Receptor-stimulated Phospholipase AP Activation Is Coupled to Influx of External Calcium and Not to Mobilization of Intracellular Calcium in C62B Glioma Cells*

C62B rat glioma cells respond to muscarinic cholinergic stimulation with transient inositol phosphate for- mation and phospholipase Az-dependent arachidonic acid liberation. Since phospholipase Az is a Caz+-sen- sitive enzyme, we have examined the role of the ago-nist-stimulated Ca2+ response in production of the ar- achidonate signal. The fluorescent indicator fura-2 was used to monitor changes in cytoplasmic Ca2+ levels ([Ca2+Ii) of C62B cells following acetylcholine treat-ment. In the presence of extracellular Ca2+, acetylcho- line induces a biphasic [Ca”+]i response consisting of an initial transient peak that precedes arachidonate lib- eration and a sustained elevation that outlasts the phospholipase Az response. The initial [Caz+]i peak is not altered by the absence of external Ca2+ and therefore reflects intracellular Ca2+ mobilization. The sustained elevation phase is dependent on the influx of external Ca2+; it is lost in Caz+-free medium and restored on the addition of Ca2+. Pretreating cells with phorbol dibutyrate substantially inhibits acetylcholine-stimulated inositol

C62B rat glioma cells respond to muscarinic cholinergic stimulation with transient inositol phosphate formation and phospholipase Az-dependent arachidonic acid liberation. Since phospholipase Az is a Caz+-sensitive enzyme, we have examined the role of the agonist-stimulated Ca2+ response in production of the arachidonate signal. The fluorescent indicator fura-2 was used to monitor changes in cytoplasmic Ca2+ levels ([Ca2+Ii) of C62B cells following acetylcholine treatment. In the presence of extracellular Ca2+, acetylcholine induces a biphasic [Ca"+]i response consisting of an initial transient peak that precedes arachidonate liberation and a sustained elevation that outlasts the phospholipase Az response. The initial [Caz+]i peak is not altered by the absence of external Ca2+ and therefore reflects intracellular Ca2+ mobilization. The sustained elevation phase is dependent on the influx of external Ca2+; it is lost in Caz+-free medium and restored on the addition of Ca2+. Pretreating cells with phorbol dibutyrate substantially inhibits acetylcholine-stimulated inositol phosphate formation and the peak [Ca2+]i response without affecting the sustained elevation in [Caz+li. This suggests that the release of internal Caz+ stores by inositol 1,4,5-trisphosphate can be blocked without interfering with Ca2+ influx. Pretreatment with phorbol also fails to affect acetylcholine-stimulated arachidonate liberation, demonstrating that phospholipase AZ activation does not require normal intracellular Caz+ release. Stimulated arachidonate accumulation is totally inhibited in Ca2+-free medium and restored by the subsequent addition of Ca2+. Pretreatment with verapamil, a voltage-dependent Ca2+ channel inhibitor, also blocks both the sustained [Ca2+Ii elevation and arachidonate liberation without altering peak intracellular Caz+ release. We conclude that the influx of extracellular Ca2+ is tightly coupled to phospholipase Az activation, whereas large changes in [Ca2+]i due to mobilization of internal Ca2+ stores are neither sufficient nor necessary for acetylcholine-stimulated phospholipase Az activation.
A wide variety of tissues respond to hormone and neuro-* This work was supported by United States Health Service Grants NS11615, HD03110, and NS20212. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "aduertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. ** To whom correspondence and reprint requests should be ad- transmitter binding with stimulated phospholipase C-dependent hydrolysis of phosphatidylinositol 4,5-bisphosphate (PIP2)' (reviewed in [1][2][3]. PIPz cleavage produces inositol 1,4,5-trisphosphate (1,4,5-IPS), which increases cytoplasmic calcium levels ([Ca2+Ji) by binding to specific intracellular receptors and mobilizing sequestered Cap+ stores (41, and diacylglycerol, which activates protein kinase C (5). An influx of extracellular Ca2+ is also stimulated by one or more possible mechanisms (3,6,7).
