Membrane Control of Ciliary Activity in the Protozoan Euplotes

The locomotory behaviour of ciliated protozoa results from the activity of the cell’s cilia. The major variables in the protozoan’s ciliary activity are the orientation of the effective stroke and the beating frequency. Ciliary orientation as used here is the direction of ciliary beat, or more specifically, the direction of the effective or power stroke. The effective stroke is defined as that component of ciliary movement which transfers the maximum force to the medium. The effective and recovery phases of the beat can generally be distinguished unequivocally by differences in shape as revealed by visual inspection of high-speed cine pictures. However, the orientation or position in space of the power stroke is more difficult to ascertain and describe. The difficulty fies in documenting with two-dimensional photography the three-dimensional form of the ciliary beat (see Kinosita & Murakami, 1967). The cilia are said to beat in reverse (ciliary reversal) when the power stroke is directed toward the anterior end of the organism so as to propel it backwards (Kinosita & Murakami, 1967). ‘Normal’ beating toward the posterior end causes the organism to swim forward. This will be termed forward beating here. The ability to change ciliary orientation is found in protozoa, the ciliated epithelia of some coelenterates and tunicate larvae.

There is evidence that ciliary orientation and frequency of beating are controlled independently. Cilia of Paramecium and Euplotes exposed to 1 mM-NiCl₂ stop beating but respond to stimulation by shifting position so as to point toward the anterior end (Naitoh, 1966; Naitoh & Eckert, 1969; Eckert & Naitoh, 1970).

Further evidence indicates that Ca is involved in the regulation of ciliary orientation. Ca produces reversal in beating cilia of extracted models of Paramecium in which the cell membrane has been rendered non-selectively permeable with Triton X-100 (Naitoh & Kaneko, 1972). In glycerinated specimens of Paramecium, Naitoh (1969) showed a quantitative relationship between Ca and ciliary orientation in the presence of optimal amounts of ATP, Mg and Zn. Kinosita (1954) found in living Opalina that ciliary reversal in response to stimulation with isotonic-KCl required extracellular Ca.

Ciliary orientation varies with membrane potential. Kinosita (1954) found in Opalina that the direction of ciliary beating shifts in a graded manner in response to spontaneous and KCl-evoked changes in membrane potential; others have found ciliary reversal correlated with depolarization, and ‘augmentation’ (extreme forward beating at high frequencies) with hyperpolarization in Paramecium (Yamaguchi, 1960; Eckert & Naitoh, 1970) and in the hypotrichs Euplotes (Naitoh & Eckert, 1969) and Stylonychia (Machemer, 1970).

It is well known that membrane potential is an important factor determining ionic movements across membranes. Eckert (1972) has proposed that ciliary reversal is coupled to membrane depolarization by an influx of Ca which occurs during a transient increase in Ca conductance of the membrane. The resulting rise in Ca concentration within the cilia is hypothesized as the immediate cause for activation of the mechanism producing the reversal (Eckert, 1972).

Much of the information about control of beating frequency comes from the work with reactivated extracted cellular models. ATP and Mg are necessary for reactivation of beating (Hoffmann-Berling, 1955; Bishop & Hoffmann-Berling, 1959; Bishop, 1962; Brokaw, 1961; Child, 1965; Seravin, 1961) and frequency has been found to vary proportionally with ATP concentration (Brokaw, 1967; Murakami & Eckert, 1972). In living cells changes in beating frequency are associated with changes in membrane potential as shown by Kinosita, Murakami & Yasuda (1965).

The first part of this present study investigates the relations of ciliary orientation and frequency to evoked changes in membrane potential. The results are discussed in the perspective of membrane-regulated Ca fluxes as the major regulatory mechanism for ciliary orientation. The second part examines the action of Ca and other ions on ciliary activity following the functional disruption of the cell membrane by extraction with the detergent Triton X. The hypotrich ciliate Euplotes was chosen for these studies because this cell is large enough for routine insertion of microelectrodes and its compound cilia (cirri) are large enough to photograph easily. Each of the cirri contains 40–120 ‘9 + 2’ cilia packed together in a hexagonal array (Roth, 1956, 1957; Gliddon, 1966).

A supply of Euplotes eurystomus was obtained from Connecticut Valley Biological Supply (Southampton, Mass.) and cultured in Chalkley’s (1930) solution (solution A) containing wheat grains. The Chalkley’s solution has the following composition: 1·7mM-NaCl, 0·05 mM-KCl, 0· 05 mM-CaCl₂, and enough Na HCO₃ to bring the solution to pH 7·0 (approximately 1 HIM). Cells were removed from the culture, washed and equilibrated for 30 min in 2 mM-KCl, 1 mM-CaCl₂ and 1 mM Tris-HCl buffer pH 7·0–7·4 (solution B) prior to the beginning of experimentation.

