GTP-γ-S Increases the Duration of Backward Swimming Behavior and the Calcium Action Potential in Marine Paramecium

The ciliated protozoan Paramecium has been used as a biological model to study the ionic mechanisms of excitability in living cells. The surface membrane in Paramecium exhibits properties of excitability similar in fundamental respects to those exhibited by metazoan nerve and muscle cells (Eckert and Brehm, 1979). To understand the mechanism by which the cells control the influx of calcium, we studied the role of guanine nucleotides on voltage-dependent calcium channels present in the marine ciliate Paramecium calkinsi.

Voltage-dependent calcium channels can be found in virtually all animal cells and they play a critical role in cell function. More than one type of calcium channel has been identified (Carbone and Lux, 1984; Armstrong and Matteson, 1985; Nowycky et al. 1985; Fedulova et al. 1985), some of which can be modulated by neurotransmitters and hormones through various metabolic pathways (Reuter, 1983; Levitan and Kaczmarek, 1987; Rosenthal et al. 1988b). Although evidence from mammalian nerve and cardiac cells indicates a role for G-proteins in regulating L-type calcium channels (Dunlap et al. 1987; Rosenthal et al. 1988a; Holz et al. 1986; Yatani et al. 1987), it still is unclear whether G-proteins modulate the T-type calcium channel.

We chose to work on Paramecium for several reasons, (i) It is the most primitive organism that shows voltage-dependent calcium channels (Oertel et al. 1977; Ehrlich et al. 1988). (ii) Its voltage-dependent calcium channels reside exclusively in the surface membrane covering the cilia (Dunlap, 1977; Machemer and Ogura, 1979). (iii) The biophysical and pharmacological properties of its calcium channels are well known. The calcium channel in Paramecium appears to be most like the T-type calcium channel found in mammalian cells, which has a small conductance, is inactivating and is not sensitive to dihydropyridines (Ehrlich et al. 1984, 1988). (iv) The duration of its backward swimming stimulated by high-potassium solution is correlated with the amplitude of the voltage-dependent calcium current (Haga et al. 1982). This correlation means that the backward swimming behavior in Paramecium can be used as a simple bioassay to estimate the activity of the voltage-dependent calcium channels. Alteration in calcium channel function as assayed by this behavioral response can then be confirmed by more direct methods, such as electrophysiological techniques.

Behavior and electrophysiological experiments described in this paper have shown that guanine nucleotides modulate calcium influx, suggesting that a G-protein-dependent process may regulate calcium channels in Paramecium. Some of this work has appeared in preliminary form (Bernal and Ehrlich, 1989a; Mcilveen and Ehrlich, 1988).

The marine ciliate Paramecium calkinsi was cultured at room temperature in 25% artificial sea water which was made as follows: stigmasterol, 5 mg ml⁻¹; caseamino acids, 0·3 g 1⁻¹; 125 mmol 1⁻¹ NaCl; 10 mmol 1⁻¹ KC1; 1 mmol 1⁻¹ sodium citrate; 5 mmol 1⁻¹ CaCl2; 15 mmol 1⁻¹ MgCl2; 10 mmol 1⁻¹ Mops-NaOH; pH7·3. Cells were grown in 25 % artificial sea water because the cells grew more rapidly at this salt concentration than in full-strength sea water (D. Cronkite, personal communication). This growing method did not modify the properties of the calcium regenerative response of Paramecium (Deitmer and Machemer, 1982). The cells were fed with Enterobacter aerogenes 12–24 h before the experiments.

The duration of the backward swimming behavior in Paramecium was determined in high-potassium solution (‘testing’ solution). Briefly, Paramecium from the normal culture medium were harvested, washed and resuspended in the ‘resting’ solution which contained: 125mmoll⁻¹ NaCl; 1 mmol 1⁻¹ CaCl2; 10 mmol 1⁻¹ Mops; pH 7·3. Some Paramecium (5–10 cells) were transferred from the resting solution to the testing solution, which contained: 62·5 mmol 1⁻¹ NaCl; 62·5 mmol 1⁻¹ KC1; 1 mmol 1⁻¹ CaCl2; 10 mmol 1⁻¹ Mops; pH 7·3. After the Paramecium had been transferred to the testing solution, they immediately started swimming backwards. This behavior was maintained for approximately 180 s, at which time they started swimming forwards again.

