Pituitary Adenylyl Cyclase-activating Peptide Activates Multiple Intracellular Signaling Pathways to Regulate Ion Channels in PC12 Cells*

Pituitary adenylyl cyclase-activating peptide (PACAP) stimulates calcium transients and catecholamine secre-tion in adrenal chromaffin and PC12 cells. The PACAP type 1 receptor in these cells couples to both adenylyl cyclase and phospolipase C pathways, but although phospolipase C has been implicated in the response to PACAP, the role of adenylyl cyclase is unclear. In this study, we show that PACAP38 stimulates Ca 2 1 influx in PC12 cells by activating a cation current that depends upon the dual activation of both the PLC and adenylyl cyclase signaling pathways but does not involve protein kinase C. In activating the current, PACAP38 has to overcome an inhibitory effect of Ras. Thus, in cells expressing a dominant negative form of Ras (PC12asn17-W7), PACAP38 induced larger, more rapidly activating currents. This effect of Ras could be overidden by intracellular guanosine-5 * - O -3-(thio)triphosphate (GTP g S), suggesting that it was mediated by inhibition of downstream G proteins. Ras may also inhibit the current through a G protein-independent mechanism, because cAMP analogues activated the current in PC12asn17-W7 cells, provided GTP g S was present, but not in PC12 cells expressing wild type Ras. We conclude that coupling of PACAP to both

Pituitary adenylyl cyclase-activating peptide (PACAP) 1 isolated from rat hypothalami is a member of the secretin-glucagon family of peptides, based on its amino acid composition (1). It comprises 38 amino acids, of which the first 27 bear remarkable homology to another member of this family, vasoactive intestinal peptide (VIP). Since its identification, PACAP has been localized to subsets of neurons within the central and peripheral nervous systems and is widely regarded as a candidate peptide neurotransmitter (2). In particular, PACAP has been identified in sympathetic, preganglionic neurons and in neurons of the adrenal medulla (3)(4)(5)(6), where it regulates catecholamine synthesis and release (7)(8)(9)(10).
Three receptor subtypes for PACAP have been identified, which display differential affinities for this peptide and specificity for VIP. The affinity of the type I receptor for PACAP is 3 orders of magnitude higher than for VIP. The type II receptor has a lower affinity for PACAP than the type I and is unable to discriminate between the two peptides. Whereas the type II receptor is exclusively coupled to adenylyl cyclase, the type I receptor has dual coupling to both adenylyl cyclase and phospholipase C (PLC) (11). Both of these receptors have been cloned and are predicted to conform to the classic seven-transmembrane domain pattern of G protein-coupled receptors. At least six splice variants of the type II receptor have been demonstrated, and these alter the precise pattern of coupling when expressed in heterologous systems (11)(12)(13). A third PACAP receptor, which is not coupled to adenylyl cyclase or phospholipase C, appears to be expressed preferentially in pancreatic beta cells (14,15).
PACAP stimulates calcium transients in a variety of cell types (16 -20), including bovine adrenal chromaffin cells (21), by stimulating both the release of Ca 2ϩ from intracellular stores and Ca 2ϩ influx. Some of these responses have been shown to involve PLC activation, but the role of adenylyl cyclase is unclear. Since PACAP stimulates inositol 1,4,5trisphosphate and cyclic AMP production in adrenal chromaffin cells (22), both pathways might be expected to contribute to the response in these cells. We therefore investigated the roles of adenylyl cyclase and PLC in the PACAP-induced [Ca 2ϩ ] i transient in chromaffin-like cells, using the rat PC12 phaeochromocytoma cell line. These cells possess the type I PACAP receptor (23) and, consistent with this, release catecholamines in response to low concentrations of PACAP (24). PACAP also causes PC12 cells to extend neurites and adopt a neuronal morphology that is distinct from that observed for nerve growth factor (23,25). This effect is independent of the activation of cAMP-dependent protein kinase (PKA), but rather depends on the activation of extracellular signal-regulated kinase 1 or 2 through a Ras-independent mechanism (23). Here we report that, in PC12 cells, PACAP activates a Ca 2ϩ -carrying inward current that is dependent upon the dual activation of both adenylyl cyclase and phospholipase C. This current appeared to be negatively regulated by Ras, partly through inhibition of a downstream G protein. PACAP also inhibited potassium current in these cells, but in contrast to the PACAP-induced inward current, this effect was mimicked by cAMP analogues and was not modulated by Ras. These data suggest that coupling of PACAP to both intracellular second messenger systems, adenylyl cyclase and phospholipase C, is required to activate the inward current associated with [Ca 2ϩ ] i transients in PC12 cells.

