A ‘TEA+−Insensitive’ Mutant with Increased Potassium Conductance in Paramecium Aurelia

Despite intensive research the chemical identity of the ionic channels of excitable membranes and the molecular mechanism of ion permeation remain elusive. One approach to the problem is to identify these channels by genetic means. Mutants with defective channels can be used as the null control in physiological investigations and chemical identifications. Paramecium has been used in this approach to the study of membrane functions through mutations because information is available on both genetics (Sonneborn, 1970) and electrophysiology (Eckert, 1972; Naitoh & Eckert, 1974) of this genus. The rationale for genetic studies of the excitable membrane in Paramecium was given by Kung (1971 b).

Three hundred lines of mutants are now available in P. aurelia (Kung et al. 1975). Among them are ones with defects in their voltage-sensitive Ca channel which result in the loss of the Ca action potential (Kung & Eckert, 1972). There are conditional mutants which reveal their defects after growing at high temperatures (Satow, Chang & Kung, 1974; Satow & Kung, 1976b). Since the ciliary motions of Paramecium are governed by the membrane potential and action potentials are coupled to reversal of the direction of ciliary beating (Eckert, 1972; Naitoh & Eckert, 1974), membrane defects are reflected as anomalies in the locomotor behaviour. Thus, mutants without action potentials are unable to reverse their ciliary beat and fail in the ‘avoiding reaction’ (the ‘pawn’ mutants, Kung, 1971a), while those with prolonged depolarizations swim backward for long distances (the ‘paranoiac’ mutants, Kung, 1971a; Satow, Hansma & Kung, 1976).

This paper reports our findings on a mutant which has a very low membrane resistance, a defect we have now traced to a K channel. The screening method used to isolate this mutant is based on one of its behavioural deficiencies, (i.e. its inability to generate spontaneous avoiding reactions in tetraethylammonium (TEA+) solutions). Details on mutagenesis, screening, behavioural phenotyping and genetic analyses of this ‘TEA+-insensitive’ mutant are given by Chang & Kung (1976).

A preliminary report of this work has appeared elsewhere (Satow, 1975).

We used the ‘TEA⁺-insensitive’ mutant (stock d4-152, genotype teaA/teaA) and wild type (stock 513) from which the mutant was derived. Both stocks belong to species 4 of Paramecium aurelia. Cells were cultured in the Cerophyl medium bacterized with Aerobacter aerogenes 20 h before use (Sonnebom, 1970). Only robust cells in logphase growth were used.

All solutions contained 1 mm-Ca(OH)₂, 1 mm citric acid, buffered to pH 7·15–7·25 with 1·0–1·2mm Tris [tris (hydroxymethyl) aminomethane]. The K solution, containing in addition 4 mm-KCl, was the ‘adaptation solution’ in which the cells were penetrated by electrodes. In most experiments, the K solution was replaced by various test solutions in which K⁺ was replaced by other cations. They were the Ca-, Mn-, Ba-, and TEA-solutions containing 2 mm of CaCl₂ or MnCl₂, 2 or 4 mm of BaCl₂ or 4 mm of NaCl or TEA-Cl (tetraethylammonium chloride, Aldrich Chemical Co.), respectively. In the series of experiments testing the effect of TEA⁺ in differents concentrations, 1, 4 and 16 mm TEA-CI were used.

The methods for intracellular recording were basically those of Naitoh & Ecker (1972). Modifications of the reference electrode and the use of continuous perfusion are described by Satow & Kung (1976a). To study the steady-state I-V relationship, 200 ms to 1 s pulses of current (less than 10⁻⁹ A) were delivered. Both the stimulating electrode and the recording electrode were filled with 500 mm-KCl with tip resistances of 70–130 MΩ.

Microelectrodes filled with 100 mm TEA-Cl were used. A depolarizing 10⁻⁹ d.c. current was applied through such an electrode for the ionophoretic injection of TEA⁺. Ten minutes of such a current is estimated to give 20 mm concentration of TEA+ in the cytoplasm, based on a transport number of 0·5 (Friedman & Eckert, 1973) and a cell volume of 1·5 x 10⁻⁷ ml. The effect of injected TEA+ lasted over 30 min, during which time membrane properties were examined.

The most outstanding electrophysiological defect of the mutant is the diminution of its response to injected current. Fig. 1 shows the responses to a moderate outward current (7 x 10⁻¹⁰ A) by wild type and the mutant bathed in the K solution. This is the ‘adaptation solution’ and there is no systematic difference in the resting level between the two strains in this solution (see Table 1). Wild type generates an action potential (the Ca regenerative response, Naitoh & Eckert, 1968a; Satow & Kung, 19766) followed by a steady-state response. The action potential of the mutant evoked by a comparable current (right) is clearly smaller in amplitude. The maximal rate of rise, as seen from the peak of the dV/dt trace, and the steady-state depolarization by the constant current are also reduced.

