zyx Cell Motility 4949-267 (1984) zy zyxwv zy Trifluoperazine-Induced Changes in Swimming Behavior of Paramecium: Evidence for Two Sites of Drug Action Tim Otter, Birgit H. Satir, and Peter Satir Department of Anatomy, Albert Einstein College of Medicine, Bronx, New York Trifluoperazine (TFP), a drug that binds to Ca2’-calmodulin (CaM) complexes, altered swimming behavior not only in living paramecia, but also in reactivated, Triton-extracted “models” of the ciliate. By comparing the responses of living cells and models, we have ascertained that two sites of drug action exist in paramecium cilia. Swimming movements were recorded in darkfield stroboscopic flash photomicrographs; this permitted accurate quantitation of velocities and body-shape parameters. When living paramecia were incubated in a standard buffer containing 10 pM TFP, their speed of forward swimming fell over several minutes and their bodies shortened. Untreated paramecia backed up repeatedly and frequently upon transfer to a solution containing barium ions (the “barium dance”), but cells preincubated in TFP did not “dance.” Instead they swam forward slowly for long periods of time without reversing and occasionally then exhibited abnormally prolonged reversals. W7 effects on swimming mimicked low doses of TFP, and the analog W5 did not visibly alter normal swimming patterns. These results suggest that TFP induces a decrease in the intracellular pCa of living paramecia, perhaps by reducing the efficiency of a calmodulinactivated calcium pump in the cell membrane. Paramecia extracted with Triton X100 and reactivated to swim forward (7 2 pCa 6) were not affected by addition of up to 40 pM TFP to the reactivation medium. We conclude that the main drug effect in living cells is probably not at the axoneme. However, at low pCa, TFP zyxwvu zyxwv Received February 10, 1984; accepted May 8, 1984. zyxw zyxwvuts Address reprint requests to Tim Otter, Cell Biology Group, Worcester Foundation for Experimental Biology, 222 Maple Avenue, Shrewsbury, MA 01545. Abbreviations used: EDTA, ethylenediamine tetraacetic acid; EGTA, ethylenebisoxyethylenenitrilo tetraacetic acid; CaM, calmodulin; TFP, trifluoperazine; ATP, adenosine triphosphate; Tris, tris(hydroxymethy1)aminomethane. 0 1984 Alan R. Liss. Inc. 250 Otter, Satir, and Satir zyxwvuts zyx zyx zyxw directly affected the ciliary axoneme to shift its behavior to one characteristic of a higher pCa: TFP inhibited backward swimming in models reactivated at pCa < 6; instead they swam forward or rocked in place. The mechanism of ciliary reversal in paramecium may therefore depend on an axonemal Ca*+-sensor, possibly bound CaM, which is affected by TFP only at low pCa, as has been postulated for other types of cilia. Key words: Paramecium, trifluoperazine, cilia, calmodulin, calcium INTRODUCTION Calcium ions control reversal of ciliary beat and swimming velocity in Parmecium [see review by Eckert et al, 19761. During reversal, depolarization of the surface membrane leads to an influx of calcium ions through voltage-sensitive channels located in the ciliary membrane [Dunlap, 1977; Machemer and Ogura, 1979; Ogura and Takahashi, 1976; Thiele and Schultz, 19811. This calcium ion influx decreases the intracellular pCa (= -log [Ca2+]),which triggers reversed beating and backward swimming. At a later time, when membrane calcium pumps restore the calcium concentration to its steady-state submicromolar level, ciliary beat renormalizes, and the paramecium resumes its normal forward swimming. The depolarization of the membrane is graded with stimulus intensity. Levels of intracellular calcium that are not high enough to induce backward swimming cause slowed forward swimming [Machemer, 1974; Naitoh and Kaneko, 1972, 19731. The mechanism by which calcium ions reverse the direction of the ciliary beat is not well understood. Several lines of evidence suggest that cilia contain a Ca2+sensitive switch that controls the pattern of active sliding of microtubules within the axoneme [Reed et al, 1982; Satir, 1982; Sugino and Naitoh, 1982; Wais-Steider and Satir, 19791. It is hypothesized that when Ca2+ is bound to the switch, the timing of sliding is altered so that corresponding changes in beat form and behavior, such as ciliary reversal, ensue [Satir, 1982; Sugino and Naitoh, 19821. Calmodulin (CaM), a Ca2+-dependent regulatory protein, is found in cell bodies and cilia of paramecium [Maihle et al, 19811, and CaM has been purified from the cilia [Walter and Schultz, 19811. Reed and Satir [1980] have shown that trifluoperazine (TFP), a phenothiazine that inhibits CaM-dependent processes [Levin and Weiss, 19761, blocks calciuminduced arrest in lamellibranch gill cilia. Effects of phenothiazines on reactivated, detergent-extracted models of lamellibranch gill cilia [Reed et al, 19821 and of Chlamydomonas flagella [Witman and Minervini, 19821 suggest that a regulatory pathway involving Ca2+-ions and CaM or CaM-like molecules is common to many types of cilia with diverse behavioral responses. Phenothiazines alter swimming behavior in Tetrahymena [Guttman and Friedman, 1962; Suzuki et al, 19811, and phenothiazine-induced alterations in swimming behavior of living paramecia have been reported in preliminary form by several laboratories [Dryl and Masnyk, 1971; Rauh et al, 1980; Satir et al, 19801. Thus, CaM is a likely candidate for the molecule in cilia that senses the intracellular pCa, activates the calcium-sensitive switch of the axoneme, and mediates, for example, ciliary reversal in Paramecium. We sought