Phosphoinositide 3-kinase gamma mediates angiotensin II-induced stimulation of L-type calcium channels in vascular myocytes.

Previous results have shown that in rat portal vein myocytes the betagamma dimer of the G(13) protein transduces the angiotensin II-induced stimulation of calcium channels and increase in intracellular Ca(2+) concentration through activation of phosphoinositide 3-kinase (PI3K). In the present work we determined which class I PI3K isoforms were involved in this regulation. Western blot analysis indicated that rat portal vein myocytes expressed only PI3Kalpha and PI3Kgamma and no other class I PI3K isoforms. In the intracellular presence of an anti-p110gamma antibody infused by the patch clamp pipette, both angiotensin II- and Gbetagamma-mediated stimulation of Ca(2+) channel current were inhibited, whereas intracellular application of an anti-p110alpha antibody had no effect. The anti-PI3Kgamma antibody also inhibited the angiotensin II- and Gbetagamma-induced production of phosphatidylinositol 3,4,5-trisphosphate. In Indo-1 loaded cells, the angiotensin II-induced increase in [Ca(2+)](i) was inhibited by intracellular application of the anti-PI3Kgamma antibody, whereas the anti-PI3Kalpha antibody had no effect. The specificity of the anti-PI3Kgamma antibody used in functional experiments was ascertained by showing that this antibody did not recognize recombinant PI3Kalpha in Western blot experiments. Moreover, anti-PI3Kgamma antibody inhibited the stimulatory effect of intracellularly infused recombinant PI3Kgamma on Ca(2+) channel current without altering the effect of recombinant PI3Kalpha. Our results show that, although both PI3Kgamma and PI3Kalpha are expressed in vascular myocytes, the angiotensin II-induced stimulation of vascular L-type calcium channel and increase of [Ca(2+)](i) involves only the PI3Kgamma isoform.

Previous results have shown that in rat portal vein myocytes the ␤␥ dimer of the G 13 protein transduces the angiotensin II-induced stimulation of calcium channels and increase in intracellular Ca 2؉ concentration through activation of phosphoinositide 3-kinase (PI3K). In the present work we determined which class I PI3K isoforms were involved in this regulation. Western blot analysis indicated that rat portal vein myocytes expressed only PI3K␣ and PI3K␥ and no other class I PI3K isoforms. In the intracellular presence of an anti-p110␥ antibody infused by the patch clamp pipette, both angiotensin II-and G␤␥-mediated stimulation of Ca 2؉ channel current were inhibited, whereas intracellular application of an anti-p110␣ antibody had no effect. The anti-PI3K␥ antibody also inhibited the angiotensin IIand G␤␥-induced production of phosphatidylinositol 3,4,5-trisphosphate. In Indo-1 loaded cells, the angiotensin II-induced increase in [Ca 2؉ ] i was inhibited by intracellular application of the anti-PI3K␥ antibody, whereas the anti-PI3K␣ antibody had no effect. The specificity of the anti-PI3K␥ antibody used in functional experiments was ascertained by showing that this antibody did not recognize recombinant PI3K␣ in Western blot experiments. Moreover, anti-PI3K␥ antibody inhibited the stimulatory effect of intracellularly infused recombinant PI3K␥ on Ca 2؉ channel current without altering the effect of recombinant PI3K␣. Our results show that, although both PI3K␥ and PI3K␣ are expressed in vascular myocytes, the angiotensin II-induced stimulation of vascular L-type calcium channel and increase of [Ca 2؉ ] i involves only the PI3K␥ isoform.
