Neuromuscular Transmission in A Sea Anemone

The quick withdrawal of sea anemones is the best-studied behavioural response in coelenterates and the one which is most often used in textbooks to illustrate the physiological properties of coelenterate neurones and muscles. The basic features of this response were analysed about 30 years ago in an important series of papers by Pantin (1935 a, b, c, d), and subsequent work has added significant details (e.g. Hall & Pantin, 1937; Pantin, 1952; Passano & Pantin, 1955; Ross, 1952, 1957; Robson, 1961). The main features of this response are the following.

1. The withdrawal is caused by rapid, symmetrical contraction of principally one muscle or group of muscles. In Metridium the major muscles involved are the longitudinal retractors of the mesenteries which depress the oral disk; in Calliactis the principal muscle is the marginal sphincter, which shuts the top of the column like a purse (Pantin, 1935a; Hall & Pantin, 1937).

2. Muscle contraction is initiated by activity in a conducting system in the column which conducts all-or-nothing events. Although the column conducting system responds as a single unit, activity at one point apparently reaching all other points, there is regional differentiation with respect to conduction velocity. The conduction velocity ranges from over 100 cm./sec. vertically in the mesenteries to 10-15 cm./sec. transversely around the base of the column (Pantin, 1953b; Robson, 1961; T = 18-20° C.).

3. The mechanical response of the muscle shows both summation and facilitation, with the resulting contraction being strongly dependent on the frequency of impulses in the column conducting system. A single stimulus to the column is usually ineffective, two stimuli within a suitably short interval are required to initiate contraction. The ineffectiveness of single stimuli appears to be due to requirements for facilitation at or near the junction between the conducting system and the muscles, for single shocks activate the conducting system and single strong stimuli directly to the muscle cause it to respond (Pantin, 1935a, d).

The considerable body of physiological information about quick withdrawal responses in anemones has previously been obtained using mechanical contraction as the measured output of the system. The nature of the conducting mechanisms and links between conduction and contraction have all been inferred from the relations between controlled electrical or mechanical stimuli and the resulting mechanical response. This paper describes some of the properties of a link between stimulation and contraction which can be measured as an electrical event, the muscle action potential.

The experiments described below were carried out with a tropical species of Calliactis. Dr C. Cuttress (Institute of Marine Biology, University of Puerto Rico) informs me that the senior synonym of this anemone is C. polypus (Forskål, 1795). Some preliminary observations on responses to stimuli were made at the Hawaii Marine Laboratory but most of the work and all of the electrical recording was done at the Eniwetok Marine Biological Laboratory on Eniwetok Atoll. Only two animals were available at Eniwetok, but these were studied intensively over a period of 4 weeks. The anemones were kept in finger-bowls in running sea water at 29-31° C. and were occasionally fed pieces of mollusc. All experiments were carried out at room temperature, 29-31° C.

It was found that electrical potentials could be recorded from a suction electrode on the surface of an anemone over the sphincter muscle or from an electrode slipped over the end of a tentacle following suitable stimulation of the anemone’s column. These potentials were obviously correlated with contraction of the sphincter or tentacle, suggesting that they were muscle action potentials. There was initially some concern that the potentials were artifacts caused by movement of tissue and electrode. To investigate this possibility, suction electrodes were used which were attached to the movable anode of an RCA 5734 mechano-electric transducer tube, the transducer forming one arm of a bridge circuit. Thus the electrode served both as a lead for electrical recording and as an attachment to the animal for mechanical recording. With this recording arrangement it was found that the electrical potentials began well before the mechanical movement, excluding the possibility that the potentials are movement artifacts. The electrode-transducer combination proved convenient for recording the time-course and amplitude of both electrical and mechanical responses, and was used in most of the experiments.