EXPERIMENTAL PROCEDURES
Materials-Tissue culture reagents were purchased from Life Technologies, Inc. PACAP38 was from Peninsula Laboratories, Inc. (Belmont, CA) and was applied in extracellular solution containing a mixture of peptidase inhibitors (10 g/ml each of chymostatin, leupeptin, and antipain and 1 M pepstatin A, all from Sigma). GTP␥S, AMP-PCP, db-cAMP, and 8-Br-cAMP were from Sigma. (R p )-cAMPS was from Biologic Life Science Institute (Bremen, Germany). Fura2-AM, H89, Ro-318220 (dissolved in Me 2 SO), U73122, and U73343 (both dissolved in CHCl 3 ) were from Calbiochem. Me 2 SO or CHCl 3 was present at Յ0.1% in the final solution and by itself did not affect membrane current. PC12asn17-W7 cells stably expressing a dominant negative form of Ras were from Glaxo-Wellcome (Stevenage) and were unresponsive to nerve growth factor (23).
Cell Culture-PC12 and PC12asn17-W7 cells were plated on collagen-coated tissue culture dishes in Dulbecco's modified Eagle's medium supplemented as described previously (23) and grown at 37°C in 6.5% CO 2 . The absence of active p21 ras in the PC12asn17-W7 cells was confirmed indirectly by examining the ability of nerve growth factor to stimulate mitogen-activated protein kinase (23) using the BIOTRAK TM mitogen-activated protein kinase enzyme assay kit (Amersham Pharmacia Biotech). For electrophysiology experiments, both cell types were plated in 24-well plates on glass coverslips coated with poly-D-lysine at a density of 1.5 ϫ 10 6 cells/well. They were incubated overnight in 6.5% CO 2 at 37°C in a minimum serum medium containing Dulbecco's modified Eagle's medium supplemented with 0.25% (v/v) fetal calf serum, 0.2% (v/v) bovine serum albumin, 2 mM glutamine, 100 mg/ml streptomycin, and 100 units/ml penicillin.
Fura-2 Measurements-Cells were harvested from tissue culture dishes when approaching confluence, centrifuged (1000 ϫ g, 5 min), and resuspended at 5 ϫ 10 6 cells/ml in Dulbecco's modified Eagle's medium containing 2 mM glutamine, 5% (v/v) fetal calf serum, 200 M sulfinpyrazone, and 6 M Fura-2/AM. The cells were incubated for 30 min at 37°C with constant stirring to load Fura-2/AM. They were then washed twice and resuspended at 10 ϫ 10 6 cells/ml in a physiological solution containing 145 mM NaCl, 5 mM KCl, 1 mM MgSO 4 , 1.2 mM NaH 2 PO 4 , 3 mM CaCl 2 , 10 mM glucose, 10 mM HEPES, 0.2 mM sulfinpyrazone, pH 7.4. After transfer to a thermostatically controlled cuvette at 37°C, fluorescence was continuously monitored at 505 nm with a Perkin-Elmer LS-50 spectrometer. Fura-2 fluorescence was excited alternately at 340 and 380 nm, and [Ca 2ϩ ] i was calculated from the ratio of fluorescence excited at the two wavelengths, using the Intracellular Biochemistry program supplied with the Perkin-Elmer package. The calculations were according to Grynkiewicz et al. (43), based on an in situ calibration that employed ionomycin and EGTA to obtain maximum and minimum ratios at saturating and minimum levels of Ca 2ϩ , respectively.
Patch Clamp Recording-Cells on coverslips were transferred to a recording chamber on the stage of an inverted microscope and superfused continuously with physiological salt solution containing 112 mM NaCl, 5 mM KCl, 1 mM MgCl 2 , 1.8 mM CaCl 2 , 10 mM glucose, 10 mM HEPES, pH 7.3. Pipettes pulled from borosilicate glass capillaries (Clark Electromedical, Reading, United Kingdom) were usually filled with a K ϩ -based solution containing 120 mM KCl, 2 mM MgCl 2 , 5 mM EGTA, and 10 mM HEPES (pH 7.3 adjusted with KOH) and had resistances of 1-3 megaohms. To block outward K ϩ currents, the KCl was replaced with equimolar CsCl, and pH was adjusted with CsOH.