To evaluate the role of guanine nucleotides in the backward swimming behavior, compounds of interest were incorporated into cells by a reversible permeabilization procedure that consists of the following steps, (i) To reduce the extracellular calcium concentration, the cells from the culture medium (5 ml) were centrifuged and resuspended three times in 2.5 ml of low-calcium solution (125 mmol 1⁻¹ NaCl; 0.1 mmol 1⁻¹ CaCl2; 0·01 mmol 1⁻¹ EDTA; 25 mmol 1⁻¹ Hepes; 1 mg ml⁻¹ bovine serum albumin; pH 7·7). A hand centrifuge was used to pellet the cells, (ii) The Paramecium were permeabilized by adding EDTA to a final concentration of 1 mmol 1⁻¹. The dye solution (6 mmol 1⁻¹ 6-carboxyfluor-escein in low-calcium solution) was added with the compounds of interest 30 s before EDTA. (Only dye solution was added in control experiments.) (iii) MgCl2 (1·2 mmol 1⁻¹) was added to reseal the cells 2 min after addition of EDTA. If the cells were allowed to remain perme⁻bilized for longer than 3 min, the yield of viable cells is dramatically reduced (C. Kung, personal communication). The cells were washed three times and resuspended in 2·5 ml of ‘wash’ solution containing 125 mmol 1⁻¹ NaCl; 1 mmol 1⁻¹ CaCl2; 0·01 mmoll⁻¹ EDTA; 25 mmol 1⁻¹ Hepes; pH7·7. Finally, (iv) the Paramecium were transferred to the resting solution and then tested for backward swimming behavior.

Only those Paramecium that fluoresced green, usually 85–90%, were studied. Care was taken to minimize the time for which the cells were observed at a wavelength that activated the dye to fluoresce, because prolonged activation of the dye often killed the cells, probably through the formation of free radicals within the cell. After it had been established that a cell had been permeabilized, the cell was observed by bright-field optics.

To isolate the calcium action potential, sodium and potassium currents were inhibited by pharmacological agents and substitution of non-permeant cations for Na⁺ and K⁺. The extracellular solution used was as follows: 125mmoll⁻¹ TEA-Cl; 10 mmol 1⁻¹ CsCl; 5 mmol 1⁻¹ 4-aminopyridine; 5 mmol 1⁻¹ 1,3-diaminopyri-dine; 15 mmol 1⁻¹ CaCl; 10 mmol 1⁻¹ Mops; pH 7·3. The electrodes were made from borosilicate thin-walled capillaries with an internal filament. Capillaries were washed by placing the glass in nitric acid (71 %) and boiling by adding drops of ethanol (95 %) for 1 h. Then the glass was washed five times with deionized water. The electrode-filling solution was as follows: 100 mmol 1⁻¹ CsCl; 2 mmol 1⁻¹ MgCl2; 10mmoll⁻¹ Mops-CsOH; 0·1mmoll⁻¹ Fast Green; pH7·0. The Fast Green (Sigma) was added to the electrode-filling solution so that the progress of the pressure injection into the cell could be followed. Electrodes filled with this solution had a resistance about 80 MΩ. To improve the electrical properties of the electrodes and the pressure injection procedure, the electrodes were beveled (Micropipete beveler model BV-10, Sutter Ins. Co., San Rafael, CA), so that the final resistance of the electrode was around 10MΩ.

Conventional intracellular recording experiments were made in single-electrode current-clamp conditions at room temperature. Because each Paramecium moves all the time, it was necessary to catch one in order to impale it for electrical measurements. This was done using a modification of the procedure described by Naitoh and Eckert (1972). Briefly, cells were suspended in a very small drop (approx. 10 μl) so they could not swim far. Our modification was that the viscosity of the medium was increased by adding about 10 μl of methylcellulose (Protoslo, Carolina Biological, Burlington, NC). In this solution the cells swam much more slowly so it was easier to impale one. A cell was found and held with an empty glass pipet and, after the methylcellulose solution had been washed away, the cell was bathed in the extracellular solution described above and impaled with an intracellular electrode. Changes of the membrane potential and injection of current were measured through the same electrode using a preamplifier in bridgemode configuration (Dagan 8500, Mineapolis, MN). The guanine nucleotides GTP, GTP-y-S and GDP-/β (Boehringer-Mannheim) were injected into the cells by pressure (Picospritzer II, General Valve Co., Fairfield, NJ) through the recording electrode. The pipets were filled with 10mmoll⁻¹ of the nucleotides dissolved in the electrode-filling solution described above. Before cellular impalement and at the end of the experiment, the permeability of electrodes was tested by applying pulses of 10–25 ms and 15–35 ×10³ Pa. In a successful experiment the cell was stained blue after injection of the drugs, which indicated that the drug was also inside the cell.

