Determination of Pure Voltage-Dependent Ca²⁺ Current In Paramecium Caudatum and Its Inhibition By Divalent Cations

Calcium ions have several important functions in excitation and motor control of ciliated protozoa (for review see e.g. Kung & Saimi, 1982; Naitoh, 1982). One such function is to carry a regenerative inward current across the ciliary membrane of ciliates and to cause reorientation and ultimately behavioural responses of the organisms. Ca²⁺ also modulates the conductance of the electrically excitable channels at the outer as well as at the inner membrane surface (Hildebrand, 1975, 1978; Hildebrand & Dryl, 1976; Brehm & Eckert, 1978b). Several experiments have shown a strong antagonism between externally applied K⁺ and Ca²⁺ in controlling excitation (Naitoh, 1968; Naitoh & Eckert, 1968a) which led to the hypothesis that membranebound calcium may determine excitability (Hildebrand, 1975).

The molecular mechanism by which Ca²⁺ controls the membrane conductance is still unknown. The influence on the surface potential has been discussed (Hook, 1979; Hook & Hildebrand, 1980; Satow & Kung, 1979; Eckert & Brehm, 1979). Recent experiments indicate a more direct participation of Ca²⁺ and other divalent cations in the gating process of voltage-sensitive Ca channels (Hagiwara & Byerly, 1981; Saimi & Kung, 1982).

Experiments which we present here support the hypothesis that other divalent cations compete with Ca²⁺ for negative binding sites at the external membrane surface and thus determine Ca²⁺ conductance in Paramecium. From measurements of Ca²⁺ inward currents which were corrected for concomitant K⁺ fluxes we can derive apparent affinity constants for different cations including Ca²⁺.

Paramecium caudatum, obtained from the Nencki Institute of Experimental Biology in Warsaw, were cultured in a hay infusion at room temperature. Two weeks after inoculation the cells were harvested by centrifugation at 170 × g for 2 min. The cells were then washed and equilibrated in the ‘standard Ca solution’ (1 mmol l⁻¹ CaCl₂, 1 mmol l⁻¹ KC1, 1 mmol l⁻¹ Tris/HCl, pH7·2–7·4). CNR-mutants (strain 16 A 604) of P. caudatum were kindly provided by Professor Dr Y. Naitoh, Ibaraki, Japan.

The ionic concentrations of the different test solutions are listed in Table 1. All chemicals (Merck, Darmstadt, F.R.G.) were of analytical grade.

In contrast to the conventional technique (cf. Naitoh & Eckert, 1972), the cells were impaled from above by straight electrodes under observation from below through an inverted microscope (Standard UPL, Carl Zeiss, Oberkochen, F.R.G.). The membrane potential was measured with an intracellular glass microelectrode filled with Imoll⁻¹ KC1 (resistance 40–60MΩ), using an Ag/AgCl-pellet as the indifferent electrode, which was connected to the bath through an agar-KCl bridge (2% agar, 3 mol I⁻¹ KC1). A second microelectrode filled with 2 mol I⁻¹ K-citrate (resistance 20–30 MΩ) was used for current injection (DePeyer & Deitmer, 1980). The tip diameter of the electrodes was about 0·5 μm.

In the feedback circuit of the voltage clamp, a high accuracy unity gain differential amplifier (BB 3627), a PID-controller (JDS 15, Baur Electronic, Meerbusch, F.R.G.) and a high voltage amplifier (BB 3582) were used. The open loop gain of this circuit was 2000× and the maximum output voltage was ± 110 V. The membrane current was monitored by a current-to-voltage converter connected to the indifferent electrode. A fast settling FET operational amplifier (BB 3550) was used to maintain virtual ground at high frequency changes of membrane currents.

Capacitive coupling between both microelectrodes outside the bath was reduced by placing a 0·5 mm steel plate between them. The potential of this plate was held on the actual potential of the voltage electrode (‘driven-shield’). Capacitive coupling in the bath was reduced by coating the voltage electrode to about 50 μm from the tip with commercially available nail enamel. When the capacitive coupling had been reduced the time to set up a certain membrane potential was less than 250 μs. Transient currents due to membrane capacitance were complete after about 700 μs.