A small drop of solution containing 3–10 cells was placed on the coverslip which was then inserted into the chamber. If the drop was large enough, the cells often moved only slightly. A microneedle was used to hold the cells with their ventral side against the coverslip. The chamber was then filled with solution B.

Voltage-recording and current-passing microelectrodes were inserted into the left margin of the cell in the area between the frontal and anal cirri. The holding needle was removed and a blunt-tipped glass needle of 2–10 μm diameter, held in a horizontal plane, was placed between the cell and the coverslip and slowly lowered vertically. This action rotated the cell so that the anal cirri were seen in profile. Often the ventral and frontal cirri were also in view in side profile.

Observations were made with a Zeiss GFL microscope equipped with Nomarski interference optics (Text-fig. 1). Light was provided by a strobe lamp (Strobex 164, Chadwick-Helmuth Co., Inc., Monrovia, Calif.). The image of the cell passed through a x 16 objective (N.A. = 0-35) and was projected with final magnification of × 80–128 onto 16 mm film (4-X Negative, Kodak) in a high-speed camera (164-4 DC Locam, Red Lake Laboratories, Santa Clara, Calif.). The frame speed was usually 200 frames/ sec. The face of a Tektronix 360 cathode-ray-tube was also projected onto the cine film by means of a prism housed in a Zeiss Basic Body II. After the camera was started and had reached the selected frame speed, a pulse of stimulating current was passed into the cell. Both the membrane potential displayed without time base on a cathode-ray- tube and the image of the cell were recorded simultaneously by superimposition on the cine film. The films were analysed with a projector equipped with single-frame advance.

Membrane potential was recorded with 0·1 M-KCl-filled microelectrodes (resistance, 100–500 MΩ) which were connected to a capacity-neutralized amplifier (NF 1, Bioelectric Instruments) and displayed on both 561 and 360 Tektronix cathode-ray oscilloscopes. A square-pulse calibrator was used to check the amplifier response before each stimulus. The indifferent electrode was a 0·5 mm glass capillary filled with

2·8 M-KCI or 0·5 M-KC1 dissolved in 2–3% agar. The bath was held at ground by connecting the indifferent electrode to the summing junction of a Philbrick P25 operational amplifier used to monitor current intensity. All liquid-wire interfaces were connected through Ag/AgCl electrodes. Current pulses for stimulation were provided by pulse generators in series with a 10¹⁰ ohm resistor and were delivered through intracellular microelectrodes filled with either 0·1 or 0·5 M-KC1. Membrane potential and stimulus current intensity were photographed from the oscilloscope face in the conventional manner. A steady membrane potential and stable response to current pulses were required before cine photography was begun. Experiments were carried out at room temperatures of 17–20 °C.

For the reactivation of extracted models many variations of extraction and reactivation solutions were tried with the aim of maximizing the duration of beating. The composition of the solutions eventually selected for this purpose, similar to that used by Eckert & Murakami (1972) for Necturus oviduct, was as follows:

1. Wash solution: K₂SO₄ 56 mM, sucrose 46 mM, Tris-maleate 20 mM, KOH 20 mM, EGTA Tris neutralized) 1 mM, MgSO₄ 0·1 mM, Tris base approximately 5 mM to bring to pH = 7·0.

2. Extraction solution: 0·01%(w/v) Triton X-100 (Octyl Phenoxy Polyethoxyethanol; Sigma Chemical Company, St Louis, Mo.) plus wash solution. The Triton X-100 was stirred for a minimum of 30 min before use.

3. Reactivation solution: ATP Na₂ 1 mM (Calbiochem) + wash solution + base to make pH 7-0.

Different Ca concentrations were made with EGTA-buffered solutions (EGTA, La Mont Laboratories), which were calculated from a dissociation constant of 4·83×10⁶ (Hagiwara & Najakima, 1966). If the value of Ca in the double-distilled water was 10⁻⁵ M, then the free Ca in the presence of 1 mM EGTA would be 2 ×10⁻⁹ M. However, since the value of free Ca was not determined and might vary, the concentration of free Ca in solutions with 1 mM EGTA was taken to be 10⁻⁸ M or less.

Cells were taken from cultures and washed twice with Chalkley’s solution in depression slides. Several cells were placed in the wash solution for 1–2 min to remove any traces of Ca, then extracted for 3–5 min at room temperature (17–20 °C) and finally transferred back to the wash solution. For detailed observations and photography the cells were placed on a coverslip which was mounted in a chamber as described above. Holding electrodes were then inserted and the cell was positioned so that its anal cirri were seen in profile as described above. Reactivation solutions then flooded the chamber and beating was photographed at 50 frames/sec. Many variations of procedure were tried including extraction in the chamber or transfer to the coverslip in the extraction solution. However, the technique detailed above gave the most prolonged duration of reactivated beating (ca. 10–30 min). Duration of reactivation was not affected by varying the extraction time between 2 to 7 min, but decreased and approached zero after 15–20 min extraction. Extraction at reduced temperatures did not enhance the duration or the extent of reactivation.

