Voltage-independent Inhibition of P/Q-type Ca 2 (cid:1) Channels in Adrenal Chromaffin Cells via a Neuronal Ca 2 (cid:1) Sensor-1-dependent Pathway Involves Src Family Tyrosine Kinase*

In common with many neurons, adrenal chromaffin cells possess distinct voltage-dependent and voltage-in-dependent pathways for Ca 2 (cid:1) channel regulation. In this study, the voltage-independent pathway was revealed by addition of naloxone and suramin to remove tonic blockade of Ca 2 (cid:1) currents via opioid and purinergic receptors due to autocrine feedback inhibition. This pathway requires the Ca 2 (cid:1) -binding protein neuronal calcium sensor-1 (NCS-1). The voltage-dependent pathway was pertussis toxin-sensitive, whereas the voltage-independent pathway was largely pertussis toxin-insen-sitive. Characterization of the voltage-independent inhibition of Ca 2 (cid:1) currents revealed that it did not involve protein kinase C-dependent signaling pathways but did require the activity of a Src family tyrosine kinase. Two structurally distinct Src kinase inhibitors, 4-amino-5-(4-methylphenyl)7-( t -butyl)pyrazolo[3,4- d ] pyrimidine (PP1) and a Src inhibitory peptide, increased the Ca 2 (cid:1) currents, and no further increase in Ca 2 (cid:1) currents was elicited by addition of naloxone and suramin. In addition, the Src-like kinase appeared to act in the same pathway as NCS-1. In contrast, addition of PP1 did not prevent a voltage-dependent facilitation elicited

involving G-protein-coupled receptors (GPCRs) 1 (1)(2)(3). One of these, the voltage-dependent pathway, in which Ca 2ϩ channel inhibition is reversed by strong depolarization, has been characterized in detail (4,5). In addition, many different but poorly characterized voltage-independent pathways for inhibition of N-and P/Q-type Ca 2ϩ channels have been studied in various neuronal cell types (1)(2)(3). These pathways regulate steadystate current magnitude but not the activation kinetics of the channels. In some cases, multiple pathways co-exist in the same neurons (1). The second messenger pathways involved have only been partially characterized. One pathway in chick dorsal root ganglion neurons (6,7) and mouse cerebellar granule cells (8) involves activation of protein kinase C (PKC). In a second pathway, in rat sympathetic neurons, muscarinic receptor inhibition of Ca 2ϩ channels (9 -11) requires a low level of intracellular Ca 2ϩ (12) and an unidentified diffusible messenger (13). In a third pathway, in chick dorsal root ganglion neurons, a voltage-independent regulation exists that requires the activity of an Src family tyrosine kinase (14). Furthermore, the ␣ 1B subunit of the N-type channel in these cells has been shown to become tyrosine-phosphorylated in response to ␥-aminobutyric acid (15). This Src kinase-mediated pathway has not been demonstrated to exist in any other cell type, or for regulation of other Ca 2ϩ channel types, and so the extent of its distribution and general significance is unclear.
We have been examining the physiological roles of neuronal calcium sensor-1 (NCS-1) protein, a member of a family of neuronally expressed EF hand Ca 2ϩ -binding proteins (16). NCS-1 and its Drosophila orthologue, frequenin, have been shown to regulate Ca 2ϩ -dependent exocytosis in neurons (17) and neuroendocrine cells (18,19). By using a dominant-negative mutant of NCS-1 (NCS-1(E120Q)), we have demonstrated a role for NCS-1 in an autocrine pathway for Ca 2ϩ channel inhibition in adrenal chromaffin cells (20). Under our recording conditions, autocrine inhibition of non-L-type Ca 2ϩ currents, due to activation of ATP and/or opioid receptors, occurred mostly through a voltage-independent pathway, and NCS-1 appeared to be required for this pathway, but expression of NCS-1(E120Q) did not interfere with prepulse facilitation. The existence of a voltage-independent pathway for Ca 2ϩ channel regulation in chromaffin cells was also shown recently in another study (21), examining Ca 2ϩ channel regulation by histamine.