to examine in greater detail the effects of drugs directed against CaM on swimming behavior and, in particular, on ciliary reversal in Paramecium. Results of some of these experiments have been summarized previously in abstract form zyx zy Effects of TFP on Paramecium 251 [Otter et al, 19831. We have investigated how trifluoperazine (TFP) or W7 [Hidaka et al, 19811 alters swimming behavior in living paramecia and in reactivated, Tritonextracted “models” of the ciliate. By comparing the response of living cells and models, we have ascertained that there are at least two sites of action of TFP in the cilium. W5 was used as a control for CaM specificity. We infer from the druginduced behaviors that, in living cells, these drugs mainly interfere with removal of calcium ions from the ciliary matrix, presumably by reducing the efficiency of a CaM-mediated membrane calcium pump; in the models, at low pCa the main drug effect is directly on the axoneme. These data suggest that Ca2+-CaM interactions could regulate swimming speed and the direction of ciliary beat in Paramecium. zy MATERIALS AND METHODS Organisms Paramecia (P. caudatum and P. tetraurelia) were grown at 27°C in 0.25% Cerophyl containing 2 mM Na2HP04 and inoculated with Enterobacter aerogenes as a food source [Sonneborn, 19701. Cells were harvested at late log or stationary phase by pelleting them gently in a hand centrifuge (125g for 60 s at room temperature). Paramecia were then washed three times by hand centrifugation in TECK buffer (in mM: Tris, pH 7.2, 10.0; EDTA, 0.1; CaC12, 1.0; KCI, 4.0). Washed cells were equilibrated in TECK buffer at room temperature for at least 30 min before beginning experiments. P. caudatum was obtained from Carolina Biological Supply (Burlington, NC) and P. tetraurelia (both wild-type and “pawn B” mutants) were provided by Dr. C. Kung (Madison, WI). Chemicals Trifluoperazine was provided by Smith, Kline, and French Laboratories (Philadelphia, PA), and W7 and W5 were obtained from Rikaken Co. (Japan). Triton X100 was from CalBiochem (La Jolla, CA) lot No. 900173, and KC1 was purchased from Fisher Scientific Co. (Fair Lawn, NJ). All other chemicals were purchased from Sigma Chemical Co. (St. Louis, MO). zy Observing and Recording Swimming Behavior We routinely observed living paramecia and reactivated, Triton-extracted models (see below) by darkfield or phase contrast microscopy. From these visual observations, we determined qualitatively how swimming behavior was or was not altered by the appropriate drugs. All quantitative measurements of behavior and examples of individual responses to the drugs were taken from photographic records, as described below. Studies on living cells and models were carried out in slide-and-coverglass preparations where the coverglass was supported on two sides by ridges of silicone grease (Dow-Corning). All coverglass pieces were cleaned exhaustively by sonication and detergent washings, rinsed thoroughly in distilled water, and stored in 100% ethanol until use. We found that such cleaning reduced the number of Triton-models that stuck to the glass during reactivation, but a significant number still became attached. The problem of models sticking to glass appeared most severe at lower pCa’s of reactivation, and the percentage of swimming cells varied accordingly. Our measurements were made only on free-swimming cells. 252 Otter, Satir, and Satir zyxwvut zyxw For recording the swimming behavior of small ciliates such as Paramecium, we have modified and improved earlier procedures of Dry1 [ 19581, Naitoh and Kaneko [ 19731, and Otter and Salmon [ 19791. As shown in Figure 1, swimming movements of P. caudatum were recorded in stroboscopic photomicrographs taken in darkfield illumination. In Figure lA, an x marks the first flash, and a small arrow indicates the direction of swimming. Swimming speed was calculated from the distance swum over several flash intervals (brackets). Multiple avoiding reactions (brief episodes of reversed swimming) give rise to star-shaped paths since successive images of the paramecium overlap (large arrows). This behavior is typical of paramecia swimming in TECK buffer containing barium ions. Depending on the timing of the flash with the rotation of a paramecium’s body, lateral (Fig. IB, “L”) or dorsoventral (Fig. lB, “D”) views of an organism may be seen. Anterior (a) and posterior (p) aspects can easily be distinguished. Prolonged backward swimming (Fig. IC) is a gyrating movement that resembles the rotation of a bowling pin. In this sequence the first image was brightest, and the photograph has captured the full lateral excursion of backward swimming. In other images of prolonged backing, the strobe flashes may not coincide fortuitously with the rate of rotation of the paramecium. The pitch of the backward spiraling movement may, therefore, appear less severe because the strobe flashed when the paramecium was near to the midline of its path (see Fig. 6). For living paramecia the interval between flashes, 7, was 0.3 s. Similar photographic records were obtained for reactivated, Triton-extracted “models” of paramecium, but for models 7 was 0.9 s. A Zeiss ICM 405 inverted microscope (Carl Zeiss, W. Germany) outfitted with an X2.5 planachromatic objective, ~ 6 . 