Calcium Entry through L-type Calcium Channels Causes Mitochondrial Disruption and Chromaffin Cell Death*

Sustained, mild K (cid:1) depolarization caused bovine chromaffin cell death through a Ca 2 (cid:1) -dependent mechanism. During depolarization, Ca 2 (cid:1) entered preferentially through L-channels to induce necrotic or apoptotic cell death, depending on the duration of the cytosolic Ca 2 (cid:1) concentration ([Ca 2 (cid:1) ] c ) signal, as proven by the following. (i) The L-type Ca 2 (cid:1) channel activators Bay K 8644 and FPL64176, more than doubled the cytotoxic effects of 30 m M K (cid:1) ; (ii) the L-type Ca 2 (cid:1) channel blocker nimodipine

land); and -conotoxin MVIIC and -agatoxin IVA were from the Peptide Institute (Osaka, Japan). The assay kit for measuring the activity of lactate dehydrogenase (LDH) and Cell Death Detection Elisaplus Kit were purchased from Roche Molecular Biochemicals. Fura-2/AM, rhodamine 123, Vybrant Apoptosis Kit, and Mitotracker Red were purchased from Molecular Probes, Inc.
Preparation and Culture of Bovine Chromaffin Cells-Bovine adrenal medullary chromaffin cells were isolated as previously described (28) with some modifications (29). To reduce the number of endothelial cells in the culture that could alter LDH measurements, cells were preplated for 30 min, and proliferation inhibitors (cytosine arabinoside, L-leucine methyl esther, and fluorodeoxyuridine) were used during the maintenance of the culture in the Dulbecco's modified Eagle's medium. For cell death studies, cells were plated at a density of 5 ϫ 10 5 cells/well on 24-well Orange plates coated with 0.01 mg ml Ϫ1 of poly-L-lysine, containing 1 ml of Dulbecco's modified Eagle's medium supplemented with 5% fetal calf serum, 10 M cytosine arabinoside, 10 M fluorodeoxyuridine, 10 M leucine methyl esther, 50 IU ml Ϫ1 penicillin, and 50 g ml Ϫ1 streptomycin. Cultures were maintained for 2-3 days at 37°C in a water-saturated atmosphere with 5% CO 2 . After 24 h, the medium was replaced by 1 ml of serum-free fresh medium and subsequently changed every 2-3 days. Cells were normally used during days 2-3, to avoid excessive growth of endothelial cells that could interfere with LDH measurements.
LDH Assay-Extracellular and intracellular LDH activities were spectrophotometrically measured by following tetrazolium reduction at an absorbance wavelength of 492 nm. Total LDH activity was defined as the sum of intracellular plus extracellular LDH activity. Released LDH was defined as the percentage of extracellular compared with total LDH activity (30).
Nuclear Staining of DNA-For microscopic nuclear DNA analysis, bovine chromaffin cells cultured in 24-well plates (5 ϫ 10 5 /well) were stained with 5 g/ml Hoechst dye 33342 in Krebs-Hepes solution for 30 min at 37°C (31). Stained cells were washed twice with phosphatebuffered saline, removed from the well by gentle shaking, and placed in a plastic tube for centrifugation at 800 rpm for 10 min. Supernatants were removed, and cells were resuspended in 50 l of phosphatebuffered saline. A sample of 15 l of the stained cells was placed on a coverslip and counted under the fluorescent microscope. Fluorescence microscopy was performed using a Nikon Elipse TE300 microscope, using the appropriate filter for an excitation wavelength of 355 nm and an emission wavelength of 465 nm. Nuclei with the apoptotic features were counted as apoptotic cells of a total of 300 cells in at least three different batches of cells. The percentage was calculated for each sample.