Membrane currents were recorded under voltage clamp using an Axopatch 200A patch clamp amplifier (Axon Instruments, Foster City, CA), filtered at 1 kHz, digitized at 2.5-10 kHz using a Digidata 1200 interface (Axon Instruments), and stored on a computer using pCLAMP (version 5) software (Axon Instruments). Series resistance, usually around 5 megaohms, was compensated by ϳ80%. Input resistance and membrane capacitance were calculated from the current transient elicited by a voltage step from Ϫ80 to Ϫ90 mV. The PC12 cells studied had an input resistance of 17 Ϯ 1 gigaohm and capacitance of 2.7 Ϯ 0.3 pF (n ϭ 70). The PC12asn17-W7 cells had an input resistance of 1.2 Ϯ 0.1 gigaohms (n ϭ 27) and capacitance of 10 Ϯ 1 pF (n ϭ 27). Currents were analyzed off-line using pCLAMP (versions 5 and 6; Axon Instruments) and Origin (version 4; MicroCal, Inc., Northampton, MA) software.
Drug solutions were applied from a homemade, gravity-driven perfusion system, which exchanged the solution around a cell in Ͻ50 ms. Effects of agents incorporated into the pipette solution were determined by recording alternately from cells with drug-free and drug-containing solutions. Results are expressed as mean Ϯ S.E. Statistical analyses were performed using paired or unpaired Student's t tests, with p Յ 0.05 considered significant. Fig. 1A, PACAP induced a biphasic increase in [Ca 2ϩ ] i in PC12 cells loaded with Fura-2/AM. Following the application of 5-500 nM PACAP38, the [Ca 2ϩ ] i increased to a peak within 1 min and then fell to a lower level that was sustained for the duration of the PACAP38 application. The sustained phase was abolished by 1 mM CoCl 2 (n ϭ 3), an inhibitor of Ca 2ϩ influx (Fig. 1A, a). The inhibitory effect was selective for the sustained response, because when PACAP38 was applied in the presence of 1 mM CoCl 2 , the transient rise in [Ca 2ϩ ] remained ( Fig. 1A, b), the peak of which was 20 Ϯ 3% (n ϭ 3) smaller than that observed under control conditions. In contrast, the sustained rise in [Ca 2ϩ ] i caused by PACAP38 was not inhibited by 1 mM CdCl 2 (n ϭ 3) or 1 M nifedipine (n ϭ 3).

PACAP Causes Ca 2ϩ Influx in PC12 Cells by Activating a Ca 2ϩ -carrying Current-As shown in
The application of PACAP38 (1 nM to 5 M) to PC12 cells voltage-clamped at negative membrane potentials activated an inward current, following an initial delay of 1-2 min (Fig. 1B,  n ϭ 93). The amplitudes of the induced currents were compa- rable among cells studied on a given day but varied widely from day to day, ranging from 10 pA to 10 nA at Ϫ80 mV, in response to 1 M PACAP38. Moreover, a second application of PACAP38 often produced a smaller response than the first. These factors complicated the analysis of dose-response relationships. Comparisons of currents activated by different concentrations of PACAP38 were therefore made using the first responses recorded from matched cells on one day, and current amplitudes were normalized against cell capacitance to control for variation in cell size. In one set of cells, the current activated by 1 nM PACAP38 (6 Ϯ 2 pA/pF, n ϭ 7) was significantly smaller (p Ͻ 0.01) than the current activated by 100 nM (20 Ϯ 4 pA/pF, n ϭ 4). In contrast, in other cells, 100 nM PACAP38 activated a current (2 Ϯ 1 pA/pF, n ϭ 3) that was not significantly different from that activated by 1 M PACAP38 (6 Ϯ 3 pA/pF, n ϭ 3). PACAP38 also induced comparable currents at 500 nM (8 Ϯ 6 pA/pF, n ϭ 3) and 5 M (13 Ϯ 9 pA/pF, n ϭ 3), suggesting that the maximum response was reached by around 100 nM PACAP38.