The movements of Euplotes are brought about by aggregates of cilia termed membra- nelles and cirri. The location of these organelles is seen in the side view of a cell as illustrated in Text-fig. 2. Going from the anterior to the posterior end, there is first a collar of membranelles, followed by seven frontal cirri, two ventral cirri, five anal cirri, and finally four caudal cirri. The activity of these different organelles varied in both free-swimming and impaled cells. The membranelles moved continuously, while the frontal cirri usually beat at a low frequency. Ventral, anal and caudal cirri were often motionless and showed only sporadic activity (see Taylor, 1920, for further details).

The position of a cirrus was expressed in terms of the ‘inclination angle’. This angle was determined by drawing a tangent to the cirrus at a point 20 % of the total length of the cirrus from the point of attachment. The intersection of this tangent line with a line drawn along the surface of the cell formed the inclination angle. The smallest angle was formed when the cirrus pointed to the posterior end of the cell. This is a modification of the method of measuring used by Okajima & Kinosita (1966), who drew a tangent at a point 7 % of the total length of the cirrus. The 7 % point could not be used here because the movements of the cirri obscured the base of the cirri during much of the beat cycle. In order to minimize the uncertainty in drawing tangent lines, inclination angles were measured at the anterior-pointing and posterior-pointing positions in the beat cycle where the proximal portions of the cirri were relatively straight. These positions usually coincided with the beginning and end of the effective stroke.

The following factors should be considered in the measurements of the inclination angle: (a) movements of the cirri were not limited to a single plane, but are somewhat three-dimensional ; (b) the cell did not lie perfectly in the plane of focus; (c) the surface outline of the cell was often vague; and (d) slight movements of the cell resulted from changes in ciliary activity. It must be emphasized that the measurements were somewhat arbitrary, and the values were most useful in describing changes in ciliary activity in a given cell.

The response of the ciliary organelles depended on the direction of transmembrane current.

The changes in the beat pattern of the anal and frontal cirri are shown before and during stimulation by pulses of current in Text-fig. 3. Prior to the stimulus, the beating of these cirri and membranelles was in the * forward ‘(i.e. rearwardly-directed, forwardswimming) direction; the anal cirri beat sporadically at a low frequency while the frontal cirri and membranelles beat continuously. After application of the stimulus pulse, and membrane depolarization, the beating direction was reversed and the frequency increased.

The activity of a cirrus was expressed by plotting the extremes in inclination angle (α) of successive cycles as a function of time (Text-fig. 3). For the anal cirri the difference in extremes (approx. 180°) was similar for forward and reversed beating. This was also noted by Okajima & Kinosita (1966). The main criterion used in the present study to distinguish the beat direction (forward or reversed) of the anal cirri was the shape of the cirrus (Text-fig. 3). The frontal cirri, in contrast, showed differences in shape in the extreme position and in the range between extremes for forward and reversed beating. Differences in extreme position of the reversed frontal cirri were less than the values recorded by Okajima & Kinosita (1966), which was consistent with the different methods of measuring. Although details of the beat cycle of the membranelles were not photographically resolved, changes in direction of beat and sometimes in frequency could be determined. As shown in Text-fig. 3 the duration of reversed beating was the same in all groups of cirri and was followed by forward beating although, in this case, the membranelles’ reversal continued approximately too msec longer. Reversal continued after the end of the pulse, and even after membrane repolarization; the duration of the reversal after the end of the stimulus depended on the amplitude of the depolarization as shown below (Text-fig. 11).

Beating frequency of the anal cirri increased from 5·5 cycles per second (c/s) in a single spontaneous cycle, to an average of 28·5 c/s within the stimulus period, while the frontal cirri increased from 18 2 to 36· 8 c/s. The frequencies became steady after one cycle and declined following the end of the pulse. A smaller decrease in frequency occurred in the frontal cirri than in the anal cirri before forward beating resumed. In both frontal and anal cirri the frequency of forward beating immediately after the end of reversed beating was higher than the frequency prior to stimulation.

Inward current of sufficient intensity produced movement in the forward beating (normal) direction and an increase in the frequency of beating (Text-fig. 4). Upon stimulation the nearly quiescent anal cirri began beating at 33 c/s in the forward direction while the frontal cirri, which were spontaneously active prior to stimulation, increased their frequency from 15 to 25 c/s. The frequency increase began at approximately the same time in all the ciliary organelles and persisted here beyond the end of the filming period (450 msec beyond end of stimulus pulse). The decline in frequency following stimulation was more rapid in the anal cirri than in the frontal cirri.