DNA Fragmentation Analysis-For the quantitative determination of cytoplasmic histone-associated DNA fragments, the Cell Death Detection Elisaplus Kit was used (32). After induction of cell death, every well was treated with the lysis buffer for 30 min. Samples were spun down at 200 ϫ g for 10 min. An aliquot from the supernatant (20 l) was transferred to the streptavidin-coated microplate for analysis. 80 l of the immunoreagent mixture prepared with 1 ⁄20 volume of anti-DNAperoxidase and 1 ⁄20 volume of anti-histone-biotin with 18 ⁄20 volumes of incubation buffer were added to each well and maintained for 2 h at room temperature with gentle shaking. Thereafter, the solution was removed by aspiration, and the wells were rinsed three times with 250 l of incubation buffer. Then 100 l of the substrate solution was added to each well and incubated for 5 min until the color developed. Measurements were made at 600 nm. Results were expressed as the absorbance of each sample and compared with basal levels.
Measurement of Cytosolic Ca 2ϩ Concentrations, [Ca 2ϩ ] c -For these experiments, cells were plated on 1-cm diameter glass coverslips at a density of 5 ϫ 10 4 cells/coverslip. Cells were loaded with Fura-2/AM (4 M) for 45 min at 37°C in Krebs-Hepes solution. Loading with the fluorescent dye was terminated by washing the coverslip containing the attached cells twice with Krebs-Hepes; then the cells were kept at room temperature for 15 min before placing them in the headstage of a Nikon Diaphot microscope to measure their fluorescence.
The fluorescence of Fura-2 in single cells was measured with the photomultiplier-based system (33), which produces a spatially averaged measurement of the [Ca 2ϩ ] c . Fura-2 was excited with light alternating between 360 and 390 nm, using a Nikon ϫ 40 fluorite objective. Emitted light was transmitted through a 425-nm dichroic mirror and 500 -545-nm barrier filter before being detected by the photomultiplier.
[Ca 2ϩ ] c was calculated from the ratios of the light emitted when the dye was excited by the two alternating excitation wavelengths (34).
Measurements of Mitochondrial Transmembrane Potential (⌬⌿m) by Fluorescence Confocal Imaging-The fluorescence of rhodamine 123 was used to measure ⌬⌿m (35). Cells were incubated for 30 min in the presence of 10 M of the dye. Rhodamine 123 was added to all of the solutions used during the experiment, in order to avoid washout of the dye. Three to six cells were studied in each experiment. Images were collected at 1 image/0.372 s. Cellular fluorescence was imaged using a MRC 6000 confocal laser-scanning microscope with excitation at 488 nm and emission at 510 nm and a ϫ 60 oil immersion objective. Pixel intensity/cell was determined using Bio-Rad software. Base-line fluorescence (F o ) was measured as normalized fluorescence emitted during the experiment. Increments in fluorescence indicate mitochondrial depolarization. All experiments were performed at 22 Ϯ 1°C in bovine chromaffin cells cultured for 2-3 days.
Mitotracker Red to Visualize Mitochondria-Cells were placed in coverslips at a density of 3 ϫ 10 5 cells. Control and treated cells (0.3 M FPL64176, 30 mM K ϩ , 5 mM Ca 2ϩ ) were loaded with the cell-permeant mitochondrial dye, Mitotracker red 100 nM at 37°C for 30 min. Once the Mitotracker probe accumulates in mitochondria, it reacts with accessible thiol groups of peptides and proteins and forms an aldehyde conjugate that is stable. Fluorescent images were taken under a confocal laser-scanning microscope with excitation at 570 nm and emission at 599 nm (36,37).
Statistical Analysis-Data were expressed as means Ϯ S.E. Statistical significance of differences between means was determined by an analysis of variance test. If significant differences were found, an appropriate multiple comparison test (Fisher PLSD) was done. In some cases, Student's t test was used (see Figs. 1 and 3-6 legends). Differences were considered significant at the level of p Ͻ 0.05.

