Protein Kinase C (cid:1) Phosphorylates the TRPC1 Channel and Regulates Store-operated Ca 2 (cid:2) Entry in Endothelial Cells*

The TRPC1 (transient receptor potential canonical-1) channel is a constituent of the nonselective cation channel that mediates Ca 2 (cid:2) entry through store-operated channels (SOCs) in human endothelial cells. We investi-gated the role of protein kinase C (cid:1) (PKC (cid:1) ) phosphorylation of TRPC1 in regulating the opening of SOCs. Thrombin or thapsigargin added to the external medium activated Ca 2 (cid:2) entry after Ca 2 (cid:2) store depletion, which we monitored by changes in cellular Fura 2 fluorescence. Internal application of the metabolism-resistant analog of inositol 1,4,5-trisphosphate (IP 3 ) activated an inward cationic current within 1 min, which we recorded using the whole cell patch clamp technique. La 3 (cid:2) or Gd 3 (cid:2) abolished the current, consist-ent with the known properties of SOCs. Pharmacological (Go¨6976) or genetic (kinase-defective mutant) inhibition of PKC (cid:1) markedly inhibited IP 3 -induced activation of the current. Thrombin or thapsigargin also activated La 3 (cid:2) -sensitive Ca 2 (cid:2) entry in a PKC (cid:1) -de-pendent manner. We determined the effects of a specific antibody directed against an extracellular epitope of TRPC1 to address the functional importance of TRPC1. External application of the antibody blocked thrombin- or IP 3 -induced Ca 2 (cid:2) entry. In addition, we showed that thrombin or thapsigargin induced phosphorylation of TRPC1 within 1 min. Thrombin failed to induce TRPC1 phosphorylation in the absence of PKC (cid:1) activation. Phosphorylation of TRPC1 and the resulting Ca 2 (cid:2) entry were essential for the increase in permeability induced by thrombin in confluent endothelial monolayers. These results demonstrate that PKC (cid:1) phosphorylation of TRPC1 is an important determinant of Ca 2 (cid:2) entry in human endothelial cells. munoprecipitation with anti-Myc Ab and Western blotting with anti- Myc and anti-TRPC1 Both antibodies detected protein at kDa in TRPC1-transfected confirming the specificity of anti- TRPC1 antibody as described 37, 38). Transendothelial Electrical Resistance Measurement— The time course of endothelial cell retraction, a measure of increased endothelial permeability, was measured using established protocols (35, 39). For transfection of the Myc-tagged TRPC1 construct, HMECs grown on gelatin-coated gold electrodes were transfected with 1 (cid:2) g/ml Myc- TRPC1 cDNA (TRPC1-overexpressing cells) or vector alone (control cells) using LipofectAMINE Plus reagent as described (24). The DNA/ LipofectAMINE Plus mixture was incubated with cells fo r 4 h inserum-free Dulbecco’s modified Eagle’s medium, followed by addition of com- plete growth medium to the cells. After 48–72 h of transfection, the cells were used for experiments. Before measuring the transendothelial elec- trical resistance, cells were placed in reduced serum medium (1% fetal bovine serum) for 1 h. Cells were then incubated with or without Go¨6976 for 1 h, after which they were stimulated with thrombin to determine the changes in electrical resistance of the endothelial mono- layer. component

In vascular endothelial cells, thrombin binding to proteaseactivated receptor-1 and cleavage of the receptor at the N terminus leads to activation of the heterotrimeric G-protein G q and a subsequent increase in [Ca 2ϩ ] i (1,2). Increased [Ca 2ϩ ] i results from the release of Ca 2ϩ from intracellular stores (endoplasmic reticulum (ER) 1 ) through inositol 1,4,5-trisphos-phate (IP 3 )-sensitive channels and activation of plasma membrane Ca 2ϩ entry store-operated channels (SOCs) as a result of Ca 2ϩ store depletion (3)(4)(5)(6)(7). Store-operated Ca 2ϩ entry coupled with depletion of intracellular Ca 2ϩ stores is essential for Ca 2ϩ signaling by thrombin because it induces endothelial cell retraction and thus leads to increased endothelial permeability (8 -11).
