Strontium-Induced Repetitive Calcium Spikes in a Unicellular Green Alga

Abstract

The divalent cation Sr²⁺ induced repetitive transient spikes of the cytosolic Ca²⁺ activity [Ca²⁺]_cy and parallel repetitive transient hyperpolarizations of the plasma membrane in the unicellular green alga Eremosphaera viridis. [Ca²⁺]_cy measurements, membrane potential measurements, and cation analysis of the cells were used to elucidate the mechanism of Sr²⁺-induced [Ca²⁺]_cy oscillations. Sr²⁺ was effectively and rapidly compartmentalized within the cell, probably into the vacuole. The [Ca²⁺]_cy oscillations cause membrane potential oscillations, and not the reverse. The endoplasmic reticulum (ER) Ca²⁺-ATPase blockers 2,5-di-tert-butylhydroquinone and cyclopiazonic acid inhibited Sr²⁺-induced repetitive [Ca²⁺]_cy spikes, whereas the compartmentalization of Sr²⁺ was not influenced. A repetitive Ca²⁺ release and Ca²⁺ re-uptake by the ER probably generated repetitive [Ca²⁺]_cy spikes in E. viridis in the presence of Sr²⁺. The inhibitory effect of ruthenium red and ryanodine indicated that the Sr²⁺-induced Ca²⁺ release from the ER was mediated by a ryanodine/cyclic ADP-ribose type of Ca²⁺ channel. The blockage of Sr²⁺-induced repetitive [Ca²⁺]_cy spikes by La³⁺ or Gd³⁺ indicated the necessity of a certain influx of divalent cations for sustained [Ca²⁺]_cy oscillations. Based on these data we present a mathematical model that describes the baseline spiking [Ca²⁺]_cy oscillations in E. viridis.

Transient elevations of the [Ca²⁺]_cy play a central role in intracellular signal transduction in plant (Bush, 1995; Trewavas et al., 1996; Webb et al., 1996) and animal cells (Petersen et al., 1994; Bootman and Berridge, 1995). Brief transient [Ca²⁺]_cy elevations with a rapid rising phase and a rapid falling phase are called [Ca²⁺]_cy spikes, reflecting the shape of these [Ca²⁺]_cy transients. Single [Ca²⁺]_cy spikes can be induced by mechanical signals, cold shock, or elicitors in plant cells. A correlation between mechanical signal strength and amplitude of the resulting [Ca²⁺]_cy spike has been shown (Knight et al., 1991, 1992), suggesting a role of [Ca²⁺]_cy spikes in signal transduction in plant cells. In animal cells repetitive transient elevations of [Ca²⁺]_cy, so-called [Ca²⁺]_cy oscillations, are well established (Tsien and Tsien, 1990; Fewtrell, 1993; Petersen et al., 1994). For plant cells only a few reports about [Ca²⁺]_cy oscillations exist. Phytohormone-induced [Ca²⁺]_cy fluctuations reported earlier were strongly damped and ceased after a few repetitions (Felle, 1988; Schroeder and Hagiwara, 1990). Only recently stable [Ca²⁺]_cy oscillations were observed in plant cells (Johnson et al., 1995; McAinsh et al., 1995; Ehrhardt et al., 1996; Bauer et al., 1997). In some cases these [Ca²⁺]_cy oscillations display a baseline spiking pattern, which means that repetitive [Ca²⁺]_cy spikes are separated by a constant [Ca²⁺]_cy baseline (Ehrhardt et al., 1996; Bauer et al., 1997). There are indications for a physiological function of [Ca²⁺]_cy oscillations in plant cells (Johnson et al., 1995; McAinsh et al., 1995; Ehrhardt et al., 1996). The mechanisms generating [Ca²⁺]_cy oscillations in plant cells are unknown.

The unicellular green alga Eremosphaera viridis responds to various stimuli with single or repetitive [Ca²⁺]_cy spikes (Bauer et al., 1997). For example, after a “light-off” stimulus a single [Ca²⁺]_cy spike occurs, which is accompanied by a parallel transient hyperpolarization of the plasma membrane. The addition of caffeine or Sr²⁺ induces repetitive [Ca²⁺]_cy spikes that are always accompanied by repetitive transient hyperpolarizations. This hyperpolarization of the plasma membrane is due to the opening of Ca²⁺-dependent K⁺ channels (Thaler et al., 1989; Förster, 1990), representing a qualitative indicator of [Ca²⁺]_cy spikes (Bauer et al., 1997). In this study the long-lasting Sr²⁺-induced repetitive [Ca²⁺]_cy spikes in E. viridis were used to elucidate the mechanism of [Ca²⁺]_cy oscillations in a green plant cell. We developed a theoretical model that serves as a basis to discuss which intracellular Ca²⁺ stores might be involved in generating [Ca²⁺]_cy spikes.

MATERIALS AND METHODS

Plant Material and Solutions

The coccal green alga Eremosphaera viridis de Bary (Algal Culture Collection Göttingen LB 228–1, Germany) was cultured and prepared for experiments as described by Köhler et al. (1983). Algal cells with diameters of at least 150 μm were used for measurements. Under standard conditions the external medium contained 0.1 mm KNO₃, 0.1 mm MgCl₂, 0.1 mm CaCl₂, and 2 mm Mes adjusted to pH 5.6 by NaOH. For measurements at low external concentration of divalent cations, the medium contained 0.1 mm KNO₃, 1 mm EGTA, and 2 mm Mes adjusted to pH 7.6 by NaOH. Under these conditions the total [Ca²⁺] was about 1.3 μm as measured by ICP-AES. Gd³⁺ and La³⁺ were added as chloride salts and Sr²⁺ was added as a chloride salt or as carbonate or gluconate. For high external [K⁺], KNO₃ was used as well as the gluconate salt. TMB₈ (Biomol, Hamburg, Germany) was used as 5 mm stock solution in water. Verapamil (Biomol) and DBHQ (Biomol) were used as 10 mm stock solutions in EtOH. CPA (Biomol) was used as 5 mm stock solution in DMSO. The resulting final concentrations of EtOH or DMSO in the external medium were ≤ 0.5% (v/v). Control experiments showed that both solvents at concentrations < 1% influenced neither the membrane potential nor cytosolic ion activities (see also Thaler et al., 1992; Sauer et al., 1993). The flow rate of the perfusion medium was adjusted to exchange the chamber volume in 1 min.

