Further Observations on Co-Ordination and the Determination of Frequency in the Peristomial Cilia of Stentor

In an earlier paper (Sleigh, 1956) it was shown that the mechanical process involved in ciliary activity functions independently of the co-ordination process. The conduction of metachronal waves is independent of mechanical factors, and involves the successive excitation of cilia along the row.

Such a transmission process could function in one of two ways. It could be the result of a continuous neuroid conduction process along the length of the peristomial row of cilia, so that cilia are stimulated to contract as the impulses pass but do not themselves take any part in the transmission of the metachronal impulses; alternatively, it could be the result of a step-by-step process passing from one cilium to the next, so that the excitation of one cilium would result in the production of an excitatory impulse responsible for the stimulation of the next cilium.

If the conduction process is of the first type, the neuroid transmission is very slow (about 600µ/sec. at 20° C.; Sleigh, 1956) compared with that found in other conducting tissues. In the second type of transmission, however, there are two parts : the intraciliary excitation process, and the interciliary conduction process. Here the actual conduction process could easily be of a speed comparable with that in nerve and heart muscle, provided that the intraciliary excitation process occupies a large part of the total time of the conduction. There is some evidence for this; the wave conduction process in Mytilus lateral gill cilia, described by Gray (1930), is much slower than that in Stentor, but in Mytilus many more cilia are involved in the formation of each metachronal wave. Also, it was shown in the author’s previous work on Stentor that the metachronal wave velocity was much increased by digitoxin; this drug lowers the threshold of excitability in heart muscle (Sollman, 1950), and might similarly affect the intraciliary excitation process.

The rate of beat of a cilium will normally be controlled by a series of excitatory impulses transmitted by one or other of the methods mentioned above. Individual isolated cilia, such as the abfrontal cilia of Mytilus studied by Gray (1930) or cilia of ciliate protozoans isolated with a small mass of protoplasm by Verworn (1889), show a spontaneous rhythmic beat. Thus the conducted excitation cannot be the only method of evoking a response in a single cilium.

The Stentor used in these experiments were from the same stock as those used earlier (Sleigh, 1956), and were cultured by the same method. The movements and metachronal characteristics of the peristomial cilia of the animal were briefly described in that paper. Measurements of the frequency of beat were made by means of the rotating mirror stroboscope, and the wavelength was measured from photographs; the product of these two measurements gives the velocity of metachronal wave transmission. The temperature was kept approximately constant during the measurements by the method previously described, and was between 18 and 20° C. in all of the experiments.

In the microdissection experiments, cuts in the peristome edge of Stentor were made with glass microneedles made for this purpose in a de Fonbrune microforge. The animal was placed in a hanging drop on a glass slide which was raised on blocks on the stage of a microscope during the operations. When the animal had completely expanded, the edge of the peristome was cut in an appropriate place by compressing the animal against the glass slide with the glass needle controlled by a de Fonbrune micromanipulator.

Several authors have claimed that a cut across a row of cilia resulted in a disorganization of the co-ordination of the row. Experiments have been carried out on ctenophores by Verworn (1889) and Parker (1905), and on ciliate protozoans by Verworn (1889) and Taylor (1920). The precise nature of the ‘disorganization’ is not given, though Verworn describes differences in rhythm, and his figures show different metachronal wavelengths on the two sides of the cut. On this evidence a cut may result in a change in either the frequency, or the wave velocity, or both.

The arrangement of the peristomial cilia of Stentor is shown in Fig. 1. Waves of activity pass along the length of the row from one end to the other, starting from the gullet region. The frequency of beat is normally the same all along the row.

In a typical experiment, several readings of the frequency and of the wave velocity of the peristomial cilia were first taken. The peristome edge was then cut across in one or two places, and the frequency and wave velocity of the various regions were again measured as quickly as possible. Further measurements were taken at intervals until the cut had healed. The cuts usually resulted in a disturbance of the normal rhythm of the system, so that the cilia on the two sides of the cut beat with a different frequency. Correlated with this difference in frequency was a difference in wavelength.