We determined the relationship between the NCS-1-dependent pathway for voltage-independent inhibition of non-L-type Ca 2ϩ channels in chromaffin cells and those previously described in neurons for the regulation of N-and P/Q-type Ca 2ϩ channels. Our results indicate a requirement for Src family tyrosine kinase activity for this pathway acting on P/Q-type Ca 2ϩ channels and that the ␣ 1A subunit of these channels is a substrate for Src family kinases. This is the first demonstration of a regulatory pathway for P/Q-type Ca 2ϩ channel inhibition involving Src family kinase activity.
Chromaffin Cell Culture and Transfection-Freshly isolated bovine adrenal chromaffin cells from adult cows (22) were plated directly onto collagen-coated 3.5-cm plastic Petri dishes for untransfected cell experiments or on non-tissue culture-treated 10-cm Petri dishes at a density of 1 ϫ 10 6 /ml for cells to be transfected the next day. The following day, cells in the non-tissue culture plates were harvested by centrifugation and resuspended in growth medium at a density of 1 ϫ 10 7 /ml, and 20 g of pEGFP (CLONTECH, Basingstoke, UK) in combination with 20 g of either pDNA3.1 (control cells) or plasmid encoding NCS-1(E120Q) was added per ml of cells. Cells and plasmids were electroporated at 250 V and 975 microfarads for one pulse, using a Bio-Rad Gene Pulser II and 4-mm cuvettes. Cells were diluted as rapidly as possible to 1 ϫ 10 6 /ml with fresh growth media, and 1 ϫ 10 6 cells were grown on 3.5-cm collagen-coated Petri dishes in a final volume of 3 ml of growth medium for a further 3-5 days. We have previously verified that GFP and NCS-1(E120Q) are efficiently co-expressed following transfection under these conditions (20).
Whole Cell Voltage Clamp Recordings-Adrenal chromaffin cells grown on 35-mm plastic dishes were washed and maintained in external bath solution containing 143 mM tetraethylammonium chloride, 10 mM BaCl 2 , 1 mM MgCl 2 , and 10 mM HEPES, pH 7.3. Where specifically indicated, some recordings were carried out in a similar buffer but containing 10 mM CaCl 2 instead of 10 mM BaCl 2 with or without the addition of 10 M nicotine. The intracellular pipette solution contained 150 mM CsCl, 2 mM Mg-ATP, 2 mM 1,2-bis(2-aminophenoxy)ethane-N,N,NЈ,NЈ-tetraacetic acid, 0.5 mM GTP, and 10 mM HEPES, pH 7.3 (23). The pipettes used had a resistance of 2.0 -4.0 M⍀ in the bath solution. Whole cell recording was carried out using an EPC-9 patch clamp amplifier (HEKA Elektronik, Lambrecht, Germany). The cells were held at Ϫ80 mV after correction for a liquid junction potential of Ϫ8 mV and depolarized under the control of PULSE software (HEKA Electronics). Capacitance and series resistance were compensated using the EPC-9 correction routines. The average series resistance and cell capacitance for cells used in this study were 8.69 Ϯ 0.42 M⍀ (in all cases Ͻ15 M⍀) and 18.9 Ϯ 0.82 picofarads, respectively (n ϭ 74). The current output was filtered at 2.9 kHz with a 4-pole Bessel filter, and data were acquired and stored at 20-s intervals. Linear leak and capacity currents were subtracted with a P/4 procedure using PULSE software, and traces are the averages of three sweeps. All experiments were carried out at room temperature (22-24°C) without superfusion during recording. All reagent applications were by direct addition to the bath buffer or pipette as indicated in the text. Ca 2ϩ channel currents were elicited by a step depolarization from the holding potential to 0 mV for 50 ms. To examine pre-pulse facilitation, a protocol was used in which an initial pre-pulse depolarization step to ϩ90 mV for 50 ms was applied. After return to the holding potential of Ϫ80 mV for 15 ms, a test pulse to 0 mV for 50 ms was applied. Data are shown as mean Ϯ S.E., and statistical differences were assessed using an unpaired Student's t test.