The electrode-transducer configuration was modified several times during the investigation; probably the most successful arrangement is shown in Fig. 1. The recording electrodes were made from drawn glass tubing and had fire-polished tips of 0·2-0·6 mm. internal diameter. The recording electrodes were joined by plastic tubing to a syringe. A chloride silver wire was fixed in the plastic tubing near the electrode end. Sea water was drawn into the electrode by withdrawing the syringe plunger until the water reached the silver wire. The electrode was then placed on the anemone and the syringe plunger was further withdrawn to create suction and hold the electrode in place. The suction was kept low to minimize distortion and damage of the tissue beneath the electrode. Potentials were recorded between the wire in the electrode and a second chloride-coated silver wire in the sea water surrounding the anemone. The electrical resistances of the fluid-filled electrodes and connecting tubing were 1-3 × 10⁵ Ω. The potentials were amplified with capacitor-coupled amplifiers with long time-constants and displayed on an oscilloscope. Electrical stimuli were given by passing 1 msec, current pulses between a plastic suction electrode, usually on the lower column, and a second chloride-coated silver wire in the surrounding sea water, the suction electrode serving as cathode. In all cases 1 min. or more was allowed between successive stimulation trials.

The use of small electrodes (to allow recording from small, discrete areas) and low suction (to avoid tissue damage) made keeping the electrode attached to the animal somewhat difficult. Attempts to make the mechanical recording nearly isometric were quite unsuccessful as the electrodes always became detached during major contractions. Consequently a rather compliant coupling between the electrode and transducer was used. This resulted in the muscle contraction being neither isometric nor isotonic, for the tension increased as the muscle shortened. In the arrangement illustrated in Fig. 1 the compliant portion of the coupling was about 5 cm. long and consisted of two strands of 0·2 mm. diameter platinum-iridium wire tightly twisted together. The output of the transducer bridge was calibrated and found to vary linearly with electrode displacement through the range of displacements encountered. Even with a compliant coupling the electrode still frequently became dislodged, especially during repetitive stimuli or bursts of after-discharge which sometimes followed single stimuli, and many more incomplete series of observations were collected than complete ones.

The movements of the tentacles and sphincter during contraction are often not unidirectional. For example, the edge of the oral disk over the sphincter often rises slightly before it moves orally during contraction. With some recording configurations the initial rising of the edge was seen as an output from the transducer of one polarity while the major oral movement gave a later and larger response of the opposite polarity. In the early experiments the position of the transducer and coupling were adjusted until the recorded mechanical response was nearly monophasic. In later experiments a link of flexible plastic tubing was inserted between the compliant coupling and the suction electrode itself. The presence of this link did not appreciably change the response of the transducer to movements which were in the direction of the long axis of the flexible tube but greatly reduced the response of the transducer to electrode movements perpendicular to this axis. With this arrangement the mechanical responses were recorded as monophasic, or nearly monophasic with only a very small initial deflexion opposite to that of the main response.

The recorded electrical responses to a given set of stimulus conditions were sometimes, and the mechanical responses were often, rather variable. The variability of the electrical responses was probably due to small changes in the electrode position and possibly also to continuing damage of the tissue beneath the electrode. In the present study attention was centered on quick muscle contractions, those which have a fixed and simple relation to external stimuli. The very slow contractions which often appear spontaneously and which are related to external stimuli in a complex manner not yet understood (Pantin, 1952) were not considered. The quick responses which were measured occurred on a background of slow, tonic changes in the shape of the anemone. The mechanical responses of coelenterate muscle have been found to depend on the initial length and tension of the muscle (Horridge, 1958; Arai, 1965). The tonic contractions altered the length and tension of the animal’s musculature and probably accounted for, at least in part, the frequent variability of the recorded mechanical responses.

Fig. 2 shows the response of the sphincter muscle to pairs of stimuli delivered to the lower column. This example and others like it indicate the following.

1. There is usually no detectable electrical or mechanical response to single stimuli. The few exceptions to this can probably all be accounted for on the basis of multiple firing in the column conducting system following single stimuli, or spon-taneous firing in the conducting system which preceded the administered stimulus by an interval so short that its facilitory effects had not completely decayed at the time of the stimulus (cf. Ross, 1952).

