Calcium Transients In Identified Leech Glial Cells In Situ Evoked By High PotaSsium Concentrations And 5-Hydroxytryptamine

Intracellular Ca²⁺ transients are widespread and are important steps in signal transduction leading to activation of cellular activity. In the nervous system, many functions of nerve cells, such as regulation and modulation of ion conductances, electrical activity and synaptic transmitter release, are mediated by a rise in intracellular [Ca²⁺] (Miller, 1987; Meldolesi and Pozzan, 1987).

In glial cells, much less is known about similar second-messenger pathways involving intracellular Ca²⁺ transients. Earlier studies have reported a rise in [Ca²⁺]_i in a cultured glioma cell line following application of transmitter substances such as 5-hydroxytryptamine (5-HT; Sugino et al. 1984). In recent years, evidence has accumulated that different kinds of stimuli, such as amino acid transmitters, norepinephrine, endothelin and mechanical deformation, may induce a rise in [Ca²⁺]_i in glioma and glial cells (Ashley et al. 1989; Hertz et al. 1989; Cornell-Bell et al. 1990; Jensen and Chiu, 1990, 1991; MacVicar et al. 1990, 1991; Salm and McCarthy, 1990). These studies have been carried out, however, on isolated glial cells in vitro, where the cells are deprived of their normal glial and neuronal environment.

In the leech central nervous system, glial cells can easily be identified (Coggeshall and Fawcett, 1964; Kuffler and Nicholls, 1966). Microelectrodes can be impaled into glial cells in the neuropile of leech ganglia to record membrane potentials and intracellular ion activities (Walz and Schlue, 1982; Deitmer and Schlue, 1987,1989). We have now developed a method of removing the ganglionic capsule and the neuronal cell bodies to expose the neuropile glial cells in situ, allowing direct measurement of intracellular ion activities using fluorescent dyes injected into a glial cell within the ganglion.

We have injected the Ca²⁺-sensitive fluorescent dye Fura-2 into the cells by passing current through a Fura-2-filled intracellular microelectrode, and have recorded intracellular Ca²⁺ transients following an increase in the external K⁺ concentration or the addition of 5-HT, a hormone and neurotransmitter in the leech nervous system (Wallace, 1981; Lent and Dickinson, 1984). Our results suggest that the elevation of [Ca²⁺]_i is, at least in part, due to Ca²⁺ influx through membrane channels and may modulate the K⁺ conductance of the glial mem-brane. A preliminary report was communicated at the joint meeting of the German and Dutch Physiological Societies (Munsch and Deitmer, 1992).

The experiments were performed on isolated ganglia of the leech Hirudo medicinalis L. Adult leeches were kept in freshwater tanks at 16 –18°C and acclimated for 3 –7 days to room temperature (20 –25°C) before experimentation. Individual ganglia (Fig. 1A) were dissected from the leech and pinned ventral side uppermost in a Sylgard-coated Perspex chamber in modified Leibowitz (L-15) tissue culture medium. This Leibowitz-15 medium (obtained from Gibco) was modified by dilution (1:3) to reach iso-osmolality and to give Na⁺ and K⁺ concentrations comparable to the physiological leech saline with the following salt solution (mmol l⁻¹): CaCl₂, 6.87; MgCl₂, 2.51; KC1, 3.32; sodium malate, 20.1; sodium pyruvate, 12.5; Hepes, 15; glucose, 15; gentamycin (10 mg ml⁻¹), 0.3%; adjusted to pH7.4 with NaOH. The ventral ganglionic capsule was removed mechanically with fine forceps. The ganglia were then incubated for 1–1.5 h in 2 mg ml⁻¹ collagenase/dispase (Boehringer-Mannheim, Germany) dissolved in modified L-15 medium at room temperature (20–25°C). After enzyme treatment the ganglia were thoroughly washed with enzyme-free medium. Neurones overlying the two neuropile glial cells of each segmental ganglion were removed by suction into a glass micropipette (tip diameter 50–100 μm), thus exposing the neuropile and revealing the cell bodies of the neuropile glial cells covered only by the inner ganglionic capsule (Fig. 1B). Ganglia were then transferred into leech saline for experimentation. Microelectrodes filled with Fura-2 (10 mmol l⁻¹ in 0.1 mol l⁻¹ KC1) were inserted into the neuropile glial cells under direct visual control. The dye was injected ionophoretically with negative constant current of 2–5 nA for 10–15 min, after which the electrode was removed.