Protein kinase C␣ (PKC␣) activation by GPCRs is known to induce increased endothelial permeability (9,11,25,26). Although this effect of PKC␣ may be the result of activation of cellular Ca 2ϩ entry, the role of PKC␣ in the regulation of SOCs is unclear, as the results of studies in several cell types have been contradictory (27)(28)(29)(30)(31)(32)(33). Thus, we addressed the role of PKC␣ in signaling TRPC1 activation and thereby triggering Ca 2ϩ entry and endothelial barrier dysfunction. Our results demonstrate that PKC␣-dependent phosphorylation of TRPC1 is an important signal contributing to SOC activation and triggering Ca 2ϩ entry. We also show that PKC␣-activated Ca 2ϩ entry is critical in mediating the thrombin-induced increase in transendothelial permeability.

EXPERIMENTAL PROCEDURES
Materials-Human ␣-thrombin was obtained from Enzyme Research Laboratories (South Bend, IN). Human umbilical vein endothelial cells (HUVECs) and endothelial growth medium (EBM-2) were obtained from Clonetics Corp. (San Diego, CA). Human microvessel endothelial cells (HMECs), a human dermal microvascular endothelial cell line, were obtained from Dr. Edwin W. Ades (National Center for Infectious Diseases, Center for Disease Control, Atlanta, GA). Trypsin and Hanks' balanced salt solution and molecular cellular and developmental biology (MCDB) media 131 were obtained from Invitrogen. Superfect transfection reagent was obtained from QIAGEN Inc. (Valencia, CA). The "metabolism-resistant" form of IP 3 , 3-deoxy-3-fluoro-D-myo-IP 3 , was from Calbiochem, and anti-TRPC1 polyclonal antibody was purchased from Alomone Laboratory (Jerusalem, Israel) or Sigma. Green fluorescent protein was purchased from Clontech. Endothelial Cell Cultures-Primary HUVECs were cultured in a T-75 flask coated with 0.1% gelatin in EBM-2 medium supplemented with 10% fetal bovine serum. HMECs were cultured in a T-75 flask in MCDB 131 medium supplemented with 10% fetal bovine serum. Cells were maintained at 37°C in a humidified atmosphere of 5% CO 2 and 95% air until they formed a confluent monolayer. Cells from each of the primary flasks were detached with 0.05% trypsin plus 0.02% EDTA, resuspended in fresh culture medium, and passaged as described below. In all experiments, unless otherwise indicated, a confluent monolayer of HUVECs was washed twice with phosphate-buffered saline and incubated in serum-free medium for 2 h before treatment with thrombin. HUVECs between passages 4 and 8 were used for all experiments as described (34,35).
Adenoviral Infection-The adenoviral construct containing a kinasedefective mutant of PKC␣ (referred to herein as Ad-dnPKC␣) was a generous gift from Dr. Viswanathan Natarajan (Johns Hopkins University School of Medicine, Baltimore, MD). Cells grown to confluence in 100-mm dishes (for determination of phosphorylation) or on 25-mm coverslips (for [Ca 2ϩ ] i measurements) were infected with 1 ϫ 10 7 plaque-forming units of the PKC␣ construct for 5 h in serum-containing medium, after which the medium was replaced with fresh medium supplemented with 10% fetal bovine serum. In parallel, monolayers were infected with adenovirus containing the ␤-galactosidase gene, lacZ (Ad-lacZ), as a control for nonspecific viral infection upon endothelial cell signaling (34). Adenovirus-infected cells were then used to determine TRPC1 phosphorylation or [Ca 2ϩ ] i as described below.
Whole Cell Patch Clamp Recording in Endothelial Cells-Patch clamp recording in the whole cell configuration was performed as described (20,35,36) on HMECs attached to a coverslip. Standard extracellular and pipette solutions (35) were provided. SOC currents were measured at a holding potential of Ϫ50 mV by depleting the cellular calcium store with 30 mol/liter 3-deoxy-3-fluoro-D-myo-IP 3 (metabolism-resistant IP 3 ) and/or 2 M thapsigargin included in the pipette solution. All experiments were performed at room temperature.
[Ca 2ϩ ] i Measurements-An increase in [Ca 2ϩ ] i was measured using the Ca 2ϩ -sensitive fluorescent dye Fura 2/AM as described (8,35). For loading of cells with Fura 2/AM, cells grown on 25-mm coverslips were incubated with 3 M Fura 2 for 15 min at 37°C. Cells were then washed twice with Hanks' balanced salt solution and imaged using an Attoflor Ratio Vision digital fluorescence microscopy system (Atto Instruments, Inc., Rockville, MD) equipped with a Zeiss Axiovert S100 inverted microscope and an F-Fluar ϫ40 1.3-numerical aperture oil immersion objective. Regions of interest in individual cells were marked and excited at 334 and 380 nm with emission at 520 nm. The 334/380 nm excitation ratio, which increased as a function of [Ca 2ϩ ] i , was captured at 5-s intervals.