Cell Death as a Function of Depolarization and Ca 2ϩ
Entry through L-type Calcium Channels: Protection by Nimodipine-To facilitate ionic manipulations, all experiments were performed in Krebs-Hepes solutions. For instance, in the experiment shown in Fig. 1a, cells were incubated in Krebs-Hepes solution containing increasing concentrations of KCl (5.9 -50 mM), with isosmotic reduction of NaCl, for 24 h at 37°C. At the end of this incubation period, the medium was collected to measure LDH released from dead cells (extracellular LDH; LDH e ); the cells that remained attached to the well were then lysed, and their LDH content was measured. The expression of LDH e as a percentage of total LDH provides an indication of the fraction of cells dying as a consequence of a given treatment, in this case depolarization of increasing strengths. Incubation of the cells for 24 h with increasing [K ϩ ] e gave a gradual release of LDH e ; maximum cell damage was observed at 30 mM K ϩ , which caused 25% cell loss; a further increase to 50 mM K ϩ did not enhance cell death. Hence, 30 mM [K ϩ ] e was chosen to perform subsequent experiments.
Cell damage was not only a function of the degree of VDCC opened upon depolarization with high [K ϩ ] e but also of the quantity of external Ca 2ϩ offered to the cells ([Ca 2ϩ ] e ). In a second experiment (Fig. 1b), the effects of increasing [Ca 2ϩ ] e on cell death, at a fixed level of depolarization (30 mM K ϩ ), were assayed. Cell death after a 24-h incubation was close to basal at 0.2-0.5 mM [Ca 2ϩ ] e (13% LDH e release); at 2-10 mM [Ca 2ϩ ] e , cell death increased to around 25%, and it declined to 20% at 20 mM [Ca 2ϩ ] e ; this effect could be due to inactivation by excess Ca 2ϩ of VDCC (38 -40).
To increase the cell damage caused by Ca 2ϩ entry through L-type VDCC, we used different L-type agonists like FPL64176, (ϩ)Bay K 8644, and (Ϫ)Bay K 8644 (41) under mild depolarizing conditions (30 mM [K ϩ ] e ) for 24 h. Fig. 1c shows the cytotoxic consequence of the incubation with these agonists at concentrations that ranged from 0.3 to 10 M. Maximum cytotoxic effects, for all drugs tested, were observed at 0.3-1 M; higher concentrations of the agonists (3-10 M) presented a decrease in LDH e ; this effect can be explained by their Ca 2ϩ antagonist action at high concentrations (41,42). Out of the three L-type agonists used, FPL64176 presented the highest cytotoxic effects. Therefore, the combination of 0.3 M FPL64176 in 30 mM [K ϩ ] e was used to induce cell death via Ca 2ϩ entry through L-type VDCC.
The cytotoxic effect of Ca 2ϩ entry through L-channels induced by FPL64176 was dependent on the [Ca 2ϩ ] e . In Fig. 1d, basal LDH e (24-h incubation in 5.9 mM K ϩ ) amounted to 8 . It is interesting that this sharp increase in cell death could be completely reversed by a 1 M concentration of the L-type Ca 2ϩ channel blocker nimodipine (Fig. 1d); LDH e after 24 h was in the range of 10 -15% (i.e. similar to that found in basal conditions (5.9 mM K ϩ )). It is worth noting that nimodipine counteracted both the LDH e increase evoked by 30 mM K ϩ in the absence (about 25%) and in the presence of FPL64176 (about 50%), suggesting that the increase in cell death was associated to increased Ca 2ϩ entry mostly through L-type Ca 2ϩ channels.
[ ] c rose initially to a peak of 905 Ϯ 120 nM (n ϭ 9) and then declined to a sustained plateau that was 1.4-fold higher (409 Ϯ 91 nM) than in low [Ca 2ϩ ] e (0.2 mM), along the rest of the 30-min depolarization period (Fig. 2d).
When measuring the total quantity of Ca 2ϩ entering the cells  4) and 302 Ϯ 62 nM⅐s (n ϭ 8), respectively. The same pattern was also observed when measuring the initial peak of [Ca 2ϩ ] c (see Fig. 3a).
