Membrane Currents of Pawn Mutants of the Pwa Group in Paramecium Tetraurelia

Ca action potentials have been reported in many excitable membranes (Fatt & Katz, 1953; Hagiwara, 1973). The Ca²⁺ entering during the action potential has a direct effect on a variety of cellular functions. In Paramecium, the Ca²⁺ reverses the direction of ciliary beat. Pawn mutants have a reduced Ca conductance and have a reduced tendency to swim backward when properly stimulated. The typical (non-leaky) pawn has no Ca conductance and does not swim backward at all.

In Paramecium the calcium current shows inactivation (Oertel, Schein & Kung, 1977; Brehm & Eckert, 1978) as in several other systems (e.g. Mounier & Vassort, 1975). Brehm & Eckert (1978) showed that the inactivation is induced by internal Ca²⁺, not voltage as for the sodium channel. Internal calcium also opens potassium channels in several systems (Meech, 1976) and this has been demonstrated in Paramecium by injection of Ca or EGTA (Satow, 1978a; Brehm, Dunlap & Eckert, 1978). However, Oertel et al. (1977) have provided evidence that Ca-induced K conductance is not involved in excitation events occurring within tens of milliseconds after depolarization in Paramecium. The role of this conductance in excitation has also been questioned by Eckert & Brehm (1979).

In this paper we describe various currents observed in the Paramecium membrane under voltage clamp, including the calcium-induced K⁺ current. We examined two mutants of the pwA complementation group (see Chang et al. 1974) – one a typical pawn, the other a leaky pawn – and compared them with the wild type to see which characteristics of the Ca channels are altered by the mutations. The pawn was found to have normal membrane characteristics except for its lack of Ca conductance, and has since been used as null control for a variety of studies (see Kung, 1979, for a review). An abstract of part of this work has been reported elsewhere (Satow, 1979).

Cells of Paramecium tetraurelia were cultured at room temperature (22 ± 1 °C) in Cerophyl medium inoculated with Enterobacter aerogenes 20 h before use (Sonne-born, 1970). Only robust cells in log-phase growth were used. There was no systematic difference in the size of the paramecia used from different strains. The two mutant strains were allelic variants of the pwA complementation group. They were a typical pawn, stock d4–500 (abbreviated as pwA₅₀₀ herein), and a leaky pawn, stock d4–132 (abbreviated as pwA₁₃₂) (Chang et al. 1974). The wild type used is stock 51s from which the mutants were derived.

A bath solution containing 1 mM-Ca (OH)₂, 0·5 mM-CaCl₂ and 1 mM-citric acid adjusted to pH 7·2 with approximately 1·3 mM-Tris is termed the ‘Ca solution’ throughout this paper. The free Ca²⁺ concentration in this solution is calculated to be 0·91 mM. Addition of 4 mM-KCl to the Ca solution yielded the ‘Ca-K solution’. This solution was used in the investigation of currents which may be carried by K+. One, 4, 8 or 16 mM-KCl was added to the Ca solution for the experiments in Fig. 6.

The methods used in intracellular recording were similar to those given by Naitoh & Eckert (1972). All experiments were performed using a voltage clamp, as described by Satow & Kung (1979). The membranes were first held at the resting levels and were then depolarized or hyperpolarized in steps. This procedure was followed to maximize the transient inward currents and to keep the membranes close to their normal states. Since the resting potentials of different strains were very similar in a given solution, the holding potentials were essentially the same for different strains (Table 1). All experiments were performed at room temperature (22 ± 1 °C).

When the wild-type membrane is first held at the resting potential level (− 31 mV in the Ca solution and − 36 mV in the Ca-K solution) and then subjected to a step depolarization, a transient inward current is observed as shown in Fig. 1 (left). The inward current reaches its peak within 3 ms and subsides rapidly afterward. An outward current follows the inward transient. For reasons not understood, the inward transient is smaller when K⁺ is present in the solution (Fig. 1, lower left). Descriptions of the inward transients of wild-type P. tetraurelia have been given by Oertel et al. (1977) and by Satow & Kung (1979).