To estimate the final concentration of the solution injected into the cell, we assumed (i) that the cell is a cylinder 100 μm long and 20 μm in diameter, (ii) that the maximum volume injected was 10 % of the cell’s volume and (iii) that the pipet concentration of nucleotides was 10 mmol ⁻¹. From these approximations, we calculated that the maximal concentration of nucleotides injected into the cell could be 3 μmol 1⁻¹.

Because backward swimming behavior is a simple bioassay for estimating the activity of the voltage-dependent calcium channels in Paramecium (Haga et al. 1982), behavioral experiments using Paramecium calkinsi were made to test the hypothesis that guanine nucleotides regulate calcium channels via G-protein modulation. The cells were reversibly permeabilized (see Materials and methods) so that the compounds of interest were incorporated inside the cell. The duration of backward swimming was measured after cells had been depolarized with K⁺ (testing solution), allowing an indirect estimation of the length of calcium channel open time in control and experimental conditions. In all the experiments presented, control cells were permeabilized and resealed as for test cells except that only 6-carboxyfluorescein was incorporated into the cell. These control treated cells and completely untreated cells swam backwards for the same time after stimulation with testing solution. The compounds shown in Fig. 1 were tested. GTP-γ-S, which binds to and activates G-proteins, had a significant effect on backward swimming. GTP-γ-S (100 μmol 1⁻¹) induced cells to maintain backward swimming for at least 900 s, at which time the effect was considered irreversible relative to the mean of the control backward swimming (180 s), and the observation was terminated. Indeed, several cells were observed to maintain backward swimming for 45 min. Because GTP-γ-S is a tetralithium salt, 100 μmol 1⁻¹ lithium was incorporated into the cell as a control. Lithium did not alter the duration of backward swimming. To determine if related compounds could reproduce the effect of GTP-γ-S, other compounds such as (GDP-β-S), (ATP-γ-S), (AMPPNP), 8-bromo cyclic AMP (8-Br-cAMP) and 8-bromo cyclic GMP (8-Br-cyclic GMP) (all at 100μmoll⁻¹) were incorporated into the cell. None of these compounds had significant effects on the backward swimming of Paramecium (Fig. 1) as determined by analysis of variance followed by Duncan’s multiple range test.

The concentration dependence of the effects of GTP-γ-S are shown in Fig. 2, which compares this response with those to GDP-β-S. It can be seen from the upper curve of Fig. 2 that GTP-γ-S at concentrations below 10⁻⁹moll⁻¹ had no effect, while at concentrations of 1 nmol 1⁻¹ the duration of backward swimming was increased fivefold. Increasing the concentration of GTP-γ-S 100 times had no further effect. Over the concentration range tested, GDP-β-S had no clear effect on the duration of backward swimming. However, when a concentration of GTP-γ-S that gives a maximal effect (1 μmol 1⁻¹) and a concentration of GDP-γ-S that was ineffective in altering the behavior (10 μmol 1⁻¹) were added to the cell simultaneously, the cell returned to forward swimming in 180s, a time no different from that of untreated cells (Fig. 3).

To determine whether the alteration in the backward swimming duration was dependent on calcium channels, the calcium channel blocker W-7 (Hennessey and Kung, 1984; Ehrlich et al. 1988) was used to inhibit the calcium influx. When W-7 (100μmoll⁻¹) was added before depolarization, the cell did not swim backwards upon depolarization, regardless of the presence of GTP-γ-S. This result indicates that extracellular calcium is needed to initiate backward swimming. When 100μmoll⁻¹ W-7 was added 5 min after the cell had been depolarized, the result was quite different. Cells without GTP-y-S inside had already begun to swim forwards, so the addition of W-7 was without effect (N=10). However, the GTP-γ-S-treated cells began to swim forwards immediately after addition of W-7 (N=10). This result suggests that, in the presence of GTP-γ-S, the calcium channel remains open and that calcium continuously enters the cell to maintain backward swimming. Thus, addition of W-7, by blocking the calcium influx, can allow the cell to return to forward swimming.

The behavior experiments described above showed that the backward swimming behavior of Paramecium can be modulated by a guanine nucleotide, but they did not show directly that the changes produced by GTP-γ-S can be attributed to changes in the electrical properties of the cell. Therefore, compounds used in the behavioral assay were tested on the calcium action potential in Paramecium.