Angiotensin II (AII) 1 receptors mediate numerous cellular functions in vascular smooth muscle cells including prolifera-tion, hypertrophy, differentiation, and regulation of vascular tone and calcium channels (1). AII have been reported to activate different pathways in vascular smooth muscles. In A7r5 smooth muscle cells, AII stimulates calcium channels via a tyrosine kinase and a undetermined phosphoinositide 3-kinase (PI3K) (2). In porcine coronary artery smooth muscle cells, AII activates the p85 subunit of PI3K via a tyrosine phosphorylation (3). In rat portal vein myocytes, AII stimulates voltagegated L-type calcium channel activity and thus calcium influx (4,5). The signaling pathway leading to this cellular response involves the G␣ 13 ␤ 1 ␥ 3 protein-coupled AT 1A receptor but does not involve tyrosine kinase. The ␤ 1 ␥ 3 dimer of the G 13 protein transduces the AII-induced stimulation of calcium channels through activation of PI3K and protein kinase C (PKC) (6,7).
PI3Ks are key enzymes involved in the regulation of multiple biological responses including regulation of ionic channel activity (7)(8)(9)(10). The primary enzymatic activity of PI3Ks is the phosphorylation of phosphoinositides at the 3Ј position of the inositol ring leading to the formation of second messengers. PI3Ks are grouped into three classes (I, II, III) depending on their substrate specificity. Class I phosphorylates PI, PI 4-P, PI 4,5-P 2 , whereas class II prefers PI and PI 4-P as substrates. Class III members phosphorylate only PI. Class I enzymes are heterodimers composed of a catalytic subunit p110␣, p110␤, p110␦, or p110␥ and a regulatory or adaptor subunit. The first three catalytic subunits are closely associated with a regulatory subunit (p85, p55, or p50), whereas p110␥ is associated with a p101 adaptor subunit (11). PI3K activities are regulated in vivo by different mechanisms. The class I p110␣, -␤, and -␦ have been reported to be regulated by receptors with intrinsic or associated tyrosine kinase activity or by small G proteins (12,13). In contrast, PI3K␥ is activated by G␤␥ released upon G protein-coupled receptor activation (14,15). By using recombinant G␤␥ dimers, the ␤ 1 ␥ 2 dimer of G proteins has been shown to be a more potent activator of PI3K␥ than the ␤ 5 ␥ 2 dimer (16). However, G protein-coupled receptors may stimulate a p85-dependent PI3K as described in human vascular smooth muscle cells (17), and PI3K␤ has also been shown to be activated by G␤␥ (15,18). Therefore, the identification of G␤␥dependent PI3Ks extends the range of intracellular processes that could be mediated by PI3K, although specific physiological processes mediated by PI3K␤ or PI3K␥ are not well described. Few reports indicate that specific isoform of PI3K can be involved in precise signaling pathway. PI3K␤ plays an important role in transmitting the mitogenic response induced by some but not all growth factors (19), whereas PI3K␥ regulates the activation of cationic current by acetylcholine in Xenopus ovocytes (20). Expression and distribution of the different PI3K isoforms in vascular smooth muscle cells are not well described, and the isoform of PI3K involved in AII-mediated response has not been identified.
The purpose of this study was to investigate which types of class I PI3K isoforms could be activated by AII to stimulate voltage-gated calcium channel activity. By using different antibodies, we identified the PI3K isoforms expressed in rat portal vein myocytes. The same antibodies were used in patch clamp experiments to identify the isoforms involved in the transduction pathway coupling AII receptors to stimulation of Ca 2ϩ channels. We have shown that rat portal vein myocytes express both PI3K␣ and -␥, but only PI3K␥ is involved in the regulation of L-type calcium channels by AII.

EXPERIMENTAL PROCEDURES
Cell Preparation-Isolated myocytes from rat portal vein were obtained by enzymatic dispersion as described previously (21). Cells were seeded at a density of about 10 3 cells/mm 2 on glass slides in physiological solution and used over the next 24 h.