Measurement of [Ca²⁺]_cy and Membrane Potential

The fluorescent Ca²⁺-sensitive dye fura-2-dextran (M_r = 10,000; Molecular Probes, Leiden, The Netherlands) was microinjected mechanically into the cytosol of the alga by a lab-made injection syringe as recently described (Plieth and Hansen, 1996). The concentration of the fluorescent dye in the cell was estimated by comparison of the fluorescence of the algal cell with the fluorescence of calibration capillaries (Plieth and Hansen, 1996). After microinjection the glass capillary was removed and a KCl-microelectrode filled with 1 m KCl was impaled to register the membrane potential in parallel with the Ca²⁺-dependent fura-2-dextran fluorescence. The ratiometric [Ca²⁺]_cy measurement was performed according to the methods of Fenton and Crofts (1990) and [Ca²⁺]_cy was determined by in vitro calibration (Grynkiewicz et al., 1985). A detailed description of membrane potential measurement and parallel [Ca²⁺]_cy measurement including ratio imaging and in vitro calibration is given by Plieth and Hansen (1996). When recorded in parallel to ratiometric [Ca²⁺]_cy measurements, computer-aided membrane potential measurements, discontinuously synchronized to [Ca²⁺]_cy measurements (Plieth and Hansen, 1996), were performed. Membrane potential measurements (Axoclamp-2B, Axon Instruments, Foster City, CA) without parallel [Ca²⁺]_cy measurements were continuously monitored on an oscilloscope, registered by an x/t-recorder, and scanned for final analysis and presentation.

Pressure Injection of Sr²⁺, Ca²⁺, Ruthenium Red, and Ryanodine

To increase directly the cytosolic Sr²⁺ activity, Sr²⁺ was microinjected into a single algal cell (Förster, 1990). A glass capillary containing 0.5 mm or 1 mm SrCl₂ was connected via a polyethylene tube to a compressed-air cylinder and was impaled into the alga. For microinjection the turgor of the algal cell was decreased by 200 mm sorbitol in the external perfusion medium and a pressure of about 0.5 to 1.0 MPa was applied to the capillary, resulting in injection rates of 10 to 50 pL min⁻¹. Internal concentrations were estimated from the volume of the cytoplasm and the rates of microinjection. The diameter of the spherical algae (≥ 150 μm) was measured for each experiment and the volume of the cytoplasm was calculated from the whole cell volume assuming 20% cytoplasm (Bethmann et al., 1995). The Sr²⁺ influx rates during microinjection ranged from 5 fmol min⁻¹ (0.5 mm at 10 pL min⁻¹) to 50 fmol min⁻¹ (1 mm at 50 pL min⁻¹). At 1 mm external SrCl₂ an initial uptake rate of 4 μm min⁻¹ corresponding to 15 fmol min⁻¹ was calculated (see below) from cation uptake measurements shown in Figure 8. Ca²⁺, ruthenium red (Sigma-Aldrich, Deisenhofen, Germany), and ryanodine (Biomol) were microinjected in the same way by glass capillaries containing 0.1 mm CaCl₂, 10 μm, 100 μm, or 1 mm ruthenium red, or 2 mm ryanodine in water. When Sr²⁺, Ca²⁺, ruthenium red, or ryanodine was microinjected, the membrane potential was measured continuously with an impaled microelectrode and was used as a Ca²⁺ indicator (Bauer et al., 1997). Since fura-2-dextran binds Sr²⁺, which results in a shift of the excitation spectrum hardly distinguishable from Ca²⁺ binding (Kwan and Putney, 1990), monitoring of [Ca²⁺]_cy via fura-2-dextran fluorescence was not applicable for Sr²⁺ microinjection experiments.

Figure 8.

Sr²⁺ uptake into E. viridis. Algal cells were incubated in the standard medium plus 1 mm SrCl₂ either containing no ER Ca²⁺-ATPase inhibitor (control, •) or containing additionally 10 μm DBHQ or CPA (○). After different incubation times (t/min) cells were separated from the medium and the [Sr²⁺] of the cell sap ([Sr²⁺]/μm) was determined. Data points of the control (•) are given as means ± se of three to six measurements; data points measured in the presence of ER Ca²⁺-ATPase inhibitors (○) are means of three measurements. Data points were described by Equation 1. With an initial [Sr²⁺] of [Sr²⁺]₀ = 2.1 μm, a fit to 19 separate data points for the control (•) yielded an equilibrium [Sr²⁺] of [Sr²⁺]_EQ = 204 ± 18 μm, and a rate constant of a = 0.021 ± 0.004 min⁻¹ (fit indicated by the solid line). When the data points measured in the presence of ER Ca²⁺-ATPase inhibitors (○) were included in the fit, parameters did not significantly change ([Sr²⁺]_EQ = 202 ± 18 μm, a = 0.022 ± 0.004 min⁻¹).

Measurement of Cellular Cation Concentrations

The cation content of total cells (cell sap plus cell wall) and of cell sap of E. viridis was determined by ICP-AES analysis as described by Bethmann et al. (1995). One-milliliter probes of a dense algal suspension were taken and incubated in 500 mL of standard medium without any addition or containing 1 mm SrCl₂, 100 or 250 μm GdCl₃, 100 μm LaCl₃, 1 mm SrCl₂ plus 10 μm DBHQ, or 1 mm SrCl₂ plus 10 μm CPA. Samples of total cells were dried overnight at 80°C and analyzed. Samples for cell sap were heated to 100°C for 30 min and centrifuged at 4750g for 30 min to remove cell walls. The supernatant representing the cell sap was analyzed.

Data Analysis and Mathematical Modeling

All results in the text are given as mean ± sd. The nonlinear regression analysis of the data in Figure 8 was done with Grafit (Erithacus Software, London, UK) based on the Marquardt algorithm. A stability analysis of Equation 2 was performed by calculating the eigenvalues (Stucki and Somogyi, 1994) for different parameter sets (Mathcad, MathSoft, Cambridge, MA). For some parameter sets yielding stable oscillations (positive eigenvalues) Equation 2 was solved numerically (Mathematica, Wolfram Research, Champaign, IL).

RESULTS

[Ca²⁺]_cy and membrane potential of E. viridis were measured simultaneously in the same algal cell. The free-running membrane potential (E/mV) under standard conditions (in 0.1 mm KNO₃, MgCl₂, CaCl₂, and 2 mm Mes adjusted to pH 5.6 by NaOH) was −84 ± 20 mV (n = 332), and the steady-state [Ca²⁺]_cy was 163 ± 42 nm (n = 50) based on in vitro calibrations.

The Effect of Sr²⁺ on [Ca²⁺]_cy and Membrane Potential

The effect of Sr²⁺ on [Ca²⁺]_cy and on membrane potential in E. viridis is shown in Figure 1. In 95% of the experiments (n = 27) the addition of 1 mm Sr²⁺ to the external medium induced repetitive [Ca²⁺]_cy spikes that were always accompanied by parallel, repetitive, transient hyperpolarizations of the plasma membrane. The repetitive transient changes of [Ca²⁺]_cy and of membrane potential continued for more than 2 h under continuous perfusion of 1 mm SrCl₂. When Sr²⁺ was removed from the external medium the repetitive changes continued in 54% of all measurements with a decreasing frequency; 1 mm SrCO₃ or 1 mm Sr²⁺ gluconate had the same effect as 1 mm SrCl₂. In the presence of Sr²⁺ a systrophe (a chloroplast translocation to the center of the cell) was frequently observed.

Figure 1.

