Ca2+ current remained in the mutant.

Transcription

1 J. Physiol. (1984), 351, pp With 5 text-figures Printed in Great Britain A MUTATION THAT ALTERS PROPERTIES OF THE CALCIUM CHANNEL IN PARAMECIUM TETRA URELIA BY R. D. HINRICHSEN AND Y. SAIMI From the Laboratory of Molecular Biology, University of Wisconsin, Madison, WI 53706, U.S.A. (Received 25 August 1983) SUMMARY 1. The membrane properties of a new mutant of Paramecium tetraurelia, dancer, were compared under voltage clamp with those of the wild type. 2. The Ca2+ current was isolated and examined using CsCl-filled electrodes and tetraethylammonium in the bath solution to block K+ channels. The amplitude of the Ca2+ transient was not altered by the mutation. However, the Ca2+ current in the mutant inactivated more slowly and less extensively: hence a larger sustained Ca2+ current remained in the mutant. 3. A change in the time course of the deactivation of the Ba2+ current was observed in the mutant. This mutational change is not likely to be the consequence of the Ca2+-channel inactivation because it is seen in the Ba2+ solution where there is little inactivation of the current. 4. Other measured properties of the Ca2+ channel, the voltage-dependent K+ current, and the resting properties of the membrane were normal in the mutant. 5. The Ca2+-activated K+ current and the Ca2+-activated Na+ current were larger in the mutant than in the wild type, consistent with a greater elevation of free intracellular Ca2+ during depolarization in the mutant. 6. It is likely that the mutation causes an alteration in the Ca2+-channel structure or in its immediate environment and thereby affects the inactivation and deactivation processes of the Ca2+ channel. 7. As would be expected from the greater Ca2+ current, the mutant tends to generate all-or-none Ca action potentials as opposed to the graded action potentials in the wild type. INTRODUCTION Mutations have been used successfully in altering or deleting specific membrane currents in Drosophila and in Paramecium (Wu, Ganetzky, Haugland & Liu, 1983; Hall, 1983; Salkoff, 1983; Kung & Eckert, 1972; Takahashi & Naitoh, 1978; Saimi, Hinrichsen, Forte & Kung, 1983). Among Paramecium mutants, pawns (Kung & Eckert, 1972) and cnrs (Takahashi & Naitoh, 1978), which lack most of the Ca2+ current, have been used as null controls of the Ca2+ current and to examine the voltage-dependent K+ current. Mutants were also used to unveil the Ca2+-dependent Na+ current (Saimi & Kung, 1980) and to assess the role of the Ca2+-dependent K+ current in the regulation of action potentials (Saimi et al. 1983).