To achieve rapid changes in membrane potential critical damping of the clamp circuit is essential (Katz & Schwartz, 1974). This, however, may lead to an undamped oscillation that ruins the cell. To protect the cell, a control mechanism was designed that immediately interrupted the feedback circuit in case of a high-frequency oscillation. In principle, this arrangement consists of a frequency-to-voltage converter in combination with a comparator circuit. The highest frequency of membrane currents that is to be tolerated can be adjusted by the voltage applied to the comparator.

At the beginning of an experiment the input resistance of the cell was measured by injecting a hyperpolarizing current of 10⁻¹⁰A (DePeyer & Deitmer, 1980). If the resistance in standard Ca solution was less than 50 MΩ or decreased by more than 25% during the experiment, the cell was disregarded.

Current-voltage relationships were determined by applying a series of voltage steps of different amplitudes. Between the voltage steps of one series the membrane was clamped to the resting potential. In every experiment one series of measurements in standard Ca solution was taken before changing to the test solution. If possible, a third series was added after changing back to the Ca solution. Solutions were changed by continuous perfusion of the experimental chamber. Perfusing 25 times the chamber volume of 0·4ml led to an exchange of about 99%.

To separate Ca²⁺ and K⁺ currents, the voltage-dependent Ca²⁺ conductance was temporarily inactivated by a preceding depolarization (Brehm & Eckert, 1978a). The amplitude of this prepulse was +50 mV relative to the resting level. Its duration and the distance to the test pulse were both 40 ms. To obtain the pure Ca²⁺ current, the K⁺ current measured under these conditions was subtracted from the overall current measured in the absence of a prepulse. Subtraction was carried out by use of a computer programme.

All experiments were carried out at 15°C. The temperature was kept constant with an accuracy of ± 0·2°C by means of a Peltier feedback device and was measured in the bathing medium at a distance of about 2 mm from the cells by use of an electronic temperature sensor (AD 590 K). The cells were adapted to 15°C for about 3 h prior to the experiments.

Membrane currents that follow a step-like depolarization of the membrane are shown in Fig. 1A. Increasing the amplitude of the voltage step first leads to a rapid Ca²⁺ inward current that reaches maximum at a depolarization of about 35 mV relative to the resting level. In addition, a delayed K⁺ outward current appears at about 25 mV and increases rapidly with further depolarization. Together with a faster activation of the K⁺ current at higher voltages this leads to an increasing overlap of both currents (cf. Eckert & Brehm, 1979). The early current, which is actually the sum of the Ca²⁺ inward current and the K⁺ outward current, reverses sign at a depolarization of about 65 mV.

To separate both currents the outward current was measured after the voltagedependent Ca²⁺ conductance had been inactivated by a depolarizing prepulse (depolarization by 50 mV for 40 ms; the delay between prepulse and test pulse was 40 ms). With a prepulse of this kind it is possible to reduce the voltage-dependent Ca²⁺ current in Paramecium to about 5% of its original value (Brehm & Eckert, 1978a; Brehm, Eckert & Tillotson, 1980). Fig. 1B shows membrane currents from the same cell for the same depolarizing steps as in Fig. 1A, but measured after a prepulse had been applied. There is no detectable inward current under these conditions. It is not expected that the extent of Ca²⁺ channel inactivation will change with the voltage of the second step, at which the currents are measured. Under this assumption the currents shown in Fig, 1B can be regarded as almost pure K⁺ currents.

At the same depolarization the amplitude of the late outward current is the same in both series (Fig. 1). Moreover, the kinetics of the K⁺ currents remain unaffected by the preceding pulse. This could be proved in experiments with the mutant strain CNR, which is devoid of a voltage-dependent Ca²⁺ conductance. The pure K⁺ currents of the mutant were not changed by the preceding pulse (data not shown).

To obtain the pure Ca²⁺ currents the traces in Fig. 1B were subtracted from those in Fig. 1A by use of a computer programme. Up to a depolarization of 105 mV distinct inward currents appear (Fig. 2), while the uncorrected early current already reverses sign at about 65 mV depolarization (Fig. 1A).

At depolarizing steps between 60 and 100 mV a delayed outwardly-directed component can be detected in the calculated current traces. This small outward current was found in each of 10 prepulse experiments with wild type cells, but was never found in experiments with the mutant. Thus, the outward component seems to depend in some way on the voltage-dependent Ca²⁺ inward current. It should be noted, however, that its maximum occurs at higher voltages than that of the Ca²⁺ current.

Small outwardly or inwardly directed currents sometimes occurred in the voltage range below 60 mV (e.g. at 34 mV, Fig. 2). These currents might be due to fluctuations in the amplitude of Ca²⁺ and/or K⁺ currents which we observed in some cells.