3 x , 16, and x 4 0 phase contrast planachromatic objective lenses, and a No. 2 phase annulus was used for all experiments. Illumination was by a conventional tungsten filament lamp for routine viewing or by a Strobex Model 136 power supply (Chadwick-Helmuth, Monrovia, CA), No. 35s xenon arc flash tube in a Zeiss 100-W lamp housing with four-lens collector, and home-made triggering circuitry for driving the power supply for stroboscopic photography. Just prior to taking a photograph, we observed specimens briefly with the “modeling” mode (60-Hz strobing frequency) of flash tube operation. Triton-extracted Models zyxw zy zyx zyxwvu P. caudatum washed in TECK buffer were extracted with 0.01 % Triton X-100 at 4°C according to Naitoh and Kaneko [19721. Specimens were reactivated at various pCa’s with 4 mM ATP according to Naitoh and Kaneko [ 19731, except that the pH of the reactivation buffer was previously adjusted to 6.4 with dilute HCI. At pH 6.4 and pCa 5, the models can swim backward for several minutes (Fig. 8). However, paramecia reactivated at pH 7.0 and pCa 5 swam backward for only a few seconds, if at all. The apparent dissociation constant of EGTA for Ca2+ (Kd) is highly sensitive to solution pH. At pH 7.0 Kd is about 0.21 pM while at pH 6.4, it is about 3.23 pM [Amos et al, 19761. Levels of Ca2+ below 1 pM are too low to promote backward swimming. At pCa 6 to pCa 7 models appeared to swim equally well at pH 7.0 or pH 6.4. We varied the pCa of reactivation by changing the ratio of Ca2+ to EGTA (5.0 mM) according to a graphical method described by Salmon and Segall [1980]. After adding the required amount of CaC12, we readjusted the pH to 6.4 before using the reactivation solution. All pH adjustments were carried out at room temperature. Aliquots of TFP (kept as a 1 mM stock in distilled water) were added to reactivation zyx zy Effects of TFT on Paramecium 253 zyxwvutsr zyxwvuts zyxwvu Fig. 1. Stroboscopic images of swimming behavior in paramecium. Forward swimming (A, brackets; B), multiple avoiding reactions (A, large arrows), and prolonged backward swimming (C) are easily distinguished by this technique. Small arrows indicate the direction of progression of the cells. For a complete description of Figure 1, see Materials and Methods ( ~ 6 3 ) . buffer and thoroughly mixed before we transferred the models to the solution. Drug addition did not alter the pH significantly ( < +0.02 pH units in 40 pM TFP). The pH of reactivation solutions measured after experimentation had increased by between 0.03 and 0.06 pH units. Such a small increase could be due to addition of stock ATP (40 pl/l ml buffer), which was stored at pH 7.0. RESULTS In Vivo Effects on Swimming Speed Paramecia (P. caudatum) swimming in TECK buffer traveled at about 1.5 mm/s in relatively straight paths, or roughly two body lengths per flash interval (Figs. 1, 2A). Incubating paramecia in TFP (5-15 pM in TECK buffer) progressively slowed their speed of forward swimming over several minutes (Figs. 2B-D, 3). Figures 2 and 3 illustrate the time course of slowing in 10 pM TFP. A group of paramecia became completely immobilized after exposure to 10 p M TFP for about 20 min (Fig. 3, inset B). The cilia of immobilized cells continued to beat rapidly, but they were not effective for swimming. These short-term effects required continuous exposure of the 254 Otter, Satir, and Satir zyxwvut zyxwv Fig. 2. Time course of slowing TFP. With increasing time (lower right corner, in minutes) in TECK buffer containing 10 WMTFP, paramecia slowed down (B-D), spun (C) and gyrated (D). Body length also decreased with time, see Figure 4; compare A with D (X30). In A and D, bar = 250 pm. cells to TFP: Paramecia washed into TFP-free medium after 7 min of exposure to 10 pM TFP swam actively 15 min later, but they swam slower than untreated control cells (Fig. 3, inset A). In addition to slowing, paramecia treated with TFP would turn, spin (Fig. 2C), and undergo partial ciliary reversal (circular paths). W7 at 40 pM elicited changes in swimming speed and behavior similar to those observed in 10 pM TFP, but 50 pM W5 did not visibly alter the normal patterns of movement for 15 min. Both wild-type and pawn B mutants of P. tetraurelia also slowed down when exposed to TFP, but we did not quantitate this response. The response of paramecia to TFP in TECK buffer varied with different preparations of organisms, with the age of the paramecia, and with cell density. Higher doses of TFP (16-40 p M ) immobilized paramecia quickly and, in addition, often elicited more complex changes in swimming behavior such as long periods of backward swimming. Cells exposed to high TFP levels became visibly damaged, lost their anterior-posterior asymmetry, and lysed within about 20 min. For a given set of experiments, we determined qualitatively the dose of TFP required to slow paramecia zyxwv zyxwvu zyxwvu zyxwvutsr zyxwvutsrqponml Effects of TFP on Paramecium A 0 a, cn 2.0 255 10pM TFP \ E E zyxwvut zyxwv Y m a, a, P 1.0 CI, S .- E .