As found with the sustained [Ca 2ϩ ] i response, the current induced at Ϫ80 mV by 1 M PACAP38 was reversibly inhibited by 1 mM CoCl 2 (n ϭ 4; Fig. 1B, a). The current induced by 1 M PACAP38 was additionally reduced by exposure to Ca 2ϩ -free medium (Fig. 1B, b), although by only 25 Ϯ 7% (n ϭ 5). Removing Na ϩ from the medium, by equimolar substitution of tetraethylammonium chloride for NaCl, had a larger effect, reducing the current by 90% from 86 Ϯ 43 pA (n ϭ 9) to 8 Ϯ 3 pA (n ϭ 10; p Ͻ 0.05). In contrast, the amplitude of the PACAPinduced current was not significantly altered at any potential when the extracellular NaCl was replaced with equimolar Na 2 SO 4 (n ϭ 3) or the K ϩ in the pipette solution was replaced with equimolar Cs ϩ (n ϭ 4).
The voltage dependence of the current induced by 1 M PACAP38 is shown in Fig. 1C. Pronounced inward rectification was apparent at negative potentials with negligible current recorded at positive potentials. When current could be measured at positive potentials, it was always inwardly directed (up to 40 mV), implying a positive reversal potential as expected for a current carried by Ca 2ϩ or Na ϩ .
Intracellular Mediators of the PACAP38-induced Current-To test the involvement of protein phosphorylation in current activation by PACAP38, phosphorylation was prevented by adding the nonhydrolyzable ATP analogue AMP-PCP (1 mM) to the pipette solution. In these conditions, 1 M PACAP38 induced a significantly smaller current at Ϫ80 mV ( Fig. 2A, a). The current induced at Ϫ80 mV by 1 M PACAP38 was also significantly reduced when the PKA was blocked, either by adding (R p )-cAMPS (2 mM) to the pipette solution ( Fig. 2A, a) or by adding H89 (50 M) to the extracellular solution ( Fig. 2A, b). The current induced by 1 M PACAP38 was also suppressed by more than 50% in the presence of the PLC inhibitor, U73122 (50 M), whereas the same concentration of U73343, an analogue of U73122 that lacks the inhibitory action on PLC, had no effect ( Fig. 2A, b). The protein kinase C (PKC) inhibitor, RO318220, also reduced the current, but by only 19% at 100 M ( Fig. 2A, b).
Despite the inhibitory effects of PKA blockers on the PACAP38-induced current, membrane-permeable analogues of cAMP, which stimulate PKA activity, were unable to mimic the action of PACAP38 even at high concentrations. As illustrated in Fig. 2B for 10 mM db-cAMP, neither this analogue nor 8-Br-cAMP induced significant current at Ϫ80 mV, although in the same cells 1 M PACAP38 did induce substantial current. These cAMP analogues did, however, mimic an inhibitory effect of PACAP38 on K ϩ currents recorded from PC12 cells (Fig. 2C).
Outward K ϩ currents activated by steps from Ϫ80 to 40 mV were reduced by 50 Ϯ 11% (n ϭ 10) in the presence of 5 M PACAP38. The same currents were reduced by 26 Ϯ 3% (n ϭ 5) in the presence of 10 mM db-cAMP and by 20 Ϯ 10% (n ϭ 5) in the presence of 10 mM 8-Br-cAMP.
Involvement of p21 ras in the PACAP38-induced Current-Studies on Drosophila muscle found that PACAP activation of inward current and modulation of K ϩ current required the co-activation of cAMP and Ras/Raf signaling pathways (26). We examined the involvement of p21 ras in the response of PC12 cells to PACAP38, using a dominant negative Ras (PC12asn17-W7) cell line. This dominant negative form of Ras was previously shown to inhibit the classical Ras-dependent stimulation of extracellular signal-regulated kinase 1/2 by nerve growth factor (23). The application of PACAP38 to PC12asn17-W7 cells induced an inward current at negative membrane potentials (Fig. 3A), as observed in PC12 cells. The amplitude of the current induced by PACAP38 (5 M, Ϫ80 mV) was, however, significantly (p Ͻ 0.03) larger in PC12asn17-W7 cells (Fig. 3B).