Membrane Current and [Ca 2ϩ ] i Measurement-Voltage clamp and membrane current recordings were made with a standard patch clamp technique using a List EPC-7 patch clamp amplifier (Darmstadt-Eberstadt, Germany). Whole-cell recordings were performed with patch pipettes having resistances of 2-5 megaohms. Membrane potential and current records were stored and analyzed using a PC computer (Pclamp system, Axon Instruments, Foster City, CA). Cell capacitance was determined in each cell tested by imposing 10-mV hyperpolarizing and depolarizing steps from the holding potential (Ϫ40 mV) and analyzing the amplitude and the time course of the recorded currents. Current density was expressed as the maximum amplitude of the current/capacitance unit (pA/pF). The kinetic of inactivation of the current was calculated with a mono-exponential function (f(t) ϭ A exp(Ϫt/) ϩ C) according to the P-clamp system. Steady-state inactivation experiments were performed by a double pulse protocol. The holding potential was Ϫ80 mV; a test pulse to ϩ10 mV was preceded by a prepulse of 10 s duration and of variable amplitude (from Ϫ80 mV up to 10 mV). Steady-state inactivation curves were fitted with a Boltzman equation.
In some experiments, PI3K antibody and G␤␥ were added to the patch clamp pipette solution. As the rate of diffusion of PI3K antibody and G␤␥ complex from the pipette to cell may be different (because of the different molecular mass of the two proteins) and the kinetics of activation of PI3K by G␤␥ may also differ from the kinetics of inhibition of PI3K by the antibody, a sandwich technique was used to fill the patch clamp pipette. The tip of the patch clamp pipette, back-filled with a solution containing the antibody plus G␤␥ protein, was dipped in a control pipette solution containing only the antibody to allow an early diffusion of the antibody.
Measurements of [Ca 2ϩ ] i were performed with an indo-1 setup described elsewhere (5). Briefly, cells were preloaded with the membranepermeable indo-1 AM (1 M) for 30 min, and 50 M indo-1 was added to the pipette solution and entered the cells after establishment of the whole-cell recording mode. [Ca 2ϩ ] i was estimated from the 405/480 nm fluorescence ratio using a calibration determined within cells.
Solutions-The normal physiological solution contained (in mM): 130 NaCl, 5.6 KCl, 1 MgCl 2 , 2 CaCl 2 , 11 glucose, 10 HEPES, pH 7.4, with NaOH. The basic pipette solution contained (in mM): 130 CsCl, 10 EGTA, 5 ATP-Na 2 , 2 MgCl 2 10 HEPES, pH 7.3, with CsOH. For the recording of Ca 2ϩ channel currents with the patch clamp technique, 5 mM BaCl 2 was substituted for CaCl 2 in the physiological solution, and bovine serum albumin (0.1%) was added in the pipette solution to increase protein infusion. AII or PDBu was applied on the cells via pressure ejection from a glass pipette. All experiments were performed at 28 Ϯ 1°C.
Western Blot-Microsomal proteins or recombinant PI3K were treated with Laemmli sample buffer containing 5% ␤-mercaptoethanol, boiled for 10 min, and separated by SDS-polyacrylamide gel electrophoresis (10% separating gel with 4% stacking gel). The resolved proteins were transferred to a polyvinylidene difluoride (PVDF, Bio-Rad) membrane (1 h at 100 V). PVDF membrane was then blocked for 1 h with 3% bovine serum albumin in phosphate-buffered saline complemented with 0.1% Tween 20 (PBS-T) and incubated overnight with the primary antibody at 2 g/ml (anti-PI3K␣, anti-PI3K␤, anti-PI3K␦, or anti-PI3K␥ antibody). After extensive washes in PBS-T, membranes were incubated for 2 h with a 0.4 g/ml dilution of peroxidase-coupled anti-rabbit or anti-goat IgG in PBS-T complemented with 3% bovine serum albumin. Specific antigen detection was performed using the Bio-Rad Opti 4CN substrate kit. Gels were analyzed with KDS1D 2.0 software (Kodak Digital Science, Paris).
Statistics-All values are given as means Ϯ S.E. Student's t test was performed to estimate the significance of the differences between mean values. A value of p Ͻ 0.05 was considered significant.