-E 4 8 12 Time (min) zyxwvu zyx zy 16 20 Fig. 3. Decrease in forward speed with increasing time in 10 pM TFP. Swimming speed measurements ( 0 )were taken from photographs such as those in Figure 2 . Points are plotted as the mean f SE for 510 organisms at each time point. In a separate experiment (O), paramecia were incubated in 10 pM TFP-TECK for 7 rnin (0),then transferred to TFP-free TECK containing 10 mM CaCI2. Fifteen minutes after transfer, these paramecia were still swimming (inset A), but control cells incubated in 10 pM TFP for 21 minutes ( A ) were completely immobilized (inset B). Cells incubated in TECK buffer for 1 h showed no change in speed (initial velocity, 1.27 f 0.05 mm-s-’; final velocity, 1.3 f 0.06 mm.s-’). Insets ( X 16). by half within 7 min (range: 7 pM to 30 pM) and did not use higher levels to study changes in swimming behavior in the living cells. Changes in Cell Shape Concurrent with a decrease in forward speed was a longitudinal contraction of the paramecium’s body (Fig. 4). As measured from darkfield photomicrographs, a normal P. caudatum was about 240 pm long and 56 pm wide at its equator. In 30 pM TFP, for example, in 10 min the length shortened to ca 160 pm, while the width increased only slightly. The ratio of length to width decreased from about 4:l to nearly 2 : 1. TFP-treated cells appeared squat and pear-shaped compared with the elongate, cigar-shaped profile of untreated cells (Fig. 2D, Fig. 4, insets). Paramecia that appeared pear-shaped all gyrated and produced abnormal swimming paths (Figs. 2D, 4C). Behavior in Ba*+-Medium Upon transfer to TECK buffer containing 10 mM Ba*+-ions, paramecia backed up quickly and often in a behavior known as the “barium dance” (Fig. l A , large arrows; Fig. 5A). However, when paramecia were preincubated in TFP (Fig. 5B) and then placed in Ba*+-containing media, they did not dance (Fig. 5C). The cells Otter, Satir, and Satir 256 zyxwvu zyxwvut zyxwvut 30pM TFP 4 T \ $ = 3 -2 -UI 2 c T I Q, -I + 1 1 I 4 zyxwvu zyxwv zyxw zyxwv I I I 1 I 8 12 16 20 24 Time (min) zyxwvutsr Fig. 4. Changes in body shape. The radio of body length to width (+) decreased over 12 min in 30 pM TFP-TECK from 4: 1 to 2: I . Most of this shape change was the product of a rapid initial decrease in Body shapes of typical paramecia are length ( 0 ) .accompanied by a slight (20%) increase in width (0). shown before TFP treatment (A) and after 6 rnin (B) or 24 min (C) in TFP-TECK. Values are plotted as mean f SE for nine measurements. Cells incubated in TECK buffer for 1 h did not change shape: (L, = 261 k 4 pm, W, = 63 f 3 Fm; Lf = 256 f 6 pm, Wt = 68 & 3 pm). Insets ( ~ 4 0 ) . partially recovered from TFP and sped up somewhat. They swam forward for long stretches of time, at speeds less than untreated controls. Any reversal was rare; however, when the TFP-treated paramecia in Ba2+ did back up, they swam backward for tens of seconds in a spiraling motion (Fig. 6B-D). When paramecia were preincubated in W7 instead of TFP, long periods of forward swimming without dancing occurred in Ba2+-medium (Fig. 5D). Treating paramecia with W5 did not visibly alter the normal dancing behavior that occurred in Ba*+-medium (Fig. 5A, inset). Drug Effects on Reactivated/PermeabilizedCells In order to determine if TFP can alter axonemal function directly, we prepared and reactivated Triton-extracted “models”-ie, permeabilized cells-of P. caudatum zyxw Effects of TFP on Paramecium 257 zyxwvutsrqp Fig. 5. Response of TFP-treated paramecia to barium ions. After a 25-min preincubation in 10 pM TFP-TECK buffer, paramecia swam slowly forward (B) and they sped up slightly upon transfer to TECK buffer containing 10 mM BaCI2 (C). They did not dance. Similarly, cells preincubated in 35 pM W7-TECK simply swam forward in Ba2+-TECK (D). For comparison, the normal dancing behavior of untreated cells is shown (A). Inset A: cells pretreated with 50 pM W5 danced normally in barium medium ( ~ 2 5 ) . zy zyxwv according to Naitoh and Kaneko [1973]. Electron microscopy indicated that cell and ciliary membranes had been partially disrupted by the detergent treatment. During extraction, paramecia contracted to about 200 pm in length, a decrease of 16 % . The models in appropriate reactivation medium were immobile in the absence of externally supplied ATP (Fig. 7). Occasionally, some individuals exhibited weak and irregular ciliary beat in this medium, presumably because there was a small amount of endogenous ATP in their cytoplasm. When models were placed in ATP-containing reactivation buffer at pCa 7, most of them swam, but some did not move. These appeared to be firmly attached to the substratum. They were normally reactivatable since their cilia were beating rapidly. We have repeated the work of Naitoh and Kaneko [1972] on model reactivation as a function of pCa, except that we have used 5 mM EGTA/Ca2+ buffers in all of our reactivation media. As shown in Figure 8, the swimming behavior of ATPreactivated permeabilized cells depended on the pCa of the reactivation buffer. At pCa 7 the models swam forward (A) at ca. 0.3 mm/sec, about one fifth as fast as intact cells. Reducing the pCa of the reactivation buffer to 6.3 slowed the forward speed (B). At pCa 6 average forward progression ceased. The models rocked in place or moved very slowly forward or backward (C). The cilia were beating, but with an altered stroke. Below pCa 6 the models swam backward at speeds up to 0.2 mm/sec (D). 258 Otter, Satir, and Satir zyxwvutsrqpo zyx Fig. 6. Prolonged reversals in barium medium. Paramecia pretreated for 6 min with 12 pM TFP (A) occasionally backed up for tens of seconds in a spiraling, bowling-pin motion when they were placed in TECK buffer containing 10 mM BaC12 (B-D; compare Fig. IC). Arrows show the direction of swimming (X30). Addition of up to 40 pM TFP to the reactivation buffer did not alter the behavior of the models at pCa greater than 6 (Fig. 9). The models simply swam according to the pCa of the reactivation buffer, and tracks identical to controls were obtained. However, below pCa 6, 40 pM TFP affected the behavior of the models. By comparing aliquots of models reactivated at pCa < 6 with models in buffer at the same pCa and containing TFP, we found that drug addition reduced both the number of backward-swimming models (Table I) and the speed of backing. For example, Figure 10 shows