Immunoprecipitation of ␣ 1A Subunit of P/Q-type Ca 2ϩ Channels and in Vitro Phosphorylation-Isolated chromaffin cells in culture were serum-starved for 24 h. The cells (10 7 cells per condition) were washed in Krebs buffer and then lysed and subjected to immunoprecipitation as described previously (24) using anti-P/Q (␣ 1A subunit of P/Q-type Ca 2ϩ channels) antibodies (25) purchased from Sigma or anti-phosphotyrosine antibodies (PY20, BD Transduction Laboratories). The immunoprecipitates were separated by 8% SDS-polyacrylamide gel electrophoresis, transferred to nitrocellulose, and immunoblotted with the specified antibodies. The in vitro phosphorylation assays were performed on immunoprecipitates that were prepared as described above. All phosphorylation reactions were incubated at 30°C for 1 h in Src kinase reaction buffer (100 mM Tris-HCl, pH 7.2, 125 mM MgCl 2 , 25 mM MnCl 2 , 2 mM EGTA, 0.25 mM sodium orthovanadate, 2 mM dithiothreitol, Upstate Biotechnology, Inc.) with the addition of 10 Ci of [␥-32 P]ATP, 3000 Ci/mmol, and 100 M unlabeled ATP, in a total reaction volume of 40 l with occasional gently mixing. The reactions were terminated by placing on ice and then washed twice with Src kinase reaction buffer plus 1% Triton X-100. The phosphorylated anti-P/Q immunoprecipitates were resolved by 8% SDS-polyacrylamide gel electrophoresis, and the gel was exposed to a phosphorscreen for 1 week. The phosphorscreen was imaged with a Molecular Dynamics Phosphor-Imager SI (Molecular Dynamics).

Voltage-dependent Inhibition of Ca 2ϩ Channels in Chromaffin Cells Is Pertussis
Toxin-insensitive-In this study, we generally recorded from adrenal chromaffin cells using a bath solution containing 10 mM Ba 2ϩ and without perfusion of the cells. Under these conditions, Ba 2ϩ stimulates exocytosis from the chromaffin cells, and the release of endogenous secretory products leads to an autocrine inhibition of Ca 2ϩ channel currents (26 -28). The inhibition can be fully relieved by application of both 100 M suramin and 10 M naloxone to block purinergic and opioid receptors, respectively (20). In these experiments, therefore, we examined the relief of tonic inhibition of Ca 2ϩ currents (facilitation). Test pulses to 0 mV were used as these gave maximal activation of Ca 2ϩ currents (20). Addition of naloxone and suramin to the bath (Fig. 1, A and B) resulted in an increase in steady-state Ca 2ϩ channel current with no change in activation kinetics as we have described previously (20). The facilitation due to naloxone and suramin developed slowly after addition of the antagonists to the bath reaching a stable peak after 3-5 min. Following this facilitation, due to maximal doses of naloxone and suramin, an additional increase in Ca 2ϩ currents could still be elicited by a 50-ms pre-pulse depolarization to ϩ90 mV which, in contrast to the effects of Voltage-dependent (prepulse) facilitation in chromaffin cells is pertussis toxin (PTX)-sensitive (26,29,30). We examined whether the voltage-independent inhibition mediated by purinergic and opioid receptors was PTX-sensitive by comparison with effects of PTX treatment on prepulse facilitation. Treatment for 24 h with 100 ng/ml PTX had no statistically significant effect on the magnitude of the Ca 2ϩ channel currents (462.3 Ϯ 97.9 pA for control and 537 Ϯ 68.8 pA for PTX-treated cells). In addition, PTX treatment had no effect on the facilitation due to naloxone and suramin addition but largely abolished the prepulse facilitation in the same cells (Fig. 2, A and  B). From a series of cells, the facilitation due to naloxone and suramin was not statistically significantly affected by PTX treatment (Fig. 2C), whereas prepulse facilitation was significantly reduced by around 80% (Fig. 2D). The effect of PTX treatment on prepulse facilitation could not be explained by an inhibition of secretion of autocrine factors as PTX treatment leads instead to an enhancement of evoked exocytosis in chromaffin cells (31,32). These data are consistent with studies on other systems where voltage-independent regulation of Ca 2ϩ channels is usually PTX-insensitive (1). In addition, they provide further evidence for the distinct nature of the pathways of voltage-dependent and voltage-independent inhibition of Ca 2ϩ currents in chromaffin cells under our recording conditions.