PI3K Isoforms Expressed in Rat Portal
Vein Myocytes-In rat portal vein myocytes, only two PI3K isoforms (PI3K␣ and PI3K␥) were detected by Western blot experiments on crude extract of myocyte membranes (Fig. 1). The antibody raised against p110␥ (PI3K␥ Ab) recognized a single band of 120 kDa, and the antibody raised against p110␣ (PI3K␣ Ab) detected a single band of 110 kDa (Fig. 1A). The two other class I PI3K isoforms (i.e. p110␤ and p110␦) were not detected in rat portal vein myocytes (Fig. 1A). It is noteworthy that the anti-p110␤ and anti-p110␦ antibodies recognized the recombinant PI3K␤ and PI3K␦, respectively (Fig. 1B). Identification of PI3K isoforms by immunochemistry on 1 day-cultured myocytes led to the same results (data not shown).
Effects of Anti-PI3K Antibodies on Ba 2ϩ Current Stimulated by Recombinant PI3K Isoforms-All experiments and measurements of the current were performed at least 5 min after the break-through into the whole-cell patch clamp configuration to allow intracellular diffusion of the pipette solution and stabilization of the Ba 2ϩ current. Ba 2ϩ current densities were calculated for a step depolarization of ϩ10 mV from a holding potential of Ϫ40 mV. The mean capacitance of the cells was 16.30 Ϯ 0.52 pF (n ϭ 160).
Intracellular application of PI3K␥ Ab (alone or in the presence of its antigenic peptide) or PI3K␣ Ab did not change the density of the basal Ba 2ϩ current (Fig. 2, A and B). Both PI3K␥ Ab and PI3K␣ Ab alter neither the kinetics of inactivation (98 Ϯ 9 ms, n ϭ 8 and 94 Ϯ 12 ms, n ϭ 6, respectively), when compared with control conditions (105 Ϯ 11 ms, n ϭ 18, Fig.  2A), nor the current-voltage relationship of the Ba 2ϩ current (data not shown). Furthermore, the steady-state inactivation of the Ba 2ϩ current was not different in the presence or absence of PI3K␥ Ab (half-inactivation potential: Ϫ30.2 Ϯ 0.7 mV, n ϭ 5 instead of Ϫ29.9 Ϯ 0.7 mV, n ϭ 6, in control condition; data not shown).
The efficacy of the different antibodies was tested on the stimulatory effects of recombinant PI3Ks on L-type Ca 2ϩ channels. PI3K␣ or ␥ antibodies were preincubated for 45 min before patch clamp recording with intracellular solution containing the recombinant PI3Ks. As shown in Fig. 3A, the Ba 2ϩ current density was increased from 7.2 Ϯ 1.2 pA/pF (n ϭ 7) in control conditions to 12.6 Ϯ 1.1 pA/pF (n ϭ 8) in the intracellular presence of 1 nM PI3K␥ (p110␥ catalytic subunit associated with p101 subunit). When PI3K␥ was preincubated with the PI3K␥ Ab, the current density was similar to that obtained in control cells (8.5 Ϯ 1.1 pA/pF, n ϭ 11; Fig. 3A). The inhibitory effect of the antibody was specific of the immune reaction since inactivation of the PI3K␥ Ab, obtained by preincubation in the presence of the antigenic peptide or by boiling the antibody, did not statistically change the Ba 2ϩ current density (12.0 Ϯ 1.1 pA/pF, n ϭ 7) compared with the Ba 2ϩ current density obtained in the presence of the enzyme alone (12.6 Ϯ 1.1 pA/pF, n ϭ 8, Fig. 3A).