that models reactivated at pCa 5.7 swam slowly backward (A), but 40 pM TFP blocked this reversal so that the models swam slowly forward (B). Models reactivated at pCa 5 swam backward more rapidly (ca. 0.15 mm/s; Figure lOC). When TFP was added, many models were seen to swim forward slowly (Fig. 10D); some cells, however, continued to move in reverse, but they slowed down. The smaller number of backward swimmers reactivated in the presence of TFP was accounted for by a group of models that swam slowly forward. Figure 11 illustrates the change from backward to forward swimming of an individual model shortly after it was reactivated in pCa 5-TFP buffer. Initially, the model swam backward for several seconds; it slowed down and then pivoted about its posterior end; finally, the zyxwvuts zyxw zy Effects of TFP on Paramecium 259 zyxw zyxw zyxwvu Fig. 7. Detergent-extracted models in reactivation buffer without ATP. Most of the models did not move (C; five consecutive strobe flashes 0.9 s apart). Viewed with phase contrast optics (A), the models retained the overall shape of living paramecia, but they have contracted slightly in length. Nomarski optics (B) reveals docked trichocysts (T), birefringent crystals (Cr), and other cytoplasmic inclusions. Numerous relatively straight cilia are extended from the cell body. Bar in A = 100 pm. (A, X 100; B, ~ 2 4 0 C, ; ~10). TABLE I. Number of Backward-Swimming Models at pCa 5* No TFP Experiment 1 Experiment 2 11 15 +40 uM TFP 1 3 *In each experiment, over 50 models were placed in pCa 5 reactivation medium with, or without, 40 pM TFP. Five minutes later, the number of models swimming backward was counted. The remaining models either swam slowly forward or rocked in place without progress through the medium. model began to swim forward slowly. Since the elapsed time of this sequence was about 1 min, relatively few models were swimming backward several minutes after the start of the reactivation experiment with TFP (Table I). DISCUSSION Responses to TFP In Vivo zyxw Short exposure of living paramecia to low doses of TFP or W7 caused longitudinal body contraction and slowed forward swimming (Figs. 2-4). Both processes 260 Otter, Satir, and Satir zyxwvut zyxw zyxwvut zyxwvu Fig. 8. Swimming in Triton-models. The swimming behavior of Triton-models depended on the pCa of the reactivation medium. Models swam forward at pCa 7 (A) or pCa 6.3 (B). At pCa 6 the models rocked in place or swam slowly forward or backward (C), and at pCa 5 , they backed up (D). For models, the interval between flashes, T , was 0.9 s. (x83). depend on changes in cell calcium. Slowing is associated with a fall in pCa at the axoneme, and body shortening in several ciliates involves a Ca2+-induced contraction of spasminlike filaments in the cell cortex [Amos, 1975; Amos et al, 1976; Huang and Pitelka, 1973; Naitoh and Kaneko, 19721. Our observations suggest that TFP induces a decrease in the intracellular pCa, which could account simultaneously for both responses. In addition, shortening of the body may itself contribute to slowing. The spacing between adjacent body cilia would be altered by body contraction, and changes in the spatial relationships could alter viscous coupling between cilia and, consequently, the pattern of ciliary beat and the velocity of swimming [Blake, 1974, 1982; Sleigh, 19761. At 10 pM TFP, the living cells slowed down and stopped but they did not swim backward. According to the calibration curve of Naitoh and Kaneko [ 19721, which we substantially confirmed, this suggests that the pCa at the axoneme fell to about 6. Because TFP had no effect on the reactivated models in the range of axonemal z zy zy 4 0 J zyxwvutsr zyxw Fig. 9. Reactivation in TFP at pCa 2 6. Decreasing the pCa of reactivation from 7 to 6 slowed the forward speed of models (A-C); at pCa 6, they made little or no progress through the medium (C). When TFP was included in the reactivation buffer, swimming tracks appeared essentially identical to those of control paramecia at the same pCa (D, +40 pM TFP; E and F, +25 pM TFP). (A,D X30; B,C,E,F X83). h) c 91 262 Otter, Satir, and Satir zyxwvut zyxwvutsrqp zyxwvuts zyxwvutsrqp Fig. 10. Inhibition of reversed swimming in Triton-models. When models were reactivated at pCa 5.7 (ca 5 pM [Ca2']), they swam slowly backward (A). Addition of 40 pM TFP to the same reactivation medium (B) blocked this reversal, and the models swam forward slowly or rocked in place. Models reactivated at pCa 5 without drug backed up quickly (C). Fewer models backed up in TFP at pCa 5 and some swam forward slowly (D). In each frame, the first flash is brightest and an arrow indicates the direction of swimming ( X 78). calcium concentrations that caused slowed forward swimming (pCa 7-pCa 6), the in vivo effects were apparently not at the axoneme. One specific TFP-sensitive site in living paramecia might be the calcium pump of the plasma membrane; TFP inhibits the activity of CaM-activated calcium pumps in the presence of calcium ions in other types of cells [Larsen and Vincenzi, 19791. While no specific protein has yet been identified as the calcium pump in paramecium, Rauh and co-workers 19801 have found a TFP-sensitive Ca2+-ATPase in paramecium cilia, and paramecia maintain low levels of calcium in their cytoplasm by actively extruding Ca2+-ions [Browning and Nelson, 1976; Hildebrand, 19781. Binding of TFP to the postulated CaM-activated Ca2+-pump could prolong backward swimming and renormalization of ciliary beat after a reversal by inhibiting the active removal of