When analyzing cell lesions as LDH e released, we observed that the amount of Ca 2ϩ entering the cell during 30 min, measured as the integral of the curve of [Ca 2ϩ ] c , correlated well with the extent of cell death observed 24 h later (see Fig. 3b). These results prove that cell lesion is a function of the quantity of Ca 2ϩ entering through VDCC and that it depends not only on the [Ca 2ϩ ] e but also on the inactivation state of those channels.
Cytoprotection Afforded by Nimodipine Is Related to the Reduction of [Ca 2ϩ ] c -We had previously observed (Fig. 1d) that 1 M nimodipine completely counteracted LDH release induced by Ca 2ϩ overload through L-type Ca 2ϩ channels. In order to determine if the protection afforded by nimodipine was related to a reduction in the levels of [Ca 2ϩ ] c , we performed experiments in single Fura-2/AM-loaded cells. Fig. 4 shows how nimodipine restored the [Ca 2ϩ ] c to basal levels after the [Ca 2ϩ ] c was raised by FPL64176, 30 mM K ϩ , both at low (0.2 mM), (Fig.  4a) and high (5 mM) (Fig. 4b) (Fig. 4c). This drastic reduction in the [Ca 2ϩ ] c could explain the total cytoprotection observed previously with nimodipine (see Fig. 1d).
Since nimodipine is selectively blocking L-type Ca 2ϩ channels, it seemed obvious that N-and P/Q-type Ca 2ϩ channels would not be contributing to the Ca 2ϩ signals or to the LDH e release in cells exposed to Ca 2ϩ overload through L-channels. This possibility was verified by measuring both [Ca 2ϩ ] c changes and cell death in the presence of non-L-type Ca 2ϩ channel blockers (-conotoxin GVIA for N-channels plus -conotoxin MVIIC for N/P/Q-channels). When measuring cell damage as a function of LDH release in depolarized cells treated with FPL64176 at different [Ca 2ϩ ] e (0.2-20 mM), we curiously observed significant cytoprotection by N/P/Q-type blockers at low [Ca 2ϩ ] e (0.2-2 mM) but not at higher concentrations (i.e. above 5 mM; see Fig. 1d). However, the increase of [Ca 2ϩ ] c induced by FPL64176, 30 mM K ϩ in low and high [Ca 2ϩ ] e was not significantly modified in the presence of the non-L-type channel blockers (data not shown).
Prolonged Ca 2ϩ Entry through L-type Channels Leads Primarily to Necrotic Cell Death-To further understand the basic mechanism by which chromaffin cells were dying upon their sustained depolarization, we compared the amount of necrosis measured as LDH release and apoptosis measured as histoneassociated DNA fragments in cells exposed for 24 h to FPL64176, 30 mM K ϩ , 5 mM Ca 2ϩ . Fig. 5a shows that the cytotoxic effect of Ca 2ϩ entry through L-channels induced significant LDH e release but no significant increase in the number of apoptotic cells. These results suggest that the elevation of [Ca 2ϩ ] c during prolonged depolarizations, in the presence of an L-type channel activator, constitutes predominantly a necrotic signal.
During cell necrosis, dramatic changes can be observed in the mitochondrial structure (9). The morphological changes in the mitochondria network induced by prolonged (24-h) Ca 2ϩ entry through L-channels could be observed in cells stained with the mitochondrial dye Mitotracker red. Chronic treatment with FPL64176 induced total disruption of the mitochondrial mesh (Fig. 5c), indicating cell necrosis. In contrast, the control cell in Fig. 5b presents the mitochondrial network completely preserved.