These defects of the mutant persist in different bath solutions. Fig. 2 shows the responses evoked by injected outward current from paramecia bathed in five solutions containing K⁺, Na⁺, TEA⁺, Ca²⁺, or Mn²⁺ as the major cation. Wild type generates action potentials when stimulated by injected current in all five solutions (Fig. 2, top). As in P. caudatum (Naitoh & Eckert, 1968 a), Mn²+ does not inhibit this Ca response. In all five solutions, the Ca regenerative response of wild type evoked by small to moderate current (< 7 x 10⁻¹⁰ A) is faster and reaches higher peaks than that of the mutant. The regenerative responses of the mutant and wild type are graded to the stimuli. The steady-state depolarization of the mutant membrane after the spike is smaller than that of wild type in all solutions.

At supramaximal currents (> 10⁻⁹ A), the mutant can often give spikes equal to those of wild type in the peak potential and in the maximal rate of rise (see Fig. 5). This and other evidence presented below show that the Ca activation is not affected in this mutant unlike the case of ‘pawn’ mutants (Kung & Eckert, 1972; Satow & Kung, 19766 ; Schein, Katz & Bennett, 1976).

Steady-state potential displacement by 200 msec injected current is plotted against the current strength. In paramecium, the I-V curves are sigmoidal (Naitoh & Eckert, 1968 a; Kung & Ecker, 1972). The resistance of the mutant membrane is smaller than that of the wild-type membrane at all points, when tested in the K solution (Fig. 3 A). In the Ca solution where no K⁺ is added, both the mutant and the wildtype membrane appear more resistive than in the K solution (Fig. 3 B). The difference between the two strains remains clear for depolarizing currents, but ceases to be statistically significant for moderate or large hyperpolarizing currents.

The zero-current input resistance of the resting membrane is estimated with small current injections. The results are summarized in Table 1. The resting membrane resistance is reduced to about half by the mutation. The comparison of the resistances of the two strains is valid since they are of the same size and rest at the same potential level.

With the application of strong hyperpolarizing current, (e.g. 10⁻⁹ A) the difference of the chord resistance between the two strains is not observed in the Ca solution (Fig. 3 B) and the TEA solution (Fig. 6), where [K]_out = 0, but it is seen in the K solution (Fig. 3 A). This result suggests that the mutant defect is on a K conductance.

When K solution is replaced with TEA solution (4 mm, see above for complete composition), wild-type animals respond with a train of action potentials as shown in Fig. 4 (top). The amplitude and frequency of the spikes vary. Both the up-and the downstrokes of the action potentials are steep, much like the response to Ba solution (Kung & Eckert, 1972; Satow et al. 1974; and Fig. 8) and unlike the response to the Na solution (Satow & Kung, 1974). Each spike corresponds to a transient period of ciliary reversal and so to an avoiding reaction in the free-swimming paramecium. ciliary reversal is coupled to action potential through Ca²⁺ (Naitoh & Kaneko, 1972; Eckert, 1972). Thus, these spontaneously generated action potentials in the TEAsolutions are a form of Ca regenerative response. Fig. 4 (bottom) shows the response of the mutant to identical treatment. Excitation is conspicuously absent. Increasing the concentration of TEA⁺ in the solution up to 16 mm does not trigger any action potential. Corresponding behavioural tests have shown little or no avoiding reactions to TEA⁺, hence the name ‘TEA-insensitive’ when it was first described (Kung et al. 1975).

Friedman & Eckert (1973) showed that TEA⁺ blocked K conductance and thus reduced the short-circuiting leakage and the repolarizing rate in P. caudatum. We have found a similar effect in the wild-type P. aurelia. Less than 10⁻¹⁰ A is often sufficient to trigger an action potential in an animal bathed in 4 mm TEA⁺. The threshold cannot be accurately assessed because the membrane is spontaneously active in the TEA solution. Fig. 5 (top row) shows these triggered action potentials in wild type. The amplitude of the triggered action potential is sharply graded with the stimulus, approaching an all-or-none response. Increasing outward current strength increases the rate of rise, but not the peak potential.

Although in the mutant there is no spontaneous response in the TEA solution we can evoke responses with outward currents (Fig. 5, bottom row). Unlike the wildtype response, the mutant response increases gradually with the injected current.

The peak potential and maximal rate of rise are smaller in the mutant for small moderate stimuli (1–5 x 10⁻¹⁰ A), but approach those of wild type for very large currents (10⁻⁹ A). Fig. 5 also shows that the steady-state potential displacement by a given current is smaller in the mutant than in wild type.

