INTRODUCTION

In several nonexcitable cell types activation of cell membrane receptors by hormones or neurotransmitters results in a biphasic calcium signal. An initial Ca²⁺ peak produced by calcium release from intracellular inositol (1,4,5)-trisphosphate (IP₃)¹-sensitive Ca²⁺ stores is followed by a sustained Ca²⁺ plateau due to Ca²⁺ entry from the extracellular space (1, 2, 3, 4, 5). The hypothesis that it is the decrease in the Ca²⁺ concentration in the internal store which causes Ca²⁺ influx into the cell was first proposed in 1986 by Putney and has been termed "capacitative calcium influx" (1, 2). Accordingly, not only hormonal stimulation, but also Ca²⁺ pool depletion following treatment with inhibitors of the Ca²⁺-ATPase such as di-tert-butyl-hydroquinone (t-BHQ) or thapsigargin (6) or with Ca²⁺ ionophores (7) leads to activation of Ca²⁺ entry.

In mast cells (7), RBL cells (8), Xenopus oocytes (9), and Jurkat T-cells (10), capacitative calcium influx is mediated by ion channels (I_CRAC) which have a high selectivity for calcium and is inhibited by La³⁺. At present it is unclear, however, if Ca²⁺ influx through I_CRAC is a common mechanism or if there exist different Ca²⁺ influx channels in different nonexcitable cell types. In pancreatic acinar cells capacitative calcium influx has been described by means of fluorescence measurements (11, 12) and we have recently shown that genistein blocks Ca²⁺ entry in mouse pancreatic acinar cells (13).

We describe here that calcium store depletion by the agonist acetylcholine (ACh), IP₃, the Ca²⁺-ATPase inhibitor t-BHQ, or the Ca²⁺ ionophore ionomycin activates a calcium conducting nonselective cation channel which can also be blocked by genistein but not by La³⁺. Capacitative Ca²⁺ entry as measured by fluorescence methods was also completely inhibited by genistein, whereas La³⁺ had no effect on Ca²⁺ influx in mouse but completely inhibited Ca²⁺ influx in rat pancreatic acini. This indicates the presence of different Ca²⁺ influx pathways in different animal species. We conclude from our data that in mouse pancreatic acinar cells the "calcium release-activated nonselective cation current" (I_CRANC) is responsible for capacitative calcium entry.

EXPERIMENTAL PROCEDURES

Mouse and rat pancreatic acinar cells were prepared from male CD-1 mice and male Wistar rats, respectively, as described previously (14). Acinar cells from male Wistar rats were prepared in the same way (14).

Patch-clamp experiments were performed in the tight-seal, whole cell, and cell-attached configuration (15) at room temperature (24 ± 2 °C) in a standard bath solution containing in mM: 140 NaCl, 4.7 KCl, 1.3 CaCl₂, 1 MgCl₂, 10 HEPES, 10 glucose, pH 7.4. Patch pipettes were manufactured from borosilicate glass capillaries and had resistances of 2 to 4 M when filled with a standard buffer containing in mM: 125 K⁺-Asp, 15.5 NaCl, 1 MgCl₂, 10 HEPES, 10 EGTA, 30 KOH, 2.5 mM Mg-ATP, 70 nM free Ca²⁺, pH 7.2. In some experiments BAPTA (10 mM) was used instead of EGTA, which did not influence the results. Patch-clamp experiments were recorded with a computer controlled EPC9 patch-clamp amplifier (HEKA; Lambrecht, Germany). Cell capacitance and series resistance were calculated with the software-supported internal routines of the EPC9 and compensated before each experiment. Data were sampled at 1 kH on the computer hard disk after low pass filtering at 400 Hz. In whole cell experiments voltage ramps were applied every 2 s to the cells (140 to 100 mV, slope 1 V/s) near the reversal. Either the resulting I/V curve is shown or current values at 60 mV (reversal potential of Cl currents) were extracted from every ramp and presented as time courses. Single-channel measurements were done in the whole cell mode with the same pipette solution as described above. In the cell-attached configuration pipettes were filled with the standard bath solution. Single-channel data were collected on a video tape by use of a modified pulse code modulator (Sony, PCM 501), and were later resampled (2 kHz) on a "80486" personal computer hard disk after low pass filtering at 600 Hz and were analyzed with laboratory written software.