2 398 R. D. HINRICHSEN AND Y. SAIMI Although there are suspected cases where the mutations occur in the channel structure (Hall, 1983; Salkoff, 1983), the identification of the mutational lesions is difficult. Besides the channel structure, it is possible that the defect occurs in processing or modifying enzymes (Reuter, 1983), or in a 'cytoplasmic' factor required for channel function (Kostyuk, 1980; Hagiwara & Byerly, 1981). Though very specific effects are seen in the membrane currents, these could be secondary effects of mutations other than those that alter the functional channel unit. The Ca2+ channel in Paramecium is activated by a depolarization of the membrane and inactivated by internal Ca2+ entered through the channels (Brehm & Eckert, 1978; Brehm, Eckert & Tillotson, 1980). Similar behaviour of the Ca2+ channels has been reported on the insect muscle and the molluscan neurones (Plant & Staden, 1981; Eckert & Tillotson, 1981; Ashcroft & Stanfield, 1982; Chad, Eckert & Ewald, 1984; Eckert & Ewald, 1983). We report here a new mutant which has a weak inactivation and an altered deactivation of the Ca2+ channel. This mutation appears to affect the structure of the functional Ca2+-channel unit. METHODS Strains and culture. We used primarily the wild type, 51s, and the mutant dancer, stock d4-623, genotype Dn/Dn, ofp. tetraurelia. Well fed cells (Saimi etal. 1983) were chosen forelectrophysiological studies. Procedures and 8et-up of electrophysiology. The basic procedures of recording were the same as described previously (Saimi & Kung, 1980, 1982). The electrodes were filled with 0 5, 3 M-KCI or 4 m-cscl, depending on the type of experiment. For the voltage clamp the electrode resistance was MC; for some current injection experiments it was more than 50 Mil with 0-5 M-KCl. The open loop gain of the voltage-clamp circuit was x 400 to x The estimated series resistance to the cell was about 100 kq in Solution B, and about 200 kq in Solution A. The capacitative surge usually had a time constant of less than 60 ss, and the voltage settled within 500 Us. The current was monitored through a virtual ground I-V converter that had a time constant of ss, depending on the requirement. The membrane was usually held at -40 mv (near the resting level) so that the inward current was largest upon depolarization (see Saimi & Kung, 1982). The solutions used had ion compositions in addition to 1 mm-hepes and 0-01 mm-edta2- (ethylenediaminetetraacetic acid) as follows (in mm): Solution A: 4 K' +1 Ca2+; Solution B: 4 Cs tetraethylammonium (TEA') + 1 Ca2+; Solution C: 4 Cs TEA+ + 1 Ba2+; Solution D: 8 Na+ + Solution C. Solutions B-D were used in conjunction with 4 M-CsCl electrodes to suppress much of the K+ current. All the salts of reagent grade came from the Cl- forms, and the ph was adjusted to with OH-. The experiments were done at room temperature ( TC). Correctionfor leakage current. Leakage current correction was done by plotting the linear currents by small hyperpolarizations on a I- V graph, then extrapolating the linear portion. The current was measured from the linear base line (e.g. see Fig. 5). The leakage conductance was about 25 ns. We noticed a small (- 1 na at 2 ms with a 40 mv step), fast decaying (T < 2 ms) current whose size was voltage-dependent but asymmetric with respect to the direction of the voltage step. This did not, however, interfere significantly with the measurement of the maximal peak Ca2+ current. At higher voltages where the inward current was smaller this error could become significant (e.g. up to 25 % at 0 mv). This asymmetric current did not contaminate the later currents, e.g. the sustained inward current. The capacitative surge (see above) was too fast to affect the measured Ca2+ current. For the measurement of the tail current we subtracted the capacitative surge electronically. In these cases the error due to the asymmetric current was insignificant since the tail current was larger. I8olation of the Ca2+ current. Upon depolarization, the voltage-dependent Ca2+ current, the voltage-dependent K+ current, the Ca2+-activated K+ current, and the Ca2+-activated Na+ current may be activated (see Kung & Saimi, 1982). Since this study concerns a genetic alteration of the

3 A Ca2+-CHANNEL MUTANT IN PARAMECIUM 399 Ca2+ current, it is important to isolate the Ca2+ current from the other currents. We were able to eliminate much of the voltage-dependent K+ current by the use of electrodes filled with 4 M-CsCl. Cs' presumably diffused from the electrode and blocked the K+ current (Tillotson & Horn, 1978; Brehm et al. 1980). Omitting K+ and adding TEA+ (from Aldrich) in Solution B further inhibited the K+ current (10 mm-tea+ outside by itself reduces the voltage-dependent Kt current to about 20 %, Y. Saimi, unpublished; Saimi & Kung, 1982). With the above procedure, the residual current in the pawn mutants, which have little or no Ca2+ current, was actually inward (as much as 1 na) after subtraction of the linear leakage current. Thus, we estimated the inhibition to be almost complete at lower voltages (< 0 mv) and greater than 90 % even at higher voltages. By the time we began recording about 2 min after the perfusion of Solution B, the suppression of the K+ current was maximal. A similar procedure with TEA' outside and the CsCI electrodes eliminated > 95 % of the K+ current in P. caudatum (Hennessey & Kung, 1984). The Ca2t-activated currents are known to activate more slowly and to have much slower and more prominent tail currents than the other currents (see Fig. 5; also Kung & Saimi, 1982). As shown in the Results, these currents are stronger and may activate faster in dancer than in the wild type because of the greater Ca2+ current. Although we see a very small inward tail current in Solution B from dancer (see Fig. 5D), the amplitude or time course of this tail did not change by substitution of Cs+ with choline or tetramethylammonium (TMA+) (data not shown). Therefore we conclude that neither Cs+, choline nor TMA+ is permeable through those Ca2+-activated channels. The same conclusion is reached for TEA+. The Ca2+-activated Na+ current activates faster than the Ca2+-activated K+ current (Saimi & Kung, 1980). In dancer the Na+ current was detected as early as at 20 ms in the presence of Na+ (data not shown). Ca2+ permeability through those Ca2+-activated channels is unlikely but not ruled out. The ion, which carries the small inward tail current seen in Fig. 5D, a current unchanged by Cs+ or choline substitution, remains unidentified. This small inward tail was not seen with shorter pulses. In summary, the conditions for isolation of the Ca2+ current are: the use of 4 M-CsCl electrodes, a K+-free, Cs+-, TEA+-containing bath (Solution B), and the limit of the time range up to 20 ms and the voltage range up to 0 mv. The sustained current, however, changed slightly after 20 ms (by less than 1 na from 20 to 90 ms). The isolated current consists of an inward transient and a small sustained inward current (Fig. 1). This isolated inward current appears to be the Ca2+ current because: (1) there was such sustained inward current even after elimination of TEA+ and Cs+ from the bath solution, leaving Ca2+ as the only cation except H+, (2) both transient and sustained inward currents decrease with a reduction in the external Ca2+ concentration (T. Hennessey, unpublished), (3) in paums with and without a dancer background such an inward current was small, (4) the augmented Ca2+-activated K+ and Na+ currents in dancer were consistent with its larger sustained inward current being the Ca2+ current, which serves to trigger the former currents (see Results), and (5) the sustained inward current is as susceptible as the peak Ca2+ current is to a drug, W-7 (Hennessey & Kung, 1984). Spatial control of the voltage. According to the ionic theory the peak voltage of the action potential (Vp) should not exceed the voltage (VO) where the total membrane current (except for the capacitative surge) becomes zero under voltage clamp. If the membrane is not sufficiently voltage clamped, VO could be larger than V (Fukuda, Fischbach & Smith, 1976). Since the Ca2+ channels of Paramecium are thought to resize in the ciliary membrane (Eckert & Brehm, 1979), one might wonder whether these Ca2+ channels are in fact controlled by the voltage clamp. We therefore compared the Vp and VO within individual cells. The action potentials were made all-or-none in the presence of 8 mm-bacl2 added to Solution A (Naitoh, Eckert & Friedman, 1972) as well as the use of CsCl electrodes in some cases. In typical cases, these voltages agreed quite well. As averaged, VO was very close to Tp (Vo- Tp: +0-6 ± 1-2 mv; mean ± S.D., n = 8 from five wild-type cells). Similar results were obtained with 3 M-KCl electrodes ( *3 mv; n = 6 from four wild-type cells). The second argument for the voltage control is that the deactivation of the Ca2+ or Ba2+ current followed the same time course whatever the amplitude was. These findings argue against an inadequate spatial clamp, at least for the voltage control of the Ca2+ channel. Finally, it is noteworthy that the transient Ca2+ currents in Solutions A and B were the same. Suppression of the K+ currents certainly should have favoured better spatial control of the voltage.