2. The recorded electrical response is primarily positive with a following negative component, part of which may be due to capacitative coupling to the amplifier. The size of the electrical response depends largely on the interval between stimuli. The largest potentials recorded from the sphincter during two-shock experiments were of the order of 1 mV.

3. The electrical response clearly precedes mechanical contraction. In a number of trials using stimuli separated by 100 msec.—an interval short enough to give large mechanical responses—the latency between the onset of the electrical potential and the onset of contraction averaged 12 msec. (S.D. 4 msec., one determination from each of twenty-four preparations). This value for the electrical to mechanical latency cannot be regarded as very reliable. The mechanical response usually does not rise abruptly from the base-line so its onset cannot be determined very accurately. Use of a more sensitive arrangement for mechanical recording would probably allow some-what earlier detection of the onset of contraction. Furthermore the electrical record represents events at a discrete locus while the mechanical record can be influenced by movements anywhere in the animal. If contraction of all parts of the sphincter is not synchronous, contraction of the portion beneath the electrode might be preceded by contraction of other areas, resulting in a deceptively short value for the measured electrical-to-mechanical latency.

The responses of the sphincter following column stimulation have a sharp threshold and are not changed by changes in the intensity of supra-threshold stimuli except in those cases where strong stimuli initiate multiple firing. This result follows from Pantin’s (1935a) demonstration that conduction in the column of sea anemones is all-or-nothing in character. The relation between the stimulus interval and the size of the evoked responses is shown in Figs. 3 and 4. These figures are from the series for which the most points were obtained ; other, shorter series all gave essentially the same result. The size of the recorded responses increases with decreasing stimulus intervals to a minimal interval below which there is abruptly no response to the second shock. This minimal interval presumably results from refractoriness in the column conducting system. The size of the minimal interval was rather variable, and ranged from 100 msec, to about 15 msec, in different experiments with stimuli 1·2-1·5 times threshold. Over half of the measured values for minimum effective stimulus interval (14 out of 27) were 25 msec, or less. Since the stimuli were not much above threshold, this must represent relative refractoriness; the absolute refractory period is probably somewhat shorter.

A pair of closely spaced stimuli, both of which excite the column conducting system, cause obvious contraction of the tentacles as well as contraction of the marginal sphincter. Tentacles near the lateral edge of the oral disk shorten, inner tentacles both shorten and bend orally. An example of the electrical and mechanical responses of a tentacle to paired stimuli is shown in Fig. 5. Responses recorded from tentacles differ from those of the sphincter in two ways.

1. Tentacles, in the resting state, appear to be poised at a more excitable level than is the sphincter. They usually produce small electrical potentials to single supra-threshold stimuli and there is occasionally a very small muscle contraction as well. The responsiveness of the tentacles can be seen in the whole anemone even without electrical or mechanical recording. A single stimulus which excites the column conducting system usually causes slight contraction of a scattered few of the tentacles on the oral disk. Similar tentacle responses in Calliactis parasitica were used by Pantin (1935 a) as evidence that the column conducting system is activated by single stimuli.

2. The electrical potential from a tentacle following the second of a pair of stimuli is sometimes similar to that seen from the sphincter but often it is more complex and dominantly negative (Fig. 9). This difference may be due to the activity being sometimes conducted into an electrode over a tentacle rather than ending beneath the electrode as is the case for sphincter recordings.

The largest potentials recorded from a tentacle in two-shock experiments were of the order of 600 μV. The electrical-to-mechanical latency in tentacles appears to be about the same as that of the sphincter (average = 13 msec, with stimuli too msec, apart, S.D. = 4 msec., 1 determination from each of 29 different preparations). Fig. 6 shows that the relation between tentacle responses and the interval between a pair of stimuli is essentially the same as that for the sphincter. The only significant difference seen in this and other similar series results from the responsiveness of the tentacles to single stimuli. The second stimulus of a pair separated by 600 msec, or more usually evokes a small response in a tentacle like that seen to single stimuli, while such widely spaced stimuli are ineffective in exciting the sphincter.