The permeability ratios of I_CRANC were calculated as,

(Eq. 1)

(E_rev = reversal potential (V), T = temperature (K), P_ca = relative permeability for calcium, P_cat = relative permeability for cations, [Ca]_o = calcium concentration outside the cell, [cat]_i = concentration of monovalent cations inside the cell) E_rev was measured, P_Ca was set to 1, and P_cat was calculated.

Isolated single cells and acini were loaded with 3-4 µM fluo-3/AM or fura-2/AM for 30 min at room temperature. After dye loading the cell suspension was stored at 4 °C and used for experiments within 5 h. Measurements with fluo-3 were done with a confocal laser scanning microscope (Zeiss Axiovert 35 microscope equipped with the confocal laser scanning and imaging system Bio-Rad MRC-600) on single cells from multiple cell clusters (up to 5 cells) as described in greater detail elsewhere (13). The internal [Ca²⁺] is given as the mean fluorescence of fluo-3.

Fura-2 ratiometric measurements were made with an imaging system (Zeiss Axiovert 135 equipped with an apparatus from T.I.L.L. Photonics, Munich) on single cells from multiple cell clusters. Cells were excited alternately at 345 and 380 nm wavelength and the emission was collected for 50-200 ms above 510 nm wavelength. The results are given as the ratio of the fluorescence intensities at the different wavelengths (345/380).

Genistein and lavendustin A were purchased from Calbiochem-Novabiochem (Bad Soden/Germany). Fluo-3/AM and fura-2/AM were obtained from Molecular Probes (Eugene, OR). t-BHQ was from Aldrich (Steinheim, Germany) and all other chemicals from Sigma (Deisenhofen, Germany).

RESULTS

In order to activate capacitative Ca²⁺ entry we depleted intracellular Ca²⁺ stores of isolated mouse pancreatic acinar cells by extracellular addition of Ca²⁺-ATPase inhibitor t-BHQ (1 µM), ACh (1 µM), ionomycin (10 µM) or by intracellular addition of IP₃ (10 µM). Cells were dialyzed in the whole cell mode with the Ca²⁺ chelators EGTA (10 mM) or BAPTA (10 mM) through the patch pipette (free Ca²⁺ clamped to 70 nM) to prevent increases in free [Ca²⁺] leading to activation of previously described Ca²⁺-dependent Cl and nonselective cation currents (16).

Under activating conditions, currents began to develop, but with different time courses depending on the test substances used for store depletion (Fig. 1, a and b). A Ca²⁺-free pipette solution induced a slow response (t₉₀ = 484 ± 137 s), most likely by depletion of stores through Ca²⁺ leaks. ACh, t-BHQ, or intracellular application of IP₃ accelerated the current development significantly (t₉₀ = 240 ± 90 s, mean of the pooled data from all three conditions) and ionomycin induced the fastest response (t₉₀ = 90 ± 17 s). To obtain current-voltage curves (I/V curves) we applied voltage ramps (140 to 100 mV, 1 V/s). The linearity of the I/V curve (Fig. 1c) indicates voltage independence. The maximal conductance of the current was variable from cell to cell. The mean was 253 ± 211 pS/pF (n = 17) in the case of t-BHQ activation. The conductance did not differ significantly when currents were activated by ACh (220 ± 99 pS/pF, n = 5), IP₃ (228 ± 111 pS/pF, n = 6), or ionomycin (195 ± 109 pS/pF, n = 13).