4 400 R. D. HINRICHSEN AND Y. SAIMI RESULTS When a paramecium is properly stimulated, the consequent depolarization of the membrane in the form of a Ca action potential causes the animal to swim backwards through the reversal of the beat direction of the cilia. The extent of backward swimming varies with the stimulation from a subtle pause in the forward swimming to long, continuous backward swimming. The mutant, dancer, described here shows rather sharp and uniform jerks for several body lengths in some solutions. All behavioural characteristics are consistent with the all-or-none amplitude and prolonged duration of the action potential in the mutant as described below, while the action potential in the wild type is graded rather than all-or-none (Naitoh et al. 1972). The detailed description of the behaviour and genetics of this mutation will be given elsewhere (R. D. Hinrichsen, Y. Saimi & C. Kung, unpublished). Action potentials and other membrane properties. When a current is injected into the wild-type cell, a graded action potential with only one strong peak is evoked whose height depends on the intensity of the current (Naitoh et al. 1972; Fig. 1 A). Similar experiments on dancer, on the other hand, reveal nearly all-or-none action potentials (Fig. 1 B). Unlike the case of the wild type, multiple action potentials of equal height are elicited in the mutant. Other membrane properties (mean+ S.D.) in the wild type (n = 4) and mutant (n = 4), such as the resting potential ( and mv, respectively), membrane resistance ( and MQ), membrane time constant (measured as a half-decay time of the potential change by a small inward current) ( and ms) and peak of the action potential by 1 na outward current (4+8 and 12+3 mv), are similar. Properties of the Ca2+ currents in the wild type and mutant. The all-or-none nature of the action potential in the mutant strongly indicates an increase in the net inward current by either the enhancement of the Ca2+ inward current or the inhibition of the K+ outward current. Under voltage clamp, the Ca2+ current in Solution B is induced by a step depolarization from the holding potential at -40 mv. The inward current peaks within 5 ms and then decays to a sustained level. Fig. 1 C and D show families of the Ca2+ currents from the wild type and mutant. The I-V relations are also plotted in Fig. 1 E and F. The most pronounced differences between the wild type and the mutant are the decay time course of the Ca2+ current and the level of the sustained inward current. The decay of the Ca2+ current is fitted by an exponential curve using the level at about 20 ms as an asymptotic level (Fig. 1 C and D). The time constant (r) thus obtained, plotted against the step voltages (Fig. 2 A-C), clearly show differences between the wild type and mutant. In both cases the r values are voltage dependent: the higher the voltage, the larger the T. The r values at higher voltages (> 0 mv) are probably underestimated because of the incomplete suppression of the outward current. Even if we allow the wild type a shift along the voltage axis, the two curves cannot be superimposed. At 18 ms the levels of the sustained inward current of the wild type and mutant also differ (Fig. 2D). This, too, is not explainable by a simple shift along the voltage axis. Other properties of the Ca2+ currents (mean + S.D.) from the wild type (n = 6) and