Immunoprecipitation and Phosphorylation of TRPC1-Cells grown to confluence in 100-mm dishes were serum-starved in phosphate-free medium for 2 h at 37°C, after which cellular proteins were labeled by incubation with 150 Ci/ml 32 P for 4 h. Following incubation with 32 P, cells were stimulated with 50 nM thrombin or 2 M thapsigargin at the indicated time intervals and lysed for 20 min on ice with 0.5 ml of radioimmune precipitation assay buffer (10 mM Tris, 150 mM NaCl, 1 mM EDTA, 1 mM EGTA, 0.5% Nonidet P-40, 1% Triton X, 1 mM sodium orthovanadate, 1 mM phenylmethylsulfonyl fluoride, 1 g/ml leupeptin, 1 g/ml pepstatin A, and 1 g/ml aprotinin) (34,35). Cell lysates were immunoprecipitated with anti-IgG or anti-TRPC1 antibody (Ab) and Western-blotted with anti-TRPC1 antibody (2 g/ml) to verify equal protein loading in each lane. To determine the specificity of anti-TRPC1 Ab, COS-7 cells transfected with Myc-tagged TRPC1 or vector alone were lysed in radioimmune precipitation assay buffer, followed by im- munoprecipitation with anti-Myc Ab and Western blotting with anti-Myc and anti-TRPC1 antibodies. Both antibodies detected protein at ϳ90 kDa in TRPC1-transfected cells, confirming the specificity of anti-TRPC1 antibody as described (35,37,38).
Transendothelial Electrical Resistance Measurement-The time course of endothelial cell retraction, a measure of increased endothelial permeability, was measured using established protocols (35,39). For transfection of the Myc-tagged TRPC1 construct, HMECs grown on gelatin-coated gold electrodes were transfected with 1 g/ml Myc-TRPC1 cDNA (TRPC1-overexpressing cells) or vector alone (control cells) using LipofectAMINE Plus reagent as described (24). The DNA/ LipofectAMINE Plus mixture was incubated with cells for 4 h in serumfree Dulbecco's modified Eagle's medium, followed by addition of complete growth medium to the cells. After 48 -72 h of transfection, the cells were used for experiments. Before measuring the transendothelial electrical resistance, cells were placed in reduced serum medium (1% fetal bovine serum) for 1 h. Cells were then incubated with or without Gö6976 for 1 h, after which they were stimulated with thrombin to determine the changes in electrical resistance of the endothelial monolayer. Data are presented as the change in the resistive (in-phase) component of the impedance normalized to its initial value at zero time.
Data Analysis-Statistical comparisons were made using unpaired Student's t test. Statistical differences of p Ͻ 0.05 were considered significant.

Regulation of Store-operated Ca 2ϩ Entry by PKC␣
Thrombin-induced Ca 2ϩ Entry-In Fura 2-loaded endothelial cells, the thrombin response consists of an immediate but transient rise in intracellular Ca 2ϩ due to release of sequestered Ca 2ϩ from the ER, followed by a secondary rise in intracellular Ca 2ϩ due to Ca 2ϩ entry. We experimentally separated the two phases (Fig. 1A) using a Ca 2ϩ add-back protocol. When administered under Ca 2ϩ -free bath conditions, thrombin elicited only the Ca 2ϩ transient due to release of sequestered Ca 2ϩ into the cytoplasm. A more sustained Ca 2 ϩ signal appeared subsequently when extracellular Ca 2ϩ was restored to the bath medium, signifying entry of extracellular Ca 2ϩ into cells. To identify the Ca 2ϩ entry pathway, we used the trivalent cation La 3ϩ or Gd 3ϩ , either of which is a known blocker of SOCs (11). Addition of La 3ϩ or Gd 3ϩ rapidly abolished the Ca 2ϩ entry phase (Fig. 1, A and B), thus indicating that Ca 2ϩ entry is mediated via SOCs. To exclude the possibility that the Ca 2ϩ entry was not the result of background or tonic Ca 2ϩ entry, we showed that there was no Ca 2ϩ entry following repletion of Ca 2ϩ in the medium in the absence of store depletion (Fig. 1A, dotted line). In another series, we observed