The responses of the tentacles and sphincter to effective stimuli are independent of where on the column the stimuli are given. The stimulus threshold is not every-where the same, however. Responses are easier to obtain with stimuli of moderate strength when the electrode is at the top of the column over the sphincter or on the thinner portion of the column below the cinclides than when it is on the central region of the column. For example, paired current pulses of 3 mA. intensity and 1 msec, duration were passed through a 2 mm. diameter suction electrode placed at different points on the columns of five anemones. With each anemone five trials were made at different loci for each of four column levels. The proportion of stimulus pairs (interval = 200 msec.) which evoked sphincter contraction varied as follows: below cinclides, 72% ; on or beside cinclides, 4% ; in middle of column, 4% ; near upper margin of column, 68%. Shocks from an electrode placed on the oral disk usually produced responses identical to those seen with column stimuli. Generally a single shock caused twitching of scattered tentacles on the oral disk, two shocks caused symmetrical contraction of the sphincter and contraction of all tentacles. Other responses could also be obtained by stimulating the oral disk, but these required stronger stimuli than the preceding. For example, single strong shocks sometimes initiated local infolding of the disk margin as well as widespread twitching of scattered tentacles. The local responses were variable in occurrence and extent and were not examined in detail.

Potentials could be recorded with an external suction electrode only from the upper margin of the column (overlying the sphincter), the tentacles, and the tentacle-bearing portion of the oral disk in one of the two anemones used for electrical recording. In the other anemone, which was somewhat smaller, potentials could sometimes also be recorded from central areas of the column above the cinclides. The potentials from the middle column, however, were always small—less than 20 μV—with stimulus conditions which gave 50-400 μV. potentials from the sphincter or tentacles.

Occasionally a single shock initiates repetitive firing in the column conducting system and causes pronounced contraction of the sphincter and tentacles (Fig. 7). As many as three electrical potentials and contraction steps, presumably representing four impulses initiated in the conducting system, have been seen following a single shock. This after-discharge was variable and unpredictable; sometimes it was frequent while later, with the same animal, it was very rare. Multiple discharge is more common following strong shocks than following just supra-threshold ones, and it is much more common after the second shock of a pair than after the first. The interval between a directly evoked response and an after-discharge impulse or between after-discharge impulses in a train is usually between 80 and 120 msec., but it can be as long as 400 msec, or as short as 15 msec. The latter value must lie very close to the refractory period of the conducting system.

During repetitive stimulation the mechanical responses of the sphincter to the individual shocks both facilitate and sum (Pantin, 1935 a; Fig. 8). The tentacles probably behave similarly, but contraction of the strong sphincter muscle dominates and obscures a mechanical record from a tentacle by about the third or fourth stimulus. The contribution made by each shock of a series to the total muscle activity is best seen in the electrical record, for the potentials are short and do not fuse ; they remain distinct events even when they occur at relatively high frequencies (Fig. 8).

The electrical potentials show marked facilitation for the first few shocks of a series, but they soon reach a plateau and begin to decline in size if the stimulation is maintained (Figs. 8, 9). The potentials from a tentacle also facilitate during repetitive stimulation and tentacle potentials usually reach a plateau height earlier in a stimulus series than do those from the sphincter (Figs. 9, 10). The fact that tentacles become maximally active earlier in a stimulus train than does the sphincter is of obvious adaptive significance. The withdrawal response of the anemone is normally initiated by mechanical stimuli which evoke a battery of impulses in the column conducting system. (Pantin, 1935a; Passano & Pantin, 1955). During such repetitive firing the tentacles will become maximally active sooner than the sphincter, and will be rapidly withdrawing before the sphincter closes the top of the column over them.