In many cases, agonist-stimulated PIPp hydrolysis is accompanied by phospholipase A2 activation, with consequent liberation of free arachidonic acid from cellular phospholipid and generation of eicosanoid second messengers (2). Although phospholipase C is generally believed to be under the direct control of one or more guanine nucleotide-binding proteins (G proteins) (8,9), the mechanism for phospholipase A2 regulation is less certain. G proteins may play a role in phospholipase Ap control (lO-lZ), but many other cellular factors are capable of modulating the enzyme's activity as well, including Ca2+ (13,14), diacylglycerol (15), protein kinase C (16)(17)(18)(19), inhibitory proteins (lipocortins) (20,21), and Na'/H' exhange (22). The physiological significance of all these (not mutually exclusive) factors is unclear, however.
Cytoplasmic Cap+ levels have long been proposed as a primary in uiuo regulator of phospholipase A2 activity (1,13,14). Purified phospholipase AP enzymes generally show a positive correlation between in vitro activity and Cap+ concentrations above basal physiological [Ca2+Ii (23)(24)(25)(26), and at least one isolated enzyme shows complete activation by Cap+ over a stimulated in uiuo [Ca2+]i range (24). Further evidence has come from studies using Cap+ ionophores or detergents to manipulate [Cap+]; in "intact" cells; the use of millimolar external Cap+ with ionophores produces phospholipase A2 activation in many cell types (2,17,27,28), including C62B cells (29). Arachidonate liberation has been observed in glomerular mesangial cells and platelets with controlled increases in [Ca2+]i over the physiological agonist-stimulated range (19,30).
Previous work in this laboratory has demonstrated that C62B rat glioma cells respond to muscarinic stimulation with rapid inositol phosphate formation followed by a transient accumulation of arachidonate peaking at 60-90 s (29,31). In this system, arachidonate liberation occurs preferentially via phospholipase Az degradation of phosphatidylinositol (31). The present study was conducted to determine the relative roles of intracellular Ca2+ mobilization and extracellular Ca2+ influx in cholinergically stimulated phospholipase Az activation in these cells.

Methods
Cell Culture"C62B cells were cultured in BME supplemented with 5% fetal calf serum, 1 mM glucose, 2 mM L-glutamine, 50 units/ml penicillin, and 50 pg/ml streptomycin in a humidified environment of 5% C02 and 95% air at 37 "C. The cells were grown to confluence in 75-cm2 Corning culture flasks and harvested by trypsinization as described previously (29); the only modification was the use of a calcium-free medium (11 mM glucose, 20 mM HEPES, 10.2 mM trisodium citrate, 4 mM potassium chloride, 110 mM sodium chloride, pH 7.2) for harvesting.
Cytoplasmic Ca2+ Measurement-Aliquots of harvested cells were allowed to settle onto sterile glass coverslips in 100-mm Corning tissue culture dishes, and the cells were grown to confluence over 2-10 days in supplemented BME as described above. Coverslips were mounted into a chamber and washed three times with unsupplemented BME, and the attached cells were incubated with 5 /IM fura-2/AM in BME for 17 min at 37 "C. The cells were washed three times with Hanks' balanced salt solution (HBSS, containing 137 mM sodium chloride, 5.4 mM potassium chloride, 0.8 mM magnesium sulfate, 0.4 mM sodium phosphate dibasic, 0.4 mM potassium phosphate monobasic, 5.6 mM glucose, 1.3 mM calcium chloride, pH 7.2) prior to measurement of fura-2 fluorescence. Fura-2 loading under these conditions produced a diffuse fluorescence with no visible intracellular localization of the dye.
