Membrane Potential Responses to Thermal Stimulation and the Control of Thermoaccumulation in Paramecium Caudatum

Since Mendelssohn’s pioneering work (Mendelssohn, 1895, 1902) it has been known that Paramecium accumulate in regions with temperatures close to their culture temperature (the optimum temperature region) (see Jennings, 1906, for a review). More recently it has been shown that specimens increased their forward swimming velocity when moving towards their optimum temperature region (Tawada and Oosawa, 1972). In contrast, they exhibited frequent changes in swimming direction when moving away from this region (Nakaoka and Oosawa, 1977). These motile responses are the major causes of their accumulation in the optimum temperature region.

Specimens swimming across a temperature gradient are subjected to a difference in temperature between the anterior and posterior ends and to a change in temperature with time. By examining the accumulation of P. caudatum in solutions of differing viscosity and temperature gradients, Tawada and Miyamoto (1973) demonstrated that the rate of change in temperature around the specimens was an important aspect of thermal stimulation involved in causing the accumulation.

It is well known that locomotion mediated by ciliary beating in Paramecium is under the control of the cell surface membrane potential. That is, membrane hyperpolarization leads to an increase in forward swimming velocity (escape response) as a result of an increased ciliary beat frequency, while membrane depolarization leads to a slowing down of the forward swimming rate, a change in swimming direction or backward swimming (avoidance response) as a result of inactivation of ciliary beating and/or the reversal of the ciliary beating direction (Eckert, 1972; Naitoh, 1974, 1982; Machemer, 1975, 1988; Machemer and Sugino, 1989). These observations suggest that a specimen ascending a temperature gradient will experience a membrane hyperpolarization before it reaches the optimum temperature region and a membrane depolarization after it has passed through this region. In other words, a rise in temperature will cause a membrane potential response, the polarity of which will depend on the temperature around the specimen relative to the optimum temperature. The response must be hyperpolarizing if the temperature is lower than optimal and depolarizing if it is higher than optimal.

The responses of Paramecium membrane potential to thermal stimulation have been reported by several authors (Hennessey et al. 1983; Nakaoka et al. 1987; Inoue and Nakaoka, 1990). The primary objectives of the present paper are (1) to examine whether there is a polarity reversal that is dependent on the ambient temperature relative to the culture (optimum) temperature and (2) if this is the case, to determine the underlying mechanism.

Specimens of Paramecium caudatum (strain G3, mating type V) were cultured in a bacterized (Klebsiella pneumoniae) lettuce infusion at 25±2°C. After their final feeding, the culture specimens were kept immersed in a constant-temperature water bath (±0.1 °C) for more than a day. The temperature of the water bath was regarded as the ‘culture temperature’, T_c. Prior to experimentation the specimens were washed with a standard saline solution (4 mmol l⁻¹ KC1, 1 mmol l⁻¹ CaCl₂ and 1mmol l⁻¹ Tris–HCl or Mops–KOH buffer; pH7.2), and then maintained in this solution for more than 30 min at T_c.

To examine the locomotor activity of specimens in the boundary between two regions with different temperatures, a thin-walled (0.1 mm) glass capillary (40 mm length, 1mm diameter) containing about 50 specimens suspended in standard saline was placed on two copper blocks (20 mm × 20 mm × 5 mm) 0.5 mm apart. In less than 1 min, the temperature of the suspension in the capillary tube reached an equilibrium level dependent on the temperature of the copper block beneath it. When the temperature of the two blocks differed, a temperature gradient was established in the capillary immediately above the gap between the two copper blocks. This gradient was stable over the experimental time-frame. The temperature of each block was stabilized (±0.1 °C) with the aid of an electronically controlled Peltier module. The lower temperature was regarded as the ‘experimental temperature’, T_e. The difference in temperature between the two regions was always 10°C, irrespective of T_c. Fig. 1 shows a typical temperature profile at the boundary. The maximum temperature gradient was 2.5°mm⁻¹. The temperature profile did not change significantly for different values of T_e

To examine locomotor responses of specimens to a sudden rise in temperature, about 50 individuals suspended in standard saline solution (45 μl) were introduced into a thin-walled (0.1mm) rectangular glass vessel (15mm×l5mm×0.2mm). The vessel was then put on a laminated nichrome heater with a copper block (15 mm×l5 mm×5 mm) beneath both. When the temperature of the suspension in the vessel had been kept constant at a particular T_e (±0.1 °C) for more than 2 min with the aid of an electronically controlled Peltier module placed beneath the block, the temperature was suddenly and temporarily raised by applying an electric current pulse to the nichrome heater.