A small depolarization resulted in a reversal of some ciliary organelles while others continued to beat in a forward direction. This was seen in Text-fig. 5 where the anal and ventral cirri beat in reverse in response to a depolarization of about 6 mV while the frontal cirri and membranelles continued to beat in a forward direction. However, the frontal group showed a reduction in frequency and in amplitude of beating. In this case three of the five anal cirri beat once in reverse while the other two beat twice.

Data on the response to weak depolarizations in four cells is summarized in Table 1. In three of the four cells ciliary organelles at the posterior end showed reversal at lower membrane potentials (i.e. lower threshold) than those at the anterior. The forward-beating frontal cirri showed a consistent reduction in frequency and, in some cases, a decrease in beating amplitude. The depolarization necessary to produce ciliary reversal of all the organelles varied between 3 and 11 mV (Table 2).

Because the frontal cirri and membranelles were beating continuously and the anal cirri spontaneously underwent single forward beats, it was difficult to establish differences of threshold between the various groups of cirri in response to hyperpolarization. Threshold for forward beating of the anal cirri varied between 15 and 29 mV in four different cells. Activation of the anal cirri was accompanied by an increase in frequency of the frontal cirri.

Because of their size (approximately 50μm long), almost planar beat and distinct differences in the shape of the cirrus when beating in forward and reversed directions, the anal cirri were used for a more detailed analysis of the ciliary response to stimulation.

The five anal cirri were usually quiescent except for occasional slow forward beats. Upon activation these cirri beat out of phase with one another, and individual cirri showed differences in threshold, number of beats and frequency. These differences were most apparent with small depolarizations and became less noticeable as the maximal response was approached. Responses of these cirri were examined at different stimulus intensities of inward and outward current.

Text-fig. 6 shows responses of the cirri to pulses of outward current of increasing intensity. Extremes of inclination angle (α) of a single cirrus are plotted along with membrane potentials, and tracings are shown of the first and second beating cycles of that cirrus.

By amplitude is meant the width of the envelope (formed by the entire length of the cirrus) of the ciliary cycle. With small depolarizations a small movement in the anterior direction or an incomplete cycle resulted. No recovery stroke occurred; the cirrus remained at the new position or moved back stiffly to its original position. Larger depolarizations caused a greater movement toward the anterior and a complete recovery stroke followed the initial effective stroke. Beyond this, the amplitude of the beat showed little increase with increasing depolarization.

When the extreme positions of the proximal segment of the cirrus were plotted (Text-fig. 6), little variation was observed. This was the case for complete reversal cycles during the course of a depolarization and for increasing depolarizations. Following the end of the stimulus pulse, there was a reduction in the posteriorly-directed extreme of the cirrus.

Cycles were counted starting from the position of the cirrus at the onset of the stimulus ; half cycles were counted as a whole cycle. The number of reversed beats increased with greater depolarization (Text-fig. 8). This resulted from a greater persistence of reversed beating beyond the stimulus (also see Textfigs. 6 and 11) and an increased frequency of beating.

Frequency was measured as the average for all the cycles, except the first, falling within the stimulus period. If only one cycle of beat occurred, the reciprocal of its duration was taken as its frequency. After the first cycle a steady frequency was maintained for the stimulus duration. After the stimulus ended, frequency declined.

Beating frequency increased with membrane depolarization until a maximum was reached after a steady-state change of about 10–20 mV (Text-figs. 9 and 13). The maximum frequencies varied between 17 and 31 c/s. By comparison the frequency of cilia in Paramecium increased from 13 to 37 during a spontaneous Ba-induced action potential (Kinosita et al. 1965).

As membrane depolarization increased, the time between onset of stimulus and the first consistent movement of the cirrus decreased (Text-fig. 10). The minimum latency (not shown in Text-fig. 10) obtained with high current intensities was 5–10 msec. This value was consistent with the latencies of 22–36 msec (Kinosita et al. 1965) and 40 msec in Ni-paralysed preparations (Eckert & Naitoh, 1970) elicited by submaximal stimulation of Paramecium.

The amplitude of the forward beat showed little variation with increasing hyperpolarization as indicated by the tracings of the first and second evoked cycles in Text-fig. 7. The amplitude was reduced slightly and the envelope directed more posteriorly in the second cycle. Okajima & Kinosita (1966) have reported a reduction in amplitude and a more posteriorly-inclined beating axis when the cirri beat at higher frequencies.

Little variation in the extremes was observed with increasing membrane hyperpolarization (Text-fig. 7). Spontaneous cycles showed smaller extremes.

Increasing hyperpolarization produced an increase in the number of cycles (Text-fig. 8) which varied widely between different cells.

