). When a single shock evokes contractions and bursts of electrical activity, these usually follow the SSI pulse, rarely the TCNN pulse. Stimulation of the SS1 alone causes tentacle contraction in responsive anemones.
Fast tentacle contractions always follow the second of two closely-spaced TCNN pulses: the TCNN shows facilitation (Pantin, 1935a). An SSI pulse, however, does not facilitate subsequent pulses in either the SSI or TCNN.
There are two pathways for activation of tentacle contractions. The TCNN pathway is mechano-sensitive and normally requires facilitation. The SS1 pathway is mechano- and chemosensitive, only requires a single SSI pulse to evoke contraction, but is very labile. It is proposed that the TCNN and the SS1 do not excite the ectodermal muscles directly, but via a multipolar nerve net.
Multiple conducting systems coordinate the behaviour of sea anemones. There are at least four separate systems (McFarlane, 1982). Pulses from three of these (the through-conducting nerve net, TCNN, the ectodermal slow system, SS1, and the endodermal slow system, SS2) can be recorded extracellularly from tentacles.
The tentacles show both feeding movements and protective reflexes. Their contractions range from local movements after gentle stimulation, to rapid symmetrical contractions after strong stimulation anywhere on the body. Tentacles have an ectodermal longitudinal muscle layer (responsible for most local and symmetrical movements) and a weak endodermal circular muscle layer. They have a well-developed ectodermal nerve net but few, if any, endodermal nerve cells (Van Marie, 1977).
Fast contractions in Calliactis parasitica are coordinated by a specialized nerve net, the TCNN. A single TCNN pulse does not normally evoke fast muscle contraction but does facilitate the action of a subsequent pulse (Pantin, 1935a). Josephson (1966) recorded muscle action potentials preceding fast tentacle contractions in C. polypus.
Single shocks to the column of Calliactis parasitica sometimes make the tentacles contract. Pantin (1935a) noted that, ‘Light stroking of the column may also produce occasional slight upward contractions of individual tentacles scattered round the disc’. When describing the results of electrical stimulation of the column, he stated that, ‘Again, in responsive animals, a single stimulus may produce a reaction in the form of a slight waving of tentacles scattered round the disc’. Davenport (1962) suggested this showed that TCNN pulses can sometimes elicit tentacle contractions without facilitation.
I show in this paper that tentacle contractions in response to single shocks are usually coordinated by the SSI, not by the TCNN. The tentacle muscles may receive dual excitatory innervation or the TCNN and SSI may act via a single interposed conducting system. These results further extend the list of known SSI-coordinated actions: pedal disc detachment, excitation of circular muscles, and activation of TCNN pacemakers in Calliactis parasitica (McFarlane, 1969b, 1976, 1983), inhibition of inherent contractions of oral disc muscles in Urticina eques (McFarlane & Lawn, 1972), and activation of swimming pacemakers in Stomphia coccinea (Lawn, 1976).
MATERIALS AND METHODS
Calliactis parasitica (pedal disc diameters, 2–4 cm) were supplied by the Plymouth Marine Laboratory. They were left attached to Buccinum shells, kept at 16–20 °C in a 70–1 tank of artificial sea water, and fed weekly. Three separate batches of ten animals were used; each batch was replaced after 4 weeks.
Recordings were made with polyethylene suction electrodes attached to tentacles. Recording and display apparatus consisted of Isleworth Electronics A103 preamplifiers, a Telequipment D1011 oscilloscope, a Datalab DL902 transient recorder, and a Linseis LY1700 plotter. Stimuli (1 ms duration) were given through a suction electrode on the column. Experiments were performed under dim light.
Direct mechanical records of tentacle contractions were not made as most movements were too weak to be detected by the available equipment. Instead, times of contraction were determined from ciné films made on Eastman 4-X negative film. The bright lights used did not appear to modify contractions. Each stimulus triggered an electronic flash to mark the event on the film. Film speed was 24 or 32 frames per second; this provided sufficient accuracy to determine whether contractions followed TCNN pulses or SS1 pulses.
Contractions following column stimulation
A single shock anywhere on the column sometimes evoked tentacle movements in responsive animals, regardless of whether or not a recording electrode was attached to a tentacle. The response ranged from a barely detectable twitch of a few tentacles, through to a large fast contraction of all the tentacles. Contractions appeared to fall into two groups. In one, movement closely followed the stimulus (an ‘early’ contraction), in the other it came 0·5–1 s after the shock (a ‘delayed’ contraction). This subjective distinction between early and delayed contractions was very clear and was later supported by analysis of ciné films. Fig. 1 shows that an early contraction follows a single TCNN pulse whereas a delayed contraction follows a single SS1 pulse.