Fluorescence imaging of glial cells was begun 5-10min after Fura-2 injection. Changes in intracellular Ca²⁺ concentration were measured by determining the ratio of Fura-2 fluorescence at 350nm excitation to that at 380nm excitation. Experiments were performed using a Deltascan dual-excitation spectrofluorimeter (LTI, Tornesch, Germany) in which shutters, monochromator settings and data aquisition were controlled by computer software and by interfaces from LTI. The cells were excited alternately at 350 nm and 380 nm (bandwidth 4–8 nm) with a 75 W xenon arc lamp through the epifluorescence port of a Nikon Diaphot microscope with an oil-immersion CF-Fluor 20Ç objective. Fluorescence intensity was measured at 510 nm using a photon-counting photomultiplier tube, and measurements were limited to a field of view slightly larger than the injected glial cell (diameter 60–100 μm, Fig. 2) by a rectangular diaphragm.

[Ca²⁺]_i was computed from an in situ calibration method using the equation described by Grynkiewicz et al. (1985): [Ca²⁺]_i=K_d [(R-R_min)/(R_max-R)] (F₀/F_s). Minimum and maximum fluorescence ratios at 350nm and 380nm (R_min, 0.385; R_max, 0.945) were determined by applying leech salines to the preparation containing either 0.1% Triton X-100 or 50 μmol l⁻¹ digitonin, first in the presence of 10 mmol l⁻¹ EGTA and then with 5 mmol l⁻¹ Ca²⁺. The fluorescence ratio measured in normal leech saline is R. F₀/F_s is the ratio of fluorescence at 380 nm excitation in the absence of Ca²⁺ to that at saturating Ca²⁺ levels and was determined automatically from the calibration. The apparent dissociation constant (K_d) between Fura-2 and Ca²⁺ was determined from a single measurement in situ to be 177 nmol l⁻¹ using a Ca²⁺ buffer solution containing 100 nmol l⁻¹ Ca²⁺ (pCa7; after Tsien and Rink, 1980). It is possible that during the calibration procedures the cells lost dye after exposure to digitonin or Triton. However, relatively stable values were obtained for at least 5 min during these calibrations, indicating that, even if some dye had leaked out of the cell, this did not influence the ratio measurement. Large cells, such as the leech neuropile glial cells, may be difficult to buffer to stable [Ca²⁺]_i using ionophores (Cohan et al. 1987).

Once the dye had leaked out of the cell it would be expected to wash away rapidly, as was the case in intact ganglia. Dye leakage and/or bleaching was also observed during experiments and this could amount to 40–70% of the fluorescence intensity loss (Fig. 3). However, this intensity loss occurred over a period of 30–60 min and should not alter the ratio of fluorescence.

Before calculating [Ca²⁺]_i, the ratios were corrected for background fluorescence. The background fluorescence was obtained from an exposed glial cell that had not been injected with Fura-2, but with all other conditions identical to the fluorescent dye measurements. The value was stored by the computer program and recalled for each calculation of [Ca²⁺]_i.

In most experiments the responses of [Ca²⁺]_i to the same, repeatedly applied stimuli tended to decrease in amplitude during the course of the experiment. In some, but not all, of these experiments, this was accompanied by a slow upward movement of the fluorescence ratio, presumably indicating a slow rise in the resting [Ca²⁺]_i. This phenomenon was concomitant with a large decrease in absolute fluorescence, and a shift in the spectra to lower wavelengths (Fig. 3).

One of the advantages of using a monochromator system is that it enables the excitation spectra of the dye-loaded cell to be measured (Fig. 3). From such spectra the shape and peak of the emission can be obtained, giving an indication of the relative dye quantity injected into the cell and the approximate resting [Ca²⁺]_i.

All experiments were performed 5–10 times, except where otherwise indicated.

The preparation of the neuropile glial cells was the same as described above. Ganglia with exposed glial cells (Fig. 1B) were pinned in an experimental chamber (volume 0.2 ml) and superfused continuously at a flow rate of 1.5 ml min⁻¹. For determination of the membrane input resistance, two microelectrodes, filled with 4 mol l⁻¹ potassium acetate, were inserted into a neuropile glial cell, one for recording voltage and the other for injecting constant current pulses. In the experiments in which the relative K⁺ conductance was to be determined, only one electrode was inserted into a neuropile glial cell. The extracellular [K⁺] was repeatedly raised from 4 to 10 mmol l⁻¹, for Imin, and the amplitudes of the resulting membrane depolarizations were used as a measurement of the relative K⁺ conductance.