If the membrane of the leaky pawn, pwA₁₃₂, is subjected to step depolarizations, inward currents smaller than those of wild type can be observed when the step depolarization is from the holding level of about −30 mV to beyond − 10 mV (Fig. 1, centre). The maximal peak inward current (I_max), measured directly from the zero current level without adjusting for leakage or rectifying currents (see below), is observed when the membrane potential is stepped up to +3 mV in cells bathed in the Ca solution. Still smaller inward current is seen when the cell is bathed in the Ca-K solution (Fig. 1, lower centre). The I_max of pwA₁₃₂ is about 25% of that found in the wild type regardless of the bath solutions (Table 1).

As shown in Fig. 1 (right), no transient inward current is seen in the typical pawn, pwA₅₀₀, beyond an early downward oscillation associated with the step depolarizations. Outward currents are recorded from this membrane at a time (2·2 ms) when maximal inward current is seen in the wild type.

The time course of the transient inward current can be characterized by the time (^max) when the inward current is maximal. The T_max’s of wild type and pwA₁₃₂ are both about 2 ms in the Ca solution and about 23 ms in the Ca-K solution (Fig. 1). The mean and standard deviations of T_max’s are listed in Table 1. There is clearly no significant difference between the wild type and pwA₁₃₂ in their T_max’s.

To compare the voltage sensitivity of the Ca conductance in pwA₁₃₂ with that in the wild type, V–I_peak relations in the Ca and the Ca-K solutions are plotted in Fig. 2. As judged by the voltage at which the I_max is observed, V_max, the voltage sensitivity of the pttb4₁₈₂ membrane is similar to that of the wild type (Fig. 2; Table 1).

Assuming that pwA₁₃₂ completely lacks the voltage-sensitive Ca conductance but is otherwise normal, the total Ca²⁺ current (I_ca) of the wild type and pwA₅₀₀ can be measured by subtracting the outward current in pwA₁₃₂ from the inward currents of wild type and puA_iai at peak time (cf. Oertel et al. 1977). Averaged I_ca of wild type and pwA₁₃₂ in the Ca solution are shown in Fig. 3 A. At +5 mV steps, I_ca is 7·8 nA in the wild type and 3-5 nA in pwA₁₃₂. Chord conductance (G_Ca) calculated using an estimated E_Ca + 115 mV (from an estimated internal concentration of free Ca⁺ of 10 ⁻⁷ M, see Eckert, Naitoh & Machemer, 1976), is 70 nmho/cell in the wild type and 32 nmho/cell in pwA₁₃₂. The conductance voltage plot is another estimate of the voltage sensitivity of the channels involved (Fig. 3 B). The the voltage at which G_Ca is half maximal, has been obtained from the conductance-voltage plot for each specimen using the I_ca values obtained by subtracting the same base line derived from the mean values of outward currents of pwA₅₀₀. The mean is − 8·0 mV (S.D. 3·1, n = 5) for the wild type and −7·2 mV (S.D. 3·9, n = 7) for pwA₁₃₂. These figures are not significantly different.

Anomalous rectification has been observed in Paramecium (Naitoh & Eckert, 1968a ; Satow & Kung, 1976a; Schein, Bennet & Katz, 1976; Oertel, Schein & Kung, 1978). The current is carried by K⁺ (Oertel et al. 1978) and so was measured in the Ca-K solution (4 mM-K). The sustained inward currents induced by step hyperpolarizations in the wild type are very similar to those in pwA₅₀₀ (Fig. 4, upper panel). In this study, the earlier portion of the inward currents is not analysed. The steady-state inward currents at 1 s after the beginning of hyperpolarizing steps (Fig. 4, middle panels with slow sweep speed) are plotted against the step voltages (Fig. 4, lower). The voltage dependencies of the anomalous rectification channels of the wild type and pwA₅₀₀ are indistinguishable. There is no statistically significant intergroup difference between the inward currents recorded from the two strains at any of the hyperpolarizing steps.

