Effects of temperature on the electrical properties of the Paramecium membrane were investigated under constant current and voltage-clamp stimulation. With cells cultured at 18 °C, the resting potential was largely stable with experimental temperatures varied over the range of 10–25 °C, whereas action potential amplitude and membrane input resistance were inversely related to temperature increases. During voltage clamp, the early calcium current was increased, the time-to-peak decreased, and the early conductance increased with temperature. Similar modifications of the culturing temperature did not affect the resting potential, input resistance and stimulus-response relationship of the action potential, but the early conductance was reduced with increase in temperature. Possible effects of long- and short-term temperature changes upon intraciliary calcium concentration are discussed.
Ciliate protozoa are sensitive to thermal shocks but they can eventually adapt within a comparatively wide range of temperatures (cf. Machemer & de Peyer, 1977 ; NeNelson & Kung, 1978). Recent experiments have confirmed earlier behavioural tests showing that Paramecium detects gradients in temperature and congregates within preferred temperature zones (Mendelssohn, 1902; Nakaoka & Oozawa, 1977; Kawabuko & Tsuchiya, 1981). Raised temperature may produce avoidance responses similar to those elicited with depolarizing stimuli (Koehler, 1939; Hildebrand, 1978; Hennessey & Nelson, 1979). The threshold of thermal avoidance has been found to be dependent upon the culturing temperature (Hennessey & Nelson, 1979).
While behavioural effects of changes in temperature are obvious, the mechanisms by which temperature interferes with the membrane function and the cellular motor response are not well understood. Lowering of the experimental temperature produces very small depolarizations of the membrane (Yamaguchi, 1960) but depresses the rate of the action potential in Paramecium (Machemer, 1974) and Stylonychia (de Peyer & Machemer, 1977). Negative temperature shifts reduce the frequency and alter the direction of ciliary beating (Machemer, 1972) presumably due to an interference with active calcium extrusion (Browning & Nelson, 1976).
The present experiments were carried out to explore the effects of culturing and experimental temperatures on the passive and active membrane properties. Our emphasis has been to characterize the combined roles of short-term and long-term temperature in modifying basic membrane parameters.
Paramecium caudatum (average length 220 μm) was reared in hay infusion at 10, 18 and 25 °C. For experimentation cells were harvested approximately 2 weeks after inoculation and washed in experimental solution containing 1 mm-CaCh and 1 mm-KC1 buffered with Tris-HCl at pH 7·4. They were collected at temperatures between 18 and 20 °C within 30 min prior to experimentation. An experimental cell was placed in a temperature-regulated 1-ml chamber. The actual temperature was monitored within 5 mm of the specimen using a calibrated NTC resistor.
The electrophysiological procedures have been described previously (Naitoh & Eckert, 1972). The membrane potential and input resistance of the cell were determined at the beginning of each experiment. The membrane potential was measured as the potential difference between an intracellular and an extracellular microelectrode, each filled with 1M-KC1 (resistance 40–60MΩ). A third electrode, filled with 2M-K-citrate, was inserted into the cell for current injection. The input resistance of the cell membrane was measured by injecting a 10−10 A hyperpolarizing constant current pulse.
The voltage clamp was performed by means of a conventional feedback system using a high-gain differential amplifier (AD 17IK). The holding potential was kept equal to the resting potential. The membrane current was monitored through a current-voltage converter connected to the bath and held at virtual ground. Three types of experiments were performed : in the first section cells which had been cultured at 18 °C for at least 2 weeks were tested at short-term experimental temperatures (te) between 10 and 25 °C. Experiments in the second section were performed at te = 18 °C using cells grown at culturing temperatures (tc) of 10, 18 and 25 °C. The third type of experiments served as controls, with the experimental temperature kept equal to the culturing temperature.
