Protein Kinase C-dependent Phosphorylation of Transient Receptor Potential Canonical 6 (TRPC6) on Serine 448 Causes Channel Inhibition*

TRPC6 is a cation channel in the plasma membrane that plays a role in Ca2+ entry following the stimulation of a Gq-protein coupled or tyrosine kinase receptor. A dysregulation of TRPC6 activity causes abnormal proliferation of smooth muscle cells and glomerulosclerosis. In the present study, we investigated the regulation of TRPC6 activity by protein kinase C (PKC). We showed that inhibiting PKC with GF1 or activating it with phorbol 12-myristate 13-acetate potentiated and inhibited agonist-induced Ca2+ entry, respectively, into cells expressing TRPC6. Similar results were obtained when TRPC6 was directly activated with 1-oleyl-2-acetyl-sn-glycerol. Activation of the cells with carbachol increased the phosphorylation of TRPC6, an effect that was prevented by the inhibition of PKC. The target residue of PKC was identified by an alanine screen of all canonical PKC sites on TRPC6. Unexpectedly, all the mutants, including TRPC6S768A (a residue previously proposed to be a target for PKC), displayed PKC-dependent inhibition of channel activity. Phosphorylation prediction software suggested that Ser448, in a non-canonical PKC consensus sequence, was a potential target for PKCδ. Ba2+ and Ca2+ entry experiments revealed that GF1 did not potentiate TRPC6S448A activity. Moreover, activation of PKC did not enhance the phosphorylation state of TRPC6S448A. Using A7r5 vascular smooth muscle cells, which endogenously express TRPC6, we observed that a novel PKC isoform is involved in the inhibition of the vasopressin-induced Ca2+ entry. Furthermore, knocking down PKCδ in A7r5 cells potentiated vasopressin-induced Ca2+ entry. In summary, we provide evidence that PKCδ exerts a negative feedback effect on TRPC6 through the phosphorylation of Ser448.

TRPC6 is a cation channel in the plasma membrane that plays a role in Ca 2؉ entry following the stimulation of a G qprotein coupled or tyrosine kinase receptor. A dysregulation of TRPC6 activity causes abnormal proliferation of smooth muscle cells and glomerulosclerosis. In the present study, we investigated the regulation of TRPC6 activity by protein kinase C (PKC). We showed that inhibiting PKC with GF1 or activating it with phorbol 12-myristate 13-acetate potentiated and inhibited agonist-induced Ca 2؉ entry, respectively, into cells expressing TRPC6. Similar results were obtained when TRPC6 was directly activated with 1-oleyl-2-acetyl-sn-glycerol. Activation of the cells with carbachol increased the phosphorylation of TRPC6, an effect that was prevented by the inhibition of PKC. The target residue of PKC was identified by an alanine screen of all canonical PKC sites on TRPC6. Unexpectedly, all the mutants, including TRPC6 S768A (a residue previously proposed to be a target for PKC), displayed PKC-dependent inhibition of channel activity. Phosphorylation prediction software suggested that Ser 448 , in a non-canonical PKC consensus sequence, was a potential target for PKC␦. Ba 2؉ and Ca 2؉ entry experiments revealed that GF1 did not potentiate TRPC6 S448A activity. Moreover, activation of PKC did not enhance the phosphorylation state of TRPC6 S448A . Using A7r5 vascular smooth muscle cells, which endogenously express TRPC6, we observed that a novel PKC isoform is involved in the inhibition of the vasopressin-induced Ca 2؉ entry. Furthermore, knocking down PKC␦ in A7r5 cells potentiated vasopressin-induced Ca 2؉ entry. In summary, we provide evidence that PKC␦ exerts a negative feedback effect on TRPC6 through the phosphorylation of Ser 448 .
Ca 2ϩ is a second messenger in all cell types. While the intracellular concentration of Ca 2ϩ ([Ca 2ϩ ] i ) is tightly controlled and normally maintained at low levels, increases mod-ulate cellular functions such as secretion, gene transcription, and the activation of a variety of effectors (1). TRPC6 allows Ca 2ϩ to enter cells from the extracellular medium when it is stimulated by the phospholipase C/inositol-1,4,5-trisphosphate (IP 3 ) 2 pathway. This channel is a member of the TRPC family (transient receptor potential canonical), which includes seven members (TRPC1 to TRPC7). Many pathophysiologies arise when TRPC6 expression or activity is dysregulated. Focal segmental glomerulosclerosis (FSGS) is a channelopathy associated with TRPC6 and is caused by missense mutations, including gain-of-function mutations (2,3). In some cases of idiopathic pulmonary arterial hypertension, the expression of TRPC6 has been reported to be higher in pulmonary artery smooth muscle cells (4,5). An up-regulation of TRPC6 in hepatocytes has also been linked to liver cancer (6). These examples stress the importance of understanding the mechanisms responsible for the regulation of TRPC6.