Microelectrodes were connected to the input of a high-impedance voltage follower and to the bridge amplifier input of a Neurohel V87 (Meyer, Munich, Germany) by chlorided silver wires. The indifferent electrode was a Ag/AgCl wire embedded in 2% agar. The signals were displayed on an oscilloscope and recorded with a pen recorder.

The normal bathing fluid (physiological, isotonic leech saline) had the following composition (mmolI⁻¹): NaCl, 85; KC1, 4; CaCl₂, 2; MgCl₂, 1; Hepes, 10; adjusted to pH 7.4 with NaOH. In solutions with a raised K⁺ concentration, equimolar amounts of Na⁺ were replaced by K⁺ to maintain iso-osmolality. In nominally Ca²⁺-free solutions (no EGTAadded), Mg²⁺ was substituted for Ca²⁺. 5-Hydroxytryptamine (5-HT) was dissolved in the appropriate saline to give a final concentration of 50 or 100 μmol l⁻¹.

5-HT and digitonin were purchased from Fluka (Neu-Ulm, Germany). Fura-2 pentapotassium salt was purchased from Molecular Probes (Eugene, OR) and Triton X-100 (t-octylphenoxypolyethoxyethanol) was obtained from Sigma (St Louis, MO). Tissue culture medium and antibiotics were obtained from Gibco (Eggenstein, Germany).

A neuropile glial cell loaded with Fura-2 is shown in Fig. 2. Fura-2 was distributed relatively evenly throughout the cell soma and within the cell processes. The neuropile glial cell body was 60–100 μm in diameter, and the numerous branching processes of the glial cell extended over 350–400 μm to fill half the neuropile space. Cells filled with Lucifer Yellow exhibited a very similar shape and branching pattern.

The resting [Ca²⁺]_i was 32±23nmol l⁻¹ (mean±s.D.), ranging from 5 to 79 nmol l⁻¹ for seven individually calibrated cells. Changes in [Ca²⁺], induced by elevation of extracellular [K⁺] or by addition of 5-HT are presented in terms of fluorescence ratio changes rather than absolute values of [Ca²⁺]_i, since many cells lost the dye within 5–15 min of addition of Triton or digitonin.

The rise in [Ca²⁺]_i in response to an increase in the external K⁺ concentration from 4 to 20 mmol I⁻¹ is shown in Fig. 4. During a 60s exposure to 20 mmol l⁻¹ K⁺ (Fig. 4A) there was a sharp increase in [Ca²⁺]_i after a delay of about 20-25 s. This delay was partly due to the wash-in of saline with the increased K⁺ concentration. When the external K⁺ concentration was restored to normal (4 mmol l⁻¹), [Ca²⁺]_i rapidly decreased. The return of [Ca²⁺]_i to the control level appeared to occur with at least two time constants, as indicated by the changes in the fluorescence ratio, with half-times of 15 and 70s. Membrane potential measurements of these glial cells have shown that in 20 mmol l⁻¹ K⁺ the glial membrane depolarizes by about 30mV, from −70 to −40mV (inset of Fig. 4A; see also Deitmer and Szatkowski, 1990).

Comparison of a brief and a prolonged exposure (for 7 min) to the high-K⁺ saline (Fig. 4B) revealed that [Ca²⁺]_i remained elevated throughout the exposure period and decreased only very slowly with a half-time of approximately 8 min at high external [K⁺]. Return to the control K⁺ concentration initiated a decrease in [Ca²⁺]_i apparently with a single time course having a half-time of 24s. This shows that [Ca²⁺]_i would remain high for at least 15–20 min in a high-K⁺ saline.

The fast rise of [Ca²⁺]_i during a K⁺-induced membrane depolarization suggests that the rise might be due to Ca²⁺ influx through voltage-gated Ca²⁺ channels in the glial cell membrane. This was tested in an experiment where external Ca²⁺ was removed prior to an increase in external K⁺ concentration (Fig. 5). In the presence of external Ca²⁺, high external [K⁺] again evoked a large, fast rise in [Ca²⁺]_i but the intracellular Ca²⁺ transient produced in the nominal absence of external Ca²⁺ was relatively slow and much smaller. Since the removal of external Ca²⁺ occurred only 1 min before the increase in external [K⁺] elicited the small rise in [Ca²⁺]_i, it is unlikely that stored intracellular Ca²⁺ had been affected. The resting [Ca²⁺]_i remained virtually unchanged during the 10 min absence of external Ca²⁺.