Calculation of the wave velocity indicated that it was approximately the same on the two sides of the cut in each experiment. The difference in wave velocity between the distal region and the proximal region varied between – 3·7 and + 4·3 %, with a mean from seventeen experiments of –0·36% and a standard error of the mean of 0·57 %. This is close to the normal variation in measurements of wave velocity, and the mean difference is not significantly different from zero. Although the cut interrupted the wave conduction process, this interruption did not change the wave velocity in any part of the peristome.

The frequency differences were much larger; the maximum recorded was 19·1 %, while the mean difference in frequency on the two sides of the cut was 8·4%. The differences within individual experiments were statistically significant in nearly every case, but the differences were very variable. When several cuts were made in the row each region showed a different frequency; examples of results from five experiments of this type are shown in Table 1.

Within each isolated region all cilia beat with the same frequency, and it was usually found that the frequency in the proximal region was the same as the frequency along the whole row before the operation. Immediately distal to the cut, new waves of activity of a different frequency were propagated towards the distal end of the row. The cilia on the distal side of the cut beat more slowly than those on the proximal side in twenty-two cases out of twenty-nine, while in the other seven cases the distal frequency was greater than that proximal to the cut. The magnitude of the difference in frequency could not be correlated with the position or size of the cut. The cut healed in 15 min. to an hour or so, depending on the size of the cut. On healing, the distal frequency reverted to the original value and the frequency was the same throughout the row. This change in frequency usually took place in a single jump, like that which occurred when the cut was made, but occasionally it was more gradual as physical continuity was re-established.

These experiments indicate that the existence of a common frequency over the whole row depends on physical continuity for the whole length of the row. This physical continuity can be broken by a cut in the row, but it has also been observed to be broken by a sharp bend in the peristome edge, such as might be caused by the contraction of one or two myonemes in the Stentor body wall, or by the peristome edge being caught up by an object in the environment. Such a discontinuity invariably resulted in a decrease in the frequency of the distal part. On other rare occasions the uniform frequency along the row may break down without any obvious cause. In all these cases there is apparently some failure of the transmission mechanism, a failure which separates the cilia distal to the ‘break’ from the proximal cilia. A new frequency is set up in the isolated region, which is the same for its whole length and which seems to be determined quite arbitrarily, because it bears no particular relation to the frequency with which the cilia were beating previously.

The observation that the conduction process can stop and start at any place tends to suggest that the transmission is not by a continuous neuroid conduction of the first type above, but by the second method involving a series of alternate excitation and conduction stages, in which the cilia are involved in the transmission process.

The cilia in the gullet region appear to have a shorter metachronal wavelength than those in the main part of the row, as shown in Fig. 1. Since the frequency is the same throughout, this means that the metachronal wave velocity in the gullet region is less than that farther along the row.

Measurements of the metachronal wavelength of peristomial cilia taken as close as possible to the gullet were compared with measurements made in other parts of the row. In every case it was found that the wavelength was shorter in the gullet region than it was distally, and that the first wave was the shortest. The wavelength increased with each successive wave until it reached a maximum and thereafter remained the same throughout the distal part of the row. Results from seven sets of these observations are given in Table 2.

This variation in the rate of conduction will not fit in with either of the transmission mechanisms outlined above. However, a further investigation of the cilia of the gullet region revealed the reason for this variation in wavelength and wave velocity. The cilia in this region are smaller and closer together than those in the distal region; as they become larger and farther apart the wavelength increases. Counts of the number of cilia concerned in the formation of each metachronal wave showed that the number of cilia per wave was the same whatever the wavelength, even though the wavelength in the distal region was about twice that of the first proximal wave measured. Seven sets of observations illustrating this are given in Table 3.

These data can be compared with those from other experiments. The average number of cilia per wave in forty-two normal animals was 6·4 and the average frequency in 361 normal animals at 18·5° C. was 27·7 beats/sec., so that the average number of cilia stimulated per second by a single metachronal impulse is about 178 and the average phase difference between successive cilia is about 5·6 msec. The average wavelength in 303 normal animals was 23·8µ, and thus the average interciliary distance was 3·9µ. The figures for the distal region shown in Table 3 agree closely with the average figures, but the corresponding values for the wavelength and the interciliary distance in the gullet region are much smaller than the average, although the number of cilia per wave is approximately the same.