Ion exchange experiments revealed a permeability sequence for the current of Rb⁺ > K⁺ = Na⁺  N-methyl-D-glucamine (NMDG⁺) as determined by changes in the reversal potentials (Fig. 1d, n = 5). A Cl component could be excluded since no changes in current were observed at the reversal potential of monovalent cations (0 mV, Fig. 1c). Furthermore, substitution of bath NMDG-Cl with NMDG-aspartate had no effect on outward currents (Fig. 1e, n = 4). The whole cell current was also able to carry Ca²⁺ and Ba²⁺ ions (Fig. 2a). Referring to the more negative reversal potentials measured after Na⁺ substitution by Ca²⁺ (V_rev(Na/Ca) = 23 ± 6.5 mV, n = 7), an apparent permeability ratio for the cations inside versus Ca²⁺ outside the cell of 13:1 was calculated. Because I_CRANC does not discriminate between Na⁺ and K⁺, this permeability ratio equals the ratio for both Na⁺:Ca²⁺ and K⁺:Ca²⁺.

To test if the described current is responsible for capacitative calcium influx in pancreatic acinar cells we compared pharmacological properties of I_CRANC and Ca²⁺ influx measured as fluo-3 fluorescence with a confocal microscope. In fluorescence experiments the plateau phase of the calcium signal is exclusively produced by extracellular calcium entering the cell (13, 17). Both I_CRANC and calcium influx were inhibited by genistein (Figs. 2, b and c, and 3, a and c). Therefore it seemed likely that Ca²⁺ influx occurred through the same genistein-inhibitable calcium pathway. Other tyrosine kinase inhibitors such as tyrphostin 25 (100 µM, n = 4, data not shown) showed 50% inhibition of both I_CRANC and Ca²⁺-influx measured with fluorescence while lavendustin A and tyrphostin B56 (n = 5, 100 µM each) had no effect. The dihydroxy analog of genistein (daidzein, 50 µM, n = 10, data not shown) which lacks the ability to regulate tyrosin kinase was also ineffective on capacitative Ca²⁺ influx. The assumption that Ca²⁺ influx as measured with fluo-3 occurred through I_CRANC was further substantiated by the finding that flufenamic acid (100 µM) (Fig. 3b) inhibited both calcium release-activated current and Ca²⁺ entry measured with fluorescence.

La³⁺ (Figs. 2c, 3c, and 4a) and other di- and trivalent cations such as Gd³⁺ (100 µM, n = 3), Cd²⁺ (1 mM, n = 3), and Mn²⁺ (1 mM, n = 3) (data not shown) had no effect on both Ca²⁺ influx and current. In particular the lack of effect of La³⁺ on I_CRANC measured in the presence of sodium (fig. 3c, left) or in the presence of calcium (Fig. 2c), and on the capacitative calcium entry measured with fluorescence techniques (Figs. 3c and 4a) is remarkable because in other systems La³⁺ is known as an inhibitor of capacitative Ca²⁺ influx and in higher concentration also of Ca²⁺ efflux due to Ca²⁺ pump inhibition (18). We therefore tested the effect of La³⁺ in more detail. We took into consideration that the maintenance of a Ca²⁺ plateau, which is usually interpreted to show Ca²⁺ influx, should also occur if both Ca²⁺ influx and Ca²⁺ extrusion was inhibited by La³⁺. In this case the effect of La³⁺ could not be taken as indication for Ca²⁺ influx following Ca²⁺ release. We therefore tested a wide range of La³⁺ concentrations (100 nM, 1 µM, 10 µM, 100 µM, and 250 µM) to inhibit agonist-stimulated Ca²⁺ entry without Ca²⁺ extrusion at low concentrations and to inhibit also Ca²⁺ extrusion at higher concentrations (5, 17, 18). In no case did La³⁺ reduce fluo-3 fluorescence whereas subsequent application of genistein in the presence of La³⁺ ([La³⁺], 100 or 250 µM) always abolished the calcium plateau (Fig. 3c). It therefore appears to be unlikely that La³⁺ inhibited Ca²⁺ entry in this concentration range. To further test the effect of La³⁺ on Ca²⁺ influx we performed the experiments shown in Fig. 4a, indicating that following hormonal depletion of Ca²⁺ stores in the absence of Ca²⁺, readdition of Ca²⁺ caused Ca²⁺ influx which was not inhibited by La³⁺.

