The effect of temperature on the outward currents in the soma of molluscan neurones in voltage-clamp conditions

The main purpose of this investigation is to describe the effects of temperature on the outwardly directed currents transient in molluscan neurones during long-lasting depolarization in voltage-clamp conditions. The outward currents are presumably carried by potassium ions. These currents are suppressed by tetraethylammonium (Hagiwara & Saito, 1959; Neher & Lux, 1972) which specifically blocks potassium channels (see Hille, 1970, for review). Besides this, it was recently shown that spike activity or voltage-clamp pulses caused changes in extracellular potassium concentration in the vicinity of molluscan neurones (Neher & Lux, 1973).

When rather large long-lasting voltage-clamp steps are applied to the somatic membrane a relatively fast inactivation of delayed outward currents is observed. After the delayed outward current reaches its peak, it decays towards a final value (Alving, 1969; Leicht, Meves & Wellhoner, 1971; Connor & Stevens, 1971a; Gola & Romey, 1971 ; Magura, Krishtal & Valeyev, 1971 ; Neher & Lux, 1971). Similar observation was made on supramedullary neurones of puffer fish (Nakajima, 1966; Nakajima & Kusano, 1966) and on stretch receptor neurones (Nakajima & Onodera, 1969).

The time course and magnitude of potassium currents in molluscan neurones are strongly dependent upon the conditioning level of the membrane potential (Connor & Stevens, 1971a; Gola & Romey, 1971; Magura, Krishtal & Valeyev, 1971). When a conditioning hyperpolarization is used, the subsequent depolarization causes the appearance of a fast transient outward current in some neurones (Hagiwara, Kusano & Saito, 1961; Connor & Stevens, 1971b; Neher, 1971a). They are activated below − 40 mV. Activation of delayed outward current occurs when the membrane potential is more positive than − 40 mV. It is assumed that the appearance of the fast outward current is a result of the removal of inactivation of the fast potassium channels. These channels are, on several criteria, operationally distinct from those of the delayed outward current (Connor & Stevens, 1971b; Neher, 1971a).

The tentative conclusion of this paper is that the temperature changes cause, not only a direct effect on the mechanisms controlling potassium currents in the soma of molluscan neurones, but also probably change in surface membrane potential in the vicinity of potassium pathways.

The experiments were performed on giant nerve cells in the visceral complex of Helix pomatia ganglia. The preparation has been described in an earlier report (Magura, 1967).

Ringer solution used in experiments had the following composition (mM): NaCl, 80; KC1, 4; CaClg, 10. The pH was adjusted to 7·5 with Tris-HCl. Temperature regulation was carried out by passing Ringer solution through reservoirs where temperatures were maintained by thermoelectric devices. The temperature of the bath was measured by a thermistor placed close to the preparation.

Two micro-electrodes filled with 3 M-KCl were inserted simultaneously into the same nerve cell. One micro-electrode (resistance 4–5 MΩ) was used to feed current the other (resistance 5–7 MΩ) to measure the membrane potential. The apparatus used for voltage clamping was similar to that described by Chamberlain & Kerkut (1969). Voltage-clamp devices suitable for molluscan neurones have also been described by Gedulding & Gruener (1970) and Gola & Romey (1971).

In our experiments the observed rise time of the membrane voltage pulse was usually less than 500 μsec and the decay time of the capacitive charging current between 1 and 2 msec (the membrane capacitance was 10–20 nF).

The effects of cooling upon the early inward and delayed outward currents is demonstrated in Fig.,1. Decrease in value and in rate of rise of delayed outward current was observed when the temperature was lowered from 20 to 10 °C. A better estimation of the effect of cooling upon the rate of rise of delayed outward current could be obtained from the record at the inward current equilibrium potential. Somewhat similar changes in early inward current occurred.

A long-lasting, rather large depolarization of the somatic membrane produces a delayed outward current which declines from its maximum to a final value. Fig. 2(a) shows that inactivation of delayed outward current is less distinct when the temperature is lowered (see also Fig. 3). The peak value of delayed outward current, rather than its final value, is depressed more effectively by cooling.

In Fig. 2(b, c) the peak value (I_max) of delayed outward currents and their final value (I_∞) taken from an experiment on the same cell at 20 and 10 °C was plotted as a function of the membrane potential. Difference in the effect of cooling on I_max and on I_∞ is clearly seen (see also Fig. 5). The curve for at 20 °C is somewhat steeper than that at 10 °C.