The I-V relation in 4 mm TEA solution is plotted for both strains (Fig. 6). Estimated zero-current input resistance of the mutant is clearly smaller than that of wild type (Table 1). Clear reduction in chord resistance in the mutant is seen for large depolarizing, but not for hyperpolarizing currents. This asymmetry is also observed in the Ca solution (Fig. 3 B), which like the TEA solution is devoid of K⁺. 16 mm TEA solution depolarizes both wild type and the mutant but the difference in resistance of the two strains remains clear. This difference persists when 1 mm TEA solution is used.

Externally applied TEA⁺ cannot eliminate the difference between wild type and the mutant. Therefore, we examined the effects of internally applied TEA⁺. Iono-phoretically injected TEA⁺ blocks the downstroke of Ca response and increases the input resistance of the wild-type P. aurelia bathed in the K solution. This is similar to the finding of Friedman & Eckert (1973) in P. caudatum. The effect is dependent on the amount of TEA⁺ injected. With 10⁻ A d.c. ionophoretic current, clear effects are first observed after 3 min. After 10 min of injection (about 20 mm TEA⁺ injected, see Materials and Methods) the steady-state potential displacements by test pulses increase to maximum (Fig. 7). The kinetics of the internal TEA⁺ effects cannot be accurately measured since the electrode characteristics are not uniform. Internal TEA⁺ also reduces the resting potential and the rate of rise of the Ca action potential. The reasons for these changes are not clear.

Injected TEA⁺ has dramatic effects on the mutant. It increases the membrane resistance of the mutant to that of wild type, similarly treated. The distinct difference between the two strains before TEA⁺ injection in terms of the action potential and the after-spike steady-state response is no longer observed. Fig. 7 shows the I-V relation of the two strains of paramecia bathed in the K solution after a 10 min injection of TEA⁺: no significant difference remains between the two curves. Comparing Fig. 7 with Fig. 3 A we see that the input resistance of both strains is clearly increased by internal TEA⁺. The low-current (10⁻¹⁰A) input resistances of two wild-type cells after this TEA⁺ injection were 115 and 90 MΩ (de-) and 90 and no MΩ (hyperpolarizing direction). Those of two mutant cells were 112, too MΩ and 82, 100 MΩ, respectively.

Naitoh & Eckert (1968b) showed that Ba²⁺ could carry the action current efficiently and block the K leakage in P. caudatum, resulting in all-or-none electrogenesis when the cell came into contact with Ba²⁺. Such spikes were also recorded in wild-type P. aurelia (Kung & Eckert, 1972; Satow et al. 1974). Fig. 8 (top) shows this response of wild type when 2 mm-Ba solution replaces the K solution in the bath. Fig. 8 (bottom) shows the response of the mutant. The down-stroke of the action potential is often steeper in the mutant than in wild type, although variations within each strain is large. Each of the spikes is correlated with a rapid avoiding reaction. Neither the mutant nor wild type ever fails to respond to the Ba solution behaviourally or electrophysiologically.

Fig. 9 shows the I^_V relation with hyperpolarizing current of wild type and the mutant in 4 mm-Ba solution. The depolarization half in this solution cannot be measured accurately, owing to spontaneous activity. The input resistances to small hyperpolarizing current (10⁻¹⁰A) of wild type and the mutant are indistinguishable, being 9 3·3 ± 20·9 MQ and 92·1 ± 17·4 MΩ, respectively.

The physiological defects of the mutant are best explained by mutational increase of K conductance. The major evidence is as follows: (1) K⁺ efflux limits the Ca response in Paramecium; larger K⁺ efflux in the mutant could explain the reduction of Ca action potentials (Figs. 1, 2 and 5). (2) The resting resistance of the mutant is half that of wild type. Membrane resistance of the mutant measured with large inward current is smaller than that of wild type only when K⁺ is the major external cation (Figs. 3 and 6). (3) Internally applied TEA⁺ counteracts the mutational effects, increasing the membrane resistance of the mutant to equal that of wild type (Fig. 7).

In response to small or moderate currents the active depolarizations of the mutant are clearly smaller and slower than wild type (Figs. 1, 2 and 5). However, large outward current evokes Ca responses in the mutant which approach the peak potential and the maximal rate of rise of those of wild type (Fig. 5). This result suggests that the Ca activation mechanism of the mutant is not impaired. Other evidence showing that Ca activation is normal in the mutant is the observation of Ba spikes (Fig. 8). Apart from a faster repolarization, the Ba spikes of the mutant are the same as those of wild type. Hagiwara et al. (1974) showed that Ba²⁺ carries the action current through the Ca channel and blocks the early K⁺ current in barnacie muscle. Since Ba²⁺ carries more current, the action potentials are all-or-none in character. Similar findings and interpretations have been made on P. caudatum (Naitoh & Eckert, 1968b, 1972) and P. aurelia (Satow et al. 1974; Kung & Eckert, 1972). Since normal Ba spikes can be generated in the mutant (Fig. 8), there is no reason to believe that the Ca activation mechanism is defective as in the cases of the ‘pawn’ mutants (Kung & Eckert, 1972; Satow et al. 1974; Satow & Kung, 1976b; Schein et al. 1976).