The behaviour of the specimens in the experimental chamber was recorded on videotape, and their locomotor activities were determined by analyzing the tapes frame-by-frame with the aid of an ultrasonic digitizer and a microcomputer. The locomotor activities determined were: (1) the swimming velocity (V_s), which is the distance along the swimming path accomplished by a specimen in unit time; (2) the linearity of swimming (L), which is the ratio of the linear distance covered by a specimen in unit time to its corresponding distance along the swimming path; and (3) the turning frequency (F_t, the frequency of the avoidance response per unit time).

The responses of a specimen’s membrane potential to thermal stimulation were examined using conventional electrophysiological techniques (Naitoh and Eckert, 1972). The temperature of the experimental vessel for electrophysiology was kept constant at the appropriate T_e (±0.2 °C) with the aid of an electronically controlled Peltier module placed beneath the vessel.

A microheater of the Nicklas (1973) type was made for thermal stimulation of a specimen impaled by microelectrodes (Fig. 2). The middle portion of a 0.1mm thick, U-shaped piece of tungsten wire was electrolytically polished to 0.05 mm in diameter in the central region. This thin region of the wire was covered with solder glass, while the rest of the wire was covered with silicon paste and cashew nutshell liquid for its electrical insulation. The insulated tungsten wire was glued onto a glass capillary mounted on an acrylic holder for placement on a micromanipulator.

When a current pulse (50 ms duration) was applied to the microheater, heat was generated primarily in this thin region of the wire. This caused a transient rise in the temperature of the surrounding solution. Time courses of the rise in temperature at different distances from the tip of the heater were monitored by a small thermocouple (0.15 mm tip diameter, 50 ms time constant) and are shown in Fig. 3A. The rise was faster and larger in the region closer to the tip of the heater. By the time that the temperature in a region 50 μm from the tip had reached its maximum, there had still been little rise in the temperature of a region 250. μm from the tip. Using this procedure, one end of an animal can be subjected to a substantial thermal shock while the other end is relatively unaffected by the stimulus.

The relationship between the peak rise in temperature and the square of the intensity of the current applied to the heater was almost linear. The square of current intensity is proportional to the heat generated at the heater tip. An example of this relationship in a region 50 μm from the heater tip is shown in Fig. 3B. All the experiments were performed at a controlled room temperature of 23 ±3 °C.

When specimens swimming in the lower-temperature (T_e) region of the capillary encountered the boundary with the higher-temperature region, they used an avoidance response to change their swimming direction. This avoidance response was more common when T_e was closer to T_c (25°C). This is well illustrated in Fig. 4, where the strength of the avoidance response (P_a) is plotted against T_e. P_a is defined as the ratio of the number of specimens that failed to enter the boundary because of the avoidance response to the total number of specimens encountering the boundary.

The temperature of the experimental chamber containing the specimens was slowly changed from 25°C (T_c) to T_e at a rate of 2.5°min⁻¹ and then kept at T_e for 1 min. The specimens were then subjected to a sudden 7.1 °C transient rise in their surrounding temperature at a maximum rate of 2.5°s⁻¹. This rate is comparable to that experienced by a specimen in the assay capillary when it enters the temperature boundary from the lower-temperature region at a swimming velocity of 1 mm s⁻¹ (the mean swimming velocity of specimens in standard saline solution at 25°C).

The specimens responded to this thermal stimulus with an avoidance response, which appears in Fig. 5 as an increase in turning frequency (F_t) and a decrease in the linearity of swimming (L). The avoidance response was less conspicuous when T_e was below T_c, and no avoidance response was observed at a T_e of 15 °C (Fig. 5A). The strength of the avoidance response was slightly reduced when T_c was above T_c (compare Fig. 5C with Fig. 5D).

The swimming velocity of the specimens (V_s) decreased concomitantly with the avoidance response, then increased as the response ceased (Fig. 5C). The increase in V_s was clearly observed when the avoidance response was infrequent (e.g. when T_e was lower than T_c: Fig. 5A,B).

The response of a specimen’s membrane potential to a sudden rise in the surrounding temperature was examined in two groups; one cultured at 25 °C and the other at 15 °C. The microheater was placed about 100 μm from both ends of a specimen impaled by microelectrodes, so that the whole surface of the specimen was subjected to a virtually simultaneous rise in temperature when an electric current pulse (0.8 A, 1 s) was applied to the heater. The time course of the rise in temperature is shown in Fig. 6A. The maximum rate of rise was 4.7°s⁻¹.

When T_ewas equal to or greater than T_c, specimens responded with a membrane depolarization upon which spikes were sometimes superimposed (Fig. 6A, 25°C, 30°C; Fig. 6B, 15°C, 25°C). When T_e was lower than T_c, the membrane was transiently depolarized, then hyperpolarized for a sustained period (Fig. 6A, 15°C; Fig. 6B, 5°C). Because the specimens deteriorated when kept at a T_e higher than T_c, their responses became less conspicuous with time (e.g. Fig. 6A, 30°C and Fig. 6B, 25°C).