The effect of cooling upon a delayed outward current may be considerably altered by a preceding shift of the membrane potential (Figs. 3-5). Fig. 3 shows that the effect of cooling, to depress the peak value of the delayed outward current, was increased when a conditioning hyperpolarization was used. (It is also seen that a conditioning hyperpolarization increases the fraction of the outward current system which is available.)

Fig. 4 also demonstrates an increase in the effect of cooling on the peak value of delayed outward current and on the peak of membrane conductance when a conditioning hyperpolarization is used. In this cell we did not observe the appearance of the fast outward current after a conditioning hyperpolarization. As can be seen in Fig. 4 the conditioning hyperpolarization significantly increases the slope of the current voltage relation of the peak value of delayed outward current at 20 °C.

The Q₁₀ for G_k(max) was approximately 1·5 without a conditioning hyperpolarization. Conditioning hyperpolarization increased the Q₁₀ for G_k(max) to 2.

The effect of cooling upon the delayed outward current is quite different from that demonstrated in Figs. 2 and 3 when a conditioning depolarization is used. As can be seen from Fig. 5(a), with a conditioning depolarization of −32 mV, cooling increased the peak value of the delayed outward current. (Note that its final value decreased.) In such conditions the inactivation of the delayed outward current was more distinct at 10 °C than at 20 °C. Cooling caused a decrease in the peak value of outward current in this neurone when a conditioning hyperpolarization was applied (− 65 mV). Fig. 5 (b) shows that the depressing effect of cooling on the peak value of the delayed outward current was most distinct with complete removal of the steady-state inactivation (at a conditioning potential more negative than −100 mV). When the conditioning potential was more positive than − 50 mV the cooling increased the peak value of the outward current. Fig. 5 (b) also demonstrates that the effect of cooling upon the final value of the delayed outward current was negligible.

The results shown in Fig. 5 can be explained if one supposes that cooling not only depresses the peak value of the delayed outward current but decreases its steady-state value of inactivation (see Effect of cooling on the fast transient outward current).

The time course of the delayed outward current inactivation is exponential with a time constant (τ) of 300 –800 msec (at 20 °C). The Q₁₀ of (τ⁻¹ was about 2 ·5. These observations seem difficult to reconcile with the results presented in Fig. 6. This Figure demonstrates that the effect of cooling on the time course of removal of the delayed outward current inactivation (during conditioning hyperpolarization) was negligible. Somewhat similar results were obtained when fast outward current was investigated (Fig. 9).

In some molluscan neurones a conditioning hyperpolarization causes the appearance of a fast transient outward current during the test depolarization (Hagiwara, Kusano & Saito, 1961; Connor & Stevens, 1971b; Neher, 1971a). Fig. 7(a, b) demonstrates the dependence of the peak value of the fast outward current upon the conditioning level of the membrane potential at 10 and 20 °C. The potential was changed from −50 to −110 mV. Test depolarization was constant (− 35 mV). At such depolarization the activation of the delayed outward current system was negligible. During a test depolarization the inactivation of outward current was nearly complete. (Note the appearance of a transient outward current at 10 °C when the conditioning potential was − 50 mV.)

It is clearly seen from Fig. 7(b) that the effect of cooling depends on the conditioning value of the membrane potential. The depressing effect of cooling was most distinct when the conditioning potential was −100 mV, when nearly complete removal of fast outward current inactivation occurred. When the conditioning potential was more positive than −65 mV, cooling slightly increased the peak value of the fast outward current. Somewhat similar effect of cooling on the delayed outward current was described above (Fig. 5).

These results were normalized and inactivation curves were obtained (Fig. 7c). They show that cooling causes the inactivation curve to shift in the direction of a more positive membrane potential. A similar effect has been observed when the external calcium concentration is increased (Neher, 1971b).

One may assume that cooling causes a decrease in the negative surface potential in the vicinity of the potassium pathways and removes their inactivation. Simultaneously cooling depresses the potassium conductance. This second effect of cooling was dominant with a conditioning pulse more negative than − 65 mV.