Ca2+-ions from the ciliary matrix (see Fig. 6). W7 (40 pM) mimicked low doses of TFP (5-15 pM), and 50 pM W5 (an hydrophobic analog of W7 that binds only weakly to CaM) did not alter swimming. These results support the conclusion that the primary target of low doses of these drugs is a CaM-activated process. zyxwv zy zyxwvu z zyxwv zyx zyxwvutsr Effects of TFP on Paramecium 263 b f \ '0 Fig. 11. Swimming track of a single model reactivated in buffer containing 40 pM TFP. In four successive photographic flash sequences (1-4), the model backed up (b), turned (t), and then swain forward (Q.The right-hand portion of the figure shows body tracings with the initial and final positions of the cell's anterior marked by a bar. Images I and 4 gave intermediate body positions as well. In the left-hand diagram, the paramecium is represented by an arrow (anterior = arrowhead). Bold arrows show actual photographed positions and light arrows are inferred body positions between flash sequences, assuming a constant velocity equal to that during the preceding flash sequence. View the figure with stereo glasses to juxtapose the two drawings and visualize the swimming path. An alternative hypothesis is that intracellular pCa fell because TFP partially opened the voltage-sensitive Ca2+-channels of the ciliary membrane. Apparently this was not the case since pawn mutants of P. tetraurelia also slowed down in TFP, and the voltage-sensitive CaZf-channels are inoperative in pawn mutants [Satow and Kung, 19801. Since pretreating wild-type paramecia with TFP markedly reduced the frequency of barium-induced reversals, it appears that TFP inhibited opening the calcium channels. Such inhibition might be a secondary effect, since the calcium channels inactivate at reduced cellular pCa [Brehm and Eckert, 19781. Comparable doses of these same drugs alter other calcium-dependent processes in paramecium with a similar time course. For example, secretion of trichocysts is inhibited by treatment with 16 pM TFP or 35 pM W7 for 5 min, while the analog W5 has no effect [Garofalo et al, 19831. A drop in pCa in the trichocyst normally causes expansion and release of the trichocyst contents. Since the drugs do not inhibit Ca2+-dependent expansion of membrane-free isolated trichocysts in vitro, their primary site of action on the secretory process is probably at the trichocyst membrane [Garofalo, 19831. zyxw Responses to TFP in Triton-Models In contrast to its effects in living cells, TFP did not alter the swimming of Triton-models reactivated at pCa > 6. TFP partially inhibited ciliary reversal in Triton-models by interfering with the axonemal mechanism that produces reversed beating, at pCa < 6, so the observed inhibitory effect on ciliary motility of models was strictly Ca2+-dependent. At low pCa, the drug-treated models swam as though the pCa of reactivation was raised, even though the pCa was buffered with 5 mM 264 Otter, Satir, and Satir zyxwvu zyx EGTA. This change in swimming direction cannot be explained by the same target of drug action as in the living cells, where TFP addition induced an apparent decrease in pCa. This experiment demonstrates that at least some of the extracted cells retained the components needed to change the direction of beating of their reactivated cilia. The Ca2+-sensitive components of the axoneme must bind tightly enough to remain attached during extraction, but not so tightly that they are inaccessible to TFP. The effects of TFP on ciliary reversal in paramecium closely resemble TFP effects on mussel gill axonemes observed by Reed and co-workers [1982]. Gill cilia arrest in a Ca2+-dependent manner as pCa falls from 7 to 6 , and they remain arrested below pCa 6 [Walter and Satir, 19781. While arrest occurs over two decades of calcium concentration, TFP restores ciliary beating only when the pCa is below 6 . Thus, the Ca2+-dependent behavioral responses of both the gill cells and paramecia are sensitive to TFP only at low pCa. In order to explain how the Ca2+-binding properties of CaM might be related to the TFP sensitivity of ciliary arrest in gill cells, Reed and co-workers [ 19821 have proposed a TFP-Ca2+-CaM “shunt” mechanism. In the proposed TFP shunt, TFP binds to and inactivates Ca2+-CaMcomplexes when multiple Ca2+-ions are bound, but TFP does not inactivate the initial Ca2+-CaM complexes that form above pCa 6 . By applying a similar analysis to paramecium cilia, we suggest that slow forward swimming results from formation of initial Ca2+CaM complexes at the axoneme and that ciliary reversal corresponds to multiple Ca2+-ions complexed with CaM at the calcium-sensitive switch in the axoneme. Drug sensitivity alone is not an adequate criterion to say that a given process is CaM-dependent. Other CaM-like molecules with drug-binding properties may also participate in the processes that we have described [Moore and Dedman, 19821. However, since CaM is known to be present in paramecium’s cilia [Maihle et al, 1981; Walter and Schultz, 19811, it is the best candidate for the specific target of drug action in our studies. At present, we do not know where CaM is localized within paramecium’s cilia in vivo and whether its distribution changes during extraction with Triton detergent. Studies on the distribution of CaM in cilia suggest that some of the CaM is loosely bound or free in the ciliary matrix, but some CaM must be bound tightly to the axoneme so as to survive treatment with detergent followed by washes with EGTA or buffers at pCa 7 or above [Gitelman and Witman, 1980; Stommel et al, 19821. In