Early contractions appeared to involve more or less simultaneous contraction of all tentacles. The movement was often barely visible and was seen in less than 25 % of individuals tested and it followed less than 10 % of shocks applied. The tentacles did not bend in this response, but just shortened slightly (Fig. 1 A).
Delayed contractions were more variable in extent and strength, but they were more common than early contractions: they were seen at some time in more than 90 % of individuals. Members of a group of 10 anemones were stimulated daily for 2 weeks with three test shocks (10 V, 1ms) 3 min apart. Each day between three and six anemones (mean 4·9) showed delayed contractions (to at least one of the test shocks), although generally the contraction was slight and detectable only as an inward movement of the inner cycle of tentacles. Only one anemone responded every day and one never reacted at all. Individual variations in responsiveness were not obviously related to nutritional state. No way was found to make unresponsive anemones respond.
The strength of the delayed response, and the number of tentacles involved, varied considerably. When the nature of the contraction could be determined, it was generally a bending movement (Fig. 1B, C). The commonest movement (50 % of observed responses) was an inward sweep of the inner cycle tentacles. An outward movement of the outer cycle tentacles occurred in 15 % of observed responses. Rarely (less than 3 % of responses) the contraction was powerful and produced marked shortening of all tentacles. The contraction in the remaining cases was too small to be clearly defined; sometimes only a few scattered tentacles moved, sometimes all tentacles gave a barely perceptible twitch. In less than 2% of all responses the shock elicited both an early and a delayed contraction.
Unlike early contractions, delayed contractions could be seen to spread around the oral disc as waves of tentacle movements. Clearly the conducting systems coordinating the two responses conduct at different rates or have different pathways. At 18 °C, the TCNN in the oral disc conducts at 60–100 cm s−1 (Pantin, 1935b) whereas the SSI conducts at 15cms−1 (McFarlane, 1969a).
Electrical activity associated with delayed contractions
Electrical events that accompany tentacle contractions will be termed Tentacle Contraction Pulses (TCPs), regardless of whether they follow TCNN or SS1 activity. The nature of these potentials is not known, they are simply taken as evidence for contraction at the recording site.
A single, low-intensity, shock to the column elicited a TCNN pulse and an SS1 pulse. If no delayed contraction was evoked then the recorded SSI pulse had no detectable after potentials (Fig. 2A). When delayed contractions did occur, electrodes picked up pulses shortly after the SSI pulse (Fig. 2B, C, D), but only from tentacles that moved. These TCPs came 0–150 ms after the SS1 pulse, they never preceded it when delayed contractions were seen. The TCPs in Fig. 2B were associated with an inward sweep of the inner cycle of tentacles. The recording electrode was later moved to a tentacle in another cycle; no TCPs were recorded when only the inner cycle tentacles moved. Fig. 2D shows large TCPs, recorded from the same site as in Fig. 2B, but when all the tentacles gave a large contraction. Fig. 2B and 2D represent the extremes of the SS1-associated contractions.
There was considerable variation in both TCP amplitude and in the delay between the SS1 pulse and the TCPs. These variations occurred between recording sites and also between recordings at the same site. In Fig. 2C the delay was short at recording site 2 but long at site 1, whereas only 3 min later this situation was reversed (Fig. 2D). At site 1 the TCPs were small in Fig. 2C but only 3 min later they were large (Fig. 2D).
Electrical activity associated with early contractions
When early contractions were seen it was found that the TCPs closely followed the TCNN pulse, not the SS1 pulse. Fig. 3A shows three consecutive stimuli recorded at one site. The first shock gave no response, the second evoked a delayed contraction, and the third gave an early contraction. With the delayed contraction TCPs followed the SSI pulse. With the early contraction TCPs followed the TCNN pulse. Rarely the early contraction was large and then the TCPs after the TCNN pulse were also large (Fig. 3B). Large early contractions may result from the facilitating effect of a preceding spontaneous TCNN pulse. Fig. 3B shows one of the rare occasions when a single shock evoked both early and delayed contractions. Two recording electrodes were used, on different tentacles. At one recording site the early TCPs were larger than the delayed TCPs, at the other site the opposite was true. It is not known if this reciprocal relationship is a common feature.