the blocking effect of a specific inhibitor of PKC␣ on Ca 2ϩ entry via SOCs (Fig. 1C). The marked inhibitory action of Gö6976 on Ca 2ϩ entry was evident because the Ca 2ϩ release from ER stores was equivalent in control and treated cells, i.e. Ca 2ϩ store depletion was the same under both conditions. Expression of dnPKC␣ (a kinase-defective mutant of PKC␣) also resulted in a ; I, data shown as the mean Ϯ S.E. of maximum steady-state current density (pA/picofarad (pF)) above the basal level under the conditions indicated (n ϭ 4 -7). *, statistically significant reduction (p Ͻ 0.001) in the IP 3 -induced current after treatment with Gö6976 versus the control (Cont; IP 3 ); #, significant reduction (p Ͻ 0.001) in the SOC current in dnPKC␣-transfected cells versus the vector (pCMV) control. In A-G, the dotted lines mark the zero current level. In A and C-G, the La 3ϩ concentration was 1 mM. The downward deflection in all current traces represents inward current. marked reduction in thrombin-induced Ca 2ϩ entry via SOCs (Fig. 1D). As a procedural control, we showed that cells infected with a control construct (Ad-lacZ) gave normal responses to thrombin (Fig. 1D). The adenoviral vector used for these experiments guaranteed nearly 90% transfection of the endothelial monolayer (34). The summary graph in Fig. 1E demonstrates that 1) Ca 2ϩ store depletion induced by thrombin was statistically indistinguishable in all groups and 2) Ca 2ϩ entry caused by Ca 2ϩ repletion was significantly reduced (Ͼ50% of the control) by genetic or pharmacological inhibition of PKC␣. To demonstrate that TRPC1 is a functional component of SOCs in endothelial cells, we used anti-TRPC1 Ab against the extracellular epitope of TRPC1, which has been recently shown to block TRPC1 function in platelets (38). Ca 2ϩ entry was measured following depletion of stores using thrombin. Anti-TRPC1 Ab pretreatment significantly inhibited the influx of Ca 2ϩ following Ca 2ϩ store depletion; however, control IgG had no effect (Fig. 1F). In these experiments, anti-TRPC1 Ab did not significantly reduce the Ca 2ϩ release transient, but markedly inhibited Ca 2ϩ entry, which fell to ϳ60% of the control. These findings indicate that normal Ca 2ϩ entry via SOCs has a requirement for TRPC1 in endothelial cells.
Thapsigargin-induced Ca 2ϩ Entry-Thapsigargin, an inhibitor of the Ca 2ϩ -ATPase found in the ER (40), was used to activate SOCs by inducing Ca 2ϩ store depletion independently of GPCRs (3,4,6,10,30). Ca 2ϩ entry in response to thapsigargin was monitored in Fura 2-loaded endothelial cells with or without Gö6976 pretreatment (Fig. 2, A and B) or in endothelial cells expressing either dnPKC␣ or the lacZ control (Fig. 2,  C and D). Inhibition of PKC␣ attenuated the store depletioninduced Ca 2ϩ entry by 52 Ϯ 5% (Gö6976) and 58 Ϯ 6% (dnPKC␣) (Fig. 2, A-D), indicating that PKC␣ contributes to SOC activation independently of GPCR pathways. IP 3 -induced Store-operated Current-IP 3 can induce Ca 2ϩ store depletion and should thereby activate Ca 2ϩ entry via SOCs. Poor selectivity for divalent cations and sensitivity to La 3ϩ or Gd 3ϩ are characteristic properties of store-operated channels in many tissues (4,41,42). Using the whole cell patch clamp technique (43), we internally applied IP 3 and measured the resulting SOC that we operationally defined as the La 3ϩ -sensitive inward current at Ϫ50 mV. Continuous current recordings showed the rapid development of an La 3ϩ (1 mM)-sensitive inward current upon internal dialysis with IP 3 or perfusion of thapsigargin. The action of La 3ϩ was reversible as shown (Fig. 3A). Without IP 3 /thapsigargin in FIG. 4. Effect of anti-TRPC1 antibody on store-operated currents. IP 3 -induced activation of SOCs is shown as whole cell continuous current recordings at a holding potential of