Tetrahymena, CaM is clearly part of the cilium [Suzuki et al, 19811, but its site of attachment to the axoneme is not certain. Structural studies localize CaM in the interdoublet nexin links [Ohnishi et al, 19821, while fractionation of Tetrahymena cilia indicates that CaM copurifies with 14s dynein on sucrose gradients [Jamieson et al, 19791. Blum and co-workers [ 19801 have proposed that CaM binds to the dynein arm in vivo and activates it when calcium is present. We are currently examining in detail the location of CaM in paramecium cilia using immunolabelling techniques. Regardless of our interpretation of what TFP is doing, addition of TFP in vivo shifted the behavior of the cells to a type characteristic of a lower axonemal pCa, while in Triton-models reactivation in TFP either had no effect (pCa 2 6 ) or shifted the behavior to a type that is characteristic of a higher pCa. Specifically, concentrations of TFP that altered behavioral responses in whole cells, where the axoneme is presumed to be at pCa 2 6 , were ineffective in eliciting swimming changes in the models at the same pCa ’s. Taken together these results suggest that there must be at least two different sites of action of TFP in cilium. We have argued that TFP inhibits a membrane calcium pump in living cells, but in the models, the drug must ultimately zyxwvu zyx zyxwvuts zyxwvu Effects of TF” on Paramecium 265 zy influence the calcium switch of axoneme. If we assume that TFP effects indicate processes that are CaM-controlled, our results suggest that at least two types of CaM exist in cilia: a form tightly bound to the axoneme that regulates sliding efficiency and reversal and a form that controls Ca*+-pumping. This second CaM could be permanently bound to the pump site or it could be a soluble matrix protein that binds reversibly to a membrane receptor in an equilibrium that depends on pCa. Our drug studies imply that the pump must normally be activated at physiological pCa ( 3 7 ) and that the primary response of the axoneme to increased calcium is not reversal, but slowed forward swimming. Exactly how Ca2+-dependent processes within one cilium are properly coordinated upon stimulation and recovery is an intriguing problem that deserves further study. zyxwv ACKNOWLEDGMENTS We thank Dr. E.D. Salmon for loaning his Strobex power supply and xenon arc lamp for use in these studies, Mr. Bill Sims for help in constructing the flash triggering circuitry, and C. Hubertus for carefully typing the manuscript. Thanks to P. Detmers and R.S. Garofalo for many enthusiastic and thoughtful discussions. Postdoctoral fellowship support to T.O. came from the American Cancer Society PF2130 and USPHS F32 GM 08986. Grant support was from USPHS HL 22560 (P.S.) and NSF PCM 8207900 (B.H.S.). NOTE ADDED After our manuscript had been submitted for review, Nakaoka et a1 [1984] published results of their studies on the regulation of ciliary beat frequency in Paramecium. Several of their findings directly support our data and conclusions. For example, TFP or W7 did not affect Ca2+-induced body contraction in Paramecium models. However, both Nakaoka et a1 [1984] and Watanabe and Ohnishi [1982] have reported that ciliary reversal in Triton-extracted paramecia is not inhibited by calmodulin antagonists. These results do not agree with our findings, and the difference is probably due to variations in the methods of extraction and reactivation. REFERENCES zyxw Amos, W.B. (1975): Contraction and calcium binding in the vorticellid ciliates. In Inoue, S . , and Stephens, R.E. (eds.): “Molecules and Cell Movement.” New York: Raven Press, pp. 41 1-436. Amos, W.B., Routledge, L.M., Weis-Fogh, T., and Yew, F.F. (1976): The spasmoneme and calcium dependent contraction in connection with specific calcium binding proteins. Syrnp. Soc. Ex. Bid. 30:273-301. Blake, J. (1974): Hydrodynamic calculations on the movement of cilia and flagella. I. Paramecium. J. Theor. Biol. 45: 183-203. Blake, J. (1982): Mechanics of ciliary transport. Prog. Clin. Biol. Res. 80:41-45. Blum, J.J., Hayes, A,, Jamieson, G.A., Jr., and Vanaman, T.C. (1980): Calmodulin confers Ca’+sensitivity on ciliary dynein ATPdse. J. Cell Biol. 87:386-397. Brehm. P., and Eckert, R. (1978): Calcium entry leads to inactivation of calcium channel in Paramecium. Science 202: 1203--1206. Browning, J.L., and Nelson, D.L. (1976): Biochemical studies of the excitable membrane of Pararnecium aurelia. Biochim. Biophys. Acta 448:338-35 1. Dryl, S. (1958): Photographic registration of movement of protozoa. Bull. Pol. Acad. Sci. 6:429-430. 266 Otter, Satir, and Satir zyxwvu zyxwvu Dryl, S . , and Masnyk, S. (1971): Observations on the effects of chlorpromazine on the motility of Paramecium aurelia, stock 51. J. Protozool. 18(Suppl.):91. Dunlap, K. (1977): Localization of calcium channels in Paramecium cuudufum. J. Physiol. (Lond.) 277:119-133. Eckert, R., Naitoh, Y . , and Machemer, H. (1976): Calcium in the bioelectric and motor functions of Puramecium. Symp. Soc. Exp. Bid. 30:233-255. Garofalo, R.G. (1983): “A Role for Calmodulin in the Regulation of Secretion in Paramecium tefruurelia.” Ph.D. Thesis, Albert Einstein Coll. of Med., Univ. Microfilms, Ann Arbor,MI. Garofalo, R.G., Gilligan, D.M., and Satir, B.H. (1983): Calmodulin antagonists inhibit secretion in Paramecium. J. Cell Biol. 96: 1072-1081. Gitelman, S.E., and Witman, G.B. (1980): Purification of calmodulin from Chlamydomonas: Calmodulin occurs in cell bodies and flagella. J. Cell Biol. 87:764-770. Guttinan, H.N., and Friedman, W. (1962): Protozoa as pharmacological tools: Phenothiazine tranquilizers. Trans. NY Acad.Sci. 26:75-89. Hidaka, H., Asano, M., and Tanaka, T. (1981): Activity-structure relationships of calmodulin antagonists: Naphthalene sulfonamide derivatives. Mol. Pharmacol. 