Stimulation of the SSI alone can elicit tentacle contractions
The SS1 has a higher threshold than the TCNN but can, however, be stimulated alone via a shallow ectodermal flap cut in the column (McFarlane, 1969b). Flap stimulation at intensities below SS1 threshold never evoked delayed contractions or TCPs. Stimuli at, or above, SS1 threshold evoked delayed contractions in responsive animals (Fig. 4A). The second shock in Fig. 4A produced a rare event, the apparent reflection of a contraction wave. Here the delayed contraction began, as usual, as two waves of tentacle movement that spread in opposite directions around the oral disc. Then, the waves did not cancel out but passed each other and this gave the impression of a reflected wave. The waves stopped when they collided on meeting again at their point of origin on the disc. The recording shows a second SSI pulse and following TCPs, presumably associated with the second contraction.
Although, as in the case of the double contraction described above, two SS1-activated contractions can occur close together, there is normally a marked decline in responsiveness with repetitive stimulation (Fig. 4B-D). The effect of repetitive stimulation varied considerably from animal to animal and from time to time in the same animal.
Comparison of evoked and spontaneous tentacle contractions
Tentacle contractions may be evoked by TCNN or SS1 pulses, but tentacles can also contract spontaneously, without preceding TCNN or SSI activity. Such contractions involve single tentacles: they do not spread. They occur regardless of whether or not an electrode is attached to the tentacle. Fig. 5 shows recordings made with two electrodes on the same tentacle. Both electrically-stimulated and ‘spontaneous’ SSI pulses may cause contractions and TCPs (Fig. 5A, B). The SS1 is probably not spontaneously active (McFarlane, 1973) : the ‘spontaneous’ SS1 pulses may arise from mechanical stimulation. The tentacles did, however, sometimes contract in the absence of SS1 pulses (Fig. 5C, D) and the movement was accompanied by TCP-like electrical events. Note that these spontaneous events spread, comparatively slowly, from tip to base of tentacle. The conduction velocity of the SS1 pulses recorded in Fig. 5B, and hence the rate of spread of evoked contractions, is about 18cms −1, whereas the spontaneous events in Fig. 5C and 5D appear to spread at 7cms−1 and 2 cm s−1 respectively. The latter value may be an underestimate as it is difficult in Fig. 5D to say where the TCP in R2 begins.
Lack of interaction between SSI and TCNN pathways
The SS1 and TCNN pathways to the tentacle ectodermal muscles have different properties. The TCNN pathway normally requires facilitation whereas the SS1 pathway operates in response to a single pulse, not to the second of two closely spaced pulses.
A single TCNN pulse rarely causes tentacle contraction but does facilitate the action of a closely-following TCNN pulse (Pantin, 1935a). Pantin believed that this facilitation occurred at the neuromuscular junction. TCPs always follow the second of two TCNN pulses 100–1500 ms apart: the amplitude and duration of the TCPs is related to the interval between stimuli (Fig. 6A). When the interval was short, TCPs were mainly single large biphasic or triphasic potentials similar to the pulses that accompany fast sphincter muscle contraction (Josephson, 1966). At longer intervals, the TCPs were less synchronous and resembled the pulses seen during SS1-evoked contractions.
There was no evidence for heterofacilitation between the SS 1 and TCNN pathways to the muscle. A single SSI pulse did not facilitate the action of a closely-following SSI pulse (Fig. 6B). Although a single TCNN pulse facilitated the TCNN pathway (Pantin, 1935a), it did not enable a subsequent SSI pulse to evoke contraction (Fig. 6C). Again, a single SSI pulse did not facilitate the TCNN pathway to the muscle (Fig. 6D, E).
Lability and variability of the response
The likelihood that a single SSI pulse will elicit a contraction declines with repeated stimulation. This is not due to sensory adaptation as the stimuli are electric shocks. It is not due to muscular fatigue as the muscle can still be activated via the TCNN. It is an activity-dependent decline in responsiveness, presumably at a site somewhere between the SS1 and the muscle. Habituation occurs in Anthopleura elegantissima (Logan, 1975) but it is not known which conducting system is involved.
Individual Calliactis parasitica differed considerably in their responses to single stimuli. Fleure & Walton (1907) commented on differences in responsiveness between different individuals of several actinian species and mentioned that captivity may modify characteristic activities. Perhaps the delayed contraction is less labile and less variable in freshly-caught animals.