In all experiments of this type there was a small, but significant, rise in [Ca²⁺]_i following an increase in the external K⁺ concentration. Since no Ca²⁺ buffer was used in these experiments, it is possible that the remaining Ca²⁺ in the saline, presumably between 1 and 10 μmol l⁻¹, could provide enough Ca²⁺ influx into the cells to produce this residual intracellular Ca²⁺ response.

In other experiments, 1 mmol l⁻¹ Co²⁺ or 0.5 mmol l⁻¹ Cd²⁺ was used to block Ca²⁺ channels in the glial membrane. These Ca²⁺ channel blockers considerably reduced the rise in [Ca²⁺]_i in the presence of 20 mmol l⁻¹ K⁺, though a small, slow intracellular Ca²⁺ transient remained even in the presence of these Ca²⁺ antagonists. Both Co²⁺ and Cd²⁺ also produced an apparent upward movement of the baseline of the fluorescence ratio. This might be due to the decreasing viability of the cells in solutions containing Co²⁺ and Cd²⁺ or to some interference of these divalent cations with the fluorescent dye.

Responses of [Ca²⁺]_i to 50μmol l⁻¹ 5-HT, a common transmitter and hormone in the nervous system of the leech (Lent and Dickinson, 1984), are shown in Fig. 6. Compared to the intracellular Ca²⁺ transient elicited by increasing the external K⁺ concentration, the rise in [Ca²⁺]_i following addition of 5-HT was significantly slower and smaller (Fig. 6A). After wash-out of 5-HT, [Ca²⁺]_i returned to control levels at the same rate as when the external K⁺ concentration was reduced from 20 to 4 mmol l⁻¹. In most experiments [Ca²⁺]_i also appeared to decrease with two time constants following 5-HT removal.

In some experiments (fewer than 20%) there appeared to be no, or only a very small, rise in [Ca²⁺]_i in the presence of 5-HT, although these cells did show a normal [Ca²⁺]_i response to increased [K⁺], The reason for the unresponsiveness of some cells to 5-HT is not yet known.

During prolonged exposure to 5-HT the [Ca²⁺]_i of responding glial cells remained high for many minutes (Fig. 6B) and only returned to the control level after the removal of 5-HT. After an exposure to 5-HT for 7.5 min, [Ca²⁺]_i returned to normal within 1–2min, apparently with a single time course. Hence, the [Ca²⁺]_i responses to 5-HT were quite like those to increased [K⁺], except for their smaller amplitude and slower rate of rise.

We also looked for the possible source of Ca²⁺ leading to the [Ca²⁺]_i response to 5-HT. In nominally Ca²⁺-free saline, the Ca²⁺ transient was much reduced, but never completely abolished (Fig. 7). After re-addition of external Ca²⁺, 5-HT again elicited a typical intracellular Ca²⁺ transient.

It has been reported that 5-HT hyperpolarizes the membrane by increasing a K⁺ conductance in these glial cells (Walz and Schlue, 1982). This was confirmed in the present study by measuring the effects of 5-HT on the membrane input resistance and the effects of brief exposures to 10 mmol l⁻¹ K⁺ (Fig. 8). The membrane input resistance decreased by 44% from 0.41±0.07 MΩ (N=12) to 0.23±0.07 MΩ (N=8) after the addition of 100μmol l⁻¹ 5-HT (Fig. 8A). The membrane depolarization in response to 10 mmol l⁻¹ K⁺ increased about twofold from 9.2±2.8mV (N=9) to 18.4±1.2mV (N=8) in the presence of 50μmol l⁻¹ 5-HT (Fig. 8B). The increased responsiveness of the cell membrane to K⁺, together with the reduced membrane input resistance, indicates that the K⁺ conductance of the glial membrane was considerably enlarged by 5-HT. Both changes, that in membrane input resistance and that in the membrane responsiveness to K⁺, were readily reversible after exposures to 5-HT lasting up to 3–4min, but were longer lasting when 5-HT was applied for 5 min or more. The time courses of the 5-HT effects on the glial membrane properties might be linked to the time courses of the Ca²⁺-related intracellular second-messenger pathway.

This study shows that [Ca²⁺]_i can be measured in identified glial cells in situ with the Fura-2 method. Rises in [Ca²⁺]_i induced by membrane depolarization and 5-hydroxytryptamine indicated the existence of voltage- and ligand-gated Ca²⁺ channels in the membrane of neuropile glial cells. Additional electrophysiological data suggested that the 5-HT-induced hyperpolarization of the glial membrane might be due to a Ca²⁺-dependent K⁺ conductance.