Sr²⁺-induced repetitive [Ca²⁺]_cy spikes and repetitive transient hyperpolarizations. The [Ca²⁺]_cy/nM (bottom) and the membrane potential (E/mV, top) were recorded simultaneously in a single algal cell. The addition of 1 mm SrCl₂ to the external medium (bar on top gives the perfusion protocol) induced repetitive [Ca²⁺]_cy spikes and parallel repetitive transient hyperpolarizations of the plasma membrane. Sampling frequency was 1/3 s⁻¹.

The discrete [Ca²⁺]_cy spikes in E. viridis had a duration of 24 ± 8 s (n = 28) with a rapid rising phase and a rapid falling phase, and were separated by intervals with a constant baseline of [Ca²⁺]_cy. The [Ca²⁺]_cy spikes had an amplitude of about 365 ± 71 nm (n = 36). Amplitudes of up to 900 nm were observed (see Fig. 4). The baseline value of [Ca²⁺]_cy during Sr²⁺-induced oscillations was 168 ± 43 nm (n = 27). This did not differ significantly from the steady-state value of 163 ± 42 nm in the absence of SrCl₂ mentioned above. The duration of a single transient hyperpolarization of the plasma membrane was 35 ± 9 s with an amplitude of −181 ± 7 mV (n = 68). The frequency of Sr²⁺-induced [Ca²⁺]_cy oscillations ranged from up to 0.8 min⁻¹ (see Fig. 1) to 0.2 min⁻¹ (see Fig. 5) and less. A quantitative analysis showed that lower frequencies (< 0.3 min⁻¹) were correlated with cytosolic fura-2-dextran concentrations above 10 μm (2 μm referring to the whole cell).

Figure 4.

Gd³⁺ inhibited Sr²⁺-induced repetitive [Ca²⁺]_cy spikes and repetitive transient hyperpolarizations. Repetitive [Ca²⁺]_cy spikes (bottom) and repetitive transient hyperpolarizations (E/mV, top) were induced by 1 mm SrCl₂. The additional perfusion of 1 mm GdCl₃ for 5 min (bars on top) reversibly inhibited the repetitive [Ca²⁺]_cy and potential changes. Sampling frequency was 1/1.5 s⁻¹.

Figure 5.

The effect of different external [K⁺] on Sr²⁺-induced repetitive [Ca²⁺]_cy spikes and repetitive transient hyperpolarizations. A, Repetitive [Ca²⁺]_cy spikes (bottom) and repetitive transient hyperpolarizations (E/mV, top) were induced by the addition of 1 mm SrCl₂ to the medium. The external [K⁺] was increased from 0.1 mm KCl (standard medium) to 1 mm, to 10 mm, and finally to 100 mm (bars on top). Sampling frequency was 1/3 s⁻¹. B, The amplitude of the transient hyperpolarizations (E/mV) was plotted against the logarithm of the external [K⁺]: log ([K⁺]_ex/M). A linear regression analysis (solid line, r = 0.999) yielded a slope of 53.8 ± 0.4 mV for each 10-fold increase in [K⁺]. Data points are given as means ± se (n = 15) with se being smaller than the symbol size.

To investigate the dose dependency, the effect of different external [SrCl₂] on the latency period and on the frequency of repetitive transient hyperpolarizations of the plasma membrane was determined (Table I). These experiments were performed as membrane potential measurements only, to avoid a frequency decrease by fura-2-dextran. The addition of 0.01 mm SrCl₂ had no effect on the membrane potential of E. viridis. The addition of increasing [SrCl₂] caused increasing depolarizations (Table I, ΔE). After a latency period, τ, which decreased at larger [SrCl₂], repetitive transient hyperpolarizations were induced in 69, 95, and 97% of the measurements at 0.1, 1, and 10 mm SrCl₂, respectively (Table I). The frequency, ν, increased, whereas the amplitude of the repetitive transient hyperpolarizations decreased at increasing [SrCl₂] (Table I). A comparable decrease of amplitudes was observed at increasing external [Mg²⁺] for transient hyperpolarizations induced by darkening (Sauer et al., 1994). Routinely, we used 1 mm Sr²⁺ because this concentration was sufficient to induce long-lasting repetitive [Ca²⁺]_cy spikes with a high probability and frequency.

Table I.

The effect of different external Sr²⁺ concentrations ([SrCl₂]/mm) on the membrane potential

[SrCl2]/mm	ΔEa	Pb	τc	νd	Amplitudee	n
	mV	%	min	min−1	mV
0.01	0	0	–	–	–	4
0.1	7  ± 2	69	4.2  ± 1.5	0.2  ± 0.05	−191  ± 4	23
1	19  ± 4	95	0.5  ± 0.3	0.5  ± 0.2	−181  ± 7	273
10	38  ± 10	97	0.3  ± 0.4	0.5  ± 0.2	−167  ± 4	35

^aThe depolarization upon Sr²⁺ addition (see Figs. 1 and 6). 

^bProbability to induce repetitive transient hyperpolarizations. 

^cLag phases after addition of Sr²⁺ before repetitive transient hyperpolarizations started. 

^dFrequency of repetitive transient hyperpolarizations. 

^eMaximum amplitude of the hyperpolarizations. 

To test the effect of a direct increase in cytosolic [Sr²⁺], SrCl₂ was microinjected into the cytoplasm of the algal cell. The microinjection of SrCl₂ (n = 15) always resulted in hyperpolarizations (Fig. 2). At lower Sr²⁺ injection rates (Fig. 2A) repetitive transient hyperpolarizations were induced with a frequency comparable to those observed at 1 mm external SrCl₂ (compare Figs. 1 and 2A). At higher Sr²⁺ injection rates (Fig. 2B) a massive hyperpolarization of the plasma membrane occurred. The frequency of the transient hyperpolarizations increased to more than 1 min⁻¹, resulting in a “fusion” of the repetitive potential spikes to give rise to a fluctuating permanent hyperpolarization.

Figure 2.

Microinjection of Sr²⁺ induced hyperpolarizations. A, With 0.5 mm SrCl₂ inside the injection pipette and small injection rates (about 10 pL/min) repetitive transient hyperpolarizations were induced as long as pressure was applied (small syringes on top indicate the duration of pressure application). B, SrCl₂ (1 mm) inside the injection pipette and larger injection rates (about 50 pL/min) resulted in a nearly permanent hyperpolarization as long as pressure was applied (duration indicated by the small syringe above). When a Sr²⁺ injection was followed by the external perfusion of 1 mm SrCl₂ (duration indicated by white bar), repetitive transient hyperpolarizations were observed. During long-lasting microinjection experiments, a continuous depolarization of the plasma membrane was frequently observed. The membrane potential was registered continuously.

The effect of different concentrations of CaCl₂ or MgCl₂ in the external medium on repetitive transient hyperpolarizations induced by 0.1 or 1 mm external SrCl₂ was tested. Neither the probability to induce hyperpolarizations nor their duration or frequency were influenced by the external concentration of Ca²⁺ (or Mg²⁺) in the range from about 1 μm Ca²⁺ (1 mm EGTA, pH 7.6; n = 10) up to 1 mm Ca²⁺ (n = 8) or 10 mm Mg²⁺ (n = 7).