Intracellular perfusion of PI3K␣ also induced an increase in the current density (11.3 Ϯ 1.3 pA/pF, n ϭ 6 versus 6.5 Ϯ 1.2 pA/pF, n ϭ 10 in control conditions, Fig. 3B) similar to that induced by PI3K␥. The current density returned to control value when the recombinant PI3K␣ was preincubated with the PI3K␣ Ab (7.0 Ϯ 0.8 pA/pF, n ϭ 5, Fig. 3B). By contrast, the PI3K␣-induced increase in current density was not altered when this enzyme was preincubated with the PI3K␥ Ab (10.9 Ϯ 0.5 pA/pF, n ϭ 5, Fig. 3B). The specificity of the PI3K␥ Ab used in functional experiments was also tested by checking its ability to cross-react with recombinant PI3K␣ in Western blot experiments. Recombinant PI3K␣ was recognized by the PI3K␣ Ab but not by the PI3K␥ Ab, whereas recombinant PI3K␥ was recognized by the PI3K␥ Ab but not by the PI3K␣ Ab (Fig. 3C). These results indicate that the antibodies can specifically recognize the different isoforms of PI3Ks and inhibit their effects on calcium channels.
Effects of Anti-PI3K Antibodies on Ba 2ϩ Current Stimulated by AII, ␤␥ Subunits of G Protein, and Phorbol Ester-The effects of AII on Ba 2ϩ currents were expressed as a percentage of stimulation rather than in variation of current density because cells were their own control, and therefore the capacitance of the cells did not influence the results. AII was applied for 5 min after the rupture of the patch, and the peak stimulatory effect was obtained 1-2 min after the beginning of AII perfusion. In control conditions, AII (0.1 M) stimulated the peak Ba 2ϩ current by 30.4 Ϯ 3.5% (n ϭ 17, Fig. 4, A and B), as reported previously (4). In the intracellular presence of PI3K␥

FIG. 2. Effects of anti-PI3K antibodies on basal Ba 2؉ currents.
A, typical Ba 2ϩ currents elicited by depolarizations to ϩ10 mV from a holding potential of Ϫ40 mV in control conditions and in the intracellular presence of 10 g/ml PI3K␥ Ab or PI3K␣ Ab as indicated. B, Ba 2ϩ current densities measured with control pipette solution (Control) and with pipette solution containing 10 g/ml PI3K␥ Ab or PI3K␣ Ab alone or in the presence of the antigenic peptide (10 g/ml). Ba 2ϩ currents were measured 4 -5 min after breakthrough into the whole-cell recording mode. Data are given as the means Ϯ S.E. with the number of experiments in parentheses.
Ab, the stimulation of the Ba 2ϩ current by AII was suppressed (2.6 Ϯ 2%, n ϭ 9, Fig. 4, A and B). This inhibition was linked to the immune function of the antibody as the infusion of the boiled antibody or the antibody preincubated with the antigenic peptide did not prevent the stimulation of the Ba 2ϩ current by AII (respectively, 29.7 Ϯ 1.9%, n ϭ 4, data not shown and 28.4 Ϯ 1.1%, n ϭ 3, Fig. 4B). In contrast, in the intracellular presence of PI3K␣ Ab, the AII-induced increase in current amplitude (27.3 Ϯ 3.8%, n ϭ 6) was similar to that obtained in control conditions (Fig. 4, A and B).
Previous results have shown that AII mobilizes G␤␥ to transduce the signal leading to the stimulation of the Ba 2ϩ currents through PI3K and PKC (6, 7). Intracellular infusion of purified G␤␥ protein (0.1 M) from bovine brain mimicked the effect of AII and stimulated the Ba 2ϩ current density from 7.1 Ϯ 0.9 pA/pF (n ϭ 15) in control conditions to 11.8 Ϯ 2.7 pA/pF (n ϭ 10; Fig. 5A). This effect was maximal within 5 min. In the presence of PI3K␥ Ab, the G␤␥ protein was ineffective, as the current density was not statistically different from the current density obtained in control conditions (Fig. 5A). In contrast, intracellular application of the PI3K␣ Ab had no effect on the G␤␥-induced stimulation of the Ba 2ϩ current density (Fig. 5A).
