Introduction

Bovine chromaffin cells express neuronal voltage-dependent Ca2+ channels of the subtypes L (α1D, Cav1.3), N (α1B, Cav2.2), and PQ (α1A, Cav2.1) [22], having similar biophysical and pharmacological properties to those expressed by neurons [20, 23, 59]. Why these cells express L as well as non-L (N/PQ) Ca2+ channels is an intriguing question. There are reports suggesting different modulation of channel subtypes and different roles in controlling exocytosis, according to various experimental conditions and patterns of chromaffin cell stimulation. The following findings illustrate this point: (1) Ca2+ channel currents (I Ca) are modulated by neurotransmitters in a voltage-dependent (N/PQ channels) or voltage-independent manner (L channels) [1, 18, 19, 30]; (2) depending on the experimental conditions, L channels may be preferentially coupled to exocytosis [40] or PQ channels may dominate the control of secretion [36]; (3) cell stimulation with action potentials (APs) produce a secretory response mostly controlled by PQ channels, while secretion stimulated with depolarizing pulses is controlled by other channel subtypes [14]; (4) L channels are tightly coupled to endocytosis while N/PQ channels are more weakly coupled to endocytosis [60, 63]; (5) L channels are more resistant to voltage-dependent inactivation while N/PQ channels are more prone to such inactivation [31, 70]; and (6) L channels undergo Ca2+-dependent inhibition at a rate slower than N/PQ channels [32].

The Ca2+-dependent inactivation of Ca2+ channels was first suggested by Hagiwara and Nakajima [28], subsequently proven by Brehm and Eckert [9] and Tillotson [68] in mollusc neurons, and later on extended to various excitable cells [27]. It is interesting to note that not all Ca2+ channel subtypes are equally prone to Ca2+-dependent inactivation. For instance, the cardiac and smooth muscle L channel is quickly inactivated upon depolarization [25, 37, 52] while the neuronal L channel inactivates more slowly [71]. Because the L channel inactivates more slowly in bovine chromaffin cells [32], it may be that Ca2+ entry through this channel could modulate N/PQ channels. In this study, we tested this hypothesis by measuring the whole-cell inward Ca2+ channels current (I Ca) in voltage-clamped bovine adrenal chromaffin cells. We used the L channel activator FPL64176 (FPL) [44, 75] to enhance the I Ca flowing through L channels to test its consequences on the inhibition of the I Ca flowing through N and PQ channels. The results support the hypothesis implying the existence of a Ca2+-mediated cross-talk between L and N/PQ Ca2+ channels.

Materials and methods

Isolation and culture of bovine chromaffin cells

Bovine adrenal glands were obtained from a local slaughterhouse. Chromaffin cells were isolated by digestion of the adrenal medulla with collagenase following standard methods [39] with some modifications [50]. Cells were suspended in Dulbecco's modified Eagle's medium (DMEM) supplemented with 5% fetal calf serum, 10 µM cytosine arabinoside, 10 µM fluorodeoxyrudine, 50 IU mL−1 penicillin, and 50 µg mL−1 streptomycin. Cells were plated on 13-mm diameter polylisine-coated glass coverslips at a density of 5 × 104 cells per coverslip. Cells were kept in a water-saturated incubator at 37°C in a 5% CO2–95% air atmosphere and used 2–5 days thereafter.

Electrophysiological recordings

ICa were recorded in voltage-clamped cells under whole-cell configuration (WCC) or perforated-patch configuration (PPC) [26, 34] of the patch-clamp technique [29]. During recording, cells were continuously perfused (at a rate of 1 mL min−1) with an external solution containing (in millimolar): 145 NaCl, 1.2 MgCl2, 2 CaCl2, 10 glucose, and 10 HEPES, pH adjusted to 7.4 with NaOH. In some experiments, equimolar Ba2+ was used instead of Ca2+ as charge carrier. For perforated-patch experiments, electrodes were filled with internal solution containing the following (in millimolar): 135 CsGlutamate, 10 HEPES, 1 MgCl2, and 9 NaCl, pH 7.2 adjusted with CsOH. For this configuration, an amphotericin B stock solution was prepared every week at 50 mg mL−1 in dimethyl sulfoxide (DMSO), stored at −20°C, and protected from light. Fresh perforated-patch pipette solution was prepared every day by addition of 10 µL stock amphotericin B solution to 1 mL intracellular solution. This solution was sonicated thoroughly, protected from light, and kept in ice. Patch pipettes had their tips dipped in amphotericin-free solution for 2–10 s and backfilled with freshly mixed amphotericin-containing solution. For whole-cell experiments, cells were dialysed with an internal solution containing (in millimolar): 10 NaCl, 100 CsCl, 20 TEA-Cl, 14 EGTA, 20 HEPES, 5 MgATP, and 0.3 NaGTP, pH 7.2 adjusted with CsOH. When necessary, EGTA was omitted from the internal solution. For PPC, recordings started when the access resistance decreased below 25 MΩ, which usually happened within 10 min after sealing [61]. For WCC, the total pipette access resistance ranged from 2 to 3 MΩ.