Mn²⁺ was used to study the influx of divalent cations during Sr²⁺-induced repetitive [Ca²⁺]_cy spikes (Fig. 3). Mn²⁺ is known to bind to fura-2 with about a 40-fold higher affinity than Ca²⁺, and it quenches the fluorescence of the dye at all excitation wavelengths (Kwan and Putney, 1990). The addition of 0.1 mm MnCl₂ to the external medium resulted in a continuous decrease of the fluorescence intensity at both excitation wavelengths (340 and 380 nm, Fig. 3, middle). This quenching of the fluorescence indicates a continuous Mn²⁺ influx into the cytoplasm of the cell. It resulted in an apparent increase of [Ca²⁺]_cy (Fig. 3, bottom), whereas membrane potential was hardly influenced (Fig. 3, top).

Figure 3.

Mn²⁺ quenched the fura-2-dextran fluorescence intensity. Repetitive [Ca²⁺]_cy spikes (bottom) and repetitive transient hyperpolarizations of the plasma membrane (E/mV, top) were induced by 1 mm SrCl₂ in the external medium. The additional perfusion of 0.1 mm MnCl₂ (bars on top) decreased the fluorescence intensity emitted by fura-2-dextran (middle traces, given in arbitrary units, a.u.) at both excitation wavelengths, 340 nm (shifted upward by 25 a.u. for clarity) and 380 nm. Sampling frequency was 1/6 s⁻¹.

For a further characterization of the permeability of the plasma membrane for divalent cations, the plant plasma membrane Ca²⁺ channel blockers Gd³⁺ and La³⁺ (Huang et al., 1994; Marshall et al., 1994; Rengel, 1994; Piñeros and Tester, 1995) were used. Figure 4 shows that 1 mm GdCl₃ in the external medium reversibly blocked repetitive [Ca²⁺]_cy spikes in parallel to repetitive transient hyperpolarizations of the plasma membrane. This was observed in all measurements in which 1 mm GdCl₃ or LaCl₃ was used (n = 7). Investigating the concentration dependence revealed the following behavior. The addition of 100 μm GdCl₃ or LaCl₃ to the external medium had no effect on repetitive changes of [Ca²⁺]_cy and membrane potential induced by 1 mm SrCl₂ (n = 14). At concentrations of GdCl₃ or LaCl₃ of 200 μm, repetitive transient hyperpolarizations of the plasma membrane were influenced in all measurements. In 2 out of 10 measurements the hyperpolarizations were reversibly blocked, in 8 measurements the frequency was reduced by up to 70% as compared with the frequency in the absence of GdCl₃ or LaCl₃. Repetitive transient hyperpolarizations induced by 0.1 mm SrCl₂ were already reversibly blocked at 100 μm GdCl₃ or LaCl₃ (n = 5). This indicates that Sr²⁺ and La³⁺ or Gd³⁺ competitively interacted with the same plasma membrane Ca²⁺ channel. The transient hyperpolarization observed after a “light-off” stimulus was not affected by 100 μm LaCl₃ (n = 12) or GdCl₃ (n = 10). This shows that these trivalent cations do not block the plasma membrane K⁺ channel, which gives rise to the transient hyperpolarization.

Besides La³⁺ and Gd³⁺, the effect of the Ca²⁺ channel blocker verapamil was investigated. Verapamil (50 μm) reversibly inhibited repetitive transient hyperpolarizations induced by 1 mm SrCl₂ in 50% of the measurements (n = 8).

To test the relationship between the transient hyperpolarization of the plasma membrane and the [Ca²⁺]_cy spike, the external [K⁺] ([K⁺]_ex/mm) was changed during Sr²⁺-induced repetitive [Ca²⁺]_cy spikes. As shown in Figure 5A, an increase of [K⁺]_ex from 0.1 to 1 to 10 and finally to 100 mm decreased the amplitude of the transient hyperpolarizations. In contrast to this, the amplitude of the [Ca²⁺]_cy spikes (321 ± 19 nm, n = 17) was not influenced by [K⁺]_ex even at 100 mm. Figure 5B shows that a 10-fold increase in [K⁺]_ex resulted in a decrease of the amplitude of transient hyperpolarizations of about 54 mV, following the Nernst potential for K⁺. The steady-state membrane potential became less negative at [K⁺]_ex > 1 mm (n = 8). Depending on the steady-state membrane potential, the transient potential changes in the presence of 100 mm K⁺ resulted in small transient hyperpolarizations or small transient depolarizations. Even [Ca²⁺]_cy spikes accompanied by a transient depolarization of the plasma membrane did not significantly differ from those observed under standard conditions (0.1 mm KNO₃), as shown in Figure 5A. K⁺ gluconate had the same effect as KCl.

To study the involvement of internal Ca²⁺ stores, the effect of the ER Ca²⁺-ATPase blockers DBHQ and CPA (Inesi and Sagara, 1994) on Sr²⁺-induced repetitive [Ca²⁺]_cy spikes was investigated. As shown in Figure 6, repetitive [Ca²⁺]_cy spikes and repetitive transient hyperpolarizations were inhibited by 10 μm DBHQ after 2.9 ± 1.2 min (n = 8). The inhibitory effect was reversible when DBHQ was washed out for more than 20 min (n = 7). After a 5-min preperfusion of 10 μm DBHQ, the addition of Sr²⁺ (n = 8) induced one or a few repetitive transient hyperpolarizations with reduced amplitudes, but never sustained oscillations (not shown). Longer preperfusions inhibited Sr²⁺-induced transient hyperpolarizations. CPA at a concentration of 10 μm had the same effects on Sr²⁺-induced repetitive transient hyperpolarizations as 10 μm DBHQ (n = 5). The baseline [Ca²⁺]_cy level was not increased in the presence of DBHQ (Fig. 6), showing that a cytosolic Ca²⁺ homeostasis was still achieved by active Ca²⁺ transport systems that are not sensitive to DBHQ. A transient hyperpolarization induced by pressure injection of Ca²⁺ was not affected by 10 μm DBHQ, indicating that DBHQ did not act on the Ca²⁺-dependent plasma membrane K⁺ channel.

Figure 6.

The ER Ca²⁺-ATPase inhibitor DBHQ blocked Sr²⁺-induced repetitive [Ca²⁺]_cy spikes and repetitive transient hyperpolarizations. Repetitive [Ca²⁺]_cy spikes (bottom) and repetitive transient hyperpolarizations (E/mV, top) induced by 1 mm SrCl₂ were inhibited after the addition of 10 μm DBHQ to the external medium (bars on top). Sampling frequency was 1/3 s⁻¹.

TMB₈, an antagonist of InsP₃-induced Ca²⁺ release from intracellular stores in animal (Zhang and Melvin, 1993) and plant cells (Schumaker and Sze, 1987; Förster, 1990), did not block Sr²⁺-induced repetitive [Ca²⁺]_cy spikes and transient hyperpolarizations at concentrations up to 200 μm (n = 28).