To confirm that the inhibition of Ca 2ϩ channels seen in the presence of Ba 2ϩ could also be observed under more physiological conditions with other secretagogues, we carried out a series of experiments using Ca 2ϩ rather than Ba 2ϩ as the charge carrier in the bath buffer. Secretion, in these experiments, was triggered by activation of nicotinic cholinergic receptors. In the presence of 10 M nicotine (4 of 4 cells), addition of naloxone and suramin again resulted in a facilitation of Ca 2ϩ currents ( Fig. 2E) showing that tonic inhibition of Ca 2ϩ channel currents also occurred under these conditions. In some cells (4/5), a smaller facilitation was also observed in the presence of external Ca 2ϩ in the absence of secretagogue indicating that accumulation of secretory products due to basal secretion can be sufficient to give some autocrine inhibition of Ca 2ϩ channels under these conditions. As expected, addition of ATP or metenkephalin did not produce any further facilitation in the presence of secretagogues. Channel facilitation was not entirely dependent on the use of suramin as in 3 of 3 cells facilitation was detected following addition of only naloxone (Fig. 2F). Nevertheless, to ensure complete relief of autocrine inhibition, a mixture of both opioid and purinergic antagonists was used in later experiments.
The Voltage-independent Pathway Is Not Mediated by PKC Activation-We have shown previously (20) that transient expression of the EF hand mutant of NCS-1, NCS-1(E120Q), increases the Ca 2ϩ currents in chromaffin cells and in such cells naloxone and suramin have no further stimulatory effect suggesting that the voltage-independent pathway is NCS-1-dependent. The yeast orthologue of NCS-1 (33) is an activator of phosphatidylinositol 4-kinase, an important enzyme in the pathway leading to synthesis of phosphatidylinositol 4,5bisphosphate, raising the possibility that expression of the NCS-1 mutant may have disrupted phosphatidylinositol 4,5bisphosphate biosynthesis and thereby receptor signaling to PKC. To test this possibility, the effect of 100 nM PMA on cells transfected with NCS-1(E120Q) was examined to see if direct PKC activation would inhibit the Ca 2ϩ currents. PMA had little or no effect on Ca 2ϩ currents (Fig. 3A) in these cells (5 of 5 cells). The converse experiment of examining whether inhibition of PKC would relieve the autocrine inhibition in nontransfected cells revealed that addition of the PKC inhibitor bisindolylmaleimide had little effect on Ca 2ϩ currents (Fig.  3B). We also examined the effect of PMA treatment after relief of autocrine inhibition by naloxone and suramin addition (Fig.  3C). These experiments were carried out in the presence of 3 M nifedipine to rule out any confounding effects of PMA on L-type Ca 2ϩ currents. Under these conditions we did not observe any inhibition of Ca 2ϩ current following PMA treatment (5 of 5 cells). These data suggest that neither the effect of NCS-1(E120Q) expression nor the naloxone/suramin voltageindependent facilitation involves inactivation of a PKCdependent pathway.
Inhibition of Src Family Tyrosine Kinase Facilitates Ca 2ϩ Currents in Chromaffin Cells via a Voltage-independent Pathway-Voltage-independent inhibition by ␥-aminobutyric acid of N-type Ca 2ϩ channels in chick sensory neurons requires the activity of a Src family tyrosine kinase (14). The involvement of tyrosine kinase activity for GPCR-mediated inhibition of Ca 2ϩ channels has not been demonstrated so far in any other cell types. We examined whether such a requirement exists in adrenal chromaffin cells. Addition to the bath of PP1 (5 M), an inhibitor of Src family kinases (34), produced a marked facilitation of the Ca 2ϩ current (Fig. 4A) that developed slowly over 3-5 min. The facilitation due to PP1 varied in magnitude between batches of cells but was observed in all cells examined (27/27) with a mean increase in current amplitude of 343.9 Ϯ 64.6 pA. The effect of PP1 was due to facilitation of non-L-type Ca 2ϩ channels as the stimulation by PP1 was preserved and occurred to a similar extent (Fig. 4B) in the presence of the L-type Ca 2ϩ channel blocker nifedipine (3 M). Addition of PP1 also increased Ca 2ϩ current inhibited due to the presence of the secretagogue 10 M nicotine in the presence of external Ca 2ϩ rather than Ba 2ϩ (data not shown).