Electrophysiological data were carried out using an EPC-9 amplifier under the control of Pulse software (HEKA Elektronik). Cells were held at −80 mV and I Ca were generated by depolarizing voltage steps of 50-ms duration to sequentially increasing test potentials by 10-mV step increments from −50 to +50 mV. Test pulses were delivered at 10-s intervals to minimize the rundown of Ca2+ currents [16]. Cells with pronounced rundown were discarded. External solutions were exchanged by a fast superfusion device consisting of a modified multibarrelled pipette, the common outlet of which was positioned 50–100 µm from the cell. Control and test solutions were changed using miniature solenoid valves operated manually.

To determine the effects of L-type Ca2+ channel agonists under WCC or PPC, cells were perfused with external solutions containing either FPL (in most experiments) or BayK8644 (BayK; in a few experiments). All experiments were performed at room temperature (24 ± 2°C).

Materials and solutions

The following materials were used: collagenase type I from Sigma (Madrid, Spain); DMEM, bovine serum albumin fraction V, fetal calf serum, and antibiotics were from Gibco (Madrid, Spain). Amphotericin B and BayK were from Sigma (Madrid, Spain) and FPL was from RBI (Natick, MA, USA); ω-conotoxin MVIIC (MVIIC) was from Bachem Feinchemikalien (Essex, UK).

FPL at 1 mM was prepared in ethanol and BayK at 1 mM in DMSO; both solutions were kept at −20°C in aliquots and were protected from light. MVIIC was dissolved in distilled water and stored frozen in aliquots at 0.1 mM. Final concentrations of drugs were obtained by diluting the stock solution directly into the external solution.

Data analysis

Under the appropriate conditions, FPL gradually augmented I Ca along the 50-ms test depolarizing pulse (Fig. 1b). Therefore, the amplitude of I Ca was always measured at the end of the test pulse (see double arrowheads in Fig. 1a). In order to measure the component of the tail current that occurred in the presence of L-type Ca2+ channel agonists, a second cursor was approximately placed 5 ms after the end of the test pulse. I Ca was not leak subtracted, and only cells with a leak current <30 pA were included in the analysis. Comparisons between means of group data were performed by one-way analysis of variance followed by Duncan post hoc test when appropriate. A p value equal to or smaller than 0.05 was taken as the limit of significance. Data are expressed as means ± standard error (SE).

Fig. 1
figure 1

FPL (1 μM) augmented I Ca under WCC recording (a–c) and inhibited I Ca under PPC recording (d–f). HP −80 mV; 2 mM external Ca2+. a, d Example trace currents generated by test pulses to the voltages indicated at the protocol on top, before (control) and at 2 min of FPL treatment. b, e I–V relationship of I Ca, before (control) and during FPL treatment. These curves were obtained by measuring I Ca amplitude at the end of the test pulse (see double arrowhead in a); currents were generated by successive test potentials given at increasing 10-mV steps from −50 to +50 mV. c, f Pooled results of I Ca amplitude (measured as indicated by the double arrowhead in a) and I tail (measured at 5 ms after returning to HP −80 mV). Data are means ± SE of eight cells for each configuration, from at least four different cultures. ***p < 0.001