The resting membrane potential was measured as about –20 mV. Only cells that showed a stable membrane potential and no changes in the duration of the calcium action potential in the 5 min before the injection of the drug were included in this study. Injection of GTP-γ-S into the cell by pressure prolonged the duration of the calcium action potential (Fig. 4). The top panel of Fig. 4A shows a representative calcium action potential in control conditions before the injection of GTP-γ-S. One minute after injection of GTP-γ-S, the duration of the calcium action potential was at least doubled when compared with the control conditions (Fig. 4A, middle panel). Three minutes after the injection of GTP-γ-S, the duration of the action potential was three times longer than control and was still increasing (Fig. 4A, bottom panel). The time course of the change in the duration of the calcium action potential over 20 min, before and after the injection of GTP-γ-S, is plotted in Fig. 4B. Five minutes after the injection of GTP-γ-S the duration of the calcium action potential was increased fourfold and this effect was maintained over 15 min. In all cells tested, the maximal effect of GTP-γ-S was obtained within 8 min after injection and the duration of the calcium action potential increased by 322±49 %(S.D., N=5).

Because GTP-γ-S is a hydrolysis-resistant analogue of GTP, the effects of GTP were examined to determine whether this compound modified the calcium action potential in a reversible manner. As with GTP-γ -S, GTP increased the duration of the calcium action potential by 300% 6min after injection (Fig. 5). However, unlike the effect observed with the hydrolysis-resistant analogue, the duration of the calcium action potential declined after the maximum effect had been reached. By 14 min after the injection of GTP the duration of the calcium action potential had returned to the control value. Similar biphasic responses were observed in all three cells tested.

Since GTP-γ -S and GTP increased the duration of the calcium action potential, GDP-β S, which inactivates G-proteins, might be expected to have the opposite effect. Fig. 6 shows the changes in the calcium action potential 2 and 5 min after injection of GDP-β -S. Notice that, 2min after injection of GDP-β-S, both the duration and the amplitude of the calcium action potential were reduced. By 5 min after injection of GDP-β-S the calcium action potential was no longer seen and the cell showed a passive response to injection of current. In all of four experiments, GDP-β -S had similar effects.

Additional control experiments were made to eliminate the possibility that injected compounds other than guanine nucleotides may modify the calcium action potential in Paramecium. First, the electrode-filling solution (containing CsCl) without guanine nucleotides was injected into the cell by pressure. No changes in either the amplitude or the duration of the action potential were detected. Second, GTP-γ-S tetralithium salt dissolved in 10 mmoll⁻¹ Mops and 0·lmmoll⁻¹ Fast Green, pH 7·0, was injected. GTP-γ-S in this solution also increased the amplitude and the duration of the calcium action potential (data not shown). Finally, results similar to those described above were obtained when GTP-γ-S was injected using a KC1 or caesium citrate electrode-filling solution instead of the CsCl electrodefilling solution.

Sometimes the cells were slightly depolarized after injection (by about 5 mV), regardless of which compound was injected. However, in all the experiments described, we detected no consistent changes in the input resistance, judged by the response to hyperpolarizing pulses after injection of compounds. Despite the experimental manipulations to isolate the calcium action potential, it is possible that a contaminating current due to K⁺ or Cl⁻ could still be present and could modify the passive properties of the cell. However, preliminary results using voltage-clamped cells show that the calcium currents are indeed activated by injection of GTP-γ-S into the cell (Bernal and Ehrlich, 1989b).

The present findings demonstrate that both backward swimming behavior and calcium action potentials of Paramecium are prolonged when guanine nucleotides are incorporated into the cell. These results support the hypothesis that a guanine-nucleotide-dependent process modulates the voltage-dependent calcium channel in Paramecium and suggest that a G-protein can up-regulate one type of calcium channel (this paper) while down-regulating another type of calcium channel (e.g. Holz et al. 1986).

We were able to study the effects of guanine nucleotides in the intact Paramecium by incorporating several putative modulators into large numbers of cells that had been permeabilized by chelating extracellular calcium. The test compounds that were added to the bath then diffused into the cells and the cells were resealed by adding excess magnesium to the suspension. If calcium was used to reseal the cells, the cells died, presumably because of the increased intracellular calcium concentration. In addition, we found that the cells could not remain leaky for extended periods (>3min). The major drawback of resealing the cells quickly was that the distribution of the test compound most probably did not reach equilibrium. For this reason, exact intracellular concentrations of the test compounds are not known. In all cases the concentration of the test compound is stated as the final concentration in the bath and it is assumed that equilibration occurred. However, it is more reasonable to assume that the bath concentrations during the permeable state are overestimates of the final concentration of the test compounds inside the resealed cell.