Short-term temperature effects
Resting potential and input resistance
The membrane resting potential of a cell cultured at 18 °C and bathed in an experimental solution of 1 mm-CaCl2 + 1 mm-KCl was typically −36 mV at 25 °C and was not significantly affected by temperature (t-test, P⩽0·05) (Table 1, Fig. 1). There were rapid fluctuations in potential that appeared smaller during a slow decrease in temperature (Fig. 1A) than during an increase (Fig. IB). The input resistance varied with short-term modifications of the experimental temperature from 67 MΩat 10 to 41 MΩ at 25 °C corresponding to a temperature coefficient (Q10) of 0·7 (Table 2).
Responses to brief current pulses
The peak value of the action potential following brief square current pulses is sigmoidally related to the current strength (Naitoh, Eckert & Friedman, 1972). Fig. 2 shows that there was an increase in threshold of excitation from +3 mV to +5 mV when temperature was raised from 10 to 25 °C. The action potential amplitude was decreased whereas the rates of rise and decay were increased.
Responses under voltage clamp
Under voltage clamp, the effects of temperature upon the calcium conductance (Gca), which is rapidly activated, could be distinguished from the effects on the potassium conductance (GK), which activates after some delay. Cells cultured at 18 °C were tested at four short-term experimental temperatures, 10, 15, 20 and 25 °C. At all temperatures tested the early current rose up to step amplitudes of 30 mV and declined beyond depolarizations of 35 mV. With increase in temperature, the early inward current showed an increase in rate of rise (Fig. 3A), and increase in maximum value (Fig. 3B) from −5·4 nA at 10°C to −9·6 nA at 25°C corresponding to a Q10 of 1·5 (Table 2). The steady-state I /V-relationships were essentially unchanged between 15 and 25 °C showing a familiar sigmoidal course. At 10 °C, however, the delayed outward and inward currents were reduced. Accordingly, Q10 values of the steady-state inward current (following −60 mV step) and outward current ( + 35 mV step) were close to 2·5 between 10 and 15 °C, but only 1·3 over the total range of temperatures Usted (10–25°C; Table 2).
Early calcium conductance
For approximations of E′ca during maximal channel activation we assume a transient value for [Ca2+]i of 10−4M. This rough estimate does not take into account (1) differences in inward fluxes and (2) effects on transient [Ca2+]i of temperature-dependent Ca pumping. The recorded early currents are sums of voltage-activated inward Ca and outward K currents. Short-circuiting of Ca current by K current is small with depolarizations up to 20 mV. Using the late outward current for reference the short-circuiting during the peak inward current was maximally 12–22% (Figs 3, 8, 9) considering the difference in rates of activation of the Ca and K conductances. With these restrictions in mind we have determined the voltage/conductance relationship at different experimental temperatures (Fig. 4).
With cells cultured at 18 °C, average maximal conductances per cell (Gmax) of 30(7 to 330 nS were found at temperatures between 15 and 25 °C. At 10 °C, Gmax was reduced (230 nS) and the slope of the conductance/voltage curve was diminished.
Early current time course
The time interval between the onset of the depolarizing voltage step and the peak of the early inward current (‘time-to-peak’) is a measure of channel activation. For voltage-sensitive Ca channels and Na channels time-to-peak decreases with rising step amplitude. Change in the logarithm of time-to-peak with temperature (Fig. 5A) suggests that time-to-peak decayed exponentially with depolarizations of up to 30 mV ; beyond this value the decay rate was reduced. When the experimental temperature was raised, time-to-peak decreased (Table 1) without major effects on its voltagedependence. Semilogarithmic plots of the rate of current activation ( = reciprocal time-to-peak) versus experimental temperature show that this relationship is exponential (Fig. 6) with a Q10 value of 2·3 (Table 2).
Long-term temperature effects
We have explored the long-term effects of environmental temperature upon the electrical properties of the Paramecium membrane by culturing cells for at least 2 weeks at preselected temperatures and then testing them at the temperature of 18 °C. In some cases the experimental temperature was kept equal to the culturing temperature.
The resting potential and input resistance were not affected by the culturing temperature in the range 10–25 °C (Table 1).
The membrane responses to brief current pulses were also unaffected. For an experimental temperature of 18 °C the threshold of the regenerative response occurred uniformly between 4 and 5 mV. The stimulus/response relationship and the action potential amplitude (45–50 mV) were independent of the culturing temperature (Fig. 7).