Frequency approached a maximum value as the hyperpolarizing stimulus was increased (Text-fig. 9). The maximum values varied between 16 and 36 c/s for the different cells. The maximum frequency was reached with a 22–30 mV hyperpolarization and showed little gradation between quiescence and maximum frequency of beating.

As the amount of hyperpolarization increased, the latency decreased

(Text-fig. 10). High current intensities resulted in minimum latencies of 10–15 msec. This can be assumed to result from faster rates of hyperpolarization with higher stimulus currents.

With progressive increases in intracellular positive potential, ciliary reversal was maintained and extended beyond the stimulus (Text-fig. 11). However, with sufficient outward current no reversal occurred until the end of the stimulus when a brief ‘off response’ resulted (Text-fig. 11, sequence d). Thus, a positive shift in potential of about 110 mV brought the membrane to a ‘suppression potential’ (Katz & Miledi, 1967) at which the stimulus had no effect on ciliary activity. A slightly smaller depolarization of 97 mV resulted in one reversed beat at the onset of the stimulus, a 65 msec period of inactivity and finally three reversed beats of less than maximum frequency before the pulse ended (sequence e). The ‘off response’ consisted of a small number of reversed beats moving at a high frequency. To assure that the cell was undamaged by these strong stimuli, a smaller depolarization (sequence f), equal to that in sequence b, was given and was found to evoke reversed beating for the duration of the pulse. Text-fig. 12 illustrates how the magnitude of ciliary reversal, measured as the number of reversed beats within the stimulus period, varies with the steady-state value of the depolarization (here measured at end of the pulse). The number of beats in reverse increased to a maximum with a progressively more positive internal potential and then declined towards zero, which was reached at a minimum positive shift of 107 mV. No response was seen for positive shifts greater than 107 mV. The suppression potential for these cells lay between +70 to +82 mV (resting potentials varied between 20 and 26 mV). Since the membrane potential sagged during stimulation, the suppression potential could not be more precisely determined.

Frequency as a function of steady-state depolarization is shown in Text-fig. 13. The frequency increased to a maximtun with increasing internal positivity and then declined to zero. The zero frequency reflected the absence of any reversed beats in the stimulus period. The intermediate frequency values arose when the membrane potential dropped below the suppression potential and reversed beating was initiated. On termination of the pulse the frequency of the first beat was greatest and declined in subsequent beats (Text-fig. 11).

Using high intensities of inward current Naitoh (1958) evoked ciliary reversal in Opalina. In the present study of Euplotes attempts to produce ciliary reversal with inward current were unsuccessful despite hyperpolarizations of 100 mV, extended stimulus durations (several seconds without filming, 750 msec with filming) and elevation of the bath Ca to 10 mM. However, some aberrations in ciliary beating during strong hyperpolarizations were observed. These included a twisting and bending at two different points along the cirrus, usually in the first 1–2 cycles, and a sporadic stoppage of ciliary activity. In two cells reversal of the frontal cirri occurred, but this was not a consistent result. In general, the amplitude and extremes of inclination angle were similar to those in Text-fig. 7.

Prolonged and strong hyperpolarization (usually deflections of 100–200 mV for periods of greater than 300 msec) disrupted and eventually caused detachment of the anal cirri while the frontal cirri and membranelles continued to beat in a forward direction. With more stimulation the cell was killed.

The effect of large hyperpolarizations on the number of forward beats within the stimulus period is shown in Text-fig. 12. The number of beats reached a maximum and declined slightly as the magnitude of the hyperpolarization increased. Frequency also increased to a maximum and then decreased as the membrane was further hyperpolarized (Text-fig. 13).

Exposure of the cell to a Ca-free solution for about 5–10 min resulted in the loss of the depolarization-evoked reversal response. After raising the external Ca to its previous level reversed beating could not be produced and the cell soon died. However, immediately after the external Ca level was lowered to 10⁻⁶ M with EGTA, large depolarizations failed to produce reversed beating. Upon reintroduction of 10⁻³ M-Ca, depolarization again resulted in ciliary reversal (Text-fig. 14). These results indicated Ca was apparently required for reversal. Reversed beating could be evoked in 10⁻⁸M- Ca, and the amplitude of depolarization needed to evoke a response was essentially unchanged after the Ca was reduced from 10 ⁻³ M to 10⁻⁵ M.

Exposure of Euplotes to the low-Ca solution resulted in a decline in resting potential and input resistance. For the cell in Text-fig. 14 the resting potential fell from –28 mV before addition of EGTA to –10 mV in 10⁻³ M-Ca after removal of EGTA. Despite the reduction in resting potential, depolarizations (about 10 mV) still evoked the reversed beating response. On the other hand, large depolarizations (up to 30 mV) in 10⁻⁶ M-Ca did not produce reversal even before the usual decline in resting potential occurred. Thus, a decline in resting potential was not necessary in order to observe a loss of the depolarization-evoked reversal.