20:57 1-578. Hildebrand, E. (1978): Ciliary reversal in Paramecium: Temperature dependence of K+-induced excitability decrease and of recovery. J. Comp. Physiol. 127:39-44. Huang, B., and Pitelka, D. (1973): The contractile process in the ciliate Sfenfor coeruleus. I. Role of microtubules and filaments. J. Cell Biol. 57:704-728. Jamieson, G . A . , Jr., Vanaman, T.C., and Blum, J.J. (1979): Presence of calmodulin in Tefruhymena. Proc. Natl. Acad. Sci. USA 76:6471-6475. Larsen, F.L., and Vincenzi, F.F. (1979): Calcium transport across the plasma membrane: Stimulation by calmodulin. Science 204:306-309. Levin, R.M., and Weiss, B. (1976): Mechanism by which psychotropic drugs inhibit adenosine cyclic 3’-5’-monophosphate phosphodiesterase of brain. Mol. Pharmacol. 12:581-589. Machemer, H. (1974): Frequency and directional responses of cilia to membrane potential changes in Paramecium. J. Comp. Physiol. 92:293-3 16. Machemer, H . , and Ogura, A. (1979): Ionic conductances of membranes in ciliated and dcciliated paramecia. J. Physiol. (Lond.) 296:49-60. Maihle, N.J., Dedman, J.R., Means, A.R., Chafouleas, J.G., and Satir, B.H. (1981): Presence and indirect immunofluorescent localization of calmodulin in Paramecium tefruurelia. J. Cell Bid. 89:695-699. Moore, P.B., and Dedman, J.R. (1982): Calcium-dependent protein binding to phenothiazine columns. J. Biol. Chem. 257:9663-9667. Naitoh, Y., and Kaneko, H. (1972): Reactivated Triton-extracted models of Paramecium: Modification of ciliary movement by calcium ions. Science 176:523-524. Naitoh, Y., and Kaneko, H. (1973): Control of ciliary activities by adenosinetriphosphate and divalent cations in Triton-extracted models of Paramecium caudatum. J. Exp. Biol. 58:657-676. Nakaoka, Y., Tanaka, H., and Oosawa, F. (1984): Ca2+-dependent regulation of beat frequency of cilia in Paramecium. J. Cell Sci. 65:223-231. Ogura, A., and Takahashi, K. (1976): Artificial deciliation causes loss of calcium-dependent responses in Parumrcium. Nature 264: 170-172. Ohnishi, K . , Suzuki, Y., and Watanabe, Y. (1982): Studies on calmodulin isolated from E~fruhymenu cilia and its localization within the cilium. Exp. Cell Res. 137:217-227. Otter, T., and Salmon, E.D. (1979): Hydrostatic pressure reversibly blocks membrane control of ciliary motility in Puranzecium. Science 206:358-36 1. Otter, T., Satir, B.H., and Satir, P. (1983): Possible mechanisms by which calmodulin antagonists alter swimming in paramecium. Biophys. J. 41:89a. Rauh, J., Levin, A.E., and Nelson, D.L. (1980): Evidence that calmodulin mediates calcium-dependent ciliary reversal in Paramecium. In Siegel, F.L. et al (eds): “Calcium Binding Proteins: Structure and Function.” New York: Elsevier North-Holland, pp. 23 1-232. Reed, W., and Satir, P. (1980): Trifluoperazine inhibits mussel gill lateral ciliary arrest. J. Cell Biol. 87:39a. Reed, W., Lebduska, S . , and Satir, P. (1982): Effects of trifluoperazine upon the calcium-dependent arrest response of freshwater mussel gill lateral cells. Cell Motil. 2:405-427. zyxw zyx zyxwv zyxwvuts Effects of TFP on Paramecium 267 Salmon, E.D., and Segall, R.R. (1980): Calcium-labile mitotic spindles isolated from sea urchin eggs (Lytechinus variegatus). J. Cell Biol. 86:355-365. Satir, B.H., Garofalo, R.S., Gilligan, D.M., and Maihle, N.J. (1980): Possible functions of calmodulin in protozoa. Ann. NY Acad. Sci. 356:83-91. Satir, P. (1982): Mechanisms and controls of microtubule sliding in cilia. Symp. Soc. Exp. Biol. 35: 179-201. Satow, Y., and Kung, C. (1980): Membrane currents of pawn mutants of the pwA group in Purunieciunz tetraureliu. J. Exp. Biol. 84:57-71. Sleigh, M.A. (1976): Fluid propulsion by cilia and the physiology of ciliary systems. In Spenccr-Davies, P. (ed.): “Perspectives in Experimental Biology. I. Zoology.” Oxford: Pcrgamon Press, pp. 125-134. Sonneborn, T.M. (1970): Methods in Paramecium rcsearch. Methods Cell Physiol. 4:241-339. Stommel, E.W., Stephens, R.E., Masure, H.R., and Head, J.F. (1982): Specific localization of scallop gill epithelial calmodulin in cilia. J. Cell Biol. 92:622-628. Sugino, K., and Naitoh, Y. (1982): Simulatcd crossbridge patterns corresponding to ciliary beating in Paramecium. Nature 295:609-611. Suzuki, Y . , Ohnishi, K., and Watanabe, Y. (1981): Tetrahymena calmodulin. Exp. Cell Res. 137:l-14. Thiele, J., and Schultz, J. (1981): Ciliary membrane vesicles contain the voltage-sensitive calcium channel. Proc. Natl. Acad. Sci. USA 78:3688-3691. Wais-Steider, J . , and Satir, P. (1979): Effect of vanadate on gill cilia: Switching mechanism in ciliary beat. J. Supramol. Struct. 11 :339-347. Walter, M.F., and Satir, P. (1978): Calcium control of ciliary arrest in mussel gill cells. J. Cell Biol. 79: 110-120. Walter, M., and Schultz, J. (1981): Calcium receptor protein calmodulin isolated from cilia and cells of Purarnecium fetraurelia. Eur. J. Cell Bid. 24:97-100. Witman, G.B., and Minervini, N. (1982): Role of calmodulin in the flagellar axoncme: Effect of phenothiazines on reactivated axonemes of Chlamydomo~as.Prog. Clin. Biol. Res. 80: 199-204. Watanabe, Y., and Ohnishi, K. (1982): Calmodulin in the cilia of Tetrahymena. In Sakai, H. et al (eds): “Biological Functions of Microtubules and Related Structures.” New York: Academic Press, pp. 115-123. zy zyx zyxw zyxwvuts zy