A localized thermal stimulus was applied to the anterior or posterior region of a specimen using a microheater placed 50 μm from the appropriate end (Fig. 3A). Representative examples of membrane potential responses are shown on the right-hand side of Fig. 7. The upper set of traces shows the responses of specimens tested at 25 °C, while the lower set shows specimens tested at 15 °C.

When T_e was equal to T_c (25°C), responses of both the anterior and posterior ends were depolarizing (Fig. 7A). The amplitude of each response (peak voltage with reference to the resting membrane potential) increased with an increase in the stimulus intensity. The anterior response was always larger and faster than the posterior response (traces to the right of Fig. 7A). The relationship between response amplitude and stimulus intensity is shown on the left-hand side of Fig. 7A.

When T_e was 15 °C, the posterior response became hyperpolarizing, while the anterior response remained depolarizing (traces to the right of Fig. 7B). The amplitude of both the hyperpolarizing and depolarizing responses increased with an increase in the stimulus intensity before saturating. The relationship between response amplitude and stimulus intensity is shown in the left-hand side of Fig. 7B.

In the next series of experiments, the membrane potential response to a localized thermal stimulus strong enough to evoke the maximal response was examined at various values of T_e while T_c was kept constant at 25°C. The relationship between the response and T_e is shown in Fig. 8 together with examples of some individual responses.

The anterior response was always depolarizing, irrespective of T_e. The amplitude of this depolarization increased at lower values of T_c (Fig. 8A). The depolarizing response was sometimes followed by a small hyperpolarization when T_c approached 15 °C.

In contrast, the posterior response was depolarizing when T_e was equal to or greater than T_c (Fig. 8, see also Fig. 7A). The depolarizing response decreased and was followed by a hyperpolarization when T_e was lower than T_c. When T_eapproached 15 °C only a hyperpolarizing response occurred (see also Fig. 7B). The relationships between the amplitude of each component of the response and T_e are shown in Fig. 8B.

In this series of experiments, the relationship between the amplitude of the membrane potential response to a localized thermal stimulus and T_e was examined using three groups of specimens, each cultured at a different T_c (15, 20 and 25°C).

The anterior responses were only slightly affected by varying T_c (Fig. 9A), whereas the posterior responses were strongly affected (Fig. 9B). The peak depolarization for each posterior response shifted in parallel with the shift in T_c. In other words, the amplitude of the posterior depolarizing response was always maximal when T_e was equal to T_c. Furthermore, the plot of the posterior hyperpolarizing response also shifted in parallel with the shift in T_c. In other words, the posterior hyperpolarizing response only occurred when T_c was lower than T_c, regardless of the value of T_c.

T_e-dependent polarity reversal of the membrane potential response to an overall thermal stimulus and its relationship to thermoaccumulation

Specimens of P. caudatum exhibited an avoidance response to a rise in their surrounding temperature (an overall thermal stimulus) when they were in a region with a temperature (T_e) equal to or higher than their culture temperature (T_c). The response was less conspicuous when T_e was lower than T_c (Figs 4 and 5).

Specimens responded to an overall thermal stimulus with a membrane depolarization when T_e was equal to or higher than T_c, but with a membrane hyperpolarization when T_e was lower than T_c (Fig. 6). The suppression of the avoidance response at lower values of T_e is attributable to this polarity reversal of the membrane potential response.

As shown in Figs 7 and 8, a localized thermal stimulus applied to the anterior region of a specimen always produced a membrane depolarization. However, similar stimuli applied to the posterior region produced a membrane depolarization when T_e was equal to or higher than T_c, but a membrane hyperpolarization when T_e was lower than T_c.

Since the cytoplasm of Paramecium is virtually isopotential (Eckert and Naitoh, 1970), the membrane potential response of a specimen to an overall thermal stimulus is determined by the algebraic sum of the membrane conductance changes occurring in both its anterior and posterior regions. Therefore, when T_c is lower than T_c, a membrane depolarization generated in the anterior region is reduced in amplitude or overcome (reversed) by a membrane hyperpolarization generated in the posterior region. The overall response is therefore less depolarizing or even hyperpolarizing. In contrast, when T_e is higher than T_c, the overall change is always depolarizing (anterior and posterior responses are both depolarizing). Therefore, it can be said that the T_e-dependent polarity reversal of the membrane potential response to an overall thermal stimulus is caused by the T_c-dependent polarity reversal of the posterior membrane potential response, although the mechanism for the reversal remains unclear.