The upper part of Fig. 8 (a, b) shows the records of membrane potential and fast outward current in two neurones at 20 °C and 10 °C. In Fig. 8(a) semilogarithmic plots of the decaying phase of these traces show that the time course of inactivation is exponential. The time constants (τ) at 20 and 10 °C were respectively 75 and 185 msec. The Q₁₀ of the reciprocal time constant (τ⁻¹) was about 2·5. In Fig. 8(b) at 20 °C the time course of outward current inactivation was complex and was determined by two time constants. The time constant of inactivation at the beginning of the decay was 60 msec (τ₁). After 50 msec the time constant had increased to 150 msec (τ₃). At 10 °C the time course of inactivation was determined by only one time constant.

One may suggest that the time course of inactivation in Fig. 8(b) is related to the activation of two different fractions of potassium channels, where the time constants of inactivation are different (the fast potassium channels and delayed outward current channels). Such a suggestion is difficult to bring into accord with the observation that the time constant of the slow phase of outward current inactivation is often shorter than the time constant of delayed outward current inactivation (Fig. 8b). It is difficult to exclude the possibility of a slow phase of fast outward current inactivation in some neurones. Fozzard & Hiraoka (1973) recently described a somewhat similar time course of inactivation of chloride currents in cardiac Purkinje fibres.

Experiments with two pulses have indicated that fast outward current inactivation is removed exponentially with a voltage-dependent time constant (Neher, 1971a). Fig. 9 demonstrates that a lowering of temperature from 20 to 10 °C only slightly prolongs the time constant for removal of fast outward current inactivation (at a conditioning potential of −100 mV). This is clearly seen in a semilogarithmic plot (Fig. 9b). The Q₁₀ of the reciprocal time constant of inactivation removal was only 1 ·3. In this neurone the Q₁₀ of developing fast outward current inactivation was about 2·5 (at a test potential of − 30 mV). It seems reasonable to suppose that interpretation of these results requires a detailed investigation of the effect of cooling on the Hodgkin-Huxley rate coefficients (α and β) for outward current inactivation.

Upon the results of the above experiments it seems reasonable to separate the delayed outward current in molluscan neurones into two components : (i) a persistent component which is maintained during long-lasting depolarization; (ii) a transient component which inactivates relatively quickly during depolarization. Temperature changes have a more significant effect on the value of the transient component than on the value of the persistent one.

The effect of temperature on the peak value of the transient component of the delayed outward current is clearly dependent upon the membrane potential which precedes the test depolarization. Similar observation was made on the fast transient outward current. A negative shift of the preceding membrane potential (conditioning hyperpolarization) increases the effect of cooling to depress the peak of transient outward currents. It reaches a maximum with conditioning polarizations at which a complete removal of potassium inactivation is observed (about −100 mV). We suppose that cooling not only depresses the peak value of the transient potassium conductance but also decreases the negative surface potential of the membrane in the vicinity of potassium pathways. This potential is created by the negative fixed charges of the membrane (see Hille (1968) for review).

Decreasing the negative surface potential by cooling causes the inactivation curve for fast outward current to shift in the direction of more positive membrane potential. It also increases the fraction of transient potassium channels which are available. Such an effect of cooling is clearly dominant at conditioning potentials at which a significant fraction of the potassium channels are inactivated.

It seems reasonable to suppose that changes in surface potential by cooling may cause certain changes in the kinetic of potassium inactivation.

In these investigations we did not estimate the outward current density and conductance in Ω ⁻¹ cm^-2. Such estimation is difficult because of the complexity of the surface of the giant neurone soma. The invasion of the giant cells by strands and trabeculae of glial cells increases enormously the membrane surface in certain places, especially in the axon hillock region (Bullock, 1961). A calculation of the surface of the soma using its capacity (assuming a value of μF/cm²) shows that the surface of the soma is about 10 times as large as that of sphere having the same diameter (Magura, 1967; Magura, Grobova & Zamekhovsky, 1972). This value is in accord with the data obtained by light and electron microscopic examinations of sections of molluscan giant neurones (Mirolli & Talbot, 1972). If this is correct the density of potassium current through the somatic membrane of molluscan neurones is, according to our calculation, much lower than in the giant squid axon. The limiting conductance for delayed outward current is approximately 8 ·10⁻¹ Ω ⁻¹/cm² (20 °C). This value is roughly 1 /40 of that in the giant squid axon.

We thank Professor P. G. Kostyuk for his kind support and valuable advice. We should like to thank Miss Elena Dolgaya for technical assistance.