Since the number of cilia in every wave is the same, the number of cilia stimulated per second by a single metachronal impulse must be the same all along the row, and the phase difference between adjacent cilia will be constant (about 5-6 msec.). The wave velocity thus depends on the number of cilia involved in the transmission, and not on the linear distance travelled by the metachronal wave. Evidence is thus strongly in favour of a transmission mechanism in which the cilia themselves take part.

Metachronal co-ordination has been explained in the past both as a mechanical process and as a neuroid transmission process, and evidence has been put forward in favour of both theories. Kraft (1890) found that a stimulus causing increased activity in a ciliated epithelium was transmitted across a region of quiescent cilia without any mechanical activity in the quiescent region. At about the same time Verworn (1891) found that there was no wave conduction past a ctenophore swimming plate that was not allowed to beat, but that waves were conducted normally if it was allowed to beat without touching any other plate. However, Parker (1905), working on Mnemiopsis, found that conduction was not interrupted when a plate was held still, nor when a plate was removed from the row. This latter experiment resembles experiments of Verworn (1889), who noted that ctenophore swimming plates on the two sides of a cut showed different rhythms and continued to beat actively. All these experiments indicate that metachronal co-ordination is by a neuroid transmission rather than by a mechanical process.

Gray (1930) thought that a neuroid transmission mechanism was unlikely for the following reasons: (1) metachronal waves can arise or be suppressed at any point along an epithelium, and cannot be traced to a centre from which the exciting stimulus can be shown to arise; (2) isolated ciliated cells are nearly always active; (3) the rate of metachronal transmission is very much slower than that of any known nervous impulse. He therefore suggested a mechanical triggering mechanism, in which one cilium excites the next when it reaches a certain stage in its own contraction cycle. This is supported by the present work, because the phase difference between adjacent cilia is constant throughout the peristomial row in Stentor, although the interciliary distance is variable. Previous work on Stentor (Sleigh, 1956) showed, however, that the phase difference was readily changed in a viscous medium or with magnesium ions without any change in the rate of wave transmission, while digitoxin reduced the phase difference and increased the rate of wave transmission.

It seems likely therefore that some type of excitatory transmission process is involved, in spite of the reasons which led Gray to discard such a theory. This type of process would be comparable with the conduction of the contraction wave in heart muscle. Both processes represent an excitation mechanism responsible for eliciting activity in contractile structures by means of transmitted impulses.

The evidence accumulated concerning ciliary co-ordination can be interpreted in two ways, according to the two theories of transmission mentioned in the introduction: (1) conduction is a continuous neuroid process, the velocity of which is slow in the gullet region and increases to a maximum in the distal part of the row; (2) conduction is a step-by-step process involving successive excitation from cilium to cilium along the row. If the second theory is accepted, then since the phase difference between successive cilia has been found to be independent of wide variations in interciliary spacing, it follows that any time involved in interciliary conduction must be short compared with that required for the excitation of each cilium. The results obtained by Kraft and Parker fit the first type of process, and are less easily explained in the second theory, while the work of Verworn and the present work on Stentor indicate that the second type of process is involved. Gray’s objections to a neuroid transmission process apply to the first theory, but are easily explained by the second, so that it seems as if the step-by-step transmission is the more likely hypothesis.

Both excitation and conduction are involved in such a step-by-step process, and the transmission will have properties belonging to both types of process. Changes in the conduction process would not be obvious since it occupies a much smaller time than the excitation process, from which it is not at present experimentally separable. The intraciliary process may be thought of as the building up of an ‘excitatory state’, under the action of an impulse from the previous cilium; discharge of this excitatory state would be responsible for triggering off (i) the contraction of the cilium, and (ii) the next conduction phase. Wave velocity will be most easily affected by agents which affect the reactions responsible for this build-up of excitation, e.g. digitoxin, which lowers the threshold for excitability in heart muscle, and may have a similar action here.