The effect of ruthenium red (Ma, 1993) and ryanodine (Smith et al., 1988), known antagonists of the ryanodine/cADPR Ca²⁺ release channel, were investigated. Both were microinjected into the cytoplasm of algal cells either before 1 mm SrCl₂ was added or during Sr²⁺-induced repetitive transient hyperpolarizations. At cytosolic ruthenium red concentrations below 10 μm the amplitudes of the transient hyperpolarizations were decreased and they were completely inhibited at ruthenium red concentrations above 10 μm (n = 21) (Fig. 7). At cytosolic ryanodine concentrations below 100 μm, amplitudes and frequency of transient hyperpolarizations were decreased. At cytosolic ryanodine concentrations above 100 μm, Sr²⁺-induced repetitive transient hyperpolarizations were completely blocked (n = 6). The transient hyperpolarization observed after a “light-off” stimulus was influenced neither by ruthenium red (Fig. 7) nor by ryanodine at concentrations as high as several hundred micrometers.

Figure 7.

The ryanodine/cADPR Ca²⁺ release channel antagonist ruthenium red (RR) blocked Sr²⁺-induced hyperpolarizations. With 1 mm inside the pipette ruthenium red was microinjected (indicated by the syringe), resulting in a cytosolic concentration of about 100 μm. The microinjection itself caused a transient hyperpolarization. Afterward the addition of 1 mm SrCl₂ to the external medium failed to induce any hyperpolarization, whereas darkening (light protocol given by the bars below) still induced a transient hyperpolarization. The membrane potential was recorded continuously.

Measurements of Cation Uptake

To investigate whether Sr²⁺ enters E. viridis, algal cells were incubated in the standard medium containing additionally 1 mm SrCl₂, and the [Sr²⁺] ([Sr²⁺]/μm) of the total cells (cell sap plus cell wall) and the cell sap were determined after different incubation times. As shown in Figure 8, the [Sr²⁺] of the cell sap increased, reaching a stable value within 2 to 3 h. In the presence of 10 μm DBHQ or CPA, Sr²⁺ uptake into E. viridis was not significantly influenced (Fig. 8). The average (n = 14) [Sr²⁺] of the total cells was 24 μm larger compared with the concentration in the cell sap, regardless of the incubation time. This reflects a constant amount of Sr²⁺, which was bound to the cell wall. The time course of the increase of the [Sr²⁺] in the cell sap was mathematically described under the assumption that the Sr²⁺ influx was compensated by a Sr²⁺ export that linearly increases with internal [Sr²⁺]: [Sr²⁺] starts at an initial value [Sr²⁺]₀ at t = 0 and reaches an equilibrium concentration [Sr²⁺]_EQ at t → ∞ according to

1

where a is the rate constant in min⁻¹. The initial [Sr²⁺] of the cell sap was determined by ICP-AES analysis before Sr²⁺ was added to the external medium as [Sr²⁺]₀ = 2.1 ± 0.75 μm (n = 5). On the basis of Equation 1, a nonlinear regression analysis of the data points summarized in Figure 8 yielded an equilibrium [Sr²⁺] of the cell sap of [Sr²⁺]_EQ = 200 μm and a rate constant of a = 0.02 min⁻¹. From these data, using the first derivation of Equation 1, an initial Sr²⁺ influx at t = 0 of 4.0 μm min⁻¹ was calculated that corresponds to a current of about 24 pA for a spherical algal cell with a diameter of 150 μm (34 μA cm⁻²).

Besides Sr²⁺, the concentrations of La³⁺ and Gd³⁺ were measured in the total cells and in the cell sap of E. viridis after 60 min of incubation at 100 μm LaCl₃ or up to 250 μm GdCl₃ in the external medium, respectively. The [La³⁺] in the total cells was about 1.0 mm, whereas the [La³⁺] in the cell sap was below the detection limit of about 0.8 μm. Similar values were obtained for Gd³⁺. A concentration of about 0.5 mm Gd³⁺ was measured for total cells whereas the concentration in the cell sap remained below the detection limit of about 1.0 μm. This indicates that La³⁺ and Gd³⁺ were not taken up into E. viridis to a considerable amount. It also demonstrated that a proper separation of the cell wall from the cell sap was achieved. In the presence of 250 μm GdCl₃, Sr²⁺ uptake after 60 min was decreased from about 145 to 59 μm.

DISCUSSION

There are two fundamental questions about [Ca²⁺]_cy oscillations: How do they arise? And what is their physiological function? The above measurements provide access to the understanding of the mechanism of [Ca²⁺]_cy oscillations in a green plant cell. In this context Sr²⁺ was used as a tool to induce long-lasting repetitive [Ca²⁺]_cy spikes. In animal cells Sr²⁺ is frequently used to induce Ca²⁺ release from intracellular Ca²⁺ stores (Mironov and Juri, 1990; Grégoire et al., 1993) and Sr²⁺ was shown to induce repetitive Ca²⁺ spikes (Bos-Mikich et al., 1995). Regarding the physiological response to increasing [Ca²⁺]_cy in E. viridis, it should be mentioned that [Ca²⁺]_cy oscillations are accompanied by a systrophe, which is a chloroplast translocation to the center of the cell (Schönknecht et al., 1998). This systrophe is observed as a reaction to excess light and is also induced in E. viridis by blue light in the presence of external Ca²⁺ (Weidinger and Ruppel, 1985).

Sr²⁺ Uptake and Compartmentalization

Sr²⁺ was rapidly taken up into E. viridis, as shown in Figure 8. It is known that binding of Sr²⁺ to fura-2-dextran results in a fluorescence excitation spectrum very similar to the spectrum caused by Ca²⁺, with a 30-fold lower affinity of fura-2-dextran for Sr²⁺ compared with Ca²⁺ (Kwan and Putney, 1990). The baseline value of [Ca²⁺]_cy during Sr²⁺-induced oscillations did not differ from the steady-state [Ca²⁺]_cy level in the absence of Sr²⁺ (see Figs. 1, 4, 5A, and 6). Therefore, the free cytosolic Sr²⁺ activity did not exceed 1 to 2 μm. On the other hand, [Sr²⁺] of up to 200 μm were measured in the cell sap. This shows that Sr²⁺ was effectively compartmentalized into internal organelles, which is comparable to animal cells (Kwan and Putney, 1990; Mironov and Juri, 1990). As documented in Figure 2, besides being effective, the intracellular compartmentalization of Sr²⁺ was also rapid; immediately after stopping Sr²⁺ microinjection membrane potential oscillations ceased. The only organelle that can contribute to the measured intracellular steady-state concentration of 200 μm Sr²⁺ (Fig. 8) is the vacuole. All other organelles have such a small volume compared with the total volume of the cell (1.8 nL), that an accumulation of this amount of Sr²⁺ (360 fmol) in another compartment is unlikely. A vacuolar [Sr²⁺] of 250 μm (80% of the cell volume) is comparable to the vacuolar [Ca²⁺] of about 400 μm in E. viridis (Bethmann et al., 1995). The plasma membrane Ca²⁺ channel blocker Gd³⁺ (Marshall et al., 1994; Rengel, 1994; Piñeros and Tester, 1995) considerably decreased Sr²⁺ uptake. Sr²⁺ uptake and compartmentalization were not affected by the ER Ca²⁺-ATPase blockers CPA or DBHQ (Fig. 8), indicating that these inhibitors influenced neither the Sr²⁺ influx and efflux across the plasma membrane (compare Eq. 1) nor Sr²⁺ uptake into the vacuole (summarized in Fig. 10).