Full size image

Results

FPL augmented I Ca under WCC but caused its inhibition under PPC

The fact that FPL increases 45Ca2+ uptake into K+ depolarized bovine chromaffin cells [70] suggested that the compound should also augment I Ca in these cells. This 45Ca2+ increase was maximal at 1 µM FPL; therefore, we used this concentration throughout this study. Generally, when a good seal was established in each experiment and access to the intracellular milieu was achieved, the targeted cell was voltage-clamped at −80 mV holding potential (HP). Then the cell was continuously perfused with an extracellular solution containing the physiological Ca2+ concentration of 2 mM. Unless otherwise indicated, under WCC recording, the pipette solution contained 14 mM EGTA; these are the conditions used in our and many other laboratories to prevent Ca2+ channel inactivation and optimize the recording of reproducible I Ca current traces along the experiment. Under PPC recording, there was no point in adding EGTA to the pipette solution since this compound will not cross the perforated-patch pores; these are also the conditions thoroughly used to record I Ca under PPC. In general, cells with leak currents greater than 30 pA were discarded (under WCC recording). Cells were usually challenged with test potentials given at different voltages at 10-s intervals.

In the example cell of Fig. 1a, a 50-ms pulse to −10 mV generated an inward current composed of (1) an initial fast-inactivating Na+ current (we did not use tetrodotoxin in this study; however, the presence of 1 mM TTX fully suppressed the I Na component of current traces); this current was large and, therefore, it appears truncated to facilitate the visualization of I Ca; (2) a low-inactivating I Ca; and (3) a deactivating tail current (I tail) caused by Ca2+ channel closing upon returning to −80 mV (see protocol on top of Fig. 1a). After 1-min of perfusion with FPL, the current increased along the depolarizing pulse and was followed by an I tail with a pronounced slowing down of current relaxation to zero baseline. These are typical effects of FPL on L channels [38, 42]. The characteristic effect of FPL on I Ca suggested the convenience of expressing all quantitative data by measuring I Ca at the end of the test pulse, as indicated by double arrowheads in the current traces of Fig. 1a. In this manner, when using the term I Ca throughout this study, we mean the amplitude of I Ca at the end of the test pulse. I tail is the amplitude of the tail current measured 5 ms after the end of the test pulse in each cell. Figure 1b shows the typical shift to the left of the current–voltage (I–V) curve caused by activators of L channels [35, 75]. FPL augmented I Ca only at steps in the negative range of the voltage scale. Figure 1c contains a summary of the averaged changes elicited by FPL: I Ca augmented by 76% and I tail by 5.8-fold.

A quite different picture emerged when I Ca was recorded under PPC. I Ca was halved when the example cell shown in Fig. 1d was treated with FPL; a tiny slowing down of I tail was noted. FPL did not shift the I–V curve and caused its depression with respect to the control curve (Fig. 1e). The averaged data of Fig. 1f indicate that although FPL decreased I Ca by 38%, I tail was still augmenting 2.2-fold. This suggested that, in spite of I Ca inhibition, FPL was still enhancing the open time of L channels.

In conclusion, under WCC recording, FPL augmented I Ca and I tail in the expected direction for an L Ca2+ channel activator. In contrast, under PPC recording, FPL blocked I Ca and caused a lesser I tail increase.

BayK mimicked the effects of FPL on I Ca

BayK is the prototype L channel ligand that has been widely used since its discovery [65] as a tool to prolong the mean open time of such channel [57]. BayK drastically augments 45Ca2+ uptake into K+ depolarized bovine chromaffin cells [24]. Under these premises, we thought that BayK (a DHP derivative) should mimic the effects of FPL (a non-DHP derivative) on I Ca and I tail. We, therefore, performed experiments with designs similar to those used with FPL. As for FPL [70], we used 1 μM BayK, a concentration that maximally augments Ca2+ entry through Ca2+ channels in bovine chromaffin cells [24].

Under WCC, BayK augmented I Ca and I tail in the expected direction (Fig. 2a). The compound also shifted to the left the I–V curve by about 10 mV. Thus, BayK enhanced I Ca at steps in the negative range of the voltage scale and depressed it in the positive range (Fig. 2b). This was similar to FPL that augmented I Ca at negative voltage steps and mildly depressed it at positive steps (Fig. 1b). BayK augmented peak I Ca by 61% and I tail by 3.3-fold (Fig. 2c).