Ϫ50 mV. The traces show the La 3ϩ -sensitive current after intracellular application of IP 3 . A and B, current from non-transfected control cells after 15 min of external application of 15 g/ml IgG or anti-TRPC1 Ab, respectively; C, current from a cell overexpressing (OE) TRPC1 cDNA; D, depressed current in a TRPC1-overexpressing cell pretreated with anti-TRPC1 Ab; E, essentially normal current recorded in a control cell internally dialyzed with anti-TRPC1 Ab; F, data shown as the mean Ϯ S.E. of maximum steady-state current density over the basal value under the indicated conditions (n ϭ 4 -7). *, statistically significant reduction in current density in the presence of extracellular anti-TRPC1 Ab versus control IgG; #, significant reduction in current density in TRPC1-overexpressing cells due to external application of anti-TRPC1 Ab. In all current records, the downward deflection denotes inward current, and dotted lines mark the zero current level. In A-E, the La 3ϩ concentration was 1 mM. pF, picofarad. the pipette, there was no current as shown in Fig. 3B, indicating that the current was not the result of a nonspecific leak. To determine whether the SOC current is regulated by PKC␣, we pretreated cells with Gö6976 prior to membrane rupture with patch pipettes. Inhibition of PKC␣ by Gö6976 markedly reduced the SOC current amplitude from a mean value of 2.9 Ϯ 0.3 to 1.3 Ϯ 0.2 pA/picofarad (Fig. 3, C and I), indicating the role of endogenous PKC␣ in regulating SOC currents. Because Gö6976 has been shown to affect PKC␤ in several cell types (30,44), we also used LY379196, an inhibitor of PKC␤, to address the possible role of PKC␤ in modulating SOC currents. As shown in Fig. 3D, inhibition of PKC␤ had no effect on the IP 3 -induced current. To demonstrate the specific role of PKC␣, we also compared the magnitude of IP 3 -induced currents in endothelial cells expressing the pCMV vector (control) and in cells expressing dnPKC␣. As shown in Fig. 3 (E and F), expression of dnPKC␣ significantly reduced IP 3 -induced currents; current amplitude was significantly decreased from 2.7 Ϯ 0.3 to 1.2 Ϯ 0.3 pA/picofarad (Fig. 3I). Perfusion of thapsigargin (2 M) activated the SOC current to a similar amplitude as that induced by IP 3 (Fig.  3G). The current-voltage relationship in Fig. 3H shows a reversal potential close to 0 mV, indicating that the inward current was carried by cations as described previously for SOCs (4,21,45,46). As TRPC1-overexpressing cells showed the same reversal potential as control cells, the results indicate that, in control cells, TRPC1 forms the main constituent of SOCs, as has been described by others (45,47,48). These results, which show that PKC␣ is critical in regulating SOC activity, are in agreement with the findings (see above) obtained using pharmacological inhibitors.
To further corroborate the role of TRPC1 in regulating Ca 2ϩ entry observed above (Fig. 1F), we determined SOC currents in control endothelial cells as well as in TRPC1-overexpressing cells after pretreatment with anti-TRPC1 Ab (38,49). Overexpression of TRPC1 increased the amplitude of IP 3 -induced currents (Fig. 4, C and F). Pretreatment with anti-TRPC1 Ab significantly reduced the mean amplitude of IP 3 -induced SOC currents in control and TRPC1-overexpressing cells (Fig. 4, B, D, and F). In contrast, preincubation of control cells with isotype-matched IgG (15 g/ml) for 15 min had no effect on the current amplitude (Fig. 4A). In addition, anti-TRPC1 Ab had no detectable effect on IP 3 -induced SOC currents when injected into the cell cytoplasm (Fig. 4E). Altogether, the data presented in Fig. 4 reinforce our observations, obtained in Fura 2-loaded cells, that PKC␣ activity controls Ca 2ϩ entry via store-operated Ca 2ϩ channels, most importantly, through the TRPC1 channel.