Epifluorescence microscopy (32) was used to monitor changes in the fura-2 fluorescence of the C62B cell monolayer. The chamber was mounted on an inverted Nikon Diaphot-TMD microscope, and a Fluor 40 X objective was used to examine 12-25 cells at a given time. Fura-2 fluorescence was recorded at 1-s intervals using a dual excitation wavelength spectrofluorometer (Spex Industries, Edison, NJ) with excitation at 340 and 380 nm and emission at 500 nm. All experiments were performed in either the Ca'+/MgZ+-containing HBSS described above or in HBSS without added Ca2+ or M%+ and with 0.5 mM EGTA. Experiments were conducted at room temperature since the Ca2+ response was found to be identical at room temperature and 37 "C, but the higher temperature greatly accelerated dye leakage. Cells were treated with acetylcholine (2.75 mM) or drugs by the addition of 4-10 X concentrates to obtain the desired final concentration. PDBu and verapamil were added as ethanolic solutions of HBSS 20-30 min prior to acetylcholine stimulation; solvent concentrations, which never exceeded 0.5%, had no effect on the fura-2monitored Ca2+ response. After stimulation, the cells were rinsed free of agonist by 10-12 media changes and were given 20 min to recover prior to restimulation. Media changes and additions of agonist or drugs were made without an interruption in recording. In some experiments, the initial acetylcholine-stimulated Ca'+ response varied slightly from all subsequent control responses, perhaps due to incomplete dye hydrolysis or a transient buffering effect (32,33); in such cases, the initial response was discarded.
The free intracellular Ca2+ concentration ([Ca*+],) was calculated from the ratio of the fluorescence intensities at the two excitation wavelengths as described by Grynkiewicz et al. (34) using a K d of 224 nM for the fura-2. Ca" complex. The minimum and maximum ratios were determined for the cells at the end of each experiment using 10 p~ ionomycin; R,,, was obtained by equilibration with 2.5 mM EGTA in a depolarizing medium (130 mM potassium chloride, 20 mM sodium chloride, 10 mM HEPES, 1 mM magnesium chloride, pH 7.2), and R,,, was then found by the readdition of Ca'+ in the same medium until the fura-2 signal reached saturation. The excitation spectrum of the cellular fura-2 under these conditions was similar to that of free fura-2 in the same solutions. Background autofluorescence, determined by the subsequent addition of 1 mM manganese sulfate to the permeabilized cells, was subtracted from the 340 and 380 nm traces prior to calculating [Ca'+];.
Labeling Cells with P'CIArachidonic Acid and rH]Znositol-Sterile glass scintillation vials were exposed to 1.5 ml of poly-D-lysine (10 pg/ml) for 20 min and allowed to air dry. C62B cells were seeded into the vials as described previously (29). The cultures were then incubated for 18-24 h, the medium was removed by aspiration, and cells were labeled by incubation for 24 h in 0.75 ml of fresh medium containing 0.2 pCi of [ 1-"Clarachidonic acid. In some cases, 25 pCi of my0-[2-~H]inositol was also included (31).
Assays for Arachidonic Acid Liberation and Inositol Phosphte Formation-These assays were conducted following protocols described previously (29,31). Briefly, the labeling medium was removed, the cultures were rinsed, fresh medium was added, and the cells were placed in a 37 "C water bath. BME containing 30 mM HEPES, pH 7.2, was used for arachidonate experiments, whereas double-label experiments were performed using CaZ+/Mg2f-containing HBSS (15 mM HEPES, pH 7.2) with 25 mM lithium chloride. Drugs were added in 50-pl aliquots to give the desired final concentrations in a total volume of 0.75 ml. PDBu and verapamil were made up in the appropriate medium and added as described above; solvent concentrations used had no effect on stimulated arachidonate liberation or inositol phosphate formation. Incubations were carried out in the 37 "C water bath for the times indicated and were terminated by the addition of 2.8 ml of chloroform/methanol/hydrochloric acid (v/v/v, 100:200:2) followed by agitation in a sonicating water bath. Cell extracts were transferred to centrifuge tubes, 0.9 ml of water and 0.9 ml of chloroform were added, and the tubes were agitated on a Vortex shaker. The tubes were kept overnight at 0 "C, and the phases were separated by centrifugation.