PKC has been suggested to be a downstream effector of PI3K in the AII-induced stimulation of Ca 2ϩ channel pathway (7). Because phorbol esters also stimulate the Ca 2ϩ channels by directly activating PKC (21), we evaluated the effect of PI3K␥ Ab on the PDBu-induced increase in Ba 2ϩ current. In the intracellular presence of either PI3K␥ Ab or PI3K␣ Ab, PDBu (0.1 M) induced a stimulation of the Ba 2ϩ current of 32.0 Ϯ 8.0% (n ϭ 7) and 37.3 Ϯ 5.2%, respectively, which was not statistically different from the stimulation in the absence of the antibody (34.8 Ϯ 5.7%, n ϭ 7, Fig. 5, B and C). Similar results were obtained with phorbol 12-myristate 13-acetate (0.1 M) instead of PDBu (data not shown). The inactive isoform of PDBu (4␣ PDBu) did not stimulate the Ba 2ϩ current in the same batches of cells that were stimulated by PDBu (data not shown). These results suggest that the stimulatory effects of AII and G␤␥ on L-type Ca 2ϩ channels are mediated via activation of the PI3K␥, which appears to be the upstream effector of PKC.

FIG. 3. Specific effects of anti-PI3K antibodies on Ba 2؉ current stimulated by recombinant PI3K isoforms.
A, effects of PI3K␥ Ab on Ba 2ϩ current stimulated by PI3K␥. Ba 2ϩ current density in control conditions (Control) and after intracellular application of PI3K␥ (1 nM) alone or together with PI3K␥ Ab (10 g/ml) or with PI3K␥ Ab and its specific antigenic peptide (10 g/ml). B, Ba 2ϩ current density in control conditions (Control) and after intracellular application of PI3K␣ (1 nM) alone or together with PI3K␣ Ab (10 g/ml) or PI3K␥ Ab (10 g/ml). Ba 2ϩ currents were elicited by depolarizations to ϩ10 mV from a holding potential of Ϫ40 mV, and peak amplitude was

Effects of Anti-PI3K␥ Antibody on Intracellular Effects of AII:
Increase in Intracellular Calcium Concentration and Production of PI 3,4,5-P 3 -In venous myocytes, AII has been reported to promote a G␤␥-mediated increase in [Ca 2ϩ ] i depending on Ca 2ϩ influx through L-type Ca 2ϩ channels without involving the inositol 1,4,5-trisphosphate receptor (5,6). The role of PI3K␥ in this cellular response was examined by using the anti-PI3K␥ antibody in [Ca 2ϩ ] i measurements experiments. Application of 0.1 M AII on single myocytes held at Ϫ50 mV induced a peak increase in [Ca 2ϩ ] i of 98.5 Ϯ 11.0 nM (n ϭ 6, Fig.  6A). When the PI3K␥ Ab was added to the pipette solution, the peak Ca 2ϩ response activated by AII was strongly inhibited (27.5 Ϯ 7.9 nM, n ϭ 5), whereas it was not affected in the presence of the PI3K␣ Ab (Fig. 6A) or when the PI3K␥ Ab was preincubated with the antigenic peptide (data not shown). To test the possible alteration of ryanodine receptors by PI3K␥ Ab, caffeine, a pharmacological activator of ryanodine receptors, was tested on cells held at Ϫ50 mV. In the presence of the PI3K␥ Ab, caffeine (10 mM) induced an increase in [Ca 2ϩ ] i that was not significantly different from that obtained in the absence of the antibody (data not shown). These results suggest that the AII-induced increase in intracellular Ca 2ϩ concentration depends on the PI3K␥-mediated stimulation of L-type Ca 2ϩ channels.