Fig. 2
figure 2

BayK (1 μM) augmented I Ca under WCC recording (a–c) and inhibited I Ca under PPC recording (d–f). HP −80 mV; 2 mM external Ca2+. a, d Example trace currents generated by test pulses to the voltages indicated at the protocol on top, before (control) and at 2 min of BayK treatment. b, e I–V relationship of I Ca, before (control) and during BayK treatment. These curves were obtained by measuring I Ca amplitude at the end of the test pulse; currents were generated by successive test potentials given at increasing 10-mV steps from −50 to +50 mV. c, f Pooled results of I Ca and I tail amplitudes. Data are means ± SE of eight cells for each configuration, from at least four different cultures. *p < 0.05, ***p < 0.001. NS nonsignificant

Full size image

Under PPC recording, BayK had effects opposite to those found under WCC (Fig. 1). For instance, in the example trace of Fig. 2d, BayK caused a mild inhibition of I Ca, and did not augment I tail. The I–V curve underwent a meager depression with no shift (Fig. 2e). The averaged results indicate a 24% blockade of I Ca and no effect on I tail (Fig. 2f).

In conclusion, from a qualitative point of view, BayK affected I Ca in a direction similar to FPL: augmentation under WCC recording and diminution under PPC. However, from a quantitative point of view, the effects of BayK were milder; therefore, we chose FPL to perform further experiments.

FPL enhanced I Ca under both WCC and PPC recordings in conditions of L current isolation

As indicated in the Introduction section, we depart from the hypothesis that Ca2+ entry through L channels causes the inhibition of N and PQ channels. In bovine chromaffin cells, L channel current accounts for only 20% of the whole-cell current; the rest is carried through N and PQ channels [2, 18, 40]. In these cells, 2 μM MVIIC suppresses the N and PQ components of the whole-cell inward current through Ca2+ channels [19]. Therefore, we recoursed to 2 μM MVIIC to block N and PQ currents, leaving the cell with its L current intact.

Under WCC recording, MVIIC inhibited I Ca by 43%, as exemplified in the original traces of Fig. 3a (inset i). When added on top of MVIIC, FPL augmented I Ca and slowed down the relaxation of I tail (inset ii) to near the control level. On the other hand, MVIIC depressed the I–V curve; when added with MVIIC, FPL shifted to the left the I–V curve and thus augmented I Ca at step potentials in the negative voltage scale (Fig. 3b). The averaged I Ca changes appear in Fig. 3c. MVIIC reduced I Ca by 58% in the presence of the toxin, FPL augmented I Ca to about 70% of the initial control current.

Fig. 3
figure 3

FPL (1 μM) augmented I Ca and I tail under conditions of WCC recording (a–c) and PPC recording (d–f) in cells treated with ω-conotoxin MVIIC (MVIIC, 2 μM) to suppress N and PQ Ca2+ channel currents. HP −80 mV; 2 mM external Ca2+. a, d Example trace currents generated by test pulses to the voltages indicated at the protocol on top, before (control), at 2 min of MVIIC treatment, and at 2 min of combined MVIIC + FPL treatment. b, e I–V relationship of I Ca, during MVIIC treatment and during MVIIC + FPL treatment. These curves were obtained as described in Fig. 1. c, f Pooled results of I Ca and I tail amplitudes. Data are means ± SE of six cells for each configuration, from at least three different cultures. *p < 0.05, **p < 0.01, ***p < 0.001

Full size image

Under PPC recording, MVIIC reduced I Ca by 67% in the example cell shown in Fig. 3d. When added on top of MVIIC, FPL augmented I Ca and delayed I tail relaxation (see inset ii in Fig. 3d). MVIIC depressed the I–V curve in a parallel manner. When added together with MVIIC, FPL augmented I Ca to about 70% of the initial control I–V curve (Fig. 3e). Averaged results are summarized in Fig. 3f; I Ca was inhibited 80% by MVIIC. Although this inhibited current was doubled by FPL, this increase was not significantly different.

Another way of isolating the L from N/PQ currents rests on the voltage-dependent inactivation of N/PQ channels. Shifting of the HP from −80 to −50 mV causes the gradual loss of 70–80% of current; the remaining current is blocked by nifedipine and is thus associated to noninactivating L channels [31, 70]. This was reproduced here under WCC recording. In the example cell shown in Fig. 4a, I Ca declined to about 30% of initial current upon switching the HP from −80 to −50 mV. FPL gradually augmented I Ca as well as I tail to reach a plateau in about a minute. Current traces from this time course curves (a, b, c) are shown in Fig. 4b; note the progressive augmentation of I Ca (i) and the drastic slowing down of I tail relaxation (ii). Pooled data (Fig. 4c) indicate a voltage-dependent current loss of 54% and its augmentation to about 60% of the initial I Ca upon FPL treatment. Particularly impressive was the eightfold increase of I tail.