Fig. 3 summarizes the voltage dependence of the K⁺ current at the end of the voltage pulse, the early current and the purified Ca²⁺ current. Due to the increasing ratio I_K/I_Ca and small fluctuations of the K⁺ current at higher voltages, the theoretical Ca²⁺ equilibrium potential of about 150 mV relative to the resting level could not be proved experimentally.

The Ca²⁺ currents of five cells as a function of the voltage step are shown in Fig. 4. Although there is a marked scattering of the values at higher depolarizations, the currents obviously do not reverse sign in the voltage range tested.

To improve the resolution of the subtraction procedure at higher voltages, 5 mmol l⁻¹ tetraethylammonium (TEA) was applied in the standard Ca²⁺ solution to reduce the K⁺ currents (Saimi & Kung, 1982). The K⁺ current decreased to about 40%, but in addition, the resting resistance of the membrane increased to approximately 250%. As a consequence, the time to set up a certain membrane potential was much prolonged, but rapid changes in membrane potential are essential for the subtraction procedure.

When part of the external Ca²⁺ was replaced by Mg²⁺, Co²⁺ or Mn²⁺ the amplitude of the inward current was decreased (data not shown). For a given degree of substitution the magnitude of the decrease depended on which cation was tested. Thus, the effect could not simply be explained by dilution of Ca²⁺ ions as the charge carriers. We assumed that Mg²⁺, Co²⁺ and Mn²⁺ inhibit the Ca²⁺ current with different efficiencies. This inhibition was tested quantitatively at constant Ca²⁺ concentration by adding either one of these ions or Ni²⁺ at different concentrations to the standard Ca²⁺ solution (see Table 1). The maximum Ca²⁺ current at a certain concentration of the tested cation was measured and compared to that of the same cell in standard Ca²⁺ solution. To keep changes in the total concentration of divalent cations as small as possible the lowest concentrations were used that showed clear effects. In addition, the concentration at which a cation could be tested was limited by its toxicity.

From the different slopes the following ratio can be obtained: K’_NI: k’_co: k’_Mn: k’_Mg= 1: 3: 5: 6·5, that is, the affinity for Ni²⁺, for example, is five times that for Mn²⁺. As the term 1 + [Ca²⁺]/k_Ca is constant throughout the experiments, the same ratio holds for the AM values. The absolute dissociation constants, however, can only be calculated if k_Ca is known.

To obtain k_Ca, the dependence of the maximum inward current on Ca²⁺ concentration was measured (Table 2). From these data, k_Ca was determined graphically by use of a Lineweaver-Burk plot. With a confidence of 95%, k_Ca lies between 40 and 130 μlnoll⁻¹.

The relative dissociation constants of the inhibitory cations can now be brought in relation to k_Ca. The uncertainty of k_Ca hardly influences its relative value, which is 4·3 or 4·7 referred to K_NI, depending on whether a k_Ca of 40 or 130 μ mol I ^{− 1} is assumed. This is due to the fact that the absolute dissociation constants depend directly on k_Ca and change with its value to approximately the same extent. The complete ratio of dissociation constants is: K_NI: k_co: k_Ca: k_Mn: k_Mg = 1: 3:4·3 – 4·7: 5:6·5.

The following methods have so far been tried to obtain pure voltage-dependent inward currents in Paramecium. (1) K⁺ currents of a mutant devoid of a voltagedependent Ca²⁺ conductance (‘pawn’ mutant of P. tetraurelia) were subtracted from the currents of a wild type cell (Oertel, Schein & Kung, 1977; Satow, 1982). This method was used up to depolarizations of 40 mV. We found that the activation of K⁺ currents was faster in the CNR mutant than in our wild type cells. Thus, correction of inward currents by use of the mutant was not possible at voltages which elicited distinct outward currents. (2) K⁺ currents were inhibited by TEA added to the outer medium (Saimi & Kung, 1982) or injected together with Cs⁺ into the cell (Eckert & Brehm, 1979; Brehm et al. 1980). However, a K⁺ current of about 25% remained. (3) More than 95% of the K⁺ current can be blocked by external TEA when a CsCl-filled electrode is used (Hinrichsen & Saimi, 1984; Hennessey & Kung, 1984). The remaining K⁺ outward current, however, seems to affect the measurement of pure Ca²⁺ currents at higher depolarizations, because the reported inward currents show a tendency to cross the voltage axis at 60 – 70 mV below the expected E_Ca.