5 A A Ca2+-CHANNEL MUTANT IN PARAMECIUM _ E 0 C 125 ms D V.= Vml -20 mv -.?f*' 1 vmvw+- V - -O -10 mv k I~~~~~ 0 mv I a fib" A -- I Pi 11, E 5 ms Vm (mv) / Ca2+ F Fig. 1. Membrane properties of the wild type V8. dancer. A and B: current injection. Currents of varied intensity (one with -0-2, two with +0-2, three with +0 4 and four with + I 0 na) are injected for 250 ms into the wild type (A) and dancer (B). The dashed lines indicate the reference levels. Bathed in Solution A; recorded with 0.5 M-KCl electrodes. C-D: voltage clamp. Families of the Ca2+ currents induced by steps from -40 mv to voltages indicated beside each trace are shown for the wild type (C) and dancer (D). The dashed lines indicate the zero current level. The leakage is not corrected. Bathed in Solution B; recorded with 4 M-CsCl electrodes. The I-V relationships of the peak Ca2+ currents (0) and sustained inward currents (@) at 18 ms in the wild type (E) and dancer (F) are obtained from the same cells as in C and D. The leakage current is subtracted from each measurement in E and F.

6 402 R. D. HINRICHSEN AND Y. SAIMI I A -1-0*5 0 mv 0 2 I ~ ~ ~~~~~~i Time (ms) /Ca f C 0 2 Time (ms) ImV 7. (ins) 6+ 'sust Vm (mv) Vm (mv) Fig. 2. Decay of Ca21 currents during voltage steps. A and B show typical cases of the decay of the Ca2+ current at the voltages indicated by the lines. The decay of the Ca2+ current measured from a sustained current level at about 20 ms is well fitted with an exponential curve both in the wild type (A) and in dancer (B). Time zero is taken at the peak of each Ca2+ current. C shows the relationships between the decay time constants (r) and the membrane potentials in the wild type (0; n = 6 except at the lowest voltage (n = 5)) and dancer (-; n = 8). In D, the levels of the sustained inward currents (Iust. at 18 ms in the wild type (0; n = 6) and dancer (0; n = 8) are plotted against the membrane potentials. Standard deviation (bars) is not shown when smaller than the diameter of the circle (mean). Bathed in Solution B; recorded with 4 M-CsCl electrodes.