Separation of lipids was performed by evaporating the organic phases under a stream of nitrogen, redissolving the residues in 50 pl of chloroform, and applying 20-p1 portions to LKGD TLC plates. Arachidonic acid was separated from esterified lipids using a solvent system consisting of the upper phase of a mixture of ethyl acetate/ isooctane/acetic acid/water (v/v/v/v, 93:47:20:100) as reported previously (29). Radiolabeled species were visualized by autoradiography using Kodak XAR film and quantitated by scraping portions of the silica gel into scintillation vials followed by liquid scintillation spectrometry.
Inositol phosphates were separated in [3H]inositol-labeling experiments by anion exchange chromatography using HPLC or Dowex AG 1-X8 resin. For HPLC analysis (35), the aqueous phases were evaporated to 1 ml under a stream of nitrogen, diluted to 4 ml with water, and applied to a Partisil 10 SAX analytical HPLC column. The column was washed with 20 ml of water to remove free inositol, and inositol phosphates were eluted using a 70-min linear gradient of 0-1.5 M ammonium formate, pH 3.7, with a flow rate of 1 ml/min and 1-ml fractions collected. The radioactivity present in the eluent was quantitated by liquid scintillation spectrometry. Peaks of radioactivity were identified by coelution with radiolabeled standards. Dowex chromatography (36) was carried out using an ammonium formate step gradient as described previously (37).
Statistical Analysis-The data are presented as the mean k S.E. of the indicated number of experiments. Statistical differences between control and treated values were analyzed using a t test for Ca2+ experiments and a one-way ANOVA with Tukey's Studentized range test for arachidonate studies. The Role of Calcium in Phospholipase A , Activation within 10 s of agonist addition, consistent with the previously determined time course of 1,4,5-IPS formation (31). A rapid drop in [Ca'+],, which lasts about 1 min, is followed by a more gradual decline to a sustained plateau of approximately 100 nM, which can last at least 20 min in the continued presence of agonist (not shown). Addition of the muscarinic receptor antagonist atropine at any time rapidly returns [Ca'+], to basal levels (Fig. 1). Acetylcholine was used at a saturating concentration of 2.75 mM to facilitate comparison of results with previous biochemical studies (29,31,37,38). The identical dose-response curves for phospholipase C and phospholipase A? activation rule out the use of agonist concentration for separating the two responses (29).
Removing Extracellular Ca2+ Block the Second Phase of the [Ca"], Response without Affecting the Initial Peak-Acetylcholine-stimulated changes in [Ca"], were examined in Ca"free medium to dissect the contributions of intracellular and extracellular Ca" pools. Ca"-containing medium was aspirated away from fura-2-loaded C62B cells and replaced with Ca'+-free medium with 0.5 mM EGTA; agonist was then added within 30 s (Fig. 2). This produced a slight lowering of basal [Ca'+], but failed to affect the initial peak [Ca'+], response.  ( p < 0.02) without suhstantially affecting the sustained plateau levels (Fig. 3 ) . Peak increases in [Ca'+], were inhibited by 83 2 4 5 ( n = 5 ) in Ca')+-free medium and 67 k 11% ( n = 6) in 1.3 mM external Ca". The time to peak response was also delaved by PDHu pretreatment (Fig. 3), going from 6 2 1 to 19 t :i s in Ca"free medium ( n = 5) and from 10 f 1 to 18 2 5 s in the presence of external Ca2+ ( n = 6). Identical PDRu pretreatments on ['"Clarachidonate-labeled cells had no effect on The Role of Calcium in Phospholipase Az Activation   3 min (29). Switching C62B cells to Ca2+-free medium with 0.5 mM EGTA 30 s prior to agonist addition completely blocks stimulated arachidonate accumulation (Fig, 6, p < 0.001, n = 7 experiments performed in triplicate). If the inhibited cells are then switched back to medium containing 1.3 mM Ca2+ to restore ca2+ influx (Fig.  2), cholinergic arachidonate liberation is also recovered (Fig.  6). Arachidonate accumulation at 3 min under these conditions was not significantly different from control 90-s stimulated levels ( p > 0.5, n = 4 experiments performed in triplicate). Media switches alone had no effect on basal arachidonate levels.