We compared the channel properties in the wild type and ptoA₅₀₀ as revealed by a technique similar to that of Oertel et al. (1978). We first hyperpolarized the membrane by a step to activate the anomalous rectifying channels, held the membrane at this level for 720 ms, and then analysed the tail currents observed when the potential was stepped to other levels. The reversal potentials of the tail currents in the Ca-K solution are −54·8 mV (S.D. 5·6, n = 8) in the wild type and −55·8 mV (S.D. 6·3, n = 10) in pwA₅₀₀. Thus, pwA₅₀₀ appears to have normal hyperpolarization-activated channels (see Discussion on E_K).

As expected, the anomalous rectification is different in the K⁺-free solution (the Ca solution) (Fig. 5). In the range from − 30 to − 50 mV, the response is ohmic. This is best attributed to the voltage-independent leakage conductance. From −50 to − 90 mV, further hyperpolarization does not increase the inward current. This is most likely due to the efflux of K⁺ through the anomalous-rectification channel resulting in outward currents that compensate for the inward leakage current. These findings are also consistent with those derived from previous current clamp experiments (Satow & Kung, 1977). When the step hyperpolarization is beyond −90 mV, the response is again ohmic, but with a large conductance. The mechanism of this inward rectification is not known. As in the Ca-K solution, there was no difference between the inward current in the wild type and that in pwA₅₀₀. This observation reinforces the notion that the mutational effect on the Ca channels does not extend to the channels responsible for the leakage current and the anomalous rectifying current.

The outward currents induced by step depolarizations, commonly known as delayed rectifying currents, were compared between the three strains to test whether the pwA mutations affect the channels responsible for these currents. The kinetics of these currents within the first hundred ms after the depolarization steps are complex and are dependent strongly on the amount of depolarization. As shown below, these early outward currents are not significantly different in the three strains, although variations within strains are large, as was observed by Oertel et al. (1977). However, the late currents are different.

The outward current induced by membrane depolarizations is considered to be carried largely by K⁺ in Paramecium (Naitoh & Eckert, 1974; Satow & Kung, 1976a) as in other excitable systems. To investigate the K⁺-dependence of the outward currents, the membrane was subjected to a prolonged depolarization (720 ms), and the potential was then stepped to various lower levels (Fig. 6) at a range of K⁺ concentrations. Tail currents may be inward or outward depending on the level of the second step. The reversal potentials (P”_R) of the tail currents induced in this manner in the Ca-K solution (4 mM-K) are − 49·5 ± 12·1 mV in the wild type and − 42·5 ± 6·4 mV pwA₅₀₀ (see Discussion). A plot of the V_R versus the K⁺ concentration shows a slope of 50 mV for a tenfold change of concentration in both strains (Fig. 6). Thus, the outward current after a 720 ms depolarization is mostly carried by K⁺. Similar examination of the tail currents after 30 ms of depolarization at the voltage of o mV shows that the V_R is about − 20 mV in the Ca-K solution (4 mM-K) and the V_R is dependent on the amount of depolarization in both the wild type and pwA₅₀₀. Thus, the early outward current is not entirely dependent on K⁺.

The outward current associated with very long and constant depolarization (seconds) are steady with depolarizations up to about the −10 mV level (Fig. 7). Larger depolarizations, however, produce currents which do not stay steady. In the wild-type membrane, the currents develop towards plateaus with half times measured in hundreds of ms (Fig. 7, left). These added currents can be as large as 4–5 nA in some cases. In the membrane of pwA₅₀₀, the currents decline rather than rise, before assuming a steady value (Fig. 7, right).