PKC␣ Phosphorylation of TRPC1
Thrombin-and Thapsigargin-induced Phosphorylation of TRPC1-As TRPC1 is a presumptive protein constituent of SOCs in human endothelial cells, and its sequence also contains several putative phosphorylation sites (50,51), we hypothesized that thrombin activates SOCs by inducing the phosphorylation of TRPC1. We therefore determined TRPC1 phosphorylation in endothelial cells labeled with [ 32 P]orthophosphate upon stimulation with thrombin for various time periods. We first determined the specificity of anti-TRPC1 Ab using COS-7 cells transfected with Myc-tagged TRPC1 or vector alone. Cells were lysed in radioimmune precipitation assay buffer and immunoprecipitated with anti-Myc Ab, followed by Western blotting with anti-TRPC1 or anti-Myc Ab. Both antibodies detected protein at ϳ90 kDa in TRPC1-transfected cells (Fig. 5A). Thrombin induced rapid (within 1 min) and sustained (up to 5 min) phosphorylation of FIG. 5. Thrombin-and thapsigargin-induced TRPC1 phosphorylation. A, COS-7 cells overexpressing the Myc-tagged TRPC1 construct were lysed, and TRPC1 was immunoprecipitated with anti-Myc Ab, followed by Western blotting with anti-Myc or anti-TRPC1 Ab. Both antibodies recognized the TRPC1 band (ϳ90 kDa). B-E, HUVECs grown to confluence were serum-starved and then labeled with 32 P for 4 h in phosphate-free medium. Cell lysates were immunoprecipitated with control IgG or rabbit anti-TRC1 Ab. Proteins were electrophoresed and transferred to nitrocellulose membranes, after which the membranes were exposed to radiography, followed by Western blotting with anti-TRPC1 Ab. B and C, thrombin-stimulated phosphorylation of TRPC1 for indicated time periods and quantitative results by densitometry, respectively. *, significant increase (p Ͻ 0.05) in TRPC1 phosphorylation compared with unstimulated cells. D and E, thapsigargin-stimulated phosphorylation of TRPC1 for the indicated time periods and quantitative results by densitometry, respectively. In B and D, the upper panels show autoradiograms of phosphorylated TRPC1, whereas the lower panels show Western blots of TRPC1. C and E, densitometric analysis of thrombin-stimulated (n ϭ 3) and thapsigargin-stimulated (n ϭ 2) TRPC1 phosphorylation, respectively. quantified as the increase in phosphorylation of TRPC1 above the level in unstimulated cells. IB, immunoblot. TRPC1 (Fig. 5, B and C). Thrombin-stimulated phosphorylation of TRPC1 was specific because it was not seen in lysates immunoprecipitated with control IgG (Fig. 5, B and C). In addition, we determined TRPC1 phosphorylation in response to depletion of stores with thapsigargin, which bypasses the GPCR pathway. Thapsigargin also induced the phosphorylation of TRPC1 within 1 min (Fig. 5, D and E), indicating that store depletion alone can trigger phosphorylation of TRPC1.
TRPC1 Phosphorylation Requires PKC␣ Activity-Because the results presented in Figs. 1-4 point to a role of PKC␣ in regulating SOC-mediated Ca 2ϩ entry, we addressed the possibility that PKC␣ regulates SOCs by phosphorylating the TRPC1 channel protein. We therefore expressed dnPKC␣ in endothelial cells to determine its role in thrombin-induced TRPC1 phosphorylation. Endothelial cells infected with Ad-dnPKC␣ or Ad-lacZ (as a control) were metabolically labeled with [ 32 P]orthophosphate and stimulated with thrombin for 1 min. As shown in Fig. 6 (A and B), thrombin induced significant increases in TRPC1 phosphorylation in Ad-lacZ-transduced cells, but not in cells expressing dnPKC␣. These findings indicate that the thrombin-induced phosphorylation of TRPC1 requires PKC␣ activity.

Role of PKC␣ in Induction of Increased Transendothelial Permeability
As the results implicated PKC␣ in regulating Ca 2ϩ entry through SOCs, we determined the role of PKC␣ in regulating endothelial barrier function. We measured the changes in transendothelial electrical resistance in control HMEC mono-layers and in cells overexpressing the TRPC1 construct with or without Gö6976 treatment. Thrombin caused a significant decrease in transendothelial electrical resistance in untreated control and TRPC1-overexpressing cells; pretreatment with Gö6976 reduced the effect of thrombin in both the control and TRPC1-overexpressing groups by 60 Ϯ 6 and 58 Ϯ 5%, respectively (Fig. 7, A and B), thus indicating the functional role of PKC␣-induced Ca 2ϩ entry in regulating endothelial barrier function. DISCUSSION The results demonstrate that PKC␣ activation in response to thrombin induces phosphorylation of the TRPC1 protein. The functional importance of TRPC1 phosphorylation in endothelial cells is evident by the following: 1) phosphorylation of TRPC1 in response to thrombin or thapsigargin occurred within minutes and preceded Ca 2ϩ entry; 2) thrombin-induced phosphorylation of TRPC1 was blocked by overexpression of dnPKC␣; and 3) the La 3ϩ -sensitive IP 3 -induced current (20,21) through SOCs was markedly reduced when PKC␣ activity was inhibited. Ca 2ϩ store depletion has thus far been considered the triggering event to open SOCs, resulting in store-operated Ca 2ϩ entry (4,5,7,52,53). The present findings imply that phosphorylation of TRPC1 by PKC␣ can also increase Ca 2ϩ entry through SOCs in endothelial cells. We found that such modulation of SOCs accounted for at least 50% of the current via SOCs in endothelial cells based on inhibitory effects of Gö6976 (100 nM) and expression of the catalytically defective PKC␣ mutant. In this study, we considered the whole cell currents through SOCs and did not address the detailed mechanisms of their modulation, which could theoretically include increases in channel open probability and amplitude of the single channel current. Either of these mechanisms could account for the partial inhibition of the macroscopic current through SOCs observed when PKC␣ was inhibited. Alternatively, it is arguable that the phosphorylated channel is preferentially opened by store depletion. Thus, the emerging picture of store-operated Ca 2ϩ entry on the basis of the present findings includes channel opening by Ca 2ϩ store depletion plus a functional modification of the channel by protein phosphorylation that enhances channel activity.