The resulting [Ca2+Ii response resembles that obtained in the absence of extracellular Ca2+, with a return to basal levels within 2 min of agonist addition (Figs. 7 B and 2C). Pretreatment of ['4C]arachidonate-labeled cells under similar conditions produced a corresponding inhibition of acetylcholinestimulated arachidonate liberation (Fig. 8, p < 0.001).

Acetylcholine Produces a Biphasic [Ca"Ii Response-Mus-
carinic stimulation of C62B rat glioma cells produces a biphasic [Ca'+Ii response, a rapid peak in [Ca2+Ii followed by a sustained elevation, similar to that seen in other systems with receptors coupled to phospholipase C (3, 41). The two phases of the response have been linked to the mobilization of intracellular Caz+ stores and the opening of plasma membrane Ca2+ channels, respectively (3,6), and our data support this interpretation. A temporal comparison of acetylcholine-stimulated changes in [Ca"], and arachidonate accumulation does not support the idea that phospholipase A2 activity in C62B cells is regulated solely by changes in overall cytoplasmic Ca2+ levels. Although [Ca2+Ii peaks within 10 s of agonist stimulation and then declines to a sustained plateau level after about 1 min (Fig. I), arachidonate accumulation has been shown to increase slowly over the first min and peak at 60-90 s (29). T h e Role of Calcium in Phospholipase A2 Activation cells (Fig. 2) is consistent with changes in 1,4,5-IP3 levels determined previously (31). PDBu is an effective inhibitor of acetylcholine-stimulated inositol phosphate formation in C62B cells (37), and we therefore used PDBu to test the relationship among inositol phosphate formation, intracellular Ca2+ release, and phospholipase A2 activation. As in other systems (e.g. 43, 44), pretreatment with PDBu substantially inhibited the peak stimulated [Ca2+Ii response in C62B cells (Fig. 3). This result is probably a direct result of decreased 1,4,5-IP3 formation since submaximal doses of agonist or 1,4,5-IP3 have been shown to produce corresponding submaximal increases in [CaZ+Ii and only fractional release of accessible internal Ca2+ stores (42). It is also possible, however, that the phorbol ester may be affecting Ca2+ release at the level of the 1,4,5-IP3-operated channel (45). The second phase of the [Ca2+It response was not substantially affected by PDBu, suggesting that Ca" influx is far less sensitive than intracellular Ca2+ release to stimulated changes in inositol phosphate levels. This result may reflect regulation of Ca2+ influx by protein kinase C or G proteins (6, 46-48) rather than inositol phosphate metabolites (3,7). PDBu pretreatment also had no effect on stimulated arachidonate liberation under conditions that depressed inositol phosphate formation and internal Ca2+ release (Figs. 4 and 5), demonstrating that normal 1,4,5-IP3 production and [Ca2+Ii increases are not required for phospholipase A2 activation in the presence of normal Ca2+ influx. We cannot rule out the possibility, however, that a small amount of intracellular Ca2+ release or a sustained elevation in [Ca2+], is required for phospholipase A2 activation. Although PDBu has been found to increase basal and stimulated arachidonate liberation in other systems (17,18), it failed to alter these parameters significantly in C62B cells.