A single segmental ganglion of Hirudo medicinalis contains only a few large glial cells, easily identified by their position and electrophysiological properties (Kuffler and Potter, 1964; Schlue et al. 1980). Being located in the neuropile, where synaptic transmission takes place, the neuropile glial cells are wellpositioned to participate in ion homeostasis. By removing the overlying ganglionic capsule and neuronal cell bodies, we were able to perform the Ca²⁺ measurements in identified glial cells in situ. In contrast, previous fluorescence measurements of [Ca²⁺] in glial cells have only been carried out in cell cultures (Jensen and Chiu, 1990, 1991; Cornell-Bell et al. 1990; Salm and McCarthy, 1990; Cornell-Bell and Finkbeiner, 1991; Charles et al. 1991; MacVicar et al. 1991; Dave et al. 1991), where glial cells were completely removed from their original environment.

We also determined the calibration parameters in the cells in situ, because there is increasing evidence that the chemical and optical properties of fluorescent dyes inside a cell may be different from those in an aqueous solution (Konishi et al. 1988; Owen, 1991). In vitro calibration parameters may yield different resting and peak values of [Ca²⁺]_i from those determined in situ (Callewaert et al. 1991; Williams et al. 1990).

The resting value for [Ca²⁺]_i found in the present study (32 nmol l⁻¹, range 5–79 nmol l⁻¹) corresponds well with the resting [Ca²⁺]_i measured in cultured astrocytes (34±3.7nmol l⁻¹, Jensen and Chiu, 1990) and in mixed glial cultures (40–75 nmol l⁻¹, Charles et al. 1991). Higher values have been reported for cultured rat cerebellar glial cells (79 nmol l⁻¹, Ashley et al. 1989), for murine cortical astrocytes (100–200 nmol l⁻¹, MacVicar et al. 1991) and for cultured rat cerebral type-1 astrocytes (101±9nmol l⁻¹, Salm and McCarthy, 1990).

A brief elevation of extracellular [K⁺], which depolarizes the glial membrane, leads to transient rises in [Ca²⁺]_i in leech neuropile glial cells. This suggests the presence of voltage-gated Ca²⁺ channels in the membrane of these cells. Ca²⁺ channels, known to be present in cultured cortical astrocytes (MacVicar, 1984; Barres et al. 1990; MacVicar et al. 1991) and Müller glial cells (Newman, 1985), would be activated by K⁺-induced membrane depolarization. The idea that Ca²⁺ enters the cell via voltage-gated channels is supported by the observation that ions such as Co²⁺ and Cd²⁺, which are known to block Ca²⁺ channels in glial cell membranes (MacVicar and Tse, 1988; Barres et al. 1988), considerably reduce rises in [Ca²⁺]_i in neuropile glial cells.

Ca²⁺ channels activated by neurotransmitters such as 5-HT also seem to be present in neuropile glial cells. The addition of 5-HT induced an increase in glial [Ca²⁺]_i, which was dependent upon the presence of extracellular Ca²⁺. In contrast to high extracellular [K⁺], which leads to a depolarization, 5-HT hyperpolarizes the glial cell membrane (Fig. 8). Therefore, Ca²⁺ influx through voltage-gated Ca²⁺ channels seems not to be involved in 5-HT-induced increases in [Ca²⁺]_i.

5-HT-induced increases in [Ca²⁺]_i have been measured using fluorescence methods in a rat glioma cell Une (Ogura et al. 1986). Recently Dave et al. (1991) have shown that only 9% of identified cerebral type 2 astrocytes responded to 5-HT with a rise in [Ca²⁺]_i. So far, direct activation of Ca²⁺ channels in the membrane of cultured astrocytes has only been shown for putative L-type channels coupled to glutamate receptors of type 2 astrocytes (Jensen and Chiu, 1990, 1991) and norepinephrine receptors of type 1 astrocytes (Salm and MacCarthy, 1990).

By comparing the depolarization-and agonist-evoked responses of [Ca²⁺]_i in the presence and absence of external Ca²⁺, it appears that most, if not all, of the increase in [Ca²⁺]_i is due to Ca²⁺ influx across the cell membrane. In the case of the response to high-K⁺ saline, the small, slow increase in [Ca²⁺]_i in nominally Ca²⁺-free saline might be due to the small amounts of Ca²⁺ remaining in the nonbuffered saline (presumably less than 10⁻⁵mol l⁻¹).