The only compound that significantly altered the duration of backward swimming was GTP-γ-S (Fig. 1). This effect was concentration-dependent (Fig. 2). That the other compounds tested did not have a significant effect on the behavior suggests that the effect of GTP-γ-S was specific. The observation that independent increases in 8-Br-cAMP and 8-Br-cGMP did not prolong backward swimming (Fig. 1) indicates that an increase in the concentration of the cyclic compounds was not a necessary event in the cascade that is stimulated by GTP. These results also suggest that increases in GTP or GTP-γ-S concentration activated a G-protein rather than a guanylyl cyclase, because adding cyclic GMP did not mimic the effect of GTP and because the K1/2 for activation by GTP of the cyclase found in Paramecium is 70μmoll⁻¹ (Klumpp and Schultz, 1982) whereas the Kl/2 for Go, one type of G-protein, is less than 1 μmoll⁻¹ (Gilman, 1987).

To confirm our findings from the behavioral experiments that calcium channels were modulated by a GTP-dependent process, electrophysiological experiments were carried out to demonstrate directly the alteration of the calcium influx. The injection of GTP-γ-S into a Paramecium dramatically prolonged the duration of the calcium action potential (Fig. 4). In addition, GTP increased the duration of the calcium action potential transiently, as expected for a compound that can be degraded by the cell (Fig. 5). GTP could not be tested in the behavioral assay because the interval between incorporation of the compound into the cell and the assay was variable and often was greater than 5 min. As seen in Fig. 5, the effect of GTP after injection into the cell began to reverse after 6min.

The experiments using the behavioral assay did not show an inhibition of backward swimming duration when only GDP-β-S was incorporated in the cell whereas, in intracellular recording experiments, GDP-β -S injection (Fig. 6) had an effect opposite to that of GTP-γ-S on the calcium action potential (Fig. 4). However, when GDP-β-S was incorporated into the cell along with GTP-γ-S, we found that GDP-β -S had a profound effect on the behavior (Fig. 3), suggesting that GTP-γ-S and GDP-β-S can compete for binding. This result supports the expectation that the behavioral response is complex and so may underestimate some effects of the guanine nucleotides that can be demonstrated by more direct measurements.

Initially we were surprised that the calcium channels in Paramecium were activated by a GTP-dependent process while mammalian calcium channels were inhibited (Holz et al. 1986; Dolphin et al. 1988; Hescheler et al. 1987; Lewis et al. 1986). One explanation is that the Paramecium calcium channels resemble T-type calcium channels (Ehrlich et al. 1984, 1988), whereas the mammalian calcium channels that are altered by G-protein activation are L-type calcium channels. It has been suggested that T-type calcium channels in dorsal root ganglia cells are the target of G-protein modulation (Dolphin et al. 1989), because GTP-γ-S at low concentration (6μmoll⁻¹) increased the T-type calcium current by 54%, while subsequent applications of higher concentrations of GTP-γ-S abolished the calcium current with a half-time of about 1 min after the application of the increased concentration of GTP-γ-S. This inhibitory effect was also found in isolated cell patches in which 100μmoll⁻¹ GTP-γ-S was found to inhibit the activity of single T-type calcium channels. Our results are similar to the effect of the lower concentration of GTP-γ-S on calcium channels of the dorsal root ganglion cells. However, the effect of higher GTP-γ-S levels was different in Paramecium from that in the dorsal root ganglion cells. In both the behavioral experiments (Fig. 1) and in the electrophysiological experiments (data not shown) the effect of increasing the concentration of GTP-γ-S in Paramecium was to maintain the activation of the calcium signal. More comparisons are needed to understand the differences in modulation of the T-type calcium channel in these different cell types.

Although the results presented in this work are in accordance with the hypothesis that GTP-binding proteins modulate calcium channels in Paramecium, further work must be done to clarify the mechanism of this modulation. Preliminary results from our laboratory using voltage-clamped Paramecium suggest that the calcium currents are modulated by injection of guanine nucleotides into the cell (Bernal and Ehrlich, 1989b and unpublished observations). However, it remains to be determined which GTP-binding proteins are present in Paramecium and how these proteins interact with other metabolic pathways.

We gratefully acknowledge Drs Michael Forte, Laurinda Jaffe, Arnold Katz and James Watras for helpful comments on the manuscript. This work was supported by grant GM39029 from the NIH (BEE) and a fellowship from the American Heart Association, Connecticut Affiliate (JB). BEE is a Pew Scholar in the Biomedical Sciences.