Effects of Raised [K⁺]_o And [Ca²⁺]_o on Conjugation and on Electrophysiological Parameters in the Marine Ciliate Euplotes Vannus

Conjugation in ciliates comprises a cascade of cell-to-cell communication events which follow a similar pattern in the various species (Miyake, 1981 ; Miyake & Nobili, 1984). The main stages are reciprocal stimulation of cells with complementary mating types (co-stimulation) and pair formation. Ionic and electrical parameters of the membrane have been studied mainly in relation to locomotion (for reviews see Kung & Saimi, 1985; Naitoh, 1984) and their role in conjugation has received relatively little attention (Akada, 1985; Hiwatashi, 1981 and personal communication; Kitamura & Hiwatashi, 1984; Naitoh, Onimaru, Takahashi & Ishizaka, 1981). The present investigation studies the effect of potassium ions upon conjugation in the marine hypotrich Euplotes vannus. Some studies of conjugation have previously been made in this species (Lueken & Oelgemöller, 1985, 1986). Potassium was studied because in Paramecium mutants pantophobiac, with a reduced Ca²⁺-dependent K⁺ current, there is a clearly documented (although not well understood) relationship between conjugation ability and locomotory behaviour (Hinrichsen et al. 1985).

Strains were derived from the syngen Naples (Italy) of the sibling species complex Euplotes vannus/crassus (Génermont, Machelon & Demar, 1985). Standard singlet clones were D35 (mating type D) and B24 (mtB). Electrophysiological measurements were performed on D35, which, if necessary, was mixed with doublet clones (mtB) for discrimination. Techniques for cultivation and for breeding of desired clones have been reported elsewhere (Lueken, Breer & Hartkemeyer, 1981 ; Lueken, Gaertner & Breer, 1983).

For electrophysiological recordings, the basic medium (EASW) contained (in mmoll⁻¹): NaCl, 430; KCl, 10; CaCl₂, 10; MgCl₂, 53; Hepps = EPPS (FlukaAG), 10; pH 8·0–8·1. Its osmolality was 1000 ±10 mosmol kg⁻¹ H₂O (Knauer Semi-micro Osmometer Type Digital). The same value was achieved, at increased K⁺ and/or Ca²⁺ concentrations, by appropriate reduction of [NaCl]. Ionic strength could not be held constant in solutions containing raised calcium concentrations.

For cultivation of clones, standard artificial sea water (SASW) was used (in mmoll⁻¹): NaCl, 465; NaHCO₃, 2·4; KCl, 10; CaCl₂, 10·4; MgCl₂, 24·8; MgSO₄, 28·1; _PH8·0–8·l.

Conjugation test media with ion concentrations similar to those in electrophysiological measurements were achieved by adding a suitable adjusting medium in the ratio 1:1 to cells in EASW. Additional tests were performed, with the same procedure, in SASW-derived media. Osmolalities ranged from 1000 (10 mmol l⁻¹ K⁺, 5·2 mmoll⁻¹ Ca²⁺) to 1290 (80 mmol l⁻¹ K⁺, 80 mmol l⁻¹ Ca²⁺) mosmol kg⁻¹ H₂O. The lowest pH value (Knick Digital-pH-Meter, electrode Ingold 405) was 7·6 (80mmoll^−l K⁺, 80mmoll⁻¹ Ca²⁺), the highest one 8·3 (10mmoll⁻¹K⁺, 10·4 mmol l⁻¹ Ca²⁺). All values lie within the range in which conjugation remains unaffected (Hartkemeyer, 1980). Ionic strengths also differed substantially without observable effects on the behaviour of cells.

Solutions in which Li⁺, Na⁺, Rb⁺ or Cs⁺ (all as chlorides) were used to replace K⁺ always contained a standard K⁺ content of 10 mmol l⁻¹.