The precise relationship between Ca 2ϩ store depletion and SOC activation remains controversial. However, despite intensive study, the ER-derived signal coupling store depletion and SOC activation is unknown. Evidence indicates that physical docking of the ER with the plasma membrane could be involved in SOC activation (6,7,52). Recently, we showed that, upon thrombin activation, RhoA signals interaction of the IP 3 receptor with TRPC1 at the plasma membrane and triggers SOCinduced Ca 2ϩ entry in endothelial cells (35). We also showed that thrombin-induced Rho activation is dependent on PKC␣, indicating that PKC␣ contributes significantly to Rho regulation of Ca 2ϩ entry (26). There is also some evidence of involvement of both tyrosine kinase-and PKC-mediated protein phosphorylation/dephosphorylation in modulation of SOC activity (4,12,30,54). It has been shown in photoreceptor cells that the TRPC is a substrate for PKC and that phosphorylation of TRPC by PKC regulates Ca 2ϩ influx through the TRPC channel (55). This led us to investigate the potential function of PKC␣ phosphorylation of the TRPC1 channel in activation of SOCs (thus, Ca 2ϩ entry) and in endothelial barrier function. In this regard, we showed that thrombin and thapsigargin stimulation led to an overall increase in TRPC1 phosphorylation in whole cell lysates as detected by Western blot analysis (Fig. 5, B-E). Our results further revealed that Ca 2ϩ store depletion activated PKC␣, which subsequently induced the phosphorylation of TRPC1. Our evidence for this concept is that inhibition of FIG. 6. PKC␣ regulates thrombin-induced TRPC1 phosphorylation. HUVECs expressing dnPKC␣ (Ad (Adv)-dnPKC␣) or the control construct (Ad-lacZ) were serum-starved, labeled with 32 P for 4 h in phosphate-free medium, and stimulated with thrombin for 1 min. Cell lysates were immunoprecipitated with anti-TRPC1 antibody. Proteins were electrophoresed and transferred to nitrocellulose membranes, after which the membranes were exposed to radiography, followed by Western blotting with anti-TRPC1 Ab. In A, the upper panel shows an autoradiogram of phosphorylated TRPC1, whereas the lower panel shows a Western blot of TRPC1. In B, densitometric data summarize the thrombin-induced increase in phospho-TRPC1 at 1 min as a multiple of the unstimulated level. Note that thrombin produced a statistically significant increase in TRPC1 phosphorylation in the transfection control (Ad-lacZ; *, p Ͻ 0.001). In cells expressing the dnPCK␣ construct, the effect of thrombin was completely inhibited compared with the control construct (#, p Ͻ 0.001). Means Ϯ S.E. are based on three independent experiments. PKC␣ impaired Ca 2ϩ entry through SOCs after store depletion induced by thapsigargin or internally applied IP 3 .