Responses under voltage clamp
The early inward current following a depolarizing voltage step decreased with increasing culturing temperature (Fig. 8). These temperature-dependent differences in current amplitude were most conspicuous when applying a step voltage of 35 mV or more. The early current maximum showed a negative shift along the voltage axis when the culturing temperature was increased. The late l/V-relationship suggests that this shift may be due, in part, to an increase in short-circuiting outward current during excitation. It is interesting to observe that a high culturing temperature increased the late outward current, but decreased the steady-state inward current during hyperpolarization; this suggests that a temperature-sensitive conductance may be common to both types of voltage-activated current. The Q10 value of the steady-state inward current equals that of the early inward current (0·8 for −60 mV step; Table 2) while the steady-state outward current shows a pronounced inverse temperature relationship (Q10 =1·7 for +35 mV step).
Long-term and short-term temperature exposures affect the electrical parameters of the membrane in a complex manner. The early inward current increased as the experimental temperature was raised (Fig. 3), but decreased at higher culturing temperature (Fig. 8). As a result of such partial cancelling, the maximum inward current was less increased when tested at the culturing temperature (Fig. 9).
Comparison of Fig. 9 with Fig. 3 also indicates that the steady-state currents are primarily a function of the short-term experimental temperature. The Q10 values (T5) resemble those calculated from data where the culturing temperature was held constant (Q10= 1·3; Table 2).
Stimulating cells at low culturing temperatures increased the duration of the early inward current (Fig. 9A). The amount of net charges passing the membrane during the early current was considerably reduced in 25°C cells (+20mV step: 6pC) with respect to 10 °C cells (12 pC). We conclude that long-term temperature exposure does not help to stabilize the amount of net charges transferred across the membrane in the course of voltage-activated early current.
Early calcium conductance
Applying our premises on the Ca equilibrium potential (see above) modifications of the culturing temperature decreased Gmax from an average of 540 nS at tc = 10 °C to 250 nS at tc = 25 °C (Fig. 10A). Testing cells at their culturing temperatures led to unaltered early conductances of about 300 nS. As temperature was raised the G/V relationship was shifted in the negative direction of the voltage axis (Fig. 10B).
Early current time course
Modifications of the culturing temperature had no effect upon the time-to-peak of the early inward current, which corresponded to values seen upon short-term modification of the experimental temperature (Fig. 5, compare B and C with A). Accordingly, the rate of activation of the early current was a function of the short-term experimental temperature, irrespective of whether or not the culturing temperature had been modified (Fig. 6).
The membrane resting potential in Paramecium is essentially independent of temperature within the ranges tested (Table 1). For a temperature increase of 10°C estimates of Eca and EK (Ogura & Machemer, 1980), together with a constant conductance ratio Gca/GK, predict a negative shift of 1·2mV of the membrane resting potential according to the Hodgkin-Horowicz equation. The observed average shift of 2mV/10°C approximates to the theoretical value. Our results confirm earlier measurements of the resting potential of P. caudatum in extremely diluted sea water (2 mV/10 °C; Yamaguchi, 1960) and in Stylonychia mytilus (1·5 mV/10 °C; de Peyer & Machemer, 1977). Thus, for ciliates so far tested, the ratio of resting conductance of Ca to K appears to be constant within a certain range of temperatures. For the symbiotic flagellate Opalina a negative shift of the resting potential of 3·6 mV/10 °C has been observed (Ueda, 1961). Stability of the resting potential upon cooling has also been reported for squid giant axon (Hodgkin & Katz, 1949) and vertebrate myocardial fibres (Trautwein & Dudel, 1954).