The inactivation of Ca²⁺ conductance by a depolarizing prepulse (Brehm & Eckert, 1978a) combined with the subtraction procedure seems to be a suitable method to isolate Ca²⁺ inward currents up to 100 mV depolarization. All measurements can be taken from one cell and neither the inner nor the outer medium of the cell has to be changed.

This method is applicable only if amplitude and kinetics of the K⁺ current remain unchanged by the preceding pulse. The experiments with the CNR mutant show that this is true. In addition, the separation of currents will only be complete if the Ca channel is fully inactivated by the prepulse and if this inactivation is independent of the test pulse voltage. At a depolarization of 35 mV, which elicits the maximal inward current in the absence of a conditioning pulse, the Ca²⁺ current is completely inhibited by the preceding pulse (see Fig. 1). However, we do not know whether inactivation is as effective at higher voltages of the test pulse. If not, this would lead to an underestimate of Ca²⁺ currents by the subtraction procedure.

What we call ‘K⁺ current’ in this paper may include other membrane currents in addition to the voltage-dependent K⁺ outward current, such as for example a prolonged, non-inactivating Ca²⁺ current (cf. Eckert & Brehm, 1979; Hinrichsen, & Saimi, 1984). However, only those components of the total membrane current that are inactivated by the preceding depolarization will be revealed by the prepulse technique, and no membrane current that does not inactivate will affect our calculation of inactivating Ca²⁺ currents. Thus, it is not surprising that we do not find the steady-state Ca²⁺ inward current described by Eckert & Brehm (1979) and Hinrichsen & Saimi (1984).

Since the outward component is not found in experiments with the CNR mutant, it seems to be related to the Ca²⁺ inward current and one may ask whether it is a Ca²⁺-activated K⁺ current. Certainly it is not due to any decrease of Ca²⁺ current inactivation at higher test pulse voltages, because this would only decrease the calculated inward currents, but could by no means produce an outward current.

Depolarization of the membrane is known to increase Ca²⁺-activated K⁺ currents (Gorman & Thomas, 1980; Barrett, Magleby & Pallotta, 1982). If the outward component in our experiments depends on both Ca²⁺ concentration and depolarization, this could explain its occurrence at voltages higher than necessary for the maximal Ca²⁺ current (by its voltage-dependence) and its decrease with further approach to the Ca²⁺ equilibrium potential (by its Ca²⁺-dependence). Thus, it seems unlikely in our opinion that the outward component occurs at depolarizations below 60 mV. One could argue that it might become undetectable because it is cancelled by an equally large inward current. A possible candidate for that would be the steady-state Ca²⁺ inward current (Eckert & Brehm, 1979; Hinrichsen & Saimi, 1984). This, however, is highly improbable because our subtraction process eliminates the steady-state Ca²⁺ current which does not inactivate (see above).

In some systems the Ca channel is permeable to K⁺ and Cs⁺ (Reuter & Scholz, 1977; Fenwick, Marty & Neher, 1982; Lee & Tsien, 1982). As a consequence, the total current through the Ca channel reverses sign 50 – 70 mV below the theoretical equilibrium potential for Ca²⁺. In our experiments, however, there is no tendency for current reversal up to depolarizations of about 100 mV, which is 55 mV below E_Ca - Thus, there seems to be little if any K⁺ permeability of the Ca channel in Paramecium.

Furthermore, the outward component cannot be explained in terms of a K⁺ permeability of the Ca channel, because it decreases with higher depolarization, which would further increase the driving force for K⁺ currents.

We have shown that Mg²⁺, Co²⁺, Mn²⁺ and Ni²⁺ reduce the voltage-dependent Ca²⁺ current when added to the standard Ca²⁺ solution. In addition, these cations shift the resting potential to more positive values (Naitoh, Eckert & Friedman, 1972; Eckert, Naitoh & Friedman, 1972). For two reasons the voltage shifts in our experiments do not account for the inhibition of the Ca²⁺ current. First, there is no correlation between the shift in resting potential and the degree of current inhibition between the different cations tested. Second, regardless of any shift in resting potential the maximum inward current always occurs at the same absolute membrane voltage. Thus, there is no reduction in driving force which may reduce the Ca²⁺ current.