Transient Ca 2ϩ Overload through L-channels Mediates Apoptotic Cell Death Secondary to the Release of Cytochrome c and Free Radicals from the Mitochondria-During an ischemic episode, a transient lack of oxygen normally occurs. Therefore, we tried experimental conditions where Ca 2ϩ overload through L-channels (0.3 M FPL64176, 30 mM K ϩ , 5 mM Ca 2ϩ ) had a duration of 30 min; we evaluated the cytotoxic consequence of such stimulus, measuring cell death as release of LDH (immediately after the stimulus) and the number of apoptotic nuclei (48 h after the stimulus) with the fluorescent dye Hoechst. We found that Ca 2ϩ entering through L-channels for 30 min almost doubled the basal release of LDH and increased the number of apoptotic nuclei from 11 Ϯ 0.7% (basal) to 26.8 Ϯ 3.3% (depo-larizing pulse) after 48 h (Fig. 6a). These results contrast with those obtained when the toxic stimulus (0.3 M FPL64176, 30 mM K ϩ , 5 mM Ca 2ϩ ) was kept for 24 h (Fig. 5). In the latter case, no significant increase in the number of apoptotic nuclei was observed; the main lesion found was necrotic. These results indicated that shorter increases in the [Ca 2ϩ ] c were capable of activating the apoptotic cascade; the question now was to verify how this mechanism was taking place.
Transient peaks in [Ca 2ϩ ] c can secondarily increase mitochondrial Ca 2ϩ ([Ca 2ϩ ] m ), lead to the opening of the mitochondrial transition pore (43), release of cytochrome c (36) with the resultant release of oxygen free radicals (like superoxide) from the electron transport chain (44,45), and initiate the apoptotic cascade (46). To determine if the cells were dying through this mechanism, we performed two types of experiments: (i) block- ade of the opening of the transition pore by CsA (3 M) and the subsequent release of cytochrome c and (ii) use of the antioxidant enzyme SOD to prevent the action of superoxide free radicals that could be released from mitochondria. When CsA and SOD were present during the 30-min depolarizing pulse (0.3 M FPL64176, 30 mM K ϩ , 5 mM Ca 2ϩ ), the number of apoptotic nuclei returned to basal levels (10.8 Ϯ 2.3 and 7.5 Ϯ 2.4%, respectively), indicating that release of cytochrome c and free radicals was involved in inducing apoptosis via Ca 2ϩ entry through L-channels. When Ca 2ϩ entry was prevented with nimodipine (1 M), apoptotic cell death was also prevented (Fig.  6a).
Cell death measured as LDH released in the first 30 min was completely counteracted when SOD and nimodipine were present during the 30-min depolarizing pulse, but not by CsA (Fig. 6b).
These results suggest that the amount of Ca 2ϩ entering the cell with FPL64176, 30 mM K, 5 mM Ca 2ϩ during 30 min, that corresponds to 496 Ϯ 77 nM⅐s of [Ca 2ϩ ] c , mediates early necrotic cell death (LDH e ) and late apoptosis. Early necrotic cell death and late apoptosis could be prevented by nimodipine and SOD. Therefore, blockade of L-channels and superoxide radicals are involved both in early necrotic and late apoptotic cell death. The fact that CsA did not protect against early necrosis suggests that release of cytochrome c, under these experimental conditions, is primarily activating the apoptotic cascade and cell death at later stages.
Calcium studies have begun to clarify the deleterious effects that may result from mitochondrial Ca 2ϩ overload. Rapid Ca 2ϩ uptake causes mitochondrial depolarization (48), impairment of energy metabolism (49), and uncoupling of electron transport from ATP production (50,51) and cell death.
With this background in mind, we carried out experiments to determine whether Ca 2ϩ driven through L-type Ca 2ϩ channels was being seen primarily by the mitochondria and if this Ca 2ϩ was able to depolarize the mitochondrial membrane that, in turn, would lead to cell death. Fluorescence measurements of [Ca 2ϩ ] c were performed in single loaded cells with Fura-2/AM. Fig. 7a shows an original trace of an experiment where two initial control (0.3 M FPL64176, 30 mM K ϩ , 5 mM Ca 2ϩ ) depolarizing pulses (10 s) were applied, followed by a pulse in the presence of the protonophore CCCP (2 M) or after pretreatment with CCCP for 20 s. In both cases, when Ca 2ϩ uptake by mitochondria was prevented with CCCP, the [Ca 2ϩ ] c signal almost doubled that of control pulses.