Hennessey, Saimi & Kung (1982) have recently recorded a depolarization in Paramecium which occurred upon rapid increase of the environmental temperature (24°C/min). No potential response was seen upon lowering the temperature. The heat-induced potential was graded with the temperature increase and related to a raised somatic Ca conductance. The time course of this very slow response to rapid changes in temperature shows some resemblance to fluctuations in membrane potential observed during slow heating, not cooling (compare Fig. IB with A). Toyotama (1981) was able to record a depolarizing potential in Paramecium upon lowering the temperature at a rate of 1 °C/min between 25 and 20 °C. These so far heterogeneous data are not sufficient for discrimination between a ‘thermoreceptor potential’ and transient instabilities of the membrane resting potential seen during temperature changes. The available evidence suggests, however, that the membrane is differently sensitive to heating and cooling.
A raised threshold depolarization and reduction in amplitude of the action poten-Kal, observed with short-term increase in temperature (Fig. 2), is consistent with a raised membrane conductance, predominantly of K. Similar temperature effects on threshold and action potential amplitude have been reported from squid axon (Hodgkin & Katz, 1949; Tasaki & Spyropoulos, 1957) and vertebrate heart tissue (Traut-wein&Dudel, 1954; Marshall & Williams, 1956; Illanes, Carpentier & Reuss, 1970). Short-term increase in temperature increased voltage-activated early and late currents (Q10 = 1·5 and 1·3, respectively; Table 2). The observed Q10 of rates of activation of the early current (2·3) is in accordance with results on squid axon (Hodgkin & Huxley, 1952) and amphibian nerve and muscle (Schwarz, 1979). The absence of an inflection in the semilogarithmic plot of rate of activation versus experimental temperature (Fig. 6) suggests that no phase transitions in membrane lipids and/or channel proteins of Paramecium occur at temperatures between 10 and 25 °C.
Long-term exposure to different temperatures affected the maximal early conductance, which decreased with rising culturing temperatures (Q10=0·6; Table 2). A comparison of the stimulus-response curves (Figs 2, 7), however, clearly indicates the leading role of the experimental temperature in determining the rate and amplitude of the action potential. The failure of the diverging curves of early and late current (Fig. 8), to modify the stimulus-response curve (Fig. 7), is puzzling at first sight. Inspections of the current time courses in Fig. 8A show, however, that the decay rate of the early inward current rose with the current size, not with depolarization. The size of the late outward current was not related to the rate of decay of the inward current, so that K current activation does not appear to limit the inward current decay. An inactivation of the voltage-sensitive early Ca channel and activation of the late K channel by intracellular ionic Ca has been recently demonstrated in ciliates (Brehm & Eckert, 1978; Deitmer, 1983) and in metazoan tissue (Meech, 1978; Tillotson, 1979). Our data are in line with the view of a Ca-dependent feedback control of the Ca conductance; such control can obviously help to stabilize the action potential at different temperatures.
We presume that the observed rise in early peak conductance following long-term cooling (Fig. 10B) rests on an increase of the number of voltage-activated Ca channels. Since the action potential does not benefit from this increase in conductance (Fig. 7), one may ask if it is functionally meaningful. It should be remembered that excitability in ciliates serves in the regulation of the internal Ca concentration (Eckert, 1972) nd thereby the control of the ciliary motor response (Machemer & Eckert,1973). Cooling has multiple effects on electric membrane properties, the chemomechanical transduction process of the ciliary axoneme, and Ca pumping from the ciliary space, so that the relationship between ciliary motility and temperature is complex. Short-term cooling of Paramecium predominantly depresses the reversed ciliary beating following an action potential (Machemer, 1974). Such altered motor responses can modify the phobic behaviour of cells, which contributes to their orientation in temperature gradients. Long-term cooling, on the other hand, increased the early peak conductance (Fig. 10A) and raised the amount of Ca influx (Figs 8A, 9A). It is conjectured that these long-term temperature effects serve to modify the time course of stimulus-induced increments in intraciliary Ca concentration with the effect that reduced frequencies of ciliary beating are associated with extended beating periods. As a result a similar amount of work is done by the cilia so that a specific behavioural response can be performed over a wide range of environmental temperatures.
We thank Drs J. W. Deitmer, T. H. Hennessey and Y. Naitoh for reading the manuscript. This work was supported by the Deutsche Forschungsgemeinschaft, SFB 114, TP A5.