Fig. 8 is a plot of these plateau levels at 10 s of the depolarizations. Although the mean values of the plateau currents of the wild type are consistently higher than those of pwA₅₀₀ in all cases, it is only with depolarizations to levels above o mV, that the wild-type current is significantly larger than that of pwA₅₀₀ (P < 0·05 at o mV ; P < 0·01 at +5 and +10 mV). We propose that this reduction of slow outward current in pwA₅₀₀ is due to the lack of a Ca-induced K⁺ current (see Discussion). Assuming that pwA₅₀₀ suffers a complete loss of this current, the kinetics of the Ca − induced K current in the wild type can be estimated by a subtraction of the trajectory of the outward current of pwA₅₀₀ from that of the wild type, as shown in Fig. 9. In this example, the proposed Ca-induced K current rises slowly with a half-time of approximately 0 ·25 s. This rise is two orders of magnitude slower than that of the voltage-sensitive K⁺ current responsible for the major portion of the delayed rectification.

The pwA gene product may regulate the production of Ca channels, or it may code for the channel structures or define the immediate environment of the channels. Thus, the reduction of the Ca²⁺ inward transient in pwA₁₃₂ and the total lack of this current in pwA₅₀₀ may be due to a reduction in the number of normal Ca channels, or due to changes in the channel properties. Based largely on the broad range of leakiness and the lack of effects on the anomalous rectification channels of the pwA mutations, Schein et al. (1976) speculated that these mutations affect the voltage sensitivity of the Ca channels.

We have presented here a thorough study of the Ca-channel characteristics using the voltage clamp. The peak time, is an indicator of the kinetics of activation of available channels. As shown in Fig. 1 and Table 1, the T_max of pwA₁₃₂ is not signicantly different from that of the wild type in Ca-or Ca-K solutions. Moreover, the T_max’s the two strains are also the same in a Ca-Ba solution (Y. Satow & C. Kung, unpublished observations), although Ba²⁺ slows the activation significantly (Satow, 1978b). The voltage sensitivity of the Ca channels can be estimated by the V_max on the V–I_peak curves or by the on the G_Ca–V plots. Data presented in Table 1 and Figs. 2 and 3 show that V_max and of wild type and pwA₁₃₂ are not significantly different. Within the resolution of our present experimental techniques, we found no significant change in either the activation kinetics or the voltage sensitivity of the Ca channels remaining in the leaky mutant pwA₁₃₂ although the Ca conductance is clearly reduced to less than half. While one can contrive more complicated alternatives, the simplest explanation would be a reduction in the number of functional Ca channels by the pwA₁₃₂ mutation and a complete lack of them in pwA₅₀₀. This could mean that the pwA gene product is required for the production, assembly or maintenance of normal Ca channels but does not affect the functions of assembled channels. This role of the pwA gene product fits well the observation that the phenotypic changes after abrupt temperature changes of the leaky mutant require a growth period (Satow, Chang & Kung, 1974). Berger (1976) showed that a wild-type cytoplasmic factor confers excitability to pawn A. This factor may be unassembled channels or a diffusible cofactor coded by the pwA locus required for Ca-channel function.

We have further investigated the possibility that the pwA product may affect other channels besides the voltage-sensitive Ca channel. The resting permeabilities of the mutants are normal since the resting membrane potential and the resting membrane resistance of pwA₁₃₂ and pwA₅₀₀ are the same as those of the wild type (Satow & Kung, 1976b). The inward currents through the anomalous-rectification channels are also similar in the wild type and the mutants (Figs. 4–6). Thus, there is no evidence for the pwA mutations affecting any conductances within the time scale of the transient inward current other than the Ca conductance. Significant differences in the slow outward current developing in seconds after the onset of depolarizations (Figs. 8, 9) will be discussed below.