Cells were transferred by sieving (nominal aperture size 5 μm) to fresh EASW or to SASW, respectively, at least 24 h before an experiment. To establish the experimental ionic conditions at the time of mixing the clones, 15 μ1 of cell suspensions (containing 400–600 cells) of either mating type were located in separate drops in depression slides. When all scheduled ion compositions with 15 replicates for each were prepared, 30 μl of the adjusting medium - for controls, EASW or SASW, respectively - was added to unite the two drops. Up to 150 mixtures could thus be started within 10min. If stimulated cells (see below) were to be treated, 30-μ1 samples of the stimulated cell mixture was added to 30 μl of the adjusting medium. Slides were stored in a moist chamber in a constant-temperature room at 24·5°C. The ionic composition of the medium was considered to be conjugation permissive when 20–90% of the cells were paired within 8h.

Fully stimulated cells were obtained after 60–90 min if cells were densely crowded by sieving (see above), as described previously (Lueken et al. 1983).

Electrophysiological recording techniques were analogous to those of De Peyer & Machemer (1977), with the use of an inverted microscope (Leitz Diavert). Both voltage and current electrodes with resistances of 30–50 MΩ were filled with 3 mol l⁻¹ KCl. The input resting resistance was determined 160 ms after starting the injection of a small hyperpolarizing constant-current pulse (0·2 nA) of 180ms duration. Solutions were exchanged by perfusion of the experimental chamber (containing 1·5 ml) with 10 ml of the new medium. The new bath junction potentials were set to zero (they did not exceed 1 mV up to 250 mmol l⁻¹ [K⁺]_o). Experiments were performed at room temperature (18–22°C).

The drug was added together with the adjusting medium to an effective concentration of 5 μmoll⁻¹.

A newly developed method, based on horseradish peroxidase and its substrate 4-chloro-l-naphthol (Merck), was used, as described by Lueken & Oelgemoller (1985).

Cells were exposed to experimental ionic conditions at either of two stages in the conjugation sequence: immediately after mixing the complementary mating types; or after the irritation regime had been applied to a cell mixture (this enables the cells to stimulate each other fully but prevents them from pair formation).

Raising [K⁺]_o from the standard 10 mmol l⁻¹ up to 50 mmol l⁻¹ had no measurable effects upon conjugation. At 55 mmol l⁻¹ the pairing rate was reduced. At 60 mmol l⁻¹ pairs occurred only rarely. At concentrations of 70 and 80 mmol l⁻¹ pair formation was entirely blocked. When cell mixtures were tested for concanavalin A binding sites after exposure to high [K⁺]_o for several hours, negative results were always obtained. When high-K⁺ solutions were diluted to half the potassium concentration, i.e. 30–40 mmol l⁻¹, cells started stimulation activities and mated after the normal waiting period of about 1 h. This was prevented if cycloheximide was added together with the diluting medium. Raised [K⁺]_o thus seems to interfere with the stimulation process and to prevent the cells from becoming capable of uniting.

Raising [Ca²⁺]_o alone up to 70 mmol l⁻¹ did not visibly affect conjugation. But the calcium ions antagonized the inhibitory effect of potassium ions on conjugation. At 60 mmol l⁻¹ K⁺, this was achieved by elevating [Ca²⁺]_o from 10·4 to 13 mmol l⁻¹. At higher potassium concentrations more Ca²⁺ was required: inhibition of conjugation at 70 mmol l⁻¹ K⁺ was annulled by at least 24 mmol l⁻¹ Ca²⁺, at 80 mmol l⁻¹ K⁺ by at least 70 mmol l⁻¹ Ca²⁺. Combinations of higher K⁺ and Ca²⁺ concentrations could not be tested because of precipitation in the adjusting media. It should be mentioned that the Donnan ratios ([K⁺]/[Ca²⁺]¹/²) for each minimal calcium concentration in conjugation-permissive ion combinations did not remain constant but decreased with increasing K⁺ concentrations (Table 1).

In all of the conjugation-permissive K⁺/Ca²⁺ combinations the onset of pair formation was remarkably delayed, to about 6h instead of the 1 h in controls.

In studies to see which other monovalent ions had effects like potassium (always with 10 mmol I^-1 K⁺ present), it was found that Rb⁺ and Li⁺ exerted the same effects as K⁺, whereas Cs⁺ and Na⁺ did not affect the conjugation process.