To check the ability of AII and G␤␥ to stimulate endogenous PI3K␥ in venous myocytes, we measured PI 3,4,5-P 3 production in the absence and presence of PI3K␥ Ab. In freshly prepared microsomes of vascular myocytes, AII (0.1 M) or G␤␥ (200 nM) applied for 15 s induced an increase in PI3K activity of 25.1 Ϯ 3.9% (n ϭ 3) and 76.1 Ϯ 1.4%, respectively (Fig. 6B). In the presence of LY294002, an inhibitor of PI3K, AII and G␤␥ did not significantly change the basal activity of PI3K (Fig. 6B). The stimulations of PI 3,4,5-P 3 production by AII and G␤␥ were strongly inhibited when microsomes were preincubated with the PI3K␥ Ab. These effects were largely reversed when the immunological activity of PI3K␥ Ab was blocked by the antigenic peptide (10 g/ml; data not shown). These results indicate that AII and G␤␥ stimulate PI 3,4,5-P 3 production, which might be responsible for Ca 2ϩ channel stimulation. DISCUSSION Our results demonstrate the specific involvement of the ␥ isoform of the PI3K in the transduction pathway leading to calcium channel stimulation and the rise of [Ca 2ϩ ] i induced by AII in rat portal vein myocytes. This conclusion is based on experiments using specific antibodies raised against different PI3K isoforms.
Western blot analysis revealed that rat portal vein myocytes expressed only PI3K␣ and PI3K␥ and no other class I PI3K isoforms. Whereas PI3K␣ has been shown to be widely expressed, PI3K␥ has been described first in neutrophils and exocrine glands (25). Recently, expression of PI3K␥ has been reported in vascular endothelial cells (26), and we have now revealed its expression in vascular myocytes. PI3K␣ and PI3K␤ have been reported to be ubiquitously and similarly expressed, because many tissues such as porcine aortic endothelial cells, fibroblasts, and colon cancer cells express both isoforms (19,27,28). In contrast, in vascular myocytes, the ␣ but not the ␤ isoform was detected either by immunochemistry or by Western blot experiments. Therefore, rat portal vein myocytes seem to display a specific pattern of PI3K isoform expression when compared with PC12 neurons, which express all class I PI3Ks (29).
Basically, the two PI3K isoforms expressed in portal vein myocytes could be involved in the regulation of calcium channel by AII. Indeed, AII acts on G protein-coupled receptors and generates free G␤␥ dimers (6), but AII is also able to stimulate tyrosine phosphorylation (30). The present study shows that only intracellular application of PI3K␥ Ab inhibited the AIIinduced stimulation of calcium channel. This inhibition is specific, as this effect is reversed when the antibody is boiled or preincubated with the antigenic peptide used to produce the antibody. Furthermore, the PI3K␥ Ab recognized only one band in Western blot experiments performed on crude extract of myocytes and did not recognize recombinant PI3K␣. The specificity of the antibody was also shown at a functional level because PI3K␥ Ab inhibited the stimulatory effect of recombinant PI3K␥ on Ca 2ϩ channel but did not inhibit the stimulatory effects of recombinant PI3K␣ and phorbol ester. Involvement of a G␤␥-activated PI3K in the transduction pathway of AII has been suggested previously by Viard et al. (7). Here, we have clearly shown that both G␤␥ dimer-and AII-induced Ca 2ϩ channel stimulation are blocked by the anti-PI3K␥ antibody, whereas anti-PI3K␣ antibody did not affect the responses. Maier et al. (15) have reported that recombinant G␤ 1 ␥ 2 dimer is able to activate PI3K␥ in vitro. As G␤ 1 ␥ 3 has been identified as the dimer involved in AII-induced stimulation of Ca 2ϩ channels (6), we suggest that G␤ 1 ␥ 3 is also able to stimulate PI3K␥. The specificity of the anti-PI3K␥ antibody in inhibiting AII-induced regulation of Ca 2ϩ channel, although exogenous PI3K␣ and ␥ are both able to stimulate calcium channels, suggests that there is a specific pathway linking G protein-coupled receptor, endogenous PI3K␥ and Ca 2ϩ channels. This specific coupling between a receptor and a PI3K isoform has also been described by Hooshmand-Rad et al. (27) who reported that, in porcine aortic endothelial cells, platelet-derived growth factor stimulates actin reorganization through PI3K␣, whereas insulin reorganizes actin through PI3K␤.