7 A Ca2+-CHANNEL MUTANT IN PARAMECIUM mutant (n = 8), such as the maximal peak Ca21 current (Imax) (-7x0+1x0 and na), the voltage where the Imax is observed ( and -10± 2 mv), the voltage where the peak Ca2+ current is half the Imax ( and mv), and the peak time of the Imax from the onset of the voltage step ( and ms), are not significantly different. We also examined the voltage-dependent K+ current in both the wild type and the mutant with a background of the pawn mutation (pawnb) in order to erase the Ca2+ current and the Ca2+-dependent currents: thus isolating the voltage-dependent K+ current (Satow & Kung, 1980). We detected no difference between the wild type and dancer (data not shown). Properties of Ba2+ currents in the wild type and mutant. Ba2+ is apparently not a good agonist for Ca2+ for the inactivation of Ca2+ channels in various cells (Brehm & Eckert, 1978; Plant & Standen, 1981; Eckert & Tillotson, 1981; Ashcroft & Stanfield, 1982). Fig. 3A and B shows families of Ba2+ currents in both the wild type and the mutant in Solution C. The Ba2+ current is sustained. The decline of the current at 40 ms as compared to the peak is very small (4 and 5 % with a voltage step to -20 mv, and 9% and 11% I to -10 mv in the wild type (n = 4) and mutant (n = 3), respectively). This suggests that Ba2+ does not substitute for Ca2+ in causing inactivation and that there is very little voltage-dependent Ca2+-channel inactivation. The I-V relationships of the Ba2+ currents are similar in both the wild type and mutant (data not shown). However, we consistently saw a small difference in the activation of the Ba2+ current, especially at lower voltages (compare Fig. 3A and B). This might indicate that there is a small shift in the voltage sensitivity of the Ca2+ channel toward hyperpolarization, but the shift would be less than 5 mv, which is near the resolution limit of the present technique. The voltages for the half-maximal Ba2+ currents are mv and mv in the wild type (n = 4) and mutant (n = 3), respectively. Although variable because of the fast run-down of the current in Solution C, the maximal Ba2+ currents are as large as -12 na in both cases. Deactivation of the Ca2+ and Ba2+ currents. We compared the deactivations of the Ca2+ or Ba2+ currents in the wild type and mutant. In Solution B, the tails of the Ca2+ currents after a 2 ms pulse are well fitted with single exponential curves with time constants (r) at -40 mv of ms (n = 9) in the wild type and ms (n = 5) in the mutant. Since it has been suggested that, besides the deactivation process, the Ca2+ tail current could also include inactivation by Ca2+ (Standen & Stanfield, 1982), the deactivation process is more accurately measured in a Ba2+ solution where there is very little Ba2+-dependent inactivation as shown above. It also allows clearer analysis since the tail is slower in a Ba2+ solution (Saimi & Kung, 1982). At -40 mv, the tail current is well fitted with a single exponential curve in both the wild type and mutant (Fig. 3C). The r values are ms (n = 10) for the wild type and ms (n = 7) for the mutant. The time constant varies with the test voltage for the tail, but not with the step size, the duration of the pulse, or the amplitude of the Ba2+ current, even after the specimen has begun to deteriorate (data not shown). Since r is voltage-dependent, the holding level as well as the test voltage for the tail in the mutant was varied. At -45 mv T is '14 ms (n = 4) and at -50 mv, ms (n = 3). Two-pulse experiment. Double-pulse experiments have provided evidence for 403

8 404 R. D. HINRICHSEN AND Y. SAIMI A Or'; rol -20 mv B z- rwjz ; Tkmmus r -10 mv _ m_.ok0 o" _ 8 r /0 mv,x~~~4.-~~~~ 0 a~~~~~~~~ e~~~ 13 ms -30 f c Vm /tail -10+ VM 'U0 1 5 ms -34+ \ w f -if i. I Time (ms) Fig. 3. Ba2+ currents from wild type and dancer. Families of the Ba2+ currents at voltages indicated beside each trace are shown for the wild type (A) and dancer (B). The Ba2+ tail currents are shown from the wild type (upper in the inset) and dancer (lower). A voltage pulse from the holding level at -40 to -15 mv (upper) or -20 mv (lower) is returned after 15 ms (upper) or 11 ms (lower) to -40 mv. Only the later part of the current is magnified to show the tail current. Correction of the capacitative surge is done by an electronic subtraction. The dashed lines indicate the holding current levels. The decays of the Ba2+ tails are plotted against time after the pulse on a semilog graph (C) for the wild type (0) and dancer (*0). The points fall on an exponential curve in both cases. Bathed in Solution C; recorded with 4 M-CsCl electrodes.

9 A Ca2+-CHANNEL MUTANT IN PARAMECIUM Ca2+-dependent Ca2+-channel inactivation (Brehm & Eckert, 1978; Brehm et al. 1980; Plant & Standen, 1981; Eckert & Tillotson, 1981; Ashcroft & Stanfield, 1982). Fig. 4A-C shows the effect of a first pulse on the Ca2+ current induced by a second pulse in such experiments. The peak Ca2+ current of the second pulse reduces as the step size of the first pulse increases, and reaches a minimal level about + 10 mv in the wild type and 0 mv in the mutant. When the first pulse is increased further, the peak Ca2+ current regains its amplitude (Fig. 4C). Although both cases show a significant decrease of the Ca2+ current due to the first pulse, the maximal extent of the reduction is 75-80% in the wild type and only % in the mutant (n = 2 each). In both cases, the peak Ca2+ current is almost undetectable at the minimal Ca2+ current level (Fig. 4A and B: middle); instead, only a sustained component is observed. These observations are consistent with the conclusion that the Ca2+ channel inactivates less in the mutant than in the wild type. Ca2+-activated currents. Since the inactivation of the Ca2+ channel is mediated by internal Ca2+, there are two obvious possibilities to account for the differences between the wild type and mutant. First, the binding of Ca2+ to some site, possibly of the channel itself (Standen & Stanfield, 1982), may be weaker in the mutant. Secondly, the removal mechanisms of Ca2+, diffusion, sequesteration and pumping, may be more efficient in the mutant. Although definitive experiments should be done with single-channel recordings, activations of the Ca2+-activated currents can serve as indicators of the accumulation and regulation of Ca2+ inside the cell (Barish & Thompson, 1983). The Ca2+-activated K+ current is measured in Solution A with 3 M-KCl electrodes, so that the current is not suppressed (see Methods). There is a slow activation of the outward current during a 500 ms pulse at intermediate voltage steps. This current, which is greater in the mutant, appears in the I-V plot as a hump of the outward current (Fig. 5A and B). This is similar to the Ca2+-activated K+ current in molluscan neurones (Gorman & Thomas, 1980). This slow outward current is followed by a slow tail: the amplitude of the maximal tail currents after 500 ms pulses are about na in the wild type and about na in the mutant (n = 2 each). Despite the difference of the amplitude of the Ca2+-activated K+ current, the half-decay times of the maximal tails are similar in the both cases (about 40 ms; n = 2 each). The Ca2+-activated Na+ current in Solution D is also exaggerated in the mutant (Fig. 5C-D'). In the presence of Na+, the inward current increases slowly during the pulse. Even more dramatically it is followed by a large and slow tail. The maximal tail currents after 500 ms pulses are to na in the wild type (n = 2) and -4-0 to -49 na in the mutant (n = 3). At its maximal amplitude after a 500 ms pulse, the tail decays with a half-decay time of ms in the mutant and ms in the wild type. Similar inward currents are seen from both cells in 8 mm-licl but not in 8 mm-kcl or -CsCl (data not shown). These observations are consistent with a higher intracellular Ca2+ concentration in the mutant than in the wild type during a voltage pulse. 405