To examine the relationship between the voltage-dependent pathway, the voltage-independent pathway, and the facilitation due to Src kinase inhibition, two experiments were carried out. In the first (Fig. 4C), two populations of cells were examined. Control cells were treated with naloxone and suramin to determine the extent of facilitation by the voltage-independent pathway. In parallel, a separate population of cells was first treated with PP1 to inhibit Src family kinases. Facilitation was achieved in all 4 cells, and subsequently, naloxone and suramin were added. The prior addition of PP1 significantly reduced the facilitation due to naloxone and suramin by around 75%. In the second experiment (Fig. 4D), the extent of prepulse facilitation in each cell was first determined, and then PP1 was added leading to a prolonged facilitation due to the inhibitor. Following the facilitation by PP1, which occurred in all 9 cells, the pre-pulse protocol was repeated. Prior addition of PP1 did not lead to any significant change in the extent of prepulse facilitation. These data suggest an involvement of a Src family activity kinase in the voltage-independent inhibition of Ca 2ϩ currents in chromaffin cells but not in the voltage-dependent pathway.
The Inhibitory Pathway Involving Src Family Kinase Activity Is NCS-1-dependent-We have demonstrated previously (20) that the voltage-independent but not the voltage-dependent inhibitory pathway in chromaffin cells is disrupted by expression of NCS-1(E120Q), a dominant-negative mutant of NCS-1. The preceding results, suggesting a role for a Src kinase in the voltage-independent pathway, predict that the facilitation due to inhibition of Src kinase activity should be disrupted by NCS-1(E120Q) expression. To test this prediction, chromaffin cells were co-transfected with plasmids encoding GFP and NCS-1(E120Q), and the effect of PP1 was examined on transfected and untransfected cells in parallel. In these experiments, PP1 again produced a marked increase in Ca 2ϩ currents in all 7 untransfected cells examined (Fig. 5A). In contrast, in NCS-1(E120Q)-transfected cells, the PP1 facilitation was significantly reduced by around 75% compared with the control untransfected cells (Fig. 5, B and C). These data are consistent with NCS-1 acting on the same voltage-independent pathway as the Src family tyrosine kinase.
Ca 2ϩ Currents Are Facilitated by a Peptide Inhibitor of Tyrosine Kinase-To confirm the involvement of an Src-like tyrosine kinase, a structurally unrelated inhibitor was used that was applied to the cells by an alternative route in the pipette buffer. The peptide derived from amino acids 137-157 of the regulatory domain of pp60 v-Src (SrcKI) is a well characterized and specific inhibitor of Src family kinases (35,36). In these experiments, Ca 2ϩ currents were monitored every minute after the establishment of the whole cell recording. When the inhibitory peptide (SrcKI) was included in the pipette buffer at a concentration of 70 M, Ca 2ϩ current increased over time during whole cell recording and reached a stable maximum after 5 min of intracellular dialysis (Fig. 6A). In parallel recordings in which control cells, without peptide in the pipette, were monitored, there was some increase in Ca 2ϩ current over this period, but the increase was substantially greater with SrcKI peptide present in the pipette buffer (Fig. 6B). The overall increase in current due to SrkI compared with control cells (682.6 Ϯ 87 pA) was similar to the increase due to PP1 in experiments carried out at the same time (Fig. 5C). None of the tyrosine kinase inhibitors had any obvious effects on the activation kinetics of the Ca 2ϩ currents, and the major effect was to increase the steady-state magnitude of the currents, a characteristic in common with the voltage-independent facilitation following naloxone and suramin addition.