The most visible effect of the Triton X treatment was the separation of each cirrus and membranelle into many small groups of ‘clumps’ of cilia. These clumps of cilia arose from the base of the cirrus, producing a brush-like appearance of frayed cilia and groups of cilia. These often interfered optically with one another, complicating the measurement of their movements.

After 1–2 min of extraction ciliary activity was absent. At this time no resting potential could be recorded, and the input resistance had dropped to less than 1 % of its value in the living cell. It is presumed that the membrane permeability was increased greatly by the extraction with Triton X so that the concentrations of small diffusible molecules within the cell had come into equilibrium with the extracellular medium. Thus, the influence of Ca on the ciliary apparatus could be directly evaluated.

Since the anal cirri were studied in the living cell, the reactivation of this group was the chief concern here. Reactivation of ciliary activity required ATP and Mg as previous workers found (Hoffmann-Berling, 1955; Alexandrov & Arronet, 1956; Seravin, 1961 ; Child, 1965; Gibbons, 1965 ; Brokaw, 1967; Brokaw & Benedict, 1968; Eckert & Murakami, 1972). The duration of beating of the anal cirri was longest in 10⁻³M-ATP and 10.⁴ M-Mg. The membranelles beat longest in 10⁻³M-Mg. No reactivation was observed in 10⁻³ M-AMP or 10⁻³ M-ADP in the presence of 10⁻⁴ M-Mg and 10⁻⁶M-Ca. Ca and ATP alone did not reactivate beating. The differences in optimal Mg concentration for maximum beating duration of the anal cirri and membranelles have not been reported and may be peculiar to the present method of extraction. Seravin (1961) found the optimal Mg concentration to be between 10⁻⁴ and 10⁻⁵ M in Euplotes extracted with saponin. Frequency increased upon raising the Mg from 10⁻⁴ to 10⁻³ M in the presence of either 10⁻⁸ or 10⁻⁶ M-Ca. Reactivation with Mg and ATP are not enough evidence that the cell is freely permeable to ATP and not metabolizing. Eckert & Murakami (1972) found that small concentrations of ATP (10⁻⁸ M) stimulated beating in Necturus oviduct decalcified with EGTA.

Therefore, to insure that the ATP synthetic capacity of Euplotes was destroyed, extraction was carried out in the presence of 10⁻³ M-iodoacetate, a glycolytic inhibitor. After such extraction the cilia were activated in a solution of 10⁻⁴ M-Mg, 10⁻³M-ATP and 10⁻³ M-iodoacetate. Unlike 10⁻³ M-cyanide, which did not affect Euplotes, iodoacetate quickly stopped movement of the living cell. The ability of externally supplied ATP to reactivate ciliary movement was further evidence that it freely entered the cell through the extracted membrane and acted as the source of energy for ciliary movement.

There is a qualitative difference in the extreme positions of the beating cirrus in reactivation solutions containing 10⁻⁸ M-Ca compared to 10⁻⁶ M-Ca. These differences are illustrated in Plate I, which shows the beating pattern of a cell exposed first to 10⁻⁸ and then to 10⁻⁶ M-Ca. In 10⁻⁸ M-Ca the extreme anterior position of the clumps of cilia was approximately perpendicular to the cell, but in 10⁻⁶ M-Ca it was approximately parallel to the longitudinal axis of the cell. The beating pattern was more posteriorly-inclined in 10⁻⁸ M-Ca, while the pattern showed a more anterior inclination and possibly larger amplitude in 10⁻⁸ M-Ca. The pattern of movement in 10⁻⁸ M-Ca resembled forward beating, and the pattern in 10⁻⁶ M-Ca reversed beating of the cirri in the living cell. However, no distinctive effective and recovery phases of the beating cycle were observed in the extracted cirri.

Ca-induced reversed beating was associated with an increase in the duration of beating together with an increase in the number of beating cirri. Cells showed less beating in 10⁻⁸ M-Ca with increased periods of reversed beating in 10⁻⁶ M-Ca. After 2–5 min in 10⁻⁶ M-Ca only 50% of the cells beat when placed in a reactivation solution containing 10⁻⁸ M-Ca.

The Ca concentration was varied between 10⁻⁸ and 10⁻⁶ M. Table 3 shows the type of beating associated with the different Ca concentrations. Threshold differed between cells but most cells showed reversed beating between 10⁻⁷ to 10⁻⁶ M-Ca.

Gradations in degree of ciliary reversal as a function of free Ca were not observed. Gradations, if they occur, presumably extend over smaller increments of Ca than those covered in these experiments.