10 406 R. D. HINRICHSEN AND Y. SAIMI l2a-l 5 ms C L5 A. V1 /2.-* -40 V I *4wr %--1-0_-,-'- -- -, -,.i w t _\[ ww_ 2I -10+ C V1 (mv) Fig. 4. Two-pulse experiment: inactivation of Ca2+ currents by preceding pulses of various amplitudes. A voltage step (-10 mv for 20 ms) from the holding membrane potential at -40 mv is preceded by a pulse of a variable step size ( V1; from -40 up to + 60 mv for 20 ms) with an interval of 40 ms (see diagram). The leakage currents are not corrected in A and B. The Ca2+ currents after subtraction of the leakage currents which are not affected much by the first pulse are plotted in C for the wild type (0) and for dancer (0). The minima in both cases correspond to the sustained current levels. Bathed in Solution B; recorded with 4 M-CsCl electrodes. DISCUSSION The comparison of the wild type and mutant clearly shows three differences in the properties of the Ca2+ current: (1) the time course of the decline (Fig. 2), (2) the sustained level (Fig. 2), and (3) the time course of the deactivation (Fig. 3). While there is a sustained Ca2+ current flowing through the same Ca2+ channels as the peak Ca2+ current after the major inactivation (Brehm & Eckert, 1978; Brehm et al. 1980), one might wonder whether there are actually two types of Ca2+ channels in Paramecium. There are cells with two types of Ca2+ channels (Hagiwara, Ozawa & Sand, 1975; Okamoto, Takahashi & Yoshii, 1976), including another protozoan,

11 A Ca2+-CHANNEL MUTANT IN PARAMECIUM 407 A +20 B 290 ms.bo D / <LO im Vm (mv) Vm (mv) C D C'. I + Nat D' _..~~~ 250 ms Fig. 5. Ca2+-activated currents. The insets in A from the wild type and in B from dancer show slow activation of the Ca2+-activated K+ currents at -15 mv (asterisks) and the tail (arrow). The I-V curves from two cells each (different symbols) are the currents at 500 ms during voltage steps without correction of the leakage currents that are shown in the hyperpolarization half. Bathed in Solution A; recorded with 3 M-KCl electrodes. The activations of the Ca2+-activated Na+ currents at -10 mv recorded with 4 M-CsCl electrodes are shown from the wild type (C and C') and dancer (D and D'). The solution is switched from Solution B (C and D) to Solution D that has Na+ (C' and D'). Stylonychia (Deitmer, 1983). For the following reasons, we think that both the transient and sustained Ca2+ currents are through only one type of Ca2+ channels in Paramecium. (1) Pawn mutations affect both the peak level and the sustained current, if not proportionally. (2) The decline of the peak inward current follows single exponential kinetics (Fig. 2). (3) We were able to detect only one time constant from the tail currents, regardless of the duration of the pulse. (4) A drug, W-7, suppresses both currents (Hennessey & Kung, 1984). Therefore, the three altered properties of the Ca2+ current can be attributed to a single mutational defect in one class of channels in dancer. The difference in the deactivation between the wild type and mutant reflects either a change in the voltage sensitivity or a change in the kinetics of the channel gating. Neither property is, however, likely to be related to the properties of the channel