Figure 10.

A schematic model for Sr²⁺-induced [Ca²⁺]_cy spikes in E. viridis. Sr²⁺ enters the cell via Ca²⁺ channels in the plasma membrane that are competitively blocked by Gd³⁺ or La³⁺. Most of the Sr²⁺ taken up into the cell is compartmentalized into the vacuole. At steady state, the same amount of Sr²⁺ (and Ca²⁺) that enters the cell is transported out of the cell by plasma membrane Ca²⁺-ATPases. The Ca²⁺-ATPases that pump Sr²⁺ (and Ca²⁺) into the vacuole or out of the cell are not blocked by DBHQ or CPA. Inside the cell Sr²⁺ induces a Ca²⁺-release from the ER. The ER Ca²⁺ channel is inhibited by either ruthenium red (RR) or ryanodine (Ry), pointing to a ryanodine/cADPR-like Ca²⁺ channel. The rapid increase in [Ca²⁺]_cy is compensated for by Ca²⁺-ATPases of internal Ca²⁺ stores and the plasma membrane. The ER is refilled by Ca²⁺-ATPases that are blocked by DBHQ or CPA. The [Ca²⁺]_cy spike causes the opening of plasma membrane K⁺ channels, resulting in a transient hyperpolarization. This K⁺ channel is known to be blocked by Ba²⁺ or TEA (Köhler et al., 1983; Thaler et al., 1989).

Sr²⁺-Induced Repetitive [Ca²⁺]_cy Spikes in E. viridis Show a Baseline Spiking Pattern

[Ca²⁺]_cy oscillations in animal and plant cells display a variety of patterns. Besides more irregular repetitive [Ca²⁺]_cy changes, sinusoidal [Ca²⁺]_cy oscillations are observed, which differ from baseline spiking [Ca²⁺]_cy oscillations displaying discrete [Ca²⁺]_cy spikes separated by a baseline of [Ca²⁺]_cy (Fewtrell, 1993). In animal cells sinusoidal [Ca²⁺]_cy oscillations and baseline spiking [Ca²⁺]_cy oscillations are mechanistically different. Whereas for sinusoidal [Ca²⁺]_cy oscillations the agonist dose increases the amplitude but not the frequency, for baseline spiking [Ca²⁺]_cy oscillations the frequency is determined by the agonist dose, whereas the amplitude is independent from the agonist dose, and the latency period before the first [Ca²⁺]_cy spike is inversely related to the agonist dose (Thomas et al., 1996). Furthermore, Ca²⁺ signals with a baseline spiking pattern continue for a long period of stimulation. The agonist dose dependency of Sr²⁺-induced baseline spiking oscillations in E. viridis is the same. The frequency increased with increasing external [Sr²⁺] (Table I), as well as with increasing microinjected [Sr²⁺] (Fig. 2). The latency period decreased with increasing external [Sr²⁺] (Table I). Moreover, Sr²⁺-induced repetitive [Ca²⁺]_cy spikes in E. viridis lasted a very long time. Not only the pattern but also the dose dependency of the repetitive [Ca²⁺]_cy spikes observed in E. viridis was comparable to animal cells. This is the first indication to our knowledge that similar patterns of [Ca²⁺]_cy oscillations in plant and animal cells may be based on a similar mechanism.

Changes in [Ca²⁺]_cy are involved in many different signal transduction processes ( Trewavas et al., 1996; Webb et al., 1996), which raises the question of how a stimulus specificity is achieved. [Ca²⁺]_cy oscillations may encode information about the stimulus by the frequency, amplitude, or duration (Clapham, 1995; Berridge, 1997). In animal cells there is considerable evidence that frequency encoding does contribute to a stimulus-specific reaction: [Ca²⁺]_cy oscillations have been shown to depend on the type or strength of a stimulus (Petersen et al., 1994; Thomas et al., 1996), and biochemical mechanisms are documented that are able to decode [Ca²⁺]_cy oscillations (Hajnóczky et al., 1995; De Koninck and Schulman, 1998). Agonist-induced frequency modulation has not been shown in plants, but the results presented here suggest that the relevant mechanisms for frequency encoding exist in E. viridis.

In Commelina communis guard cells an increase to 100 μm of the external [Ca²⁺] induces baseline spiking [Ca²⁺]_cy oscillations, however, a further increase to 1 mm results in asymmetric and irregular Ca²⁺-induced [Ca²⁺]_cy elevations (McAinsh et al., 1995). In E. viridis the external concentration of Ca²⁺ or Mg²⁺ had no effect on Sr²⁺-induced oscillations.

Membrane Potential Oscillations Are Caused by [Ca²⁺]_cy Oscillations

The measurements presented here (Figs. 1 and 3–6) demonstrate the close correlation between [Ca²⁺]_cy spikes and transient hyperpolarizations. This correlation raises the question of whether the [Ca²⁺]_cy spikes cause transient hyperpolarizations or whether the transient hyperpolarizations cause Ca²⁺ spikes. For animal cells both possibilities are well documented. On the one hand, autonomous [Ca²⁺]_cy oscillations driven by Ca²⁺ release from internal stores may cause membrane potential oscillations due to the opening of Ca²⁺-dependent plasma membrane ion channels (Lee and Earm, 1994; Wojnowski et al., 1994; D'Andrea and Thorn, 1996). On the other hand, autonomous plasma membrane oscillations may cause [Ca²⁺]_cy oscillations due to Ca²⁺ influx via voltage-dependent Ca²⁺ channels (Li et al., 1995b; Larsson et al., 1996). Within the time resolution of our measurements (1.5–6 s) [Ca²⁺]_cy normally increased at the same time as the membrane potential started to hyperpolarize. The experiments illustrated in Figure 5 show that the plasma membrane potential as changed by external [K⁺] does not exert any significant effect on Sr²⁺-induced [Ca²⁺]_cy spikes. This is opposite to what is expected for [Ca²⁺]_cy spikes caused by an influx of Ca²⁺ across the plasma membrane. Such [Ca²⁺]_cy spikes vary with the membrane potential, but this was not observed (Fig. 5). Even when the transient hyperpolarization switched to a depolarization, the amplitude and time course of the [Ca²⁺]_cy spikes were not influenced. This clearly shows that the [Ca²⁺]_cy spikes caused the membrane potential changes rather than the other way around. The close correlation between [Ca²⁺]_cy spikes and transient hyperpolarizations in E. viridis (Figs. 1 and 3–6; Bauer et al., 1997) is due to the Ca²⁺-dependent opening of plasma membrane K⁺ channels (Fig. 5B).