With increasing Ca²⁺ concentration the maximum inward current occurs at more positive voltages. In the small concentration range tested (0·33 – 2·0 mmol l ^{− 1}) this voltage shift is always equal to that of the calculated E_Ca. Thus, there seems to be no change in the driving force for Ca²⁺ that would affect the value of k_Ca.

Naitoh & Eckert, (1968a,b) and Naitoh et al. (1972) tested the influence of different cations on the Ca²⁺ (Ba²⁺) action potential of Paramecium following current injection. They found that addition of Mg²⁺, Co²⁺ or Mn²⁺ had only minor effects on the Amplitude of the action potential even when the final concentration of these ions was several times that of Ca²⁺, or Ca²⁺ plus Ba²⁺. It was concluded that there is little if any inhibition of the voltage-dependent Ca²⁺ conductance in Paramecium by cations that effectively block Ca²⁺ currents in metazoan tissue (Brehm, Dunlap & Eckert, 1978; Eckert & Brehm, 1979; Brehm et al. 1980). We found that the resting resistance of the membrane increased after addition of Mg²⁺, Co²⁺, Mn²⁺ or Ni²⁺. This may explain the absence of an inhibitory effect on the action potential measured by these authors: an increase in input resistance would increase the passive component of an action potential during injection of a constant current and additionally as a consequence the voltage-dependent Ca²⁺ conductance. Both effects would counteract any inhibition of the Ca²⁺ (Ba²⁺) action potential.

The inhibition of the Ca²⁺ current by Mg²⁺, Co²⁺, Mn²⁺ and Ni²⁺ demonstrated in our experiments is in good agreement with the results of Browning & Nelson (1976). They found that ⁴⁵Ca²⁺ uptake upon excitation is inhibited by Co²⁺, Mn²⁺ and Ni²⁺, and to a smaller extent by Mg²⁺.

In Stylonychia, addition of Mn²⁺ (1 mmol l ^{− 1}) to standard Ca²⁺ solution reduces the voltage-dependent Ca²⁺ current, while Mg²⁺ (1 mmol^{− 1}) has no significant effect (DePeyer & Deitmer, 1980). Co²⁺ reduces the two different voltage-dependent Ca²⁺ conductances that can be separated in Stylonychia (Deitmer, 1983).

The absolute dissociation constants calculated from our results are about two orders of magnitude smaller than those Hagiwara & Takahashi (1967) found in barnacle muscle. The Ca²⁺ concentration in the barnacle medium, however, is much higher (20 mmol l ^{− 1} than in our standard Ca²⁺ solution. The higher affinity of the Paramecium membrane (or Ca channel site) is probably necessary to guarantee excitability in a medium of low Ca²⁺ concentration.

The simple model of Hagiwara & Takahashi (1967) adequately describes the saturation and inhibition of the Ca²⁺ current in Paramecium. However, changes in the outer surface potential may also be involved (Hook & Hildebrand, 1979, 1980; Eckert & Brehm, 1979). One can calculate that, due to fixed negative charges, the Ca²⁺ concentration at the membrane surface is higher than in the bulk solution, and that it does not change to the same extent when the Ca²⁺ concentration in the bulk solution is changed (McLaughlin, 1976). Thus, increasing the Ca²⁺ concentration in the bulk solution will not increase equally the concentration of charge carriers at the Ca channel. This could be the reason for the saturation of Ca²⁺ current with increasing Ca²⁺ concentration (Hagiwara, 1973; Hagiwara & Byerly, 1981).

Reduction of the surface potential is possible by both screening and binding (McLaughlin, Szabo & Eisenman, 1971). We found that the tested cations reduced the Ca²⁺ current with different efficiency. This cannot be explained just by screening, because the membrane will not distinguish between different divalent cations at some distance from its surface (Ohmori & Yoshii, 1977). Rather we have to assume the binding of competing cations with different affinities to negative binding sites. Similar conclusions have been drawn from experiments on frog muscle (Hahin & Campbell, 1983). These surface charges will not necessarily be part of the Ca channels. However, if they are not, a coupling mechanism between surface charges and Ca²⁺ conductance has to be assumed: changes in the surface potential will be detected by the voltagesensitive gates and may in this way change the excitability of the membrane.

We thank R. Backbier, H. Erkens and D. Grammig for their invaluable help throughout the construction of the experimental set-up and the latter for the design of the comparator circuit. This work was supported by the Deutsche Forschungsge-meinschaft (SFB 160).