Ca2+ Influx Is Tightly Coupled to Phospholipase A2 Actiuation-Since PDBu did not block external Ca2+ influx or arachidonate liberation, it was possible that such influx provided the Ca2+ required for phospholipase A2 activation. This hypothesis was tested by assessing stimulated arachidonate liberation using a Ca2+-free medium protocol that was shown to inhibit the influx-dependent second phase of the [Ca2+Ii response without altering peak intracellular Ca2+ release (Fig.  2). Acetylcholine-stimulated arachidonate accumulation was totally inhibited under such conditions (Fig. 6), demonstrating that Ca'+ influx is required for phospholipase AS activation and, additionally, showing that normal intracellular Ca2+ release and the resulting large changes in [Ca2+], are not sufficient to support phospholipase A2 activation. Readdition of external Ca2+ to the inhibited cells restores not only Ca2+ influx and the sustained phase of the [Ca2+Ii response (Fig. 2) but also stimulated arachidonate accumulation (Fig. 6), reinforcing the conclusion that Ca2+ influx is a crucial event in phospholipase Az activation. In addition, this latter experiment suggests that Ca2+ influx may be involved in the timing of the transient arachidonate signal; delaying Ca2+ influx for 90 s after agonist stimulation results in arachidonate liberation at a time (3 min) when the normal response has ended (29).
Corroborative date on the importance of external Ca2+ in phospholipase Az activation were obtained using verapamil, an inhibitor of voltage-dependent Ca2+ channels (39,40). Verapamil pretreatment produced a stimulated [Ca2+IL response similar to that obtained with Ca2+-free medium; the influx-dependent phase of the response was blocked without affecting the peak intracellular Ca2+ release (Fig. 7). Once again, these alterations in the [Ca2+], response were associated with essentially quantitative inhibition of acetylcholine-stim-ulated arachidonate accumulation (Fig. 8). Although verapamil is assumed to be acting directly on the voltage-dependent Ca2+ channels in these experiments (39), we cannot rule out the possibility that the dose used affects receptor-operated channels (49) or indirectly affects the voltage-sensitive channels by interfering with Na' or K+ currents (40).
The present results can be explained by a model for phospholipase A2 regulation in which the enzyme is functionally associated with plasma membrane Ca2+ channels. Activation of phospholipase A2 may require Ca2+ influx in order to generate very high Ca2+ concentrations in the microenvironment of the enzyme, as suggested for regulation of neurotransmitter release (50). Evidence exists for localized membraneassociated increases in [Ca2+Ii following muscarinic stimulation of rat parotid acinar cells (51), and such a hypothesis would explain why many phospholipase A2 enzymes require millimolar Ca2+ concentrations for maximal in vitro activation (e.g. 13,25,26). Phospholipase AP may be permanently associated with the membrane near Ca2+ channels or may bind to the membrane as a consequence of Ca2+ influx during activation (52). The Ca2+ released into the cell from sequestered intracellular stores is assumed to be functionally uncoupled from the enzyme by virtue of its location.
This model is consistent with the observed phospholipase A2 requirement for extracellular Ca" in other systems (e.g. 11,17,18). In addition, it may help to explain the discrepancy between [Ca2+Ii changes and phospholipase A2 activation noted with different agonists in platelets (53), and the recent demonstration of temperature-dependent dissociation of bradykinin-stimulated 1,4,5-IP3 and lysophosphatidylinositol formation in endothelial cells (54). Although it does conflict with experiments showing inhibition of arachidonate liberation with 8-(diethylamino)-octyl-3,4,5-trimethoxybenzoic acid (11,28), an inhibitor of intracellular Ca2+ release ( 5 5 ) , recent studies have indicated that this compound may block phospholipase AS in a Ca2+-independent manner (19). Work by Balsinde et ~l .
(56) in human neutrophils has suggested that cell compartmentalization may play a role in phospholipase Az regulation, demonstrating preferential ionophoreinduced activation of the enzyme within an undefined intracellular region. Interestingly, this same study found that plasma membrane-associatedphospholipase Ai! used phosphatidylinositol preferentially as substrate, as does the C62B cell enzyme (31).
We would stress that this model does not preclude a coincident role for G proteins (10-12), protein kinase C (16)(17)(18)(19), lipocortins (20,21), Na+/H+ exchange (22), or other factors in phospholipase Az regulation. In fact, since Ca2+ influx does not desensitize in C62B cells following acetylcholine stimulation and arachidonate liberation does, one or more of these factors may be responsible for terminating the enzyme's action in this system. Our results strongly suggest, however, that Ca2+ influx is required for muscarinic receptor activation of phospholipase Az in these cells.