Fig. 4
figure 4

FPL (1 μM) augmented I Ca and I tail under conditions of WCC recording (a–c) and PPC recording (d–f) in cells voltage-clamped at −50 mV in order to suppress N and PQ Ca2+ channel currents. HP was initially maintained at −80 mV; subsequently, HP was switched to −50 mV as indicated by the horizontal bars on top of a and d; 2 mM Ca2+ was present in the extracellular solution. a, d Time courses of I Ca and I tail. FPL was perfused during the time period indicated by the horizontal bars. Test pulses to −10 mV were applied at 10-s intervals. b, e Example I Ca (i) and I tail currents (ii) taken from experiments of a and d at the times indicated by a–c. c, f Pooled results on I Ca and I tail amplitudes calculated at the HP indicated and after 2 min perfusion with FPL. Data are means ± SE of six cells for each configuration, from at least three different cultures. *p < 0.05, **p < 0.01, ***p < 0.001

Full size image

Under PPC recording, the reduction of I Ca was also gradual and stabilized after a couple of minutes at −50 mV HP (Fig. 4d). FPL gradually augmented I Ca and I tail to reach a plateau after about 2 min. The example traces shown in Fig. 4e show the mild FPL-induced I Ca increase (i) and the more pronounced slowing down of I tail relaxation (ii). Pooled data (Fig. 4f) show a 60% decrease of I Ca at −50 mV HP and 37% increase elicited by FPL with fivefold augmentation of I tail.

In conclusion, under conditions of L current isolation, either with MVIIC or depolarized HP, FPL equally augmented I Ca and I tail under WCC and PPC recording conditions. This contrasts with the results obtained when I Ca flows through all channel types where FPL augmented I Ca under WCC recording but caused its inhibition under PPC. It seems, therefore, that this inhibition must be associated to N/PQ channels, rather than L channels.

The presence or absence of EGTA in the cytosol modifies the effects of FPL under WCC and PPC recordings

Under WCC recording, the intracellular solution contained 14 mM EGTA. Under these conditions, Ca2+ entering through Ca2+ channels will be chelated, the [Ca2+]c will not increase, and the Ca2+-dependent inhibition of Ca2+ channels will be prevented [32]. We, therefore, expected that an experiment performed in conditions similar to those of Fig. 1 under WCC recording, but using an intracellular solution deprived of EGTA, will favor the inhibition of N/PQ channel current.

The example I Ca traces of Fig. 5a show that FPL inhibits I Ca by 36% at a test potential of 0 mV and by 49% at +10 mV. There was a tiny slowing down of I tail, particularly at +10 mV. This inhibition was seen only at depolarizing test potentials, as shown by the I–V curves of Fig. 5b. Pooled data indicate that the initial I Ca of 350 pA was reduced 45% by FPL. This was a picture quite similar to that found under PPC recording with no EGTA in the cytosol (compare the I–V curves of Figs. 1e and 5b).

Fig. 5
figure 5

Effects of EGTA on the changes of I Ca elicited by FPL under WCC recording (a–c) and PPC recording (d–f) recording conditions. HP −80 mV; 2 mM external Ca2+. Under WCC recording, the pipette solution contained no EGTA; under PPC recording, before the experiment, cells were incubated during 45 min with cell permeable EGTA-AM at 37°C. Cells were subsequently washed during 15 min at room temperature with the extracellular solution. Then, the cells were patched and the experiments were performed as usual under PPC recording. a, d Example trace currents generated by test pulses to the voltages indicated at the protocol on top, before (control) and at 2 min of FPL treatment. b, e I–V relationship of I Ca, before (control) and during FPL treatment. c, f Pooled results on the effects of FPL on I Ca (measured at 0 mV test potential). Data are means ± SE of six cells for each configuration, from at least three different cultures. *p < 0.05, **p < 0.01

Full size image

Under PPC recording, EGTA could not be given in the pipette solution because it will not cross the pores of the perforated membrane patch. Hence, we recourse to the cell-permeable EGTA-AM to load the cell cytosol with EGTA before doing the experiment. Figure 5d shows example I Ca traces; a tiny augmentation of I Ca and I tail were seen upon FPL treatment both at −10 and 0 mV test potentials. FPL shifted the I–V curve to the left, indicating that FPL augmented I Ca only at hyperpolarizing potentials (Fig. 5e). Pooled data show that FPL augmented I Ca by 33% (Fig. 5f). These FPL effects were quite similar to those found under WCC recording with EGTA in the cytosol (compare the I–V curves of Figs. 1b and 5b).