TRPC1 is expressed in several cell types, including human endothelial cells (10,14,(21)(22)(23), vascular smooth muscle cells (13), salivary gland cells (4,56), human platelets (38), and airway smooth muscle cells (16). We have focused on TRPC1 in this study because human endothelial cell (HUVECs and HMECs) predominantly express this isoform; they also express TRPC3 and TRPC4, but to a considerably lesser extent (11,22,23,35). These minor isoforms could contribute to the SOC, but may be regulated differently. A recent study by Venkatachalam et al. (29) using human embryonic kidney cells as a TRPC expression system showed that PKC inhibits TRPC4-and TRPC5-mediated Ca 2ϩ entry, but that the kinase does not regulate Ca 2ϩ entry through TRPC3 channels; the latter isoform is directly activated by diacylglycerol or its analogs. Because TRPC function varies with the cell type used, and its expression level per se determines channel activity, it is possible that these isoforms are regulated by distinct mechanisms. In our studies on endothelial cells, we would have observed an inhibition of SOC activity if the predominant effect of PKC␣ were to inhibit TRPC4. Because this was not the case, and expression level of TRPC4 was low in the endothelial cells used, the role of TRPC4 in the present experiments is inferred to be minor.
It has been shown that the SOC is inhibited by transfection with antisense TRPC1; extracellular addition of anti-TRPC1 antibody that specifically recognizes TRPC1-(557-571), which is present in TRPC1 and no other protein; or deletion of the TRPC1 gene (12-14, 38, 49). We also found that pretreatment of endothelial cells in medium containing anti-TRPC1 Ab significantly inhibited SOC-induced Ca 2ϩ entry in control cells as well as in cells overexpressing TRPC1, thus indicating that TRPC1 is a critical protein constituent of functional SOCs. Recently, it has been shown that the TRPC1 S5-6 region comprises the pore-forming unit of the channel and that mutation of TRPC1-(567-793) causes a reduction in the total number of functional channels without changing the channel properties (12). Although these studies provide convincing evidence that TRPC1 forms functional SOCs, they do not explain how TRPC1 becomes activated after store depletion. The TRPC1 protein has multiple serine/threonine phosphorylation sites in the putative pore-forming region and N and C termini (51). As PKC␣ is a serine/threonine kinase, we propose that phosphorylation of TRPC1 serine/threonine sites by PKC␣ regulates TRPC1 channel opening and hence store-operated Ca 2ϩ entry.
PKC has been shown to modulate a variety of intracellular events by regulating the activity of ion channels. However, the results are extremely variable with respect to the regulation of SOCs by PKC (28 -33) because there are multiple PKC isoforms with different functions. When activated, PKC normally translocates to its target site, which, in the case of SOCs, is the plasma membrane (17). It is probable that Ca 2ϩ store depletion is involved in modifying the SOC via PKC activation, but it is unknown about how the kinase is activated.
Our results suggest that thrombin-or thapsigargin-induced phosphorylation of TRPC1 by PKC␣ strongly contributes to the regulation of SOC-induced Ca 2ϩ influx. Phosphorylation of serine/threonine residues in the TRPC1 protein appears to maintain the channel activity. The precise phosphorylation site is unknown and requires further study. We observed the optimum phosphorylation by thrombin within 1 min, with continued activity for at least 5 min under our experimental conditions. Hence, PKC␣-dependent channel phosphorylation may be required for full agonist activation of SOCs. The specific target for PKC␣ identified in this study is one or more phosphorylation sites in TRPC1.
Because Ca 2ϩ signaling can activate PKC␣, which, in turn, has an important role in SOC recruitment (9,11), we also addressed the role of PKC␣-induced Ca 2ϩ entry in mediating the increase in endothelial permeability. We measured the transendothelial electrical resistance in endothelial cells to test the hypothesis that PKC␣ activation increases endothelial permeability. PKC␣ inhibition markedly decreased the effect of thrombin on endothelial permeability in both TRPC1-overexpressing and control cells. Therefore, SOCs arguably play an important role in the mechanism of endothelial barrier dysfunction.
In conclusion, our data show that 1) PKC␣ regulates the influx of Ca 2ϩ through the SOC and 2) PKC␣ phosphorylation of TRPC1 is responsible for physiologically significant modulation of SOCs. The results indicate that PKC␣ regulates endothelial barrier function through the phosphorylation of TRPC1. Transfected cells were grown to confluence on gold electrodes, and cells were placed in reduced serum medium (1% fetal bovine serum) for 2 h. Cells were then incubated with or without Gö6976 (100 nM) for 1 h, after which cells were stimulated with 50 nM thrombin to measure the changes in transendothelial electrical resistance. A, traces representative of a single experiment showing the inhibitory effect of Gö6976 in TRPC1overexpressing (OE) cells; B, data shown as the mean Ϯ S.E. of the peak reduction in electrical resistance produced by thrombin challenge (n ϭ 5). * and #, values different from Gö6976-treated monolayers.