The results of this work support various lines of evidence that Ca plays an essential role in the reversal of ciliary beating. Evidence for this was obtained in living cells by Bancroft (1906) and Kinosita (1954). In extracted models of Paramecium Ca is required for a change in the orientation of non-beating cilia (Naitoh, 1969) and for reversed ciliary beating (Naitoh & Kaneko, 1972). In the latter study, the orientation, as judged by swimming direction and velocity, gradually reversed as the Ca concentration was raised from 10⁻⁷ M to 10⁻⁶ M and above. The maximum velocity of swimming in reverse occurred at 5 × 10⁻⁵ M-Ca. The essential role of Ca in the reversal of ciliary beating was confirmed for Euplotes in the present work. The direction of beating of the extracted anal cirri reversed when the Ca concentration was raised above approximately 10⁻⁷ M. This is significant because it indicates an effective concentration sufficiently low so that transient Ca influx across the cell membrane may be adequate to produce reversal.

In unstimulated squid axons (Baker, Hodgkin & Ridgway, 1971) intracellular free Ca concentrations of 3 × 10⁻⁷ M have been reported. Muscle contraction in barnacle fibres is initiated at free Ca concentrations of 4– 8 × 10⁻⁷ M (Hagiwara & Naka, 1964; Ashley, 1967). Syneresis of isolated myofibrils showed a threshold of 10⁻⁷ M-Ca (Weber & Herz, 1963). The present data on extracted models of Euplotes show a similar degree of sensitivity of ciliary reversal (occurring between 10⁻⁷ and 10⁻⁶ M-Ca) to Ca, and is in the range reported for Paramecium (Naitoh & Kaneko, 1972).

No gradation in the direction of the effective stroke as reported for models of Paramecium (Naitoh & Kaneko, 1972) and in living Opalina (Kinosita, 1954) was observed in Euplotes. The effective stroke of the anal cirri is confined largely to a single plane for both forward and reversed beating, while the cilia of Paramecium and Opalina beat in different planes perpendicular to the cell surface depending on the extent of reorientation of the beat. The change in the beating direction of the anal cirri is expressed as a change of shape of the quasiplanar movements of the cirrus.

Ca is required for reversed beating in extracted models, and this raises the question of how the intracellular Ca concentration is controlled in the living cell. Ca conductance changes, resulting from changes in membrane potential, could lead to an increased Ca influx and subsequent increase in intracellular Ca (Eckert, 1972). Naitoh, Eckert & Friedman (1972) have demonstrated in Paramecium graded, regenerative depolarizations, termed ‘calcium responses’ because their amplitudes increase with greater external Ca concentrations.

Increases in free intracellular Ca concentration accompanying depolarization have been demonstrated in barnacle muscle, where the peak tension developed during contraction is a function of the Ca concentration (Ashley & Ridgway, 1970).

Depolarizations of at least 3–11 mV (Table 2) were required to produce ciliary reversal. The component of depolarization due to Ca influx (as opposed to purely electrotonic depolarization) through postulated voltage-sensitive Ca channels is uncertain. However, calculations demonstrate that currents producing millivolt changes in membrane potential can make increments in intraciliary concentrations of Ca sufficient to evoke reversal of the ciliary beat (Eckert, 1972).

The changes in the response (reversed beating) with progressive depolarization were consistent with a progressive increase in intracellular Ca concentration. Reversed beating showed a threshold of membrane potential change and increased in frequency and duration with greater depolarizations (Text-fig. 6). Likewise, the intracellular Ca concentration probably increased after a given degree of depolarization and continued to increase to a maximum with greater depolarization, as has been found in barnacle muscle (Ashley & Ridgway, 1970). With greater depolarizations the reversed beating continued for some time beyond membrane repolarization (Text-fig. 11). This might result from a prolongation of the period of increased Ca concentration assuming a time-dependent process (e.g. Ca pump) for the reduction of the concentration below the threshold for reversed beating.

According to the ionic hypothesis (Hodgkin, 1964; Katz, 1969) the net flux of an ion will be reduced (at a given conductance) if the driving force on the ion is reduced (i.e. as the membrane potential approaches the equilibrium potential of the ion). The data from the extracted model suggests that the intracellular concentration of Ca in the unstimulated living cell is below 10⁻⁶ M. Assuming an intracellular Ca concentration of 10⁻⁷ M, and knowing the extracellular concentration to be 10⁻³ M, the calculated value of the Ca equilibrium potential (E_CA) is in the vicinity of + 116 mV. When the membrane potential was shifted by stimulus current to positive values less than + 70 mV, reversed beating occurred. However, further increase in positive potential produced a suppression of ciliary reversal (Text-figs. 11 and 12). This is to be expected if it is assumed that reversal requires a certain minimum increment in net Ca influx. As E_ca was approached, the influx of Ca presumably dropped below the rate required to produce the threshold concentration in the cilium and the stimulus, although suprathreshold, failed to produce ciliary reversal. The minimum suppression potential lay between + 70 and + 82 mV. A similar suppression of ciliary reversal was observed with long depolarizing currents in Opalina (Naitoh, 1958). Katz & Miledi (1967) found no release of pre-synaptic transmitter in the squid stellate ganglion and Baker et al. (1971) found Ca entry itself to be suppressed in squid giant axon when the membrane potential was raised to levels approaching the calculated Ca equilibrium potentials.