Cells behaved equally in test media derived from EASW, i.e. with adjustment of osmolality, and those derived from SASW, i.e. at increased osmolality.

If stimulated cells were exposed to at least 60mmoll⁻¹K⁺, they did not form pairs, although they possessed fully developed concanavalin A binding site fields when the treatment was started. Raised [K⁺]_o thus blocked pair formation for cells which were ready to unite. Again, Rb⁺ and Li⁺ had identical effects to K⁺, whereas Cs⁺ and Na⁺ were ineffective. The ability to form pairs remained in high [Ca²⁺]_o.

Additional Ca²⁺ applied together with the monovalent ions did not antagonize the inhibition of pair formation; cells did not pair even after more than 24h. This unexpected observation was further explored by adjusting stimulated cells to 60 mmol l⁻¹ K⁺. Cells did not pair. At different times, two representative samples were removed and tested separately. One was diluted to the conjugation-permissive K⁺ level of 30mmol l⁻¹, with cycloheximide added to prevent new stimulation: after 30min pairing occurred to the usual extent; after 60min the rate was substantially lower; after 120 min no pairing at all was observed. The other sample of each set was tested in parallel for the existence of concanavalin A binding site fields (oval fields = OFs). The percentage of cells positive for OFs diminished in correspondence with the loss of pairing ability. Thus the maintenance of pairing ability was too short for the Ca²⁺ to become effective. The antagonizing action of calcium on cells blocked by raised [K⁺]_o occurred after a delay of 5 h (see above).

The following data were obtained in experiments on vegetative as well as on conjugating cells before and after stimulation; no differences were observed between the different stages.

In EASW, the cells displayed a resting potential of −41 ±0·35 mV, N = 32, with a specific membrane resistance of 2·25 ±0·13×10⁴Ωcm², N = 8. (General electrophysiological features will be reported elsewhere.) The dependence of the resting potential on various concentrations of K⁺ or Ca²⁺, independently or in combination, is shown in Fig. 1. Elevating K⁺ concentrations up to 50mmoll⁻¹ caused no pronounced alteration of the resting potential; but above this concentration there was a linear relationship versus log [K⁺]_o, with a slope of 38 mV, as compared with the 58 mV per 10-fold change predicted from the Nernst equation if the membrane were to behave like a pure K⁺ electrode. The depolarizing action of extracellular K⁺ became evident at the same concentration as the inhibitory effects on conjugation. Also, raised [Ca²⁺]_o reduced the membrane potential; its depolarizing effect was observed at lower concentrations, from 32 mmol l⁻¹ up to 70 mmol l⁻¹. Then, deleterious effects became manifest, and at 100 mmol l⁻¹ Ca²⁺, some of the specimens disintegrated. A linear relationship between resting membrane potential and log [Ca²⁺]_o could not be established. If K⁺ and Ca²⁺ concentrations were raised together, in proportions such that inhibition of conjugation by K⁺ was annulled, the effect of calcium at all concentrations added to the effect of potassium, although to a small extent.

In contrast, antagonistic actions of the ions were exerted upon the input resting resistance (Fig. 2). After injection of equal inward current pulses at different ion concentrations (Fig. 2A) the resulting hyperpolarizations were ohmic-capacitive with a time constant of 20–25 ms. The amplitude decreased with increasing potassium concentrations (top row), but increased with increasing calcium concentrations (bottom row). If Ca²⁺ was added to stimulation-inhibiting K⁺ concentrations in amounts annulling inhibition (middle row), then hyperpolarization at each combination of concentrations occurred as in the standard medium. Correspondingly, the input resting resistances (Fig. 2B) diminished with rising [K⁺]_o within the range relevant for inhibition of conjugation, but increased if [Ca²⁺]₀ was raised. Combinations of both ions could balance the input resting resistance around its standard value with a slight dominance of the Ca²⁺ effect at higher concentrations.