Since La³⁺ had been described to inhibit capacitative Ca²⁺ influx in rat pancreatic acinar cells using fura-2 (17), we repeated the experiments with La³⁺ in rat to compare the results with mouse pancreatic acinar cells. In mouse pancreatic acinar cells La³⁺ had no effect on calcium entry as measured with fluo-3 (see Fig. 4a, left) or fura-2 (see Fig. 4a, right). However, in rat pancreatic acinar cells capacitative Ca²⁺ entry was inhibited by 100 µM La³⁺ (Fig. 4b). Our results confirm experiments described previously by Tsunoda et al. (17) on rat pancreatic acinar cells and indicate different capacitative Ca²⁺ influx pathways in mouse and rat. In contrast, genistein (50 µM) completely inhibited capacitative Ca²⁺ influx in both mouse (see Fig. 3, a, right, and c, right) and rat as measured by fluorescence (data not shown).

In some experiments (n = 6 out of n = 32 and 7 out of 52 cells) in which store depletion resulted in the activation of only a small current it was possible to recognize single channel events in the whole cell mode of the patch-clamp technique. The I/V curves of these channels were linear and reversed at 0 mV (Fig. 5a). The conductance in the whole cell mode measured with standard solutions was in the range of 40-45 pS (43 ± 3.3 pS, n = 7). In the cell-attached mode the 40-45 pS channel could be irreversibly activated with t-BHQ (Fig. 5b) or reversibly with ACh (Fig. 5c). Under these conditions the channel always coexisted with the previously described 27 pS channel (Fig. 5b) which is Ca²⁺-activated and therefore is observed under conditions at which cytoplasmic Ca²⁺ rises (19). Because the 27 pS channel has a high density (about 500 channels/cell (20)) compared to the 45 pS channel (about 70 channels/cell, calculated as the mean single cell conductance/single channel conductance assuming a P_O of 0.6 for the single channel) it is not surprising that the 40-45 pS channel was seen in the cell-attached mode in only 10% of the cells. The possibility that the two conductance levels which we found in cell-attached experiments are sublevels of one channel type can be ruled out by the finding that in all whole cell experiments only a single conductance level of 40-45 pS was observed. Excising patches with the 40-45 pS channel into a standard bath solution resulted in an immediate run-down of this channel type leaving only the 27 pS channel active (n = 4).

DISCUSSION

The results presented here indicate that capacitative Ca²⁺ entry in mouse pancreatic acinar cells is produced by a calcium release-activated Ca²⁺ conducting nonselective cation channel (I_CRANC). The main characteristic of I_CRANC is its insensitivity to La³⁺ which argues against the presence of the previously described I_CRAC (7) in mouse pancreatic acinar cells. Flufenamic acid, an inhibitor of nonselective cation channels (21), and genistein, a tyrosine kinase inhibitor, also inhibited both I_CRANC and capacitative Ca²⁺ entry. The genistein effect should be emphasized because it was shown before (13) that it inhibits calcium entry without affecting calcium release from IP₃-sensitive pools. We believe that these pharmacological similarities of I_CRANC and capacitative calcium entry are consistent with the possibility that I_CRANC is responsible for capacitative Ca²⁺ influx in mouse pancreatic acinar cells. The mechanism for the effect of genistein on calcium influx is not yet clear because genistein effects are diverse (22). The lack of effect of other tyrosine kinase inhibitors like lavendustin A or tyrphostin B56 on mouse pancreatic acinar cells led us to conclude that a more direct interaction between channel and genistein rather than the proposed inhibition of a tyrosine kinase (23, 24) is responsible for inhibition.

The nonselective cation current, which is activated by Ca²⁺ store depletion (I_CRANC) in mouse pancreatic acinar cells, differs from capacitative Ca²⁺ influx described in other systems. In mast cells and other cell types I_CRAC seems to be the dominant influx pathway for Ca²⁺ (7, 8, 9, 10). It is highly Ca²⁺ selective and can be inhibited over 90% with low concentrations of La³⁺ (10 µM) and in part by Cd²⁺ and Co²⁺ (25). Single channels producing I_CRAC could not be identified so far and noise analysis of I_CRAC made it likely that the single channel conductance is much below the resolution threshold of the patch-clamp technique (25). Comparing the characteristics of I_CRANC as described here with those of I_CRAC it appears that both currents are different although they have the same activation mechanism.