Stimulation of Ca 2ϩ channel by AII induces an increase in intracellular calcium concentration that mediates vascular contraction. This rise in intracellular calcium concentration is due to a calcium entry through L-type Ca 2ϩ channels and a subsequent calcium release through ryanodine receptor channels (5). We have shown here that the PI3K␥ Ab inhibited the AII-induced increase in cytosolic calcium concentration. The effect of the antibody is not linked to an inhibition of ryanodine receptors or to a depletion of the intracellular Ca 2ϩ stores because responses to caffeine, known to increase the open probability of ryanodine receptors, were not altered.
In several vascular and nonvascular tissues, stimulation of ionic channels by PI3K via a tyrosine kinase-dependent pathway has been described. However, in our experiments, the PI3K␣ Ab, which specifically recognizes PI3K␣ in Western blot experiments and inhibits the stimulatory effect of recombinant PI3K␣ on Ca 2ϩ channels, did not affect either the AII-mediated stimulation of Ca 2ϩ channel or the rise in [Ca 2ϩ ] i . These results suggest that the tyrosine kinase-activable PI3K␣ is not involved in AII-stimulation of Ca 2ϩ channel in our model. These results are consistent with our previous results (7) but differ from those of Seki et al. (2), which indicate that in a vascular cell line derived from embryonic rat aorta (A7r5), AII stimulates Ca 2ϩ channels via a tyrosine kinase. In neuron, L-and N-type Ca 2ϩ channels are also stimulated by a tyrosine kinaseactivated PI3K (31). Moreover, we report in the present study that high concentrations of recombinant PI3K␣ may stimulate vascular calcium channels to the same extent that recombinant PI3K␥. These results suggest that, depending on the cell type and the mediator, different endogenous PI3K isoforms may modulate Ca 2ϩ channels. Concerning the potential role of PI3K␣ in vascular myocytes, one may speculate that it could be involved in the long-lasting effects of AII. Indeed, in porcine coronary artery, AII induces tyrosine phosphorylation and translocation of p85 within 15 min (3). However, further investigations are required to study the role of PI3K␣ in rat portal vein myocytes. Because measurement of lipid kinase activity revealed that both AII-and G␤␥-stimulated PI 3,4,5-P 3 productions were inhibited by LY294002 and the PI3K␥ Ab and because these inhibitory effects were reversed when PI3K␥ Ab was inactivated by the antigenic peptide, we conclude that AII and G␤␥ increase the lipid-kinase activity of PI3K␥ in venous myocytes. The larger effect of G␤␥ compared with that of AII in stimulating PI 3,4,5-P 3 production might be related to a larger concentration of exogenous G␤␥ used when compared with the concentration of G␤␥ endogenously produced upon AT 1 receptor activation. This difference is correlated with a larger Ca 2ϩ channel current stimulation by exogenous G␤␥ than by AT 1 receptor activation. PI 3,4,5-P 3 messenger has been shown to directly activate or modulate ion channels (10,20) or to activate enzymes such as Akt/protein kinase B MAP kinase, p70 s6 kinase and PKC (10,32). A MAP kinase inhibitor (PD98059) did not prevent the PI3K␥-dependent stimulation of calcium channels, 2 whereas PKC inhibitors inhibit the effects of AII, G␤␥, and PI3K on Ca 2ϩ channels in rat portal vein myocytes (4,7). Because PI 3,4,5-P 3 may activate atypical PKCs (20,33,34), it can be postulated that atypical PKCs are involved in the regulation of Ca 2ϩ channel by AII. However, the involvement of the serine-kinase activity of PI3K (35) in the AII-induced stimulation of Ca 2ϩ channel cannot be excluded.
In conclusion, the present study shows that despite the expression of both PI3K␣ and PI3K␥ in rat portal vein myocytes, the AII-induced stimulation of L-type Ca 2ϩ channels, and the rise of [Ca 2ϩ ] i specifically involves the PI3K␥.