Ca 2ϩ Currents Facilitated by PP1 Addition Are Mainly Due to P/Q-type Ca 2ϩ Channels, and the ␣ 1A Subunit of P/Q-type Ca 2ϩ Channels Is a Substrate for Src Family Kinases-In parallel experiments, PP1 was added to control cells (Fig. 7A) or to cells pretreated with the specific blocker of P/Q Ca 2ϩ channels -agatoxin TK (37). In the presence of 10 mM Ba 2ϩ , -agatoxin TK (400 nM) had relatively minor effects on the magnitude of Ca 2ϩ currents. The facilitation due to PP1 addition was, however, abolished by -agatoxin TK (Fig. 7B) indicating that this facilitation was largely due to removal of an inhibition of P/Qtype Ca 2ϩ channels. As these data appeared to identify the major type of Ca 2ϩ channel involved in the facilitation by PP1, we examined the possibility that P/Q-type channel subunits are themselves substrates for Src family kinases. We first established that the ␣ 1A subunit of P/Q-type Ca 2ϩ channels could be immunoprecipitated from chromaffin cell extracts (Fig. 8A). A major polypeptide of around 190 kDa, a more minor polypeptide of around 220 kDa, and other minor larger forms were detected using the anti-P/Q antiserum in both cell extracts and immunoprecipitates consistent with previous work on brain extracts (25). The polypeptides were absent from the supernatants remaining after immunoprecipitation indicating efficient recovery in the immunoprecipitates. To assess phosphorylation by associated Src family kinases, the immunoprecipitates were incubated with [ 32 P]ATP in the presence and absence of PP1, and incorporated radioactivity was detected with a Phosphor-Imager. This demonstrated that a polypeptide close in size to the ␣ 1A subunit of P/Q-type Ca 2ϩ channels was phosphorylated in a PP1-sensitive manner by endogenous kinases in the immunoprecipitate (Fig. 8B). To determine whether the ␣ 1A subunit could be endogenously phosphorylated on tyrosine, anti-phosphotyrosine or control immunoprecipitates from chromaffin cells were probed with anti-␣ 1A . A 220-kDa polypeptide was detected with anti-␣ 1A that exactly co-migrated on the same blot with the larger form of the ␣ 1A subunit in the antiphosphotyrosine immunoprecipitate (Fig. 8C). The specificity of this band was revealed by the fact that it was not detected by anti-␣ 1A in a control immunoprecipitate (Fig. 8C) nor by other antisera in the anti-phosphotyrosine immunoprecipitate. These data are consistent with tyrosine phosphorylation of the 220-kDa form of the ␣ 1A subunit. DISCUSSION The findings from this study provide further evidence for the existence of distinct voltage-dependent and voltage-independent pathways for autocrine inhibition of voltage-gated Ca 2ϩ channels in adrenal chromaffin cells (Fig. 8). The voltage-inde- pendent pathway in these cells has been little studied. We have exploited the predominance of the voltage-independent pathway for Ca 2ϩ current inhibition by endogenous ATP and opioids under our recording conditions to examine the pathway involved. This was done in the current study by the use of the antagonists naloxone and suramin to block the effects of activation of purinergic and opioid receptors, but facilitation could be observed following addition of naloxone alone. Previous work (23, 26 -28, 38 -40) has confirmed that autocrine inhibition can be mimicked in perfused cultures by addition of ATP and opioid peptides. A key finding of this study is the demonstration of a single voltage-independent pathway for inhibition of mainly P/Q-type Ca 2ϩ channels that require activity of a Src family kinase. The involvement of a Src-like kinase in the regulation of N-type channels was originally described for chick dorsal root ganglion neurons (14) but had not been demonstrated in other cell types or for other types of Ca 2ϩ channels. This is, therefore, the first demonstration of a tyrosine kinasedependent pathway for Ca 2ϩ channel regulation in chromaffin cells and for P/Q-type Ca 2ϩ channels.