12 408 R. D. HINRICHSEN AND Y. SAIMI inactivation since the difference is detected in the Ba2+ solution where there is little inactivation. If the mutation causes a shift in the voltage sensitivity of the Ca2+ channel, it may be as much as 10 mv toward hyperpolarization compared to that of the wild type. The differences in the inactivation properties (see Fig. 2C and D), on the other hand, are not explained by the shift in the voltage sensitivity of the Ca2+ channel in the mutant. Dancer may be a structural mutation. Simulations of Ca2+-dependent inactivation (Standen & Stanfield, 1982; Plant, Standen & Ward, 1983; Chad et al. 1984) take into account the efficiency of the Ca2+ removal from the submembrane compartment near the Ca2+ channels. If the mutant exhibits more rapid Ca2+ removal, elevation of Ca2+ should be smaller for a given Ca2+ current. The enhanced Ca2+-activated K+ and Na+ currents in the mutant, however, do not support this idea. Also, it is not easy to reconcile the observed change in the deactivation in the mutant with a more efficient removal of Ca2+. We favour, therefore, the view that the mutation causes an alteration in the Ca2+-channel structure or in its immediate environment, e.g. the annular lipid, which in turn affects two properties of the Ca2+ channel, the inactivation and deactivation. It has been predicted that a change in the binding constant of Ca2+ to the site(s) of the Ca2+ channel, if they interact directly, affects both the time course and the extent of the inactivation of the Ca2+ channel (Standen & Stanfield, 1982). Ca2+-channel inactivation is also voltage-dependent. Although the Ca2+-channel inactivation is mediated via internal Ca2+, its kinetics are also governed by membrane voltage level. In Paramecium, the 7 of the Ca2+-current inactivation is much larger when the voltage step is higher (Fig. 2). This appears not to result from the activation of any contaminating outward current which would only hasten the decay of the inward current, or from a lower internal Ca2+ concentration, which is inconsistent with larger peak and sustained Ca2+ currents at higher voltages. Similar behaviour is seen in Aplysia neurones (Chad et al. 1984), but the inactivation in the insect muscle appears to show little voltage dependency (Ashcroft & Stanfield, 1982). The dependency of the Ca2+-induced channel-state change on the membrane potential has also been reported. The activation of Ca2+-activated K+ current is facilitated by higher voltages (Gorman & Thomas, 1980; Moczydlowski & Latorre, 1983) through a voltage-driven concentration change of Ca2+ near the binding site(s) of the channel located part way along the voltage drop across the membrane. It appears that this interesting scheme does not apply to the Ca2+-channel inactivation described here since higher voltages should hamper rather than facilitate the process. It is possible that the voltage-dependent step is not the Ca2+ binding to the channel but the transition of the channel-ca2+ complex into the inactivated configuration. We thank Professor C. Kung for his continued encouragement and help in preparation of this manuscript. We also thank Dr R. Eckert for the comments and preprints, and Drs M. Gustin, T. Hennessey and E. Richard for their comments. Special appreciation goes to Dr J. Kung who corrected the text grammatically. This work was supported by grants from N.S.F. (BNS ) and N.I.H. (GM ). REFERENCES ASHCROFT, F. M. & STANFIELD, P. R. (1982). Calcium inactivation in skeletal muscle fibres of the stick insect, Carauiu8 morobus. J. Physiol. 330,