In addition, the current activated with a delay of 26 Ϯ 3 s (n ϭ 5), which was significantly (p Ͻ 0.02) shorter than the delay of 49 Ϯ 9 s (n ϭ 7) measured in wild type PC12 cells (Fig. 3C). The inhibitory effect of PACAP38 on K ϩ current was also retained  Fig. 1B. The numbers of cells studied are indicated within or beside bars. B, application of 10 mM db-cAMP had no effect in cells displaying an inward current in response to 5 M PACAP38 before and after db-cAMP application. C, both PACAP38 (5 M) and db-cAMP (10 mM) reduced the outward K ϩ current activated by voltage steps from Ϫ80 mM to 40 mV. **, p Ͻ 0.01; ***, p Ͻ 0.001 for inhibition of current amplitude.
The effects of cAMP analogues were unaffected by the presence of the dominant negative Ras. Although they failed to activate significant inward current at Ϫ80 mV (Fig. 3A), 10 mM db-cAMP or 8-Br-cAMP reduced the outward current at 40 mV (Fig. 3E) by 18 Ϯ 6% (n ϭ 5) and 20 Ϯ 10% (n ϭ 3), respectively.
Modulation by GTP␥S-G proteins are uncoupled from receptors by the binding of GTP␥S, which is expected to cause a nonselective activation of G proteins in the cell. When 300 M GTP␥S was added to the pipette solution, it activated an inward current at negative membrane potentials (Fig. 4). This current was distinct from that induced by PACAP38, because it was not blocked by 1 mM CoCl 2 , and it reversed direction near 0 mV (Fig. 4A). GTP␥S additionally modulated the PACAP38dependent current. As expected for a response involving a G protein-coupled receptor, 10 mM GTP␥S reduced the inward current activated by 1 M PACAP38 at Ϫ80 mV in PC12 cells, from 17 Ϯ 3 pA/pF (n ϭ 5) to 2 Ϯ 1 pA/pF (n ϭ 7; p Ͻ 0.001). In contrast, 300 M GTP␥S potentiated the PACAP38-induced current and shortened the delay to current activation in wild type PC12 cells. This can be seen in Fig. 4B, where a second application of PACAP38, presented several minutes after dialyzing the cell with GTP␥S, produced a larger current than when it was first applied shortly after obtaining the whole-cell configuration. After equilibration with GTP␥S, 2 M PACAP38 activated an inward current at Ϫ80 mV that was 877 Ϯ 298% (n ϭ 4) larger than at the start of recording. At the same time, the latency to current activation was reduced from 130 Ϯ 30 s (n ϭ 4) to 40 Ϯ 10 s (n ϭ 7; p Ͻ 0.01). The enhancing effect of GTP␥S on the PACAP38-induced current was not observed in PC12asn17-W7 cells (Fig. 4C), where the current amplitude at Ϫ80 mV was 4 Ϯ 2 pA/pF (n ϭ 5) under control conditions and 8 Ϯ 3 pA/pF (n ϭ 18, p Ͼ 0.05) after equilibration with 300 M intracellular GTP␥S.
In PC12 cells, as observed under control conditions, 10 mM db-cAMP had essentially no effect on membrane current recorded with pipette solution containing 300 M GTP␥S ( Fig.  4D; n ϭ 3). In contrast, in PC12asn17-W7 cells exposed to 300 M GTP␥S in the recording pipette, 10 mM db-cAMP induced an inward current at Ϫ80 mV ( Fig. 4E) with an amplitude of 409 Ϯ 180 pA/pF and latency of 20 Ϯ 10 s (n ϭ 7). As found for the PACAP38-dependent current, the current activated by db-cAMP was abolished by 1 mM CoCl 2 (Fig. 4E). DISCUSSION As previously found in adrenal chromaffin cells (21), PACAP38 caused a biphasic increase in [Ca 2ϩ ] i in PC12 cells. The initial transient rise may have reflected Ca 2ϩ release from intracellular stores, while the sustained increase was due to Ca 2ϩ influx, as indicated by a selective inhibitory effect of CoCl 2 on the sustained phase. L-type, voltage-gated Ca 2ϩ channels were found to contribute to Ca 2ϩ influx in bovine, porcine, and canine (21,26,27) but not rat (29) chromaffin cells. They cannot account for Ca 2ϩ influx in rat PC12 cells, because the PACAP38-induced increase in [Ca 2ϩ ] i was little affected by CdCl 2 or nifedipine at concentrations causing complete block of L-type channels (30). Ca 2ϩ influx was more likely mediated by the CoCl 2 -sensitive inward current activated by PACAP38 in voltage-clamped PC12 cells. This current reversed direction in the voltage range expected for Ca 2ϩ or Na ϩ and was inhibited by removing either ion from the extracellular solution, implying that both ions carried the current. Since removal of intracellular K ϩ failed to alter the current, it may reflect cation FIG. 3. Effects of PACAP38 in PC12asn17-W7 cells. A, PACAP38 (5 M) induced inward current at Ϫ80 mV, while db-cAMP (10 mM) had essentially no effect. B, the current induced by 5 M PACAP38 was larger in PC12asn17-W7 cells than in wild type PC12 cells. C, the duration of the latent period between the start of PACAP38 (5 M) application and the appearance of current was shorter in PC12asn17-W7 cells. D and E, the outward K ϩ current activated by voltage steps from Ϫ80 to 40 mV was reduced by PACAP38 (5 M) and db-cAMP (10 mM) in PC12asn17-W7 cells. The numbers of cells studied are indicated within bars. *, p Ͻ 0.05 compared with PC12 cells.