In conclusion, under WCC, EGTA removal from the patch pipette turned FPL-induced I Ca augmentation into inhibition; conversely, under PPC, the presence of EGTA in the cytosol converted FPL-induced I Ca inhibition into potentiation. This supports the view that Ca2+-dependent inactivation by Ca2+ entry through L channels is responsible for the inhibition of the fraction of the current carried through non-L channels.

Effects of FPL on Ca2+ channel currents using Ba2+ (instead of Ca2+) as charge carrier, under WCC and PPC recordings

The Ca2+ inactivation of Ca2+ channels is prevented when using Ba2+ (instead of Ca2+) as charge carrier. This is true for neurons [27] as well as bovine chromaffin cells [7, 72]. By using 2 mM Ba2+ (instead of 2 mM Ca2+) as charge carrier, we expected to find similar effects of FPL on Ba2+ currents (I Ba), under both WCC and PPC recordings.

Figure 6a shows example I Ba traces obtained under WCC recording. FPL augmented I Ba by 90% and caused a pronounced delay of I tail relaxation. FPL shifted the entire I–V curve to the left, indicating I Ba augmentation at steps in the negative range of the voltage scale and inhibition in the positive range (Fig. 6b). Pooled results show that FPL augmented I Ba by 90% and I tail by as much as sixfold (Fig. 6c).

Fig. 6
figure 6

FPL (1 μM) augmented I Ca under both WCC (a–c) and PPC recording (d–f) if 2 mM Ba2+ (instead of 2 mM Ca2+) was used as charge carrier. HP −80 mV; 2 mM external Ba2+. a, d Example trace currents generated by test pulses to the voltages indicated at the protocol on top, before (control) and at 2 min of FPL treatment. b, e I–V relationship of I Ba, before (control) and during FPL treatment. These curves were made as indicated in Fig. 1. c, f Pooled results of I Ba and I tail amplitudes before and during FPL treatment. Data are means ± SE of six cells for each configuration, from at least three different cultures. **p < 0.01, ***p < 0.001

Full size image

Under PPC recording, the data came in a similar direction. For instance, the example traces of Fig. 6d show a progressive augmentation of I Ba along the depolarizing pulse and a pronounced slowing down of the tail current. However, the I–V curve was shifted to the left only in the negative range of the voltage scale where I Ba was augmented by FPL (Fig. 6e). Pooled results indicate that FPL enhanced I Ba by 80% and I tail by as much as sevenfold (Fig. 6f).

In conclusion, FPL modified I Ba (peak and tail currents) in a similar way, irrespective of whether the recordings were made under WCC or PPC.

Discussion

The central finding of this study was that FPL, an enhancer of Ca2+ entry through L-type Ca2+ channels, caused a paradoxical inhibition of the whole-cell inward current flowing through Ca2+ channels during depolarization of voltage-clamped bovine chromaffin cells. Such inhibition was initially observed only under PPC recording since, under WCC recording, FPL augmented I Ca and I tail in the expected direction. We later on found that these opposing FPL actions could be explained by the absence or the presence of the mobile Ca2+ buffer EGTA in the intracellular solution used under PPC or WWC recordings, respectively.

The initial observation on FPL-induced I Ca blockade was done under PPC recording (Fig. 1d). Thus, the possibility existed that, under this patch-clamp configuration, FPL caused a direct pharmacological inhibitory effect on some or all of the major Ca2+ channels expressed by bovine chromaffin cells, i.e., L, N, and PQ [23]. In fact, at 10 μM, FPL inhibited the majority of the whole-cell current in HEK cells expressing N channels; however, at 1 μM, the concentration used in this study, FPL did not block the N current [38].

ICa inhibition by FPL may rather be explained in a different context. For instance, under PPC, when FPL inhibited ICa amplitude, Itail remained enhanced (Fig. 1d, f). This indicates that FPL was indeed augmenting ICa through L channels. However, these channels carry only about 15–20% of the whole-cell ICa in bovine chromaffin cells, the rest being carried by N and PQ channels [2, 3, 17, 40]. Therefore, the inhibition by FPL of N/PQ current components could mask its enhancing effects on the L current component. This was demonstrated to be the case in the experiment where N/PQ currents were blocked by MVIIC (Fig. 3) or by maintaining the HP at −50 mV (Fig. 4). Under these conditions of L current isolation, FPL always augmented ICa and Itail regardless of the recording conditions, WCC or PPC (Figs. 3 and 4).