The electrical potentials which can be recorded from the surface of an anemone seem almost certainly to be muscle action potentials. They are most easily recorded in the immediate vicinity of muscles clearly involved in the withdrawal response, they precede contraction by a short interval, and they are not all-or-nothing but are continuously graded with the size of the mechanical contraction. The possibility cannot be excluded that the potentials arise from nervous activity with the increase in potential amplitude during repetitive stimulation being caused by interneural facilitation leading to an increased density of active neurones beneath the recording electrode. This alternative seems unlikely, however, since the potentials do not seem to increase in discrete steps, and the smooth form of the potentials gives no hint that they are produced by the summation of a number of short, all-or-nothing events. The dominant positivity of the potentials from the sphincter is somewhat surprising, but the shape and polarity of the potentials cannot be evaluated until the current pathways are known. It might be pointed out that action potentials recorded from the surface of crustacean muscles are often positive (e.g. Wiersma & Wright, 1947) even though internal recording indicates that the individual fibres become depolarized during activity.

For the most part the results of the present study are not surprising; they agree with and confirm Pantin’s earlier analysis of quick withdrawal by anemones. The sensitivity pattern of the column to stimuli and the areas from which potentials can be recorded in Calliactis polypus are in good agreement with the distribution of elements in the withdrawal response of the related C. parasitica. The column conducting system for this response is thought to be an endodermal nerve net which is especially well developed on the retractor faces of the mesenteries (Pantin, 1952; Robson, 1961). Processes from this net pass through the mesogloea surrounding the sphincter to reach this muscle (Robson, 1965). In the middle of the column, where the stimulus threshold in C. polypus is greatest, the endodermal nerve net is separated from the external ectoderm by a thick layer of mesogloea. The threshold is lower in the pedal regions where the column is thin with little mesogloea between ectoderm and endoderm in well-expanded animals, and it is also lower near the sphincter where the nerve fibres probably come closer to the external surface as they move peripherally to reach the sphincter. Nerve-cell processes have been found running from the endoderm of the mesenteries to the ectoderm of the oral disk in the anemone Mimetridium cryptum (Batham, 1965). Similar nervous connexions between the mesenteries and disk have not been described for Calliactis, but they probably exist since the disk is clearly in conductional continuity with the column conducting system. The sphincter and tentacles are involved in the withdrawal response and they produce easily recorded potentials. The mesenteric retractor muscles may also take part in this response (Pantin, 1935 b) but these muscles are quite far removed from the external surface of the animal so it is not expected that they would produce potentials which could be easily recorded with a surface electrode. Potentials can be recorded from the oral disk adjacent to the tentacles, indicating that disk musculature is also activated during the withdrawal response. Possibly the disk musculature acts synergistically with the sphincter to close the top of the column.

There are some interesting differences between the responses of Calliactis polypus and those reported for C. parasitica. Pantin (1935a) found that the mechanical response of the sphincter of C. parasitica to paired stimuli was maximal with stimulus intervals somewhat longer than the absolute refractory period of the column conducting system and the contraction fell off markedly with shorter intervals even when the second shock was clearly effective in activating the muscle. Pantin suggested that this decline with short stimulus intervals might be due to asynchrony in the refractory periods of elements in the nerve net; some elements innervating the musculature being refractory at short stimulus intervals while others are not. In C. polypus the mechanical and electrical responses continued to increase up to the shortest effective stimulus interval, but the stimuli were always near threshold and it is not known if the responses would have declined if stronger stimuli and shorter intervals, even closer to the absolute refractory period, had been used.