Since the contraction and transmission processes are to a large extent independent, the method by which the frequency of beat is determined still remains to be considered. Isolated cilia often show contractility; the beat of an abfrontal cilium in Mytilus is quite unrelated to that of any other cilia on the gill, and frequently exhibits long interkinetic periods. If we assume that the excitation of this cilium is intrinsic, we can visualize a fall in the threshold of excitability until the process becomes self-triggering. In Stentor, however, the constancy of the frequency along the row clearly indicates that the triggering is extrinsic—it may be assumed that this occurs at a somewhat higher threshold of excitability than that which produces spontaneous excitation, which will therefore not take place; the cilia will beat at a higher frequency than that of their spontaneous beat.

It is still true, however, that the frequency in Stentor can be significantly reduced by increasing the mechanical resistance (viscosity) of the medium; this observation appears incompatible with the hypothetical mechanism of excitation outlined above. It becomes necessary to assume that the excitation of a new contraction is delayed until the cilium has completed its contraction against whatever external resistance may be applied. However, the pacemaker of the whole system in Stentor appears to be situated in the gullet region and it is reasonable to suppose that it is one of the cilia in that region. If so, the effect of a viscous medium on the pacemaker and on the paced cilia will be similar, and the control of the distal cilia by the proximal ones may still be governed by an excitation mechanism. In this way it is possible to understand why a change in viscosity affects the frequency without altering the rate of conduction, whilst the addition of digitoxin, which changes the wave velocity by altering the excitability of the individual cilia, invariably affects the frequency as well, presumably by its action on the cycle of excitability of the pacemaker.

These ideas may be compared with the conduction system in the vertebrate heart, whose properties were discussed by Davies & Francis (1946). The different chambers of the frog heart all have their own intrinsic rhythms, decreasing in the order: sinus, atria, ventricle, bulbus cordis. Normally, the four parts all beat with an identical rhythm which originates in the sinus musculature and is conducted throughout the heart. The pacemaker system is very like that postulated for cilia, and it is interesting to note that cilia can beat at different frequencies under the control of different pacemakers, as do the chambers of the heart when isolated. Thus cuts across a row of cilia result in the setting up of new frequencies in the isolated regions; these new frequencies depend on a new rate of stimulation of the cilia initiated by a cilium at the beginning of the region. The new frequency was usually slower, but occasionally faster, than the original frequency, and it is possible that the fast controlling rate in the cilium next to the cut resulted from a local injury, since the observed discontinuities of rhythm in the absence of physical damage always resulted in a decrease in frequency at the discontinuity. Thus it seems likely that the intrinsic stimulation rate decreases from the gullet to the distal end of the row in the same way as the intrinsic rhythm of the various chambers of the heart decreases from sinus to bulbus cordis. This is borne out by the observation that the disturbance in the gullet region when the Stentor swallows food results in a temporary drop in the frequency, presumably because a cilium further down the row is acting as a temporary pacemaker.

The combined action of the pacemaker and transmission mechanisms is illustrated in Fig. 2. This shows a pacemaker cilium whose frequency is imposed by transmitted impulses on two of the cilia which follow it. The frequency of the row of cilia depends on the rate of contraction and the rate of excitation of the pacemaker cilium, the slower factor being the limiting one, and on the ability of the paced cilia-to beat as fast as the rate set by the pacemaker. The wave velocity depends on the rate of the interciliary conduction and on the time taken by the intraciliary excitation process in each participating cilium.

Bradfield (1955) explains the ciliary contraction cycle as the result of a sequence of propagated contractions in the peripheral fibrils, involving the passage of waves of excitation in both directions around the cilium simultaneously. In either direction there will be five excitation stages concerned in the initiation of the propagated contractions. The average phase difference between successive cilia in Stentor is about 6 msec. Much of this time is taken up by the excitation process within the cihum, and it is not unreasonable to assume that each of the five fibrillar excitation stages takes about 1 msec. In the light of the theory that nervous tissues arose from ciliated epithelia, it is interesting to speculate on the similarities in timing and character between the excitation process outlined above and that at a nerve synapse.

It is a pleasure to thank Prof. J. E. Harris for his critical interest and continued guidance in this work, which was carried out during the tenure of a grant from the Department of Scientific and Industrial Research.