FIG. 4. Modulation by GTP␥S.
When 300 M GTP␥S was added to the pipette solution, an inward current gradually developed at Ϫ80 mV after rupturing the membrane patch to form the whole-cell configuration, in both PC12 (A, B, and D) and PC12asn17-W7 (C) cells. A, the GTP␥S-activated current was insensitive to block by 1 mM CoCl 2 (inset) and displayed a near linear current versus voltage relationship with current reversal near 0 mV. Points and bars represent mean Ϯ S.E. of 4 -7 cells. B and C, comparison of currents induced by PACAP38 (500 nM) when applied shortly after breakthrough to the whole-cell configuration and after equilibration with 300 M intracellular GTP␥S in PC12 (B) and PC12asn17-W7 (C) cells. D and E, db-cAMP (10 mM) had no effect on wild-type PC12 cells (D) but induced an inward current in PC12asn17-W7 cells at Ϫ80 mV, which was inhibited by 1 mM CoCl 2 . channels with selectivity for Na ϩ and Ca 2ϩ . PACAP38 was also found to activate a Na ϩ -dependent current in bovine chromaffin cells (21). Activation of cation channels by PACAP would directly mediate Ca 2ϩ influx and cause membrane depolarization, leading to the opening of voltage-gated Ca 2ϩ channels. This depolarizing action of PACAP would be potentiated by its inhibitory effect on K ϩ current. Since L-type channels are one of several voltage-gated Ca 2ϩ channel subtypes expressed in PC12 and chromaffin cells (30,31), variation in the sensitivity of the secretory PACAP response to L-type Ca 2ϩ channel antagonists may reflect differences in the relative expression of these channels.
Blockade by millimolar GTP␥S of PACAP38-induced current activation, and hence Ca 2ϩ influx, is consistent with the involvement of a G protein-coupled receptor. Current activation also involved protein phosphorylation, because it was prevented by the nonhydrolyzable ATP analogue AMP-PCP. The response required activation of the adenylyl cyclase-PKA pathway, because two different PKA inhibitors prevented it. The failure of membrane-permeant analogues of cAMP to mimic this effect of PACAP38 in PC12 cells shows, however, that this pathway is not, in itself, sufficient to account for the response. This contrasts with the inhibitory effect of PACAP on K ϩ current, which could be mimicked by cAMP analogues. As suggested for adrenal chromaffin cells (21), current activation by PACAP38 also appeared to require the PLC signaling pathway, because it was inhibited by the selective PLC blocker U73122 but not its inactive analogue U73343. Downstream activation of PKC is unlikely to play a major role, because although the PKC inhibitor RO318220 reduced the response, the effect was small and observed at high drug concentrations. The PLC pathway involved in the activation of Ca 2ϩ influx by PACAP appears, therefore, to be distinct from PKC.
Taken together, our data indicate that adenylyl cyclase and PLC are both activated by PACAP to elicit an inward current in PC12 cells, so that inhibition of either pathway blocks the response. This finding is consistent with the known dual coupling of the type I PACAP receptor to the two pathways. The concentration dependence of the PACAP38-induced current is also consistent with the involvement of a type I receptor, as are previous studies showing the involvement of a type I receptor in the secretory response of PC12 cells to PACAP (24). Heterologous expression studies (11,13,32) have consistently shown that coupling of PACAP to phospholipase C through the type I receptor requires higher concentrations of agonist (Ͼ10 nM) than does activating adenylyl cyclase (Ͻ1 nM). At the concentrations of PACAP38 required to activate the current in this study, both pathways would have been stimulated.