Multiple pathways exist for the regulation of neuronal voltage-gated Ca 2ϩ channels. One inhibitory pathway involving GPCRs (1-3) is the so-called voltage-dependent pathway (4,5). Occupation of various GPCRs leads to release of ␤␥ subunits that directly interact with (41)(42)(43) and inhibit the ␣ 1 subunits of N-and P/Q-type Ca 2ϩ channels (2,44). This inhibition is released by prior strong depolarization (prepulse facilitation) leading to the description of this pathway as being voltage-dependent. Voltage-dependent regulation changes the activation kinetics of the channels as well as the magnitude of the steadystate current. The existence of such a pathway in adrenal chromaffin cells has also been known for many years (45). The existence of autocrine feedback inhibition of Ca 2ϩ channels in chromaffin cells has been shown to be due to released secretory granule contents activating both purinergic and opioid receptors (23, 26 -28, 38 -40). This pathway has generally been found to be voltage-dependent with the inhibition being partially reversed by a prepulse depolarization step, but recently voltage-independent inhibition of Ca 2ϩ channel currents has been observed in chromaffin cells (20,21).
We have demonstrated that the voltage-independent inhibition of Ca 2ϩ channel currents in chromaffin cells is unlikely to involve PKC. This is consistent with previous work examining the effect of PKC activation or inhibition in chromaffin cells which also saw no effect on Ca 2ϩ channel currents (46). Instead, we have provided evidence for a requirement for Src family kinase activity. The evidence for a crucial involvement of a Src family tyrosine kinase in the regulation of Ca 2ϩ currents in chromaffin cells comes from the effects of structurally distinct inhibitors applied externally or via the pipette solution. External application of the Src kinase inhibitor PP1 (34) increased the Ca 2ϩ currents. The requirement for Src family kinase activity was confirmed by introduction of a peptide inhibitor by inclusion in the recording pipette. This inhibitor peptide, SrcKI, is derived from the regulatory domain of pp60 v-Src and is believed to be a specific inhibitor of tyrosine kinases (35,36). The ablation of the facilitation due to naloxone and suramin blockade of purinergic and opioid receptors by prior addition of PP1 and the loss of a PP1 effect in NCS1(E120Q)-expressing cells strongly suggested the overlap of these pathways. We therefore conclude that a single voltageindependent pathway exists that is both Src-dependent and NCS-1-dependent (Fig. 9). The fact that two structurally distinct Src kinase inhibitors, PP1 and SrcKI, produced increases in Ca 2ϩ currents argues against their effects being due to some nonspecific actions of the compounds. The specificity of action of PP1 and the restriction of a Src kinase requirement to the voltage-independent pathway was demonstrated by the lack of effect of prior PP1-induced facilitation on the magnitude of pre-pulse facilitation. Previous work (47,48) has shown that stimulation of secretion from chromaffin cells is accompanied by an increase in tyrosine phosphorylation due, at least in part, to activation of Src family tyrosine kinases, suggesting that Src kinases would be active under the conditions used in the present study.
Tyrosine kinase inhibitors have been shown previously to inhibit L-type Ca 2ϩ channels in vascular smooth muscle cells (49). The stimulation of Ca 2ϩ currents seen here was due, however, to increases in non-L-type Ca 2ϩ channels as it was detectable in the presence of nifedipine. It has been demonstrated that N-and P/Q-type Ca 2ϩ channels are present at similar densities in bovine adrenal chromaffin cells (27,50) and are often regulated in parallel in response to GPCR activation (39). Under our conditions, the voltage-independent feedback inhibition of Ca 2ϩ currents was mainly due to inhibition of P/Q-type Ca 2ϩ channels as facilitation was prevented by the addition of the P/Q-type antagonist -agatoxin TK.

FIG. 9. A model is shown for the regulation of non-L-type Ca 2؉ channels in adrenal chromaffin cells with a comparison of the voltage-dependent and voltage-independent pathways.
In the presence of 10 mM extracellular Ba 2ϩ , present in the bath buffer, exocytosis will be constantly stimulated leading to release of secretory granule contents, some of which have autocrine actions. In other studies, the effects of ATP and opioid agonists on the voltage-dependent pathway have been reported. In the present study, even after addition of a saturating dose of the purinergic and opioid antagonists suramin and naloxone, further tonic inhibition could be relieved by a prepulse depolarization implying the existence of a third (X) autocrine factor, acting in the voltage-dependent pathway. The properties of the two pathways for Ca 2ϩ channel inhibition based on the data in this and an earlier study are listed in the lower part of the figure.