The pwA mutant series is important in our future research since it provides the proper null control for experiments concerning excitation and excitation-related effects. pwA₅₀₀ will be especially useful since it is a complete pawn mutant and has no detectable effects on other channel functions. pwB mutations, on the other hand, appear to be complicated since they affect the anomalous-rectification channel (Schein et al., 1976) and non-K leakage current (Satow, 1979), they have an additional phenotype of being K⁺ resistant (C. L. Shusterman & C. Kung, unpublished observations; Shusterman, Thiede & Kung, 1978) and show phenotypic reversion (Eckert & Brehm, 1979; T. Hennessey & C. Kung, unpublished observations), although they also have reduced or deleted Ca-channel activities.

It has been reported in molluscan neurones that the delayed outward current consists of two types of K⁺ current ; one voltage dependent and the other Ca dependent (Thompson, Aldrich & Getting, 1978). As described in the Introduction, there is evidence for both the voltage-dependent K conductance and the Ca-dependent K conductance in Paramecium (Satow, 1978a; Brehm et al. 1978). The latter is supposed to be induced by a high concentration of internal Ca²⁺ which enters during excitation. Therefore, it would be reasonable to expect that the Ca-induced K⁺ current is small or missing in the pawn mutants. Since pwA₅₀₀ shows no transient inward current upon membrane depolarizations (Fig. 1, Table 1), we attribute the outward currents of pwA₅₀₀ to the voltage-sensitive channel and the leakage channel. While these currents rise to levels comparable to or slightly lower than those of the wild type within some 100 ms upon moderate to strong depolarizations, they then decay rather than stay steady or increase, as in the wild type (Figs. 7, 9). This decay describes the inactivation of the voltage-sensitive K conductance, following the above argument. It is probably a secondary effect of the mutational loss of Ca channels and not an independent effect of the mutation on the delayed rectification channel. The difference between the outward currents in the wild type and in pwA₅₀₀ in response to the same depolarizations, as shown in Fig. 9C, is now assigned to the Ca-induced outward currents. These currents are probably carried largely by K⁺ as deduced from the experiments shown in Fig. 6 where the ‘tails’ trailing the outward currents have a reversal potential close to E_k, although the Ca-induced currents are only parts of the total currents. (The tail currents from pwA₅₀₀ similarly treated also have a reversal potential close to E_K, even though we believe that this mutant is lacking the Ca-induced outward current. This is probably because the currents remaining in this mutant are K+ currents, although they are voltage-dependent instead of Ca-induced.) The reversal potentials of tail currents in 4 mM-K solution after 25 ms hyperpolarization from a holding potential of − 20 to − 25 mV were demonstrated by Oertel et al. (1978) to be close to E_K in P. tetraurelia. The reversal potentials of the tail currents after 720 ms depolarization from the holding potential of − 30 to − 35 mV (Fig. 6), as well as those associated with anomalous rectification (Fig. 4) in our studies, are more negative than the E_K established by Oertel et al. (1978). The reason for this departure is not fully understood although it may be related to the fact that the resting potential of the paramecia used in this study is consistently higher (− 30 to − 35 mV in the Ca-K solution, Table 1) than that observed by Oertel et al. (1977, 1978) (− 20 to − 25 mV in a similar solution).

The Ca-induced K⁺ current, isolated from the voltage-induced K+ current and the leakage current, has kinetics of hundreds of milliseconds to seconds (Fig. 9, see Y. Satow & C. Kung, in preparation, for details). Such slow kinetics are in contrast with those of the voltage-induced events in time scales of milliseconds or tens of milliseconds. This large difference in time scales explains why the Ca-induced K⁺ current is largely not involved, in early events in excitation. The slowly rising and falling Ca-induced K⁺ current may be related to electrical events and behaviour of similar time scale such as the prolonged plateau depolarizations (Naitoh & Eckert, 1968b; Satow & Kung, 1974) and the continuous ciliary reversal (Jennings, 1906; Dryl, 1973) in Paramecium.

We thank Drs D. Oertel, S. Schein, Y. Saimi and Ms C. L. Shusterman for their critical reading of the manuscript. This work was supported by NSF grant BNS77-20440.