A similar experimental procedure was carried out to test the contribution of the endoplasmic reticulum to the buffering of Ca 2ϩ entering through L-type Ca 2ϩ channels (Fig. 7b). After the application of two control pulses (0.3 M FPL64176, 30 mM K ϩ , 5 mM Ca 2ϩ ), the endoplasmic reticulum was depleted using the combination of 10 mM caffeine, 10 M ryanodine, and 1 M thapsigargin. The depolarizing pulse given in the presence of the endoplasmic reticulum-depleting solution did not increase the intracellular Ca 2ϩ signal, indicating little or no contribution of the endoplasmic reticulum to the buffering of cytosolic Ca 2ϩ in these experimental conditions. Therefore, these results suggest that mitochondria are the main buffering system for Ca 2ϩ overload through L-channels.
In order to determine whether [Ca 2ϩ ] c elevations mediated by activation of L-channels (FPL64176, 30 mM K ϩ , 5 mM Ca 2ϩ ) were able to depolarize the mitochondria, we measured the ⌬⌿m with the fluorescent dye rhodamine 123. Confocal images were taken every 0.372 s. Fig. 8 shows experiments that illustrate how FPL64176, 30 mM K ϩ , 5 mM Ca 2ϩ induced increases in the fluorescence of rhodamine 123, indicating mitochondrial membrane depolarization; when the same cell was perfused with FPL, 30 mM K ϩ , 5 mM Ca 2ϩ in the presence of nimodipine, mitochondrial depolarization was prevented. These results FIG. 8. Ca 2؉ entry through L-channels depolarizes the mitochondrial membrane, an effect that could be prevented by nimodipine. After recording base-line fluorescence of rhodamine 123, the cells were exposed for 10 s to 0.3 M FPL64176, 30 mM K ϩ , 5 mM Ca 2ϩ ; a direct depolarization of the mitochondria could be seen as an increase of fluorescence. After washout, the same cell was again stimulated with FPL64176, 30 mM K ϩ , 5 mM Ca 2ϩ in the presence of 1 M nimodipine (added 2 min before), and the increase in fluorescence was completely abolished. The same set of stimuli were repeated in the same cell. b and c represent the sequential fluorescent images obtained in a confocal microscope of the first and second stimuli presented in a.
demonstrate that depolarization of the mitochondrial membrane occurs when an elevation of [Ca 2ϩ ] c occurs, in this case after the opening of L-type Ca 2ϩ channels. DISCUSSION We demonstrate in this study that sustained, mild K ϩ depolarization causes chromaffin cell death (Fig. 1b) through a Ca 2ϩ -dependent mechanism. Under these conditions, Ca 2ϩ can gain the cell cytosol through several of the voltage-dependent Ca 2ϩ channels described in bovine chromaffin cells. As indicated by numerous patch-clamp studies, external Ca 2ϩ enters the voltage-clamped bovine chromaffin cell during depolarization through L-type (20%), N-type (30%), and P/Q-type (50%) Ca 2ϩ channels (see Ref. 18). Despite this, it was curious that most of the Ca 2ϩ entering through L-channels, but not through N-or P/Q-channels, was responsible for the activation of the death signal, as proven by the following observations. (i) The L-type Ca 2ϩ channel activators Bay K 8644 and FPL64176 more than doubled the cytotoxic effects of 30 mM K ϩ ; (ii) the L-type Ca 2ϩ channel blocker nimodipine suppressed the cytotoxic effects of 30 mM K ϩ alone or 30 mM K ϩ plus FPL64176; (iii) conversely, toxin blockade of N-and P/Q-channels caused partial cytoprotection at 0.2-2 mM [Ca 2ϩ ] e and no protection at 5-20 mM [Ca 2ϩ ] e ; and (iv) the potentiation by FPL64176 of the K ϩ -evoked [Ca 2ϩ ] c elevation was suppressed by nimodipine.