Function of the response
A single SS1 pulse can cause contraction of tentacles, a movement at times so small as to be without obvious significance. At other times, a strong inward sweep of the inner cycle tentacles, or an outward flourish of the outer cycle tentacles, may aid in food gathering. The SS1 can be excited by both touch, especially of the lower column, and by dissolved food substances (McFarlane & Lawn, 1972), so such tentacle movements might allow the capture of nearby food. A few seconds after food is held above the oral disc, out of reach of the tentacles, the tentacles begin to twitch. Some movements are uncoordinated, and may involve direct stimulation of individual tentacles, but some are waves of contraction, presumably coordinated by the SS1.
Well-developed tentacle contractions occur in Boloceroides mcmurrichi, an anemone that swims by lashing its tentacles. Lawn & Ross (1982) recorded bursts of pulses (tentacle burst pulses - TBPs) associated with each tentacle flexion. They found no SSI but their recordings clearly show a small SS 1 -like pulse (flexion trigger pulse - FTP) just preceding each TBP burst. The FTP system is probably the SSI, here again eliciting tentacle movements. Their records show a 100–200ms delay between FTPs and TBPs, comparable with the 0–150ms delay seen between SSI pulses and TCPs in Calliactisparasitica.
Carlgren (1942) considered the tribe Boloceroidaria (e.g. Boloceroides mcmurrichi) to be more primitive than the tribe Thenaria (e.g. Calliactis parasitica). SS1-activation of tentacle contractions may be a primitive feature, one that is much reduced and of little functional significance in Calliactis. Recordings from the two species in the primitive actinian suborder, Protantheae, should be revealing: Gonac-tiniaproliféra also swims by tentacle flexions (Robson, 1971) and Protanthea simplex reacts to various stimuli with violent twitching or lashing of the tentacles (Manuel, 1981).
Organization of the ectodermal ‘nerve plexus’
The nervous structure of anthozoan ectoderm is not clearly understood. Hertwig & Hertwig (1879–80) described a ‘nerve plexus’ formed from sensory cell branches and from ganglion cells. Electronmicrographs show this plexus is 10 ‘neurites’ thick (Kawaguti, 1964; Van Marie, 1977). As the mean neurite diameter is less than 1 pm, and the fibres appear tightly packed, there must be around 100000 profiles in a transverse section of a 3 mm diameter tentacle. Such a neuronal abundance surely exceeds requirements for behavioural coordination and contrasts sharply with the smaller number of bipolar and multipolar nerve cells seen in methylene blue stained whole mounts (Robson, 1963). The status of the plexus has been questioned. Vander-meulen (1974) suggests the layer is a concentration of basal ramifications from the supporting cells. Van Marie (1977) says the plexus contains both nervous elements and supporting cell branches but that when profiles are empty (without vesicles) one cannot determine their nature.
I will consider two recent descriptions of neural structure - Peteya (1973) on Ceriantheopsis americanus and Van Marie (1977) on four anthozoans — and try to jelate them to the physiological findings. In Ceriantheopsis the plexus contains rout fibre types (A, B, C, D) in the ratios 4000:4000:4: 1 and with diameters of 0·3–1μm: 0–10·3μm: 4–6μm10–20μm. Peteya believes C and D fibres to be axons of bipolar neurones and A fibres to be axons of multipolar neurones respectively. A fibre cell bodies are 5–10 μm in diameter and each carries 3–4 neurites. Only A fibres bear neuromuscular synapses. Van Marie could not distinguish fibre classes by diameter but he gave pharmacological evidence for two distinct nerve nets. One employs a catecholamine transmitter and the other is purinergic; only the purinergic system synapses with the muscles. The TCNN in Calliactis parasitica consists of large bipolar nerve cells (Robson, 1965). These may correspond to Peteya’s C and D cells and Van Marie’s catecholamine-containing cells. The multipolar nerve cells of Calliactis (Hertwig & Hertwig, 1879–80;,Robson, 1963) may be equivalent to Peteya’s A cells and Van Marie’s purinergic cells. I propose therefore that the TCNN excites tentacle muscles indirectly by way of a multipolar nerve net. Robson (1971) proposed such an arrangement in the primitive anemone Gonactinia prolifera. In the hydromedusan Polyorchis penicillatus a network of small multipolar neurones lies between the swimming motoneurones and the epitheliomuscular cells (Singla, 1978; Spencer, 1982).