In both crustacean muscle and paramecium, the regenerative Ca responses are graded to the strength of the stimuli. The graded Ca action potentials are often contrasted with the well-known all-or-none Na action potentials in nerve. The Ca response is graded because the Ca activation is considerably slower than the Na activation in nerves : the Ca action potential is determined by both inward Ca current and the counteracting outward K current (Hagiwara, Hayashi & Takashi, 1969; Naitoh & Eckert, 1968 a).

The defective Ca action potentials of the mutant can therefore be the result of larger outward K counter current in the generation of action potential. The downstrokes of the action potentials of the mutant are often steeper than those of wild type and almost always undershoot the steady-state depolarization levels (Figs. 1, 2 and 5), suggesting increased K⁺ efflux during both depolarizing and repolarizing phases of the calcium action potential.

The sigmoidal nature of the I⁻V curves of Paramecium is not fully understood. Presumably there are voltage-insensitive ion channels as well as channels sensitive to depolarization and those sensitive to hyperpolarization. Paramecium membrane was found to be permeable to a large variety of cations based on their ability to depolarize the membrane (Naitoh & Eckert, 1968a). Browning & Nelson (1976) show that the permeability to K⁺ is higher than those to other ions in Paramecium.

Comparing the parts of the I^_V curves with moderate to high hyperpolarizing currents, it is clear that the resistance of the mutant is nearly the same as that of wild type when tested in solutions devoid of K⁺, i.e. in the Ca solution (Fig. 3B) and the TEA solution (Fig. 6). Membrane resistance in this part of the I-V curve of the mutant is clearly smaller than that of wild type when K⁺ is the major cation in the bath (Fig. 3 A). These observations are consistent with the view that the portion of the conductance increased by the mutation is K⁺ specific, since for large hyperpolarizations the membrane current is carried by the cations in the external solution.

Tetraethylammonium ion (TEA⁺) affects the K channels in excitable membranes: both the voltage dependent K channel in frog myelinated nerve fibre (Hills, 1967; Armstrong & Hille, 1972) and the resting K permeability of barnacie muscle (Keynes et al. 1973). Friedman & Eckert (1973) showed that internal TEA⁺ increases membrane resistances and block the down-stroke of the Ca action potential in P. caudatum. We found that the mutationally increased conductance can be blocked by internal TEA (Fig. 7), again suggesting that the mutation increases K conductance.

Externally applied TEA⁺, though not as effective as that internally applied, also blocks the K efflux in paramecium. By reducing the short-circuiting effect, this blockage makes the Ca activation more effective. Thus, the Ca action potentials are more effective. Thus, the Ca action potentials are more sharply graded with the strength of applied current, approaching all-or-none (Fig. 5). In fact, the generation of the action potential is so efficient that spontaneous activity is recorded (Fig. 4). In contrast, in the mutant the evoked action potentials are gradually graded (Fig. 5) and spontaneous activity is absent (Fig. 4). This suggests that the outward K current of the mutant remains high when bathed in this TEA solution. This may mean that the effect of externally applied TEA⁺ is not strong enough to compensate for the mutationally induced K conductance. Alternatively, it is possible that the mutation increases a K conductance which is less or not sensitive to external TEA⁺.

The mutant behaves similarly to wild type in the Ba solution (Fig. 8). In solutions of 4 mm Ba²⁺ or above, the resting membrane resistance and I-V relation (Fig. 9) are not different from those of wild type. These results show that Ba²⁺, like internally applied TEA⁺, is effective in blocking the mutationally altered K channel. This blockage is concentration dependent ; lower concentration (2 mm) is less effective.

The mechanism of K⁺ permeation in normal paramecium is not well understood. For example, it is not known if the resting K conductance can be activated or if the activation is mediated through depolarization or Ca²⁺ or both. Mutation affecting the relative resting K permeability in the case of the ‘fast-2’ mutant (Satow & Kung, 1976a) can give a phenotype quite different from the ‘TEA⁺-insensitive’ mutant studied here. The increased K conductance of this mutant clearly affects the outward going rectification. It is, therefore, possible that the K channel responsible for the K delayed rectification has been mutated in such a way that it is now activated even when the membrane is at rest. Future work on this and other membrane mutants may help us identify and sort out the different K channels in Paramecium.

We thank Professor S. Hagiwara for his helpful criticism in the preparation of the manuscript. Supported by PHS GM 19406 and NSF BMS 75-10433.