Excess K⁺ suppressed the initial steps in intercellular communication. Excess Ca²⁺ did not affect conjugation, but when applied together with K⁺ it annulled the K⁺-induced inhibition. Although with increasing [K⁺]_o the Donnan ratios did not remain exactly constant, this antagonism points to a cationic exchange system on the cell membrane with competition for anionic binding sites according to the Donnan principle (Jahn, 1962). A similar observation has been made for P. caudatum (Hiwatashi, 1955), but was not pursued further (K. Hiwatashi, personal communication). Recently, it has been found that ions act upon intraclonal conjugation in Paramecium in the opposite direction to that in E. vannus-. high [K⁺]_o induces pair formation, high [Ca²⁺]_o prevents it (Kitamura & Hiwatashi, 1984). Removal of Ca²⁺ from the cell surface is supposed to be the deciding event. Kitamura & Hiwatashi consider that Ca²⁺ influx is involved, as it is in ciliary reversal (Kung & Saimi, 1985). Cronkite (1976) has accumulated strong evidence for Ca²⁺ influx in chemically induced conjugation. However, induction of conjugation is also possible for mutants with membranes defective in Ca²⁺ conductance (Hinrichsen et al. 1985). For a full explanation of these phenomena more direct observations are necessary.

It seems striking that depolarizing effects of K⁺ on the resting potential become measurable at exactly the concentration that inhibits conjugation. Since additional Ca²⁺ enhances depolarization, it cannot cancel the K⁺-induced block of co-stimulation by means of its influence on the resting potential. Instead, restoration of the input resting resistance could be the crucial event (Fig. 2). The causal relationship between the K⁺/Ca²⁺-dependent membrane resistance and the conjugation-suppressive action of K⁺ could be mediated by changes in the resting conductance of K⁺ channels. This resembles the pnt (pantophobiac) mutants of Paramecium, which show a reduced Ca²⁺-dependent K⁺ current. The connection to their inability to mate properly, however, remains contradictory, since there are genotypes with normal that still have difficulty in mating, and, conversely, without that can mate normally (Hinrichsen et al. 1985). Therefore, the participation of K⁺ efflux in ciliate conjugation remains doubtful. The substitution of K⁺ by Rb⁺ or Li⁺, but not Cs⁺ or Na⁺, is not to be expected for a single K⁺ channel. Possibly there are different types of channels involved. Hypotrichs, as can be judged from the freshwater species Stylonychia mytilus (Deitmer, Machemer & Martinac, 1984; Ivens & Deitmer, 1986), bear very heterogeneous populations of K⁺ and Ca²⁺ channels which offer many possibilities for regulation of conductance(s). Investigations of the marine ciliates, Paramecium calkinsi (Deitmer & Machemer, 1982) and Fabrea salina (Dryl, Demar-Gevais & Kubalski, 1982; Kubalski, 1985), have provided no information about conjugation.

In addition, elevation of [K⁺]_o might not only change resting conductances but also the potassium equilibrium potential, or, less specifically, the surface potential (Eckert & Brehm, 1979). Reduction of input resting resistance can only be attributed to reduction of resting membrane potential in the case of potassium; in the case of calcium, an increase in resistance accompanies the depolarization.

Pair formation of fully stimulated cells was inhibited by the same minimal concentration of K⁺ as for unstimulated cells, but there was no antagonism by additional calcium ions. This can be explained by the quick loss of the stimulated status. But it remains unclear why after some hours the cells did not start a new stimulation cycle in the conjugation-permissive medium.

The observations reported here seem to justify the assumption that the distribution of ions on the membrane and/or ion movements across the membrane, especially of K⁺ and Ca²⁺, play important roles in ciliate conjugation. Further investigation will perhaps explain the relationship between the involvement of these ions in phenomena such as sexual behaviour (e.g. prevention of pair formation in mixtures of different mating types of Euplotes and induction of intraclonal conjugation in Paramecium) and locomotory events such as ciliary reversal.

The authors thank Drs Tilly Bakker-Grunwald, E. Bakker and H. Breer for valuable suggestions, Mrs Barbara Oelgemöller for skilful assistance and Mrs Gitta Moehrke for preparing the manuscript. Dr Machemer’s (Bochum) and his collaborators’ substantial help and encouragement are gratefully acknowledged. The work was supported by Deutsche Forschungsgemeinschaft, SFB 171, C7.