The Role of Ca²⁺ Release and Re-Uptake by Internal Stores

Most models describing the mechanism of [Ca²⁺]_cy oscillations in animal cells are based on a repetitive Ca²⁺ release and re-uptake by intracellular Ca²⁺ stores (Tsien and Tsien, 1990; Fewtrell, 1993; Petersen et al., 1994). We used DBHQ (Fig. 6) and CPA to demonstrate the role of Ca²⁺ release and re-uptake by intracellular Ca²⁺ stores for [Ca²⁺]_cy oscillations in E. viridis. CPA and DBHQ are well-established inhibitors of ER Ca²⁺-ATPases in animal cells (Inesi and Sagara, 1994). In plant cells DBHQ and CPA recently have been shown to act specifically on ER Ca²⁺-ATPases as well (Logan and Venis, 1995; Hwang et al., 1997; Liang et al., 1997). Both inhibitors are lipophilic and readily enter the cell (Busch and Sievers, 1993; Du et al., 1994; Trebacz et al., 1996), which explains the long duration needed for a wash out with E. viridis. In line with a specific action on ER Ca²⁺-ATPases, CPA or DBHQ influenced neither the transport of divalent cations across the plasma membrane and the tonoplast (Fig. 8) nor the steady-state [Ca²⁺]_cy (Fig. 6). Blocking ER Ca²⁺-ATPases by CPA or DBHQ prevents the re-uptake of Ca²⁺ into the ER, resulting in a rapid emptying of the Ca²⁺ store that drives [Ca²⁺]_cy oscillations. Thus, a Sr²⁺-induced Ca²⁺ release is no longer possible, i.e. [Ca²⁺]_cy oscillations come to an end.

The Role of Ca²⁺ Fluxes across the Plasma Membrane

Since plasma membrane Ca²⁺-ATPases transport Ca²⁺ out of the cell, especially during [Ca²⁺]_cy spikes, a certain Ca²⁺ influx is necessary to prevent a depletion in Ca²⁺. In animal cells Ca²⁺ fluxes across the plasma membrane have been shown to be substantial for sustained [Ca²⁺]_cy oscillations (Tsien and Tsien, 1990; Fewtrell, 1993; Petersen et al., 1994). The quenching of the fura-2-dextran fluorescence by externally added Mn²⁺ (Kwan and Putney, 1990) (Fig. 3) indicated that the plasma membrane of E. viridis had a significant permeability for Mn²⁺, which is known to permeate plant plasma membrane Ca²⁺ channels (Piñeros and Tester, 1995).

During transient hyperpolarizations Mn²⁺ quenching significantly increased (Fig. 3). This could be caused by the increasing electrical driving force for Mn²⁺ influx and does not necessarily point to a voltage-dependent conductance increase. An increasing Mn²⁺ quenching indicates an increasing Ca²⁺ influx, suggesting that a component of the [Ca²⁺]_cy spikes might arise directly via Ca²⁺ influx during the transient hyperpolarization. However, neither the external [Ca²⁺] nor the membrane potential (see Fig. 5) have a significant influence on [Ca²⁺]_cy spike amplitudes. Therefore, a Ca²⁺ influx during transient hyperpolarizations does not seem to contribute significantly to [Ca²⁺]_cy spike amplitudes.

The application of the plant plasma membrane Ca²⁺ channel blockers La³⁺ or Gd³⁺ (Huang et al., 1994; Marshall et al., 1994; Rengel, 1994; Piñeros and Tester, 1995) reversibly inhibited Sr²⁺-induced repetitive [Ca²⁺]_cy spikes and transient hyperpolarizations in E. viridis (Fig. 4). The cation uptake measurements showed that neither La³⁺ nor Gd³⁺ reached micromolar intracellular concentrations, which were reported to block ER (Klüsener et al., 1995) or tonoplast Ca²⁺ channels (Johannes et al., 1992; Pantoja et al., 1992). Therefore, La³⁺ and Gd³⁺ are very likely to act on plasma membrane Ca²⁺ channels. Accordingly, the dark-induced transient hyperpolarization that is caused by a single [Ca²⁺]_cy spike in E. viridis (Bauer et al., 1997), probably due to Ca²⁺ release from the chloroplast (Schönknecht et al., 1998), was not affected by these plasma membrane Ca²⁺ channel blockers. As the results in Figure 5 show, [Ca²⁺]_cy spikes in E. viridis are not caused by membrane potential-driven Ca²⁺ influx. However, the inhibitory effect of La³⁺ or Gd³⁺ indicates that, comparable to animal cells, a certain Ca²⁺ influx across the plasma membrane is necessary for sustained [Ca²⁺]_cy oscillations. Since La³⁺ or Gd³⁺ also block Sr²⁺ uptake, and Sr²⁺ is rapidly compartmentalized at the same time, the effect of the trivalent cations may alternatively be explained by a decrease of the cytosolic Sr²⁺ activity, resulting in cessation of oscillations. However, caffeine-induced repetitive [Ca²⁺]_cy spikes in E. viridis have recently been shown to be reversibly inhibited by La³⁺ or Gd³⁺ as well (Bauer et al., 1997), corroborating the view that a certain Ca²⁺ influx is essential for sustained [Ca²⁺]_cy oscillations.

A Mathematical Model for Sr²⁺-Induced Repetitive [Ca²⁺]_cy Spikes

On the basis of the results mentioned above, it becomes possible to describe the repetitive [Ca²⁺]_cy spikes in E. viridis theoretically by adapting a mathematical model proposed by Stucki and Somoggyi (1994).

2

The two differential equations describe the changes of the Ca²⁺ activities in the cytosol and in the store due to the repetitive release and re-uptake of Ca²⁺ by internal stores plus Ca²⁺ fluxes across the plasma membrane. This is illustrated in Figure 9A. y = [Ca²⁺]_cy is increased by a steady influx A across the plasma membrane. Ca²⁺-ATPases pump cytosolic Ca²⁺ into internal stores (c) or across the plasma membrane (b) with a rate proportional to y. The fluxes out of the store depend on the difference between the Ca²⁺ activities in the store and in the cytosol (x–y). There is a linear leak (α), and a Ca²⁺ release channel (β), which is assumed to be modulated in a cooperative manner (K, n) by cytosolic Ca²⁺ (y) and Sr²⁺ (Sr in Eq. 2) activities. The change in Ca²⁺ activity inside the store x is determined by the balance of the flux into the store (c · y) and the fluxes out of the store already mentioned. The assumption of a Sr²⁺-induced Ca²⁺ release is a feature additionally introduced to the original mathematical model of Stucki and Somogyi (1994). This Sr²⁺ dependence is established for different animal cells (Mironov and Juri, 1990; Grégoire et al., 1993). Figure 9B demonstrates that this theoretical approach is sufficient to model [Ca²⁺]_cy oscillations with a baseline spiking pattern. With the parameters chosen for Figure 9B there are no oscillations in the absence of Sr²⁺ (Sr = 0) in the cytoplasm. Only in the presence of a certain amount of Sr²⁺ is Ca²⁺ released by the Ca²⁺ release channel from the intracellular Ca²⁺ store, initiating sustained [Ca²⁺]_cy oscillations.