The partial cytoprotection afforded by combined toxins at the lower [Ca 2ϩ ] e is puzzling but can surely be interpreted in the context of the recent finding of our laboratory showing the Ca 2ϩ -dependent inactivation of L-, N-, and P/Q-channels in voltage-clamped bovine chromaffin cells (39). The elevation of [Ca 2ϩ ] c elicited by the mitochondrial uncoupler CCCP during cell depolarization causes a faster inhibition of N-and P/Qchannels, as compared with L-channels. This might explain why, in the present study, Ca 2ϩ entry through L-type Ca 2ϩ channels caused greater cell lesion than that gained through Nand P/Q-channels. Despite the fact that L-channels account for only one-fifth of the total Ca 2ϩ entering the cell during depolarization, they are capable of triggering a cell death signal with more efficacy than N-or P/Q-channels. Now the question arises as to how such Ca 2ϩ entry through L-channels causes cell death.
Prolonged depolarization induced by K ϩ caused a transient sharp rise of [Ca 2ϩ ] c followed by a sustained plateau. Although this bulk [Ca 2ϩ ] c elevation reached only around 1 M, it is certain that mitochondria see greater Ca 2ϩ transients at subplasmalemmal sites near the Ca 2ϩ channels; thus, using mitochondrially targeted aequorin, we have recently shown that Ca 2ϩ inside the mitochondria can reach near millimolar concentrations during cell depolarization (52). If these elevations of [Ca 2ϩ ] m are sustained, then the mitochondrial transition pore will open (49,53,54), and the apoptotic cascade will be activated. These findings suggest the following. (i) Cell exposure to K ϩ plus FPL64176 caused depolarization of mitochondria, surely due to mitochondrial Ca 2ϩ accumulation; (ii) CsA, a blocker of the mitochondrial transition pore, prevented the apoptotic cell death induced by K ϩ plus FPL64176; (iii) and SOD also suppressed this apoptotic signal, suggesting that mitochondrial Ca 2ϩ overload was generating free radicals to cause cell death (13,55,56).
It is interesting that nimodipine provided full protection against both necrotic and apoptotic cell death. This finding reinforces the view that nimodipine might have direct neuroprotectant effects on neurons subjected to an ischemic insult (14,57) in addition to its well known cerebrovascular vasodilatory effects (58). The neuroprotectant actions of nimodipine in clinical trials performed in patients suffering a thrombotic stroke have proven difficult to demonstrate (59 -61). However, in experimental animal models of cerebral ischemia, a clear nimodipine-induced neuroprotection has been shown (22,62), which is in line with the results of the experiments shown here.
In conclusion, our data suggest that N-and P/Q-type Ca 2ϩ channels, which suffer rapid Ca 2ϩ -dependent inhibition after cell depolarization, are unlikely to contribute to cell death upon a depolarizing stimulus. This observation is in line with previous data from our laboratory indicating that blockers of N-and P/Q-type Ca 2ϩ channels did not protect against veratridineinduced cell death (64,65). However, they disagree with the observation that N-type Ca 2ϩ channel blockers afforded protection in a rat model of cerebral ischemia (15). In contrast, L-type channels that are localized preferentially at the neuronal soma (66) inactivate more slowly and are clearly associated with cell Ca 2ϩ overload, mitochondria depolarization, generation of free radicals, and cell death. These data strengthen the view that Ca 2ϩ entry through L-channels during a cerebral ischemic condition causing neuronal depolarization (67) might be a critical determinant of delayed death of neurons located in the penumbra area. Hence, dihydropyridine blockers of L-channels should have pronounced neuroprotectant actions if given with an adequate therapeutic window to stroke patients.