Finally, the experiments in reduced external Ca demonstrate that no reversal occurs in response to depolarization in a solution where the external concentration of Ca is below a minimum (Text-fig. 14). This can be interpreted as the result of a reduction in the inward Ca current because of a reduced electrochemical gradient. Consequently, the change in the internal Ca concentration is insufficient for a reversal of beating direction.

Small depolarizations (3–6 mV) of Euplotes produced reversed beating of the anal cirri while the frontal cirri continued to beat in a forward direction (Text-fig. 5, Table 1). Such differences in responsiveness of cilia have been observed in other preparations. Naitoh (1958) found regional differences in the current threshold for reversal beating in Opalina. The differences in threshold coincided topographically with the differences in membrane resistance over the cell surface. Thus, the anterior end showed both the lowest threshold and the lowest membrane resistance. Similarly, in Paramecium Ni-paralysed cilia at the anterior end displayed a faster and greater reorientation in response to depolarization (Eckert & Naitoh, 1970). The differences in beating direction in Euplotes are not the result of variations in the amplitude of depolarization between the anterior and posterior ends since the cell is isopotential (Naitoh & Eckert, 1969). The observed differences may reflect dissimilarities in the reorientation process, in the local intracellular concentrations of Ca required for reversal, or in the local increment in intracellular concentration of Ca produced by excitation. The last possibility seems most likely in view of local differences in resistance found in OpaUna (Naitoh, 1958).

Hyperpolarization activates the anal cirri to beat in a forward direction (Textfigs. 4 and 7). The stimulus-response characteristics of forward beating differ in some ways from those of depolarization-induced reversed beating. First, activation of forward beating required a greater shift (3–6 times) in membrane potential than did reversed beating (Text-fig. 8). Secondly, the latencies were generally greater than those found after depolarization (Text-fig. 10). Thirdly, the relation between amplitude of hyperpolarization and number of evoked beats appeared to be more discontinuous than the relation associated with depolarization (Text-fig. 8) although lack of data makes this only a preliminary finding. Finally, the forward beating evoked by hyperpolarization continues, in some cells, for a much longer period than reversed beating (Text-fig. 8).

Large hyperpolarizing stimuli were applied to Euplotes to determine whether an increased electrochemical gradient acting on Ca would increase Ca influx to the extent that reversed beating would result (Text-fig. 12). No reversed beating was observed, even with current intensities which caused the cell to deteriorate. This contrasts with results in Opalina (Naitoh, 1958) in which large inward currents produced reversed beating. The negative results with Euplotes may be due to a low level of Ca conductance with hyperpolarization or simply greater susceptibility to damage by hyperpolarization.

Large increases in frequency were produced in response to both depolarizing and hyperpolarizing stimuli. The maximum frequencies for reversed and forward beating were similar in any one specimen (Text-figs. 9 and 13).

Dr A. Murakami made indispensable suggestions and Drs H. Machemer.Y. Naitoh, and K. Friedman provided advice. Dr J. Frankel and J. Ruffolo, Department Zoology, University of Iowa identified the species of Euplotes. The authors are grateful to Dr T. L. Jahn for the use of his projector and to J. Fonseca for his assistance. Dr D. Junge and Dr J. Morin made suggestions on the early form and Dr J. Sheridan assisted in the final preparation of the manuscript.

This work was supported by a U.S.P.H.S. Training Grant 5TO1 GM00448 to the Physiology Department, U.C.L.A. and U.S.P.H.S. grant NS08364 and NSF grant GB-30499 to R. Eckert. Based on a dissertation submitted by Miles Epstein in partial fulfilment of the requirements for the Ph.D. to the Zoology Department, University of California, Los Angeles.

Beating activity of the anal cirri of an extracted-cell model is shown in solutions of 10⁻³ and 10⁻⁶ M-Ca Photographs at intervals of 10 frames (i.e. 200 msec) show the movements of some fragmented and cirri first in 10⁻⁵ M (top of the figure) and then in 10⁻⁶ M-Ca (bottom of the figure). The frame speed was 50 frames/sec. Below each series of photographs are tracings of the positions of one cirrus fragment, indicated by the white triangles in the photographs. Tracings were made at intervals of 5 frames (i.e. 100 msec). The cirrus fragments traced in 10⁻⁸ and 10⁻⁶ M-Ca are probably not the same. A and P in the top tracing indicate the anterior and posterior ends of the cell.