Figure 9.

A theoretical model of the repetitive [Ca²⁺]_cy spikes in E. viridis. A, Schematic model of the Ca²⁺ fluxes involved in generating repetitive [Ca²⁺]_cy spikes in E. viridis. The [Ca²⁺]_cy, y, and Sr²⁺ activity, Sr, cause a Ca²⁺/Sr²⁺-induced Ca²⁺ release (β) from an internal Ca²⁺ store. The Ca²⁺ store is refilled by a Ca²⁺-ATPase (c), and there is a continuous Ca²⁺ efflux due to a “leak” (α). A plasma membrane Ca²⁺-ATPase removes Ca²⁺ from the cytoplasm (b), and there is a Ca²⁺ influx from the external medium (A). B, Temporal response of the system depicted in A and described by Equation 2. The [Ca²⁺]_cy, y, is plotted as a function of time, t, for: A = 0.2; b = 1; c = 2; α = 0.05; β = 15.7; K = 1; Sr = 0.1; n = 4; x(t=0) = 2.8; y(t=0) = 0.2. For the sake of simplicity no units are given and the axes are given in arbitrary units.

The theoretical model described by Equation 2 and depicted in Figure 9 is a minimal model that does not consider buffer capacities, the influx and compartmentalization of Sr²⁺, or the volumes of different compartments. However, it contains all the elements discussed above (see Fig. 10): An intracellular Ca²⁺ store that is filled by a DBHQ or CPA-sensitive Ca²⁺-ATPase, a continuous Ca²⁺ efflux from the Ca²⁺ store, which slowly depletes the store when the Ca²⁺-ATPase is blocked, and a Ca²⁺ influx pathway across the plasma membrane that is blocked by Gd³⁺ or La³⁺.

The Nature of the Ca²⁺ Release

Most [Ca²⁺]_cy oscillations investigated in nonexcitable animal cells turned out to be caused by Ca²⁺ release via the InsP₃-activated Ca²⁺ channel (Berridge, 1993; Fewtrell, 1993; Li et al., 1995a). TMB₈, which was shown to inhibit InsP₃-induced Ca²⁺ release in E. viridis (Förster, 1990), had no effect on Sr²⁺-induced [Ca²⁺]_cy oscillations, indicating that InsP₃-activated Ca²⁺ channels are not involved. In animal cells, besides caffeine, Sr²⁺ is known to induce Ca²⁺ release from internal stores via the ryanodine/cADPR Ca²⁺ channel (Meissner, 1994; Lee et al., 1995). In E. viridis caffeine and Sr²⁺ alike induce repetitive [Ca²⁺]_cy spikes and transient hyperpolarizations (Bauer et al., 1997). Ruthenium red (Ma, 1993) and ryanodine (Smith et al., 1988) are known to be specific for the ryanodine/cADPR Ca²⁺ channel, not interacting with the InsP₃-activated Ca²⁺ channel (Ehrlich et al., 1994). Ruthenium red and ryanodine were recently shown to block Ca²⁺ release in plant cells as well (Allen et al., 1995; Muir and Sanders, 1996). Both inhibitors when microinjected into E. viridis blocked Sr²⁺-induced repetitive transient hyperpolarizations (Fig. 7). Neither ruthenium red (Fig. 7) nor ryanodine affected the transient hyperpolarization observed after darkening, indicating that both inhibitors specifically blocked Sr²⁺-induced Ca²⁺ release and not the [Ca²⁺]_cy spike induced by darkening (Bauer et al., 1997). Most likely, in E. viridis Sr²⁺ induced a Ca²⁺ release from intracellular Ca²⁺ stores by activating a type of ryanodine/cADPR Ca²⁺ channel (Fig. 10).

Probably this ryanodine/cADPR Ca²⁺ channel is located in the ER (Fig. 10). This is evident from the effect of CPA or DBHQ, which specifically inhibit ER Ca²⁺-ATPases in animal (Inesi and Sagara, 1994) as well as in plant cells (Logan and Venis, 1995). As discussed above, in E. viridis CPA or DBHQ influenced neither the transport of divalent cations across the plasma membrane and the tonoplast (Fig. 8) nor the steady-state [Ca²⁺]_cy (Fig. 6). Moreover, a preperfusion of DBHQ or CPA for more than 5 min completely blocked Sr²⁺-induced oscillations, indicating a rather small volume or Ca²⁺ content of the affected Ca²⁺ store. In E. viridis as in animal cells (Kass et al., 1989; Demaurex et al., 1992), there is probably a continuous Ca²⁺ efflux from the store and when the compensating Ca²⁺ uptake by Ca²⁺-ATPases is blocked, this results in a Ca²⁺ store depletion even in the absence of Sr²⁺-induced [Ca²⁺]_cy oscillations. This depletion within a few minutes excludes the involvement of the vacuole, since the vacuole is a huge store of free Ca²⁺ in E. viridis (Bethmann et al., 1995). Little is known about Ca²⁺ release channels of the plant ER. Klüsener et al. (1995) isolated and reconstituted a voltage-dependent Ca²⁺ channel from the ER of a higher plant mechanoreceptor organ that was not affected by InsP₃ or ryanodine. In E. viridis as in other plant cells, the vacuole is the largest internal Ca²⁺ store and probably plays a key role in ion homeostasis and compartmentalization (see above). However, the [Ca²⁺]_cy oscillations described here are driven by a rather small internal Ca²⁺ store, probably the ER, and not the vacuole (Fig. 10). In good agreement, Plieth et al. (1998) recently demonstrated that the elevation in [Ca²⁺]_cy during action potentials in Chara sp. is neither caused by Ca²⁺ influx across the plasma membrane nor by Ca²⁺ release from the vacuole. A Ca²⁺ release from internal stores different from the vacuole gives rise to elevated [Ca²⁺]_cy (Plieth et al., 1998).

The Sr²⁺-induced [Ca²⁺]_cy oscillations in the unicellular green alga E. viridis show a dose-dependent frequency increase (Table I). This suggests that in E. viridis, comparable to animal cells, [Ca²⁺]_cy oscillations might encode information about external stimuli by their frequency, mediating stimulus-specific reactions by Ca²⁺-dependent signal transduction processes. It is likely that Sr²⁺-induced repetitive [Ca²⁺]_cy spikes are initialized by a Sr²⁺-induced Ca²⁺ release from the ER via a type of ryanodine/cADPR Ca²⁺ release channel. Our current working model of the different transmembrane Ca²⁺ fluxes and compartments involved in Sr²⁺-induced [Ca²⁺]_cy oscillations in E. viridis is summarized in Figure 10.

Abbreviations:

[Ca²⁺]_cy

cytosolic Ca²⁺ activity

cADPR

cyclic ADP-Rib

CPA

cyclopiazonic acid

DBHQ

2,5-di-tert-butylhydroquinone

ICP-AES

induction coupled plasma-atomic emission spectroscopy

InsP₃

inositol 1,4,5-trisphosphate

TMB₈

3,4,5-trimethoxybenzoic acid 8-diethylaminooctyl ester