Under our recording conditions, the autocrine inhibition reversed by addition of purinergic and opioid antagonists is voltage-independent and does not modify channel activation kinetics. The demonstration that prepulse facilitation can still be elicited after blockade of these receptors suggests the existence of an additional autocrine inhibitor that acts in a specific voltagedependent pathway (Fig. 9). There are many bioactive components released from chromaffin cells that could be candidates for this inhibitor (51). Much of the earlier literature on Ca 2ϩ channel regulation in chromaffin cells has emphasized the existence of voltage-dependent inhibition by ATP and opioids even though the inhibition was only partially reversed by strong prepulse depolarization (23, 26 -28, 38 -40). As noted previously for sympathetic neurons (9), the demonstration of voltage-independent pathways and the relative contribution of voltage-dependent and voltage-independent mechanisms for channel regulation appear to be crucially dependent on the composition of the pipette buffer. Voltage-independent inhibition of non-L-type Ca 2ϩ currents in chromaffin cells by histamine was demonstrated recently by others (21). Of interest was the observation that current inhibition could be switched from voltage-dependent to voltage-independent modes by changes in Ca 2ϩ buffering in the pipette solution.
In our previous work (20) on the functional roles of the Ca 2ϩ -binding protein NCS-1, we had demonstrated that overexpression of a dominant-negative mutant with a disabling mutation in the third EF hand (NCS-1(E120Q)) resulted in an increase in Ca 2ϩ current magnitude in chromaffin cells. In these cells, addition of naloxone and suramin did not result in any further increase in Ca 2ϩ currents implying that the mechanisms for Ca 2ϩ current facilitation overlapped and that the voltage-independent pathway had a requirement for NCS-1.
We have now extended those observations to show an additional overlap with the Src kinase-dependent pathway (Fig. 9).
The mechanism by which Src-like kinase activity leads to Ca 2ϩ channel inhibition is not known in detail, and until now, it had only been reported in one study (14). Direct tyrosine phosphorylation of the ␣ 1B subunit of the N-type Ca 2ϩ channel following ␥-aminobutyric acid application has been demonstrated in chick DRG neurons (15), but the exact Src kinase family member involved is unknown. The ␣ 1A subunit of P/Qtype Ca 2ϩ channels has been shown to be phosphorylated in vitro by PKC, cAMP-dependent protein kinase, and protein kinase G (25). In anti-␣ 1A immunoprecipitates from brain at least two forms of the ␣ 1A subunit were convincingly identified and were suggested to be splice variants of the channel subunit. The major 190-kDa form was a substrate for PKC and protein kinase G, whereas the 220-kDa form was selectively phosphorylated by cAMP-dependent protein kinase (25). The possibility that the ␣ 1A subunit is a substrate for Src family kinases has not been examined, although the protein has several potential tyrosine phosphorylation motifs in the intracellular loops. 2 We have demonstrated here that endogenous kinases present in immunoprecipitates of the ␣ 1A subunit phosphorylated a polypeptide of around the expected size for ␣ 1A subunit and that this is inhibited by PP1. In addition, a polypeptide corresponding to the 220-Da form of the ␣ 1A subunit was specifically detected by anti-␣ 1A in an anti-phosphotyrosine immunoprecipitate. These data suggest that the ␣ 1A subunit of P/Q-type Ca 2ϩ channels could potentially be a direct target for Src family kinases. Activation of Src kinase via GPCR pathways is well established (52), and Src and other non-receptor tyrosine kinases have been shown to interact directly with G-protein ␣ subunits and may be direct G-protein effectors (53,54) as well as being Src kinase substrates themselves (55). Further work will be required to fully characterize the Src-dependent pathway for Ca 2ϩ channel regulation and the significance of phosphorylation of the ␣ 1A subunit. Our finding of a Src-dependent pathway in adrenal chromaffin cells points to the likelihood that this is a widely expressed pathway for voltage-independent regulation of non-L-type Ca 2ϩ channels.