Various experimental data support the view that FPL was not directly inhibiting the N/PQ currents; rather, such inhibition was indirectly associated to enhanced Ca2+ entry through L channels that were kept open longer in the presence of FPL, as the following facts suggest: (1) under PPC recording, the presence of cytosolic EGTA turned the I Ca inhibition into an I Ca augmentation (Fig. 5); (2) conversely, under WCC recording, EGTA removal from the intracellular solution turned the augmented I Ca into a depressed I Ca (Fig. 5); (3) when using Ba2+ as charge carrier, which does not inactivate Ca2+ channels [7, 27], FPL augmented I Ba under both WCC and PPC recordings. Thus, FPL-induced I Ca inhibition did not depend on the patch-clamp configuration used but on whether the experimental conditions allowed an elevation of [Ca2+]c during the application of depolarizing pulses. The possibility existed that the effects of FPL were linked to the washout of a cytosolic factor that is the target of the Ca2+; such factor could be removed under WCC recording. We did experiments in six cells, recording under WCC from the beginning of seal rupture. I Ca amplitude was gradually augmented from about 40 pA at the moment of seal rupture to about 140 pA in 1 min thereafter. FPL caused a pronounced I Ca augmentation from the beginning of seal rupture but only in the presence of EGTA. This indicates that the washout of a cytosolic factor is not involved in the FPL effects of I Ca.

Now the question arises as to whether the amount of Ca2+ entering the cell during brief depolarizing pulses (50 ms in our present study) is sufficient to build up a local [Ca2+]c transient to inactivate N/PQ Ca2+ channels. The averaged [Ca2+]c peaks measured with Ca2+ probes do not detect the high [Ca2+]c transients occurring at subplasmalemmal sites during cell depolarization [13, 54, 67]. Using pulsed laser imaging, Monck and workers [45] visualized hot spots of submembranous Ca2+ evoked by 50 ms depolarizing pulses, the same stimulation pattern used in the present study. These hot spots developed in 20–50 ms and extended laterally by several micrometers. Partial or complete submembranous rings of elevated [Ca2+] appeared at 50–100 ms after channel opening, and the Ca2+ gradient collapsed at 200–400 ms after the pulse began. This was interpreted as due to a limited Ca2+ diffusion because of rapid Ca2+ sequestration by immobile buffers [45, 76].

In bovine chromaffin cells, the cytosol has a Ca2+ binding capacity of 4 mmol/L cells. The endogenous Ca2+ buffer is scarcely mobile, has a low affinity for Ca2+ (k d ~ 100 µM) and an activity coefficient of ~1/40 [74, 76]. The two-dimensional diffusion coefficient is ~40 µm2/s and shows inhomogeneities at the plasma membrane [51]. Brief openings of VDCCs generate [Ca2+]c microdomains as high as 10 or even 100 µM [8, 55]. Using mitochondria as biosensors for [Ca2+]c changes, we found that ACh or K+ pulses caused [Ca2+]c elevations of 20–40 µM at subplasmalemmal sites of bovine chromaffin cells [47]. Because of rapid diffusion of Ca2+ towards the surrounding cytosol, these Ca2+ microdomains are very much restricted in time and space [53, 54]. The presence of mobile Ca2+ buffers accelerates diffusion and oppose the development of high [Ca2+]c microdomains during cell depolarization [4, 5, 56, 64]. For instance, at 50 µM, a high-affinity low-capacity Ca2+ buffer such as fura-2 increases the apparent rate of Ca2+ diffusion by as much as four times [76].

In our present experiments, we dialyzed the cytosol with 14 mM EGTA under WCC recording or we incubated the cells with EGTA-AM for PPC experiments. Being a strong buffer [43], EGTA will prevent the buildup of high Ca2+ microdomains at subplasmalemmal sites during cell stimulation with 50 ms depolarizing pulses, as happened to be the case with fura-2 [76]. Under these conditions, it is understandable that Ca2+ induced N/PQ channel inhibition was not taking place and that the dominant FPL effects were I Ca and I tail increase and slowing down of I tail due to L channel activation.