A striking difference between C. polypus and C. parasitica is the different time-scales on which the two animals operate. The refractory period found for C. parasitica was 40-65 msec. (Pantin, 1935 a), and paired stimuli separated by 1·3 sec. or more were effective in evoking twitch-like contraction of the sphincter (Pantin, 1935 a; Fig. 7). In C. polypus the refractory period and maximal interval between effective stimuli were both shorter by a factor of 2-3. Experiments on the two species were carried out at different temperatures—18-20° C. for C. parasitica and 29-31° C. for C. polypus— but in both cases the temperature at which the animal was studied was close to environmental temperature. Muscular facilitation has been found to decay at the same rate in Bunodactis, a tropical anemone, and in Metridium senile, a temperate species, when both are measured at the same temperature (Pantin & Vianna Dias, 1952). But since facilitation decay has a positive temperature coefficient, the decay is faster in the tropical anemone at environmental temperatures. In view of the fact that genetic and physiological acclimation to temperature differences has been found in many animals (but not all) of a number of different groups (cf. Prosser & Brown, 1961), it is somewhat surprising that tropical and temperate anemones differ so much in the temporal characteristics of their responses when each is at its normal temperature.

C. polypus and C. parasitica also differ in the way in which they respond to stimulation of the oral disk. Pantin (1935a) described the responses of C. parasitica to repetitive shocks delivered to the oral disk. Typically the first stimulus had no effect, the second stimulus evoked a local response (contraction of a nearby tentacle or local disk contraction) and the third stimulus evoked a widespread response (twitch of scattered tentacles or sphincter contraction) or a local response which sometimes could be made to spread progressively further around the disk by sub-sequent stimuli. In C. polypus a single shock to the oral disk of lowest effective intensity always activated the whole disk as evidenced by twitching of widely scattered tentacles. The activity also presumably reached the column conducting system, foi a second shock shortly after the first caused sphincter contraction. In C. polypus the oral disk and column seem to contain a continuous all-or-nothing conducting system. Pantin chose to explain the local and through-conducted responses in C. parasitica as being due to activity in a single conducting system. He interpreted the examples of progressively increasing spread with repetitive stimulation to mean that facilitation was required before there was transmission between conducting units at the boundaries of adjoining sectors, a process which he termed ‘intemeural facilitation’. The difference between the responses reported for C. parasitica and those found in C. polypus might be explained by suggesting that some of the junctions between conducting elements in C. parasitica require facilitation to become transmissive while most or all of the junctions in C. polypus are permanently facilitated (cf. Josephson, Reiss & Worthy, 1961). But the situation is probably not as simple as that. With C. polypus strong electric shocks and mild mechanical stimulation, such as is produced by lightly stroking the disk with a probe, do evoke local disk contraction. The local responses are not confined to the area immediately stimulated, and the distance to which excitation spreads following mechanical stimuli is to some extent graded with the intensity of stimulation. The presence of through-conduction and local responses in the same tissue argues for the presence of two conducting systems in that tissue. Horridge (1958) found that electrical stimulation of the oral disk of Cerianthus could evoke local or through-conducted responses, with the local responses having the lower threshold. These results suggest the presence of two conducting systems in the oral disk of this anthozoan as well. Whether the presence of multiple conducting systems in the oral disk is a general feature of Anthozoa, or indeed whether it is found in the most studied anthozoan, C. parasitica, would seem to be yet undetermined.

Isolated tentacles from the anemones Radianthus and Anemonia and both isolated and in situ tentacles from Tealia have been found to give contractions to single, directly applied electrical shocks, with the size of the contraction being graded with stimulus intensity (Davenport, 1962). Judging from the results with C. polypus it seems likely that such stimulation activates the tentacles in ways other than through the conducting system which mediates the quick-withdrawal response of the whole anemone, perhaps by directly exciting the musculature. When the through-conducting system is activated by column stimuli the tentacles give tiny contractions or no contractions due to single stimuli, and the size of the contraction with repetitive stimuli is independent of stimulus strength above threshold.

This study was supported by Public Health Service Research Grants NB 05263 and NB 06054 and by a grant to the Eniwetok Marine Biological Laboratory from the U.S. Atomic Energy Commission. I wish to thank Mr S. C. March for invaluable assistance during the course of this work.