13 A Ca2+-CHANNEL MUTANT IN PARAMECIUM 409 BARISH, M. E. & THOMPSON, S. H. (1983). Calcium buffering and slow recovery kinetics of calcium-dependent outward current in molluscan neurones. J. Physiol. 337, BREHM, P. & ECKERT, R. (1978). Calcium entry leads to inactivation of calcium channel in Paramecium. Science, N. Y. 202, BREHM, P., ECKERT, R. & TILLOTSON, D. (1980). Calcium-mediated inactivation of calcium current in Paramecium. J. Physiol. 306, CHAD, J., ECKERT, R. & EWALD, D. (1984). Kinetics of calcium-dependent inactivation of calcium current in voltage-clamped neurones of Aplysia californica. J. Phyaiol. 347, DEITMER, J. (1983). Ca channels in the membrane of the hypotrich ciliate Stylonychia. In The Physiology of Excitable Celil, ed. GRINELL, A. & MOODY JR, W. J. New York: A. R. Liss, Inc. (in the Press). ECKERT, R. & BREHM, P. (1979). Ionic mechanisms of excitation in Paramecium. A. Rev. Biophy8. Bioeng. 8, ECKERT, R. & EWALD, D. (1983). Inactivation of calcium conductance characterized by tail current measurements in neurones of Aplysia californica. J. Phygiol. 345, ECKERT, R. & TILLOTSON, D. L. (1981). Calcium-mediated inactivation of the calcium conductance in caesium-loaded giant neurones of Aply8ia californica. J. Physiol. 314, FUKUDA, J., FISCHBACH, G. D. & SMITH JR, T. G. (1976). A voltage clamp study of the sodium, calcium and chloride spikes of chick skeletal muscle cells grown in tissue culture. Devl Biol. 49, GORMAN, A. L. F. & THOMAS, M. V. (1980). Potassium conductance and internal calcium accumulation in a molluscan neurone. J. Physiol. 308, HAGIWARA, S. & BYERLY, L. (1981). Calcium channel. A. Rev. Neuro8ci. 4, HAGIWARA, S., OZAWA, S. & SAND, 0. (1975). Voltage clamp analysis of two inward current mechanisms in the egg cell membrane of a starfish. J. gen. Physiol. 65, HALL, J. C. (1983). Genetic analysis of behavior in insects. In Comprehenwive Insect Physiology, Biochemistry, and Pharmacology, ed. KERKUT, J. A. & GILBERT, L. I. Oxford: Pergamon Press (in the Press). HENNESSEY, T. M. & KUNG, C. (1984). An anticalmodulin drug, W-7, inhibits the voltage-dependent calcium current in Paramecium caudatum. J. exp. Biol. (in the Press). KOSTYUK, P. G. (1980). Calcium ionic channels in electrically excitable membrane. Neuroscience 5, KUNG, C. & ECKERT, R. (1972). Genetic modification of electric properties in an excitable membrane. Proc. natn. Acad. Sci. U.S.A. 69, KUNG, C. & SAIMI, Y. (1982). The physiological basis of taxes in Paramecium. A. Rev. Physiol. 44, MOCZYDLOWSKI, E. & LATORRE, R. (1983). Gating kinetics of Ca2+-activated K+ channels from rat muscle incorporated into planar lipid bilayers: evidence for two voltage-dependent Ca2+ binding reactions. J. gen. Phy8iol. 83, NAITOH, Y., ECKERT, R. & FRIEDMAN, K. (1972). A regenerative calcium response in Paramecium. J. exp. Biol. 56, OKAMOTO, H., TAKAHASHI, K. & YOSHII, M. (1976). Two components of the calcium current in the egg cell membrane of the tunicate. J. Physiol. 255, PLANT, T. D. & STANDEN, N. B. (1981). Calcium current inactivation in identified neurones of Helix aspersa. J. Physiol. 321, PLANT, T. D., STANDEN, N. B. & WARD, T. A. (1983). The effects of injection of calcium ions and calcium chelators on calcium channel inactivation in Helix neurones. J. Physiol. 334, REUTER, H. (1983). Calcium channel modulation by neurotransmitters, enzymes and drugs. Nature, Lond. 301, SAIMI, Y., HINRICHSEN, R. D., FORTE, M. & KUNG, C. (1983). Mutant analysis shows that the Ca2+ induced K+ current shuts off one type of excitation in Paramecium. Proc. nain. Acad. Sci. U.S.A. 80, SAIMI, Y. & KUNG, C. (1980). A Ca-induced Na-current in Paramecium. J. exp. Biol. 88, SAIMI, Y. & KUNG, C. (1982). Are ions involved in the gating of calcium channels? Science, N. Y. 218, SALKOFF, L. (1983). Drosophila mutants reveal two components of fast outward current. Nature, Lond. 302,

14 410 R. D. HINRICHSEN AND Y. SAIMI SATOW, Y. & KUNG, C. (1980). Ca-induced K+-outward current in Paramecium tetraurelia. J. exp. Biol. 88, STANDEN, N. B. & STANFIELD, P. R. (1982). A binding-site model for calcium channel inactivation that depends on calcium entry. Proc. R. Soc. B 217, TAKAHASHI, M. & NAITOH, Y. (1978). Behavioural mutants of Paramecium caudatum with defective membrane electrogenesis. Nature, Lond. 271, TILLOTSON, D. & HORN, R. (1978). Inactivation without facilitation of calcium conductance in caesium-loaded neurones of Aply8ia. Nature, Lond. 273, Wu, C-F., GANETZKY, B., HAUGLAND, F. N. & Liu, A.-X. (1983). Potassium current in Drosophila: different components affected by mutation of two genes. Science, N. Y. 220,