As discussed above, the bovine chromaffin cell cytosol contains strong Ca2+ buffers that severely restrict the diffusion of the Ca2+ entering through Ca2+ channels. Therefore, we must admit that, to cause inactivation of N/PQ channels, the Ca2+ has to enter the cell through L channels located close by. This effect might occur under physiological stimulation of chromaffin cells. However, it could better be unmasked by FPL-induced enhanced Ca2+ entry through L channels. From a functional point of view, this possible colocalization of L and non-L channels can be further investigated using FPL as a pharmacological tool to selectively augment Ca2+ entry through L channels. A challenging question to explore is whether L, N, and PQ channels form functional clusters units to regulate diverse exocytotic responses tailored to specific needs. In this context, it is interesting to note that some authors have found hot spots of Ca2+ entry and release in bovine chromaffin cells [62]. Furthermore, active exocytotic zones have been found in bovine chromaffin cells [66, 73] and a functional polarization of secretory sites have been revealed using confocal fluorescence microscopy, also in bovine chromaffin cells [12]. However, this clustering of Ca2+ channel subtypes at specific plasmalemmal sites makes it difficult to understand how the Ca2+ entering through a given channel type may have a selective function distinct from the Ca2+ entering through other channel type. Also, the question emerges on why Ca2+ entering through N/PQ channels (about 80% of the I Ca) is not eliciting the inactivation of such channels. Unfortunately, compounds that prevent the inactivation of N/PQ channels (as the case is for L channels with FPL and BayK) are not available to help answer these questions. Further insight into the possible colocalization of Ca2+ channel subtypes could be achieved through the judicious use of various concentrations of EGTA (slow Ca2+ chelator) and BAPTA (rapid Ca2+ chelator). This approach has been useful to study Ca2+ channel colocalization with the secretory machinery (reviewed in García et al. [23]). However, the extrapolation of this strategy to demonstrate colocalization of Ca2+ channel subtypes may be difficult in the context of the present experiments. Immunofluorescence experiments with specific antibodies for α1 subunits of L, N, and P/Q channels could also provide some information on the possible colocalization of these channels. An alternative explanation may be linked to the different molecular mechanisms underlying the regulation of Ca2+ channel subtypes by the Ca2+-binding protein calmodulin and/or intracellular messengers [11, 15, 37, 49]. It may be that the lower Ca2+ affinity of calmodulin lobe or another lower affinity Ca2+-binding protein experiences activating concentrations in the microdomains around isolated L channels and then dissociates and travels to affect N/PQ channels.

Finally, the possible specialization of the various Ca2+ channel subtypes expressed within the same cell to perform specific functions deserves a comment. Since the discovery that bovine chromaffin cells expressed multiple Ca2+ channels, efforts have been made to find out a kind of specialization for each channel subtype to drive a specific cell function. For instance, concerning exocytosis, L channels [6, 33], PQ channels [36], or all channels [41, 69] drive the secretory process. This disparity may be due to different stimuli and experimental conditions used [23]. The possible specialization of Ca2+ channel subtypes in bovine chromaffin cells has been explored in functional aspects other than exocytosis. For instance, Ca2+ entry through L channels, but not through N and PQ channels, causes mitochondrial disruption and cell death by apoptosis [10]; also, Ca2+ entry through L channels are preferentially coupled to endocytosis [60, 63]. However, other studies do not attribute a special function to a given channel subtype. This is the case for tyrosine hydroxylase activation [58], SNAP-25 expression [21], ERK phosphorylation [48], or mitochondrial Ca2+ uptake [46]; these studies concluded that these cell functions depended more on critical [Ca2+]c levels than on Ca2+ entry through a given Ca2+ channel subtype. The results of our present study, however, clearly indicate that the Ca2+ entering through L channels have a clear-cut function, the modulation of N and PQ channel activities during cell depolarization. To our knowledge, this is the first report suggesting this function for the L-type voltage-dependent Ca2+ channel.

To conclude, the L-type Ca2+ channel activator FPL caused opposing effects on I Ca, i.e., augmentation under WCC recording and inhibition under PPC recording. The experiments done in this study to clarify these apparent paradoxical effects suggest that Ca2+ entry through L channels causes a Ca2+-dependent inhibition of N and PQ channels. FPL seems to be a good tool to further explore functional aspects of this novel regulatory action of Ca2+ flux through L-type channels.