Single channels activated by Ca²⁺ store depletion were identified in vascular endothelium (26). These channels were relatively Ca²⁺ selective (permeability ratio Ca/Na = 10/1) and had a conductance of 11 pS (with 10 mM Ca²⁺). Also in A431 cells single channels activated by store depletion were found which had a relative low conductance in the presence of high Ca²⁺ or Ba²⁺ concentrations (200 mM Ca²⁺, conductance 2 pS or 160 mM Ba²⁺, conductance 16 pS, respectively) (27). These channels clearly differ from the I_CRANC channel in mouse described here.

Nonselective cation channels have been discussed as candidates for mediating Ca²⁺ influx in nonexcitable cells (28). While it had been described for several systems that agonist activated nonselective cation channels allow calcium influx into cells (29, 30, 31) our study demonstrates that depletion of Ca²⁺ stores can directly activate those channels.

Whereas I_CRAC is inhibited by several di- and trivalent ions such as Co²⁺, Cd²⁺, and most effectively by La³⁺ (25), these ions are ineffective in mouse pancreatic acinar cells. In particular the lack of effect of La³⁺ on I_CRANC measured as electrical current and as calcium fluorescence with both fluo-3 and fura-2 should be emphasized. La³⁺ at low concentrations (micromolar) has been reported to inhibit capacitative calcium entry, while at higher concentrations (millimolar) it inhibited the calcium extrusion mechanism in lacrimal acinar cells (18). Moreover it was shown in rat pancreatic acinar cells (17) and guinea pig pancreatic acinar cells (5) that La³⁺ (25-250 µM) inhibited capacitative calcium entry. In mouse pancreatic acinar cell 1 mM La³⁺ inhibited Ca²⁺ extrusion and the authors assumed that Ca²⁺ influx was inhibited, too (32). Direct evidence for this assumption was not given, however. In contrast to these results we did not find any inhibition of Ca²⁺ entry by La³⁺ up to 250 µM in mouse pancreatic acinar cells with different experimental protocols including different depletion methods (ACh and t-BHQ) and different detection methods (measurements of electrical current and of fluorescence with fluo-3 or fura-2). The reason for the discrepancy between the data in the literature and our data are most likely due to species differences which can also be concluded from our own results which compare the effects of 100 µM La³⁺ in rat and mouse (Fig. 4). These differences in the La³⁺ sensitivity between rat and mouse make it likely that rat but not mouse pancreatic acinar cells use I_CRAC for capacitative Ca²⁺ influx as it had been assumed already by Bahnson et al. (33).

If in addition to I_CRANC mouse pancreatic acinar cells would also contain I_CRAC we would have expected at least, in part, inhibition of capacitative Ca²⁺ influx by La³⁺. It therefore appears that in mouse pancreatic acinar cells I_CRAC is not present or in such small amounts that it is undetectable by our methods. In conclusion our data indicate that a Ca²⁺ conducting nonselective cation channel is activated following depletion of intracellular IP₃-sensitive Ca²⁺ stores in mouse pancreatic acinar cells.

The characteristics of I_CRANC are different from I_CRAC previously described in mast cells and other cell types (34) in that the latter is highly Ca²⁺ selective, completely inhibited by La³⁺, and in part by Cd²⁺ and Co²⁺. Evidence suggests that mammalian Ca²⁺ influx channels are homologues of insect trp and trpl channels (35, 36, 37). Whereas trp is selective for Ca²⁺, has a high La³⁺ sensitivity, and is activated by Ca²⁺ store depletion (and therefore seems to have similarities with I_CRAC, 38), trpl is a nonselective cation channel conducting also Ca²⁺ and Ba²⁺ ions. It has a lower La³⁺ sensitivity compared to trp but does not seem to be activated by Ca²⁺ store depletion (35, 38, 39). Whether I_CRANC in pancreatic acinar cells shares genetic homologies with these insect channels remains to be determined in future studies.