The findings reported here provide an explanation for previous conflicting observations on the role of adenylyl cyclase in mediating [Ca 2ϩ ] i transients. Inhibitors of either PKA (19,20,33,34) or PLC (12,16,17,21,35) pathways have been reported separately to block [Ca 2ϩ ] i transients in a variety of cell types. However, other studies that found that analogues of cAMP or forskolin could not mimic the effect of PACAP were interpreted as evidence against a role for the cAMP pathway in mediating the [Ca 2ϩ ] i response (16,17,27).
The small GTP-binding protein Ras has been implicated in the regulation of ion channel activities (36 -39). Furthermore, co-activation of the Ras/Raf and cAMP signaling pathways was found to be necessary for the activation of inward current and modulation of K ϩ current by PACAP38 in Drosophila muscle (26). This was not the case in PC12 cells, because both the activation of inward current and inhibition of K ϩ current caused by PACAP38 were retained in cells expressing a dominant negative form of Ras. In fact, in these PC12asn17-W7 cells, PACAP38 activated a larger current with a shorter latency, suggesting that Ras may exert a tonic inhibitory influence on the current or on the signaling pathways that activate the current. Tonic regulation of Ca 2ϩ channels by Ras was previously found in sensory neurons, although in these cells it had a stimulatory effect (38). An inhibitory action of Ras was found in atrial cells, where it prevented the coupling of muscarinic receptors to K ϩ channels (36). Since Ras can be activated by ␤␥ subunits of heterotrimeric G proteins and by Ca 2ϩ influx (40,41), its inhibitory effect on the inward current would be reinforced in the presence of PACAP38, through activation of the G protein-coupled type I receptor. Alternatively, it is possible that Ras exerts its inhibitory effect only during stimulation of the PACAP receptor.
Interestingly, the intracellular application of 300 M GTP␥S in wild-type PC12 cells mimicked the effect of the dominant negative Ras, in that it reduced the latency and potentiated the amplitude of the PACAP38-induced current. Since it failed to do this in PC12asn17-W7 cells where the response was already enhanced and accelerated, it is likely that this action of GTP␥S reflects activation of GTP-binding proteins downstream of Ras, which bypass its inhibitory effect. Interestingly, in PC12asn17-W7 cells, GTP␥S enabled the activation of inward current by db-cAMP, an effect not seen in wild-type PC12 cells. Since CoCl 2 blocked the db-cAMP-induced current, it was probably the same as the current activated by PACAP38. Since current activation by PACAP38 appeared to require the coactivation of the cAMP and PLC pathways, this suggests that the PLC pathway was active in the presence of GTP␥S. Ras may therefore have an additional inhibitory effect in PC12 cells that is not bypassed by activating downstream G proteins. Since this inhibitory effect is overcome in the presence of PACAP38, Ras itself must be under inhibitory control by a signaling pathway activated by the PACAP receptor.
Even in the presence of GTP␥S, PACAP38 was required to activate the Ca 2ϩ -carrying inward current in both PC12 and PC12asn17-W7 cells. Although GTP␥S did activate an inward current at negative potentials in the absence of PACAP38, it was distinct from the PACAP-dependent current because it reversed direction near 0 mV and was not blocked by CoCl 2 . This current resembled a GTP␥S-activated Cl Ϫ current previously identified in chromaffin cells (42), suggesting that the same current is present in chromaffin-derived PC12 cells.
In summary, the data presented here indicate that Ca 2ϩ influx initiated by PACAP in PC12 cells requires the dual activation of both intracellular signaling pathways coupled to the type I receptor, namely adenylyl cyclase and phospholipase C. The Ca 2ϩ influx pathway is additionally under the negative influence of Ras, although part of this inhibitory action can be overidden by GTP␥S, presumably acting on GTP-binding proteins that are downstream of Ras. Our data also suggest an additional inhibitory action of Ras that cannot be overcome by activating downstream G proteins. This effect must be exerted on the PLC pathway, on the adenylyl cyclase pathway downstream of PKA, or on the channel itself.