The proposed model (Fig. 7) positions a multipolar nerve net between the TCNN and the muscles. From this the neuromuscular facilitation of tentacle muscles (Pantin, 1935a) may be reinterpreted as being interneural facilitation between the TCNN and the multipolar nerve net. This may not, however, apply to the specialized fast muscle - the sphincter - studied by Pantin. Indeed, only bipolar neurones, not multipolar neurones, have been found immediately adjacent to sphincter muscle cells (Robson, 1965). The model shows multipolar neurones linking the TCNN to the endodermal circular muscles - a multipolar net overlies the circular muscles (Robson, 1965). The TCNN action on the endodermal multipolars may be both excitatory and inhibitory; Ewer (1960) showed the TCNN has a dual action on circular muscle contractions.
Some multipolar cells may be pacemakers. At times the pacemaker output may pass directly to the innervated muscles, to excite the spontaneous contractions so obvious in isolated preparations. Ross (1960) states that the evidence, on balance, supports a neurogenic origin for such contractions. At other times the pacemakers may excite the TCNN, giving rise to TCNN bursts that coordinate the activity of many muscle groups (McFarlane, 1974a). Pacemaker⟶.TCNN connections may be absent in the ectoderm (McFarlane, 1974a).
The model shows sensory input to all the conducting systems. The TCNN and SSI both respond to touch (Pantin, 1935a ; McFarlane, 1969b), the SS1 and SS2 respond to dissolved food substances or to contact with shells (McFarlane, 1970, 1976), and the SS2 may receive input from endodermal receptors that detect stress in endodermal muscles (McFarlane, 1974a). Multipolar neurones may be directly excited by sensory cells and thus cause local contractions in response to touch; this implies there is limited spread of activity through the multipolar net. Local contractions could, however, involve spread of activity through the muscle field itself - the model shows links between muscle cells.
Both slow conduction systems may connect with the multipolar neurones. Trans-mesogloeal links between the SSI and a multipolar net are suspected in Stomphia coccinea (Lawn, 1980). SS2 activity inhibits spontaneous contractions of endodermal muscles (McFarlane, 1974a) and inhibits TCNN pacemakers (McFarlane, 1974b). The SS1 excites endodermal circular muscle contractions (McFarlane, 1976), inhibits spontaneous contractions of ectodermal muscles (McFarlane & Lawn, 1972), activates TCNN pacemakers (McFarlane, 1983) and excites occasional tentacle contractions (present work). The observed variability of delayed contractions may be explained by regional and temporal changes in the functional state of the SS1 connexions with the multipolar cells. The 0–150 ms delay between SSI pulses and TCPs might be evidence for an indirect action of the SS1 on the muscles. Alternatively, the SSl may not always activate the muscle at the recording site. The delay would then suit result from a slow spread of TCPs through the muscle sheet. Fig. 7 shows junctions between muscle cells. Coupling could be electrical or mechanical, but neither have yet been demonstrated.
Nature of the SSI
It is unlikely that the SS1 is the multipolar nerve net. There are few nervous elements, other than sense cells, in the column ectoderm (Hertwig & Hertwig, 1879–80), yet the SS1 is present in this region. The SS1 could, however, be a nerve net formed by sensory cell branches.
Some hydrozoan conducting systems are nervous, but others are non-nervous and involve cell-to-cell conduction in sheets of epithelial or epitheliomuscular cells (Mackie, 1970). Non-nervous conduction has not been demonstrated in anemones; the somata of the ectodermal supporting cells might form the SS1 but they are tall and thin whereas non-nervous systems invariably involve cells only a few microns high (Anderson, 1980). Perhaps a network of supporting cell branches forms the conducting elements. Shelton (1982) made a similar suggestion, based on an unpublished electron microscopic study of Calliactis parasitica. Such a ‘fibre network’, a form of non-nervous conducting system, might be equivalent to Peteya’s B fibres. Peteya, however, thought that the B fibres of Ceriantheopsis americanus were part of an effector system, a proposal supported by the finding that synapses were only found from A fibres onto B fibres; none were seen between B fibres. Synapses between B fibres would be necessary to form a conducting system and also as the basis for the observation that SSI activity is abolished by excess magnesium ions (McFarlane, 1969a).