Membrane depolarization could also lead to an influx of Na⁺ through voltagegated Na⁺ channels known to be present in the membrane of astrocytes and oligodendrocytes (Bevan et al. 1985; Barres et al. 1988), which could activate some influx of Ca²⁺via the Na⁺/Ca²⁺ exchanger. There is no evidence for the existence of a Na⁺/Ca²⁺ exchange mechanism in glial membranes, and if Na⁺/Ca²⁺ exchange were to be responsible, a much slower increase in [Ca²⁺]_i would be anticipated.

Another source of the intracellular Ca²⁺ transients might be Ca²⁺ release from internal stores, such as endoplasmic reticulum, by receptor-mediated generation of a second messenger. Ligand-induced inositol-l,4,5-trisphosphate [Ins(l,4,5)P₃] is known to trigger release of Ca²⁺ from intracellular stores in cortical mammalian astrocytes and astrocytoma cell lines (Pearce and Murphy, 1988). In a rat glioma cell line, stimulation of phosphoinositide hydrolysis by activation of 5-HT receptors has been demonstrated (Ananth et al. 1987), and in a polyploid glioma cell line activation of 5-HT receptors induces a rise in cytosolic [Ca²⁺], possibly via Ins(l,4,5)P₃ hydrolysis (Reiser et al. 1989). Two observations, however, argue against 5-HT-induced intracellular Ca²⁺ release in leech neuropile glial cells: (1) 5-HT-evoked rises in [Ca²⁺]_i were greatly reduced in the absence of external Ca²⁺, and (2) Ca²⁺ oscillations, which are characteristic of receptor-mediated intracellular Ca²⁺ transients in astrocytes (Jensen and Chiu, 1990, 1991; Cornell-Bell and Finkbeiner, 1991; Charles et al. 1991; Salm and McCarthy, 1990), were never observed even during long exposures to 5-HT.

Addition of 5-HT hyperpolarizes the neuropile glial membrane, and it has been suggested that this results from the activation of a K⁺ conductance (Walz and Schlue, 1982). Our experiments confirm this suggestion by showing that 5-HT induced an increase in the relative K⁺ conductance and a decrease in the input resistance of neuropile glial cells. We have shown that 5-HT also evokes an increase in [Ca²⁺]_i. These changes in membrane properties and [Ca²⁺]_i are only reversible when glial cells are given short exposures to 5-HT. Applications of 5-HT for 5 min or more led to sustained rises in [Ca²⁺]_i, which might be responsible for the long-lasting effect (>20min) of 5-HT on the K⁺ conductance of the glial membrane.

These parallel effects of 5-HT on [Ca²⁺]_i and on the membrane properties indicate that the K⁺ conductance induced by 5-HT might be elicited by the rise in [Ca²⁺]_i; suggesting the presence of Ca²⁺-dependent K⁺ channels in the neuropile glial cell membrane. Indeed, Ca²⁺-dependent K⁺ channels have been reported in astrocytes (Quandt and MacVicar, 1986). In rat glioma cells, activation of 5-HT₂ receptors directly opened Ca²⁺ channels; the resulting increase in [Ca²⁺]_i caused a hyperpolarization by activating Ca²⁺-dependent K⁺ channels (Ogura et al. 1986). A similar mechanism might be involved when depolarization of the glial membrane by high external [K⁺] leads to an influx of Ca²⁺ through voltageactivated Ca²⁺ channels (MacVicar, 1984; MacVicar et al. 1991).

The rises in [Ca²⁺]_i elicited by membrane depolarization and 5-HT may play a role in the K⁺ clearance properties of glial cells by coupling a high K⁺ permeability of glial membranes to neuronal activity and the concentration of a neurohormone, such as 5-HT, in the nervous tissue (Walz, 1988, 1989). Similarly, [Ca²⁺]_i-induced membrane potential changes of the glial cell could influence pH regulation in the nervous system by affecting electrogenic Na⁺/HCO₃⁻ cotransport (Deitmer and Szatkowski, 1990; Deitmer, 1991) and other voltage-dependent conductances and transport systems in the glial membrane.

We thank Fatima Pfeiffer for providing the photomicrograph of a Fura-2-filled glial cell and H.-P. Schneider for excellent technical assistance. We are grateful to the LTI company for providing some equipment. This study was supported by grants from the Deutsche Forschungsgemeinschaft (De231/7,9).