The temperature difference between the ends of a specimen ascending a temperature gradient is so small (less than 0.5°C even in a temperature gradient as sharp as that shown in Fig. 3A) that both ends are subjected to a virtually simultaneous rise in temperature (overall thermal stimulation). It is therefore presumed that a specimen ascending such a gradient produces a membrane hyperpolarization before it reaches the region having a temperature equal to T_c, since the temperature around the specimen (T_e) is lower than T_c. Thus, the specimen continues (or even accelerates) its forward swimming towards the region of higher temperature. In contrast, the membrane of the specimen is depolarized after it has passed over the region, because T_c then exceeds T_c. The specimen consequently makes an avoidance response that returns it to the region closer to T_c. The T_e-dependent polarity reversal must be a major cause of thermoaccumulation of specimens in a region with a temperature equal (or close) to T_c.

It should be noted that the threshold rate of rise in temperature for evoking a membrane potential response was lower for overall stimulation than for localized stimulation (Figs 6 and 7). Overall thermal stimulation affects a wider membrane area for a longer period than does localized stimulation. This might be a possible cause of the lower response threshold. Other aspects of thermal stimulation that affect the membrane potential response should be examined in detail.

Mendelssohn (1895,1902) and Jennings (1906) reported that the temperature of the region where specimens of Paramecium accumulated differed depending on the temperature at which the specimens had previously been kept equilibrated for several hours. This implies that the temperature at which the polarity reversal of the membrane potential response takes places (the reversal temperature) changes in accordance with the culture temperature. Hennessey and Nelson (1979) reported that the threshold temperature for thermal avoidance in Paramecium changed in accordance with the culture temperature.

We found that a hyperpolarizing response to an overall thermal stimulus was seen whenever T_e was lower than T_c, irrespective of the value of T_c (Fig. 6). This indicates that the reversal temperature does shift according to T_c as predicted by the behavioural observations.

We also demonstrated that a plot of the amplitude of the posterior hyperpolarizing response to a localized thermal stimulus against T_e shifted in conjunction with T_c (Fig. 9B), while a corresponding plot of the anterior depolarizing response shifted only slightly (Fig. 9A). It is therefore concluded that the T_c-dependent shift of the reversal temperature is attributable to the T_c-dependent shift of the posterior hyperpolarizing response. The mechanism underlying this shift remains unclear.

The peak of the plot of the amplitude of the posterior depolarizing response against T_e also shifted in conjunction with T_c. The peak always occurred at the value of T_c that was equal to T_c (Fig. 9B). The anterior and posterior depolarizing responses seem to combine to cause the specimen to exhibit an avoidance response to a thermal stimulus. The specimen thus remains in the area with a temperature equal or close to T_c.

Martinac and Machemer (1984) found that the input resistance of a specimen of Paramecium was higher when T_e was lower than T_c, whereas the voltage-activated maximum calcium conductance was not much affected by lowering T_c (see also Inoue and Nakaoka, 1990). We found that the amplitude of the anterior depolarizing response increased as T_e decreased (Figs 8A and 9A). The increased amplitude of this response might be attributable to an increased input resistance at low values of T_e if the heat-activated conductance change is not greatly affected by lowering T_e.

Nakaoka et al. (1987) found that, in dissected fragments of P. multimicronucleatum, the membrane potential response to a thermal stimulus was hyperpolarizing in anterior fragments, but depolarizing in posterior fragments. Our results for the posterior region are consistent with their observations, while those for the anterior region are inconsistent. Although the causes of the discrepancy have not been determined, it is highly probable that regenerated membrane at the cut end of a cell fragment will have thermoreceptive properties different from those of normal membrane. This problem should be further examined. It should be noted that many pioneering workers, such as Jennings and Jamieson (1902), Alverdes (1923), Koehler (1939) and Kamada and Kinosita (1940), examined thermal and/or chemical sensitivity in fragmented Paramecium.

It is well known that mechanical stimulation of the anterior region of Paramecium produces a depolarizing mechanoreceptor potential, while mechanical stimulation of the posterior region results in a hyperpolarizing one (Naitoh and Eckert, 1969). Application of a current pulse to the microheater caused an elongation of the heater, as well as expansion and convection of the solution around the heater. This heat-mediated mechanical turbulence might activate the mechanoreceptor channels. However, tapping the microheater or squirting the bath solution against the specimen did not evoke mechanoreceptor potentials. It is, therefore, highly probable that the membrane potential responses caused by application of an electric current to the microheater are caused by heat and not by mechanical turbulence. The ionic specificity and adaptation properties underlying mechanically and thermally evoked membrane potential responses will be described elsewhere.

This work was supported by grants from the Ministry of Education of Japan; from the Mitsubishi Foundation; and from the Honda Motorcycle Foundation. We are grateful to Professor D. Macer for his critical reading of the manuscript.