The exact mechanism by which TRPC6 is activated is not clear, but it occurs when a GqPCR or tyrosine kinase receptor is activated, leading to IP 3 and diacylglycerol (DAG) formation from phosphatidyl-4,5-bisphosphate (PIP 2 ) hydrolysis by phospholipase C. IP 3 binds to its receptor, IP 3 R, on the endoplasmic reticulum leading to Ca 2ϩ release, which is the first phase of Ca 2ϩ mobilization. The second phase takes place as Ca 2ϩ channels, including TRPCs and Orai (7,8), are activated at the plasma membrane and maintain high intracellular Ca 2ϩ levels as long as the stimulation is sustained. DAG, along with Ca 2ϩ , activates protein kinase C (PKC), which regulates the activity of many proteins involved in Ca 2ϩ signaling, including IP 3 R (9, 10), L-type Ca 2ϩ channels (11), and TRP proteins (12). Trebak et al. showed that TRPC3, the closest relative of TRPC6, is phosphorylated on residue Ser 712 by PKC (13), and that the activation of PKC by phorbol 12-myristate 13-acetate (PMA) inhibits 1-oleyl-2-acetyl-sn-glycerol (OAG)-mediated TRPC3 channel activation (13).
It has been shown that TRPC6 is negatively regulated after phosphorylation of threonine 69 by PKG (14). This phosphorylation event is also essential for the anti-hypertrophic effects of phosphodiesterase 5 inhibitors (15)(16)(17). Because phosphorylation events play an important role in TRPC6 activity, we verified, in the present study, whether TRPC6 was a substrate for PKC. We showed that PKC can phosphorylate TRPC6 and that the inhibition of PKC by bisindolylmaleimide I (GF1) or Gö 6983 enhances agonist-induced Ca 2ϩ entry into TRPC6expressing cells. Mutagenesis studies showed that none of the twelve canonical phosphorylation sites for PKC exposed to the intracellular environment is involved in the phosphorylation of TRPC6. Further analysis of the TRPC6 sequence suggested that PKC␦ can phosphorylate TRPC6 on Ser 448 , a noncanonical phosphorylation site for PKC. The mutation of Ser 448 to Ala abolished the ability of PKC to phosphorylate TRPC6. In addition, agonist-induced Ca 2ϩ entry into cells expressing TRPC6 S448A was not modified by GF1. Furthermore, we showed that PKC␦ regulates vasopressin (AVP)induced Ca 2ϩ entry into A7r5 cells, which endogenously express TRPC6. In summary, we demonstrated that TRPC6 is phosphorylated on a non-canonical site by PKC in cellulo and that this phosphorylation down-regulates the activity of TRPC6.
Molecular Biology-Standard molecular biology techniques were used for to isolate, analyze, and clone DNA (18,19). Point mutations in mouse TRPC6 were introduced using a PCR-based site-directed mutagenesis strategy. The PCR fragment was subcloned into the pCR-Blunt II-TOPO vector using a Zero Blunt TOPO PCR cloning kit. The fragment was sequenced and was then introduced into HA-tagged TRPC6 in pcDNA3.1. All constructs were confirmed by sequencing from double-stranded DNA templates using the dideoxynucleotide termination method (20).
Cell Culture and Transfection-HEK293T cells and A7r5 vascular myocytes were maintained at subconfluence in Dulbecco's modified Eagle's medium (DMEM) supplemented with 10% fetal bovine serum (FBS), 50 units/ml of penicillin, and 50 g/ml streptomycin at 37°C in a humidified atmosphere containing 5% CO 2 . T6.11 cells (HEK293 stably transfected with mouse TRPC6) were cultured in the same medium supplemented with G418 (400 g/ml). Metabolic Labeling-Stably or transiently transfected cells grown in 60-mm Petri dishes were washed once with phosphate-free DMEM and incubated for 4 h in phosphate-free DMEM supplemented with 250 Ci/ml 32 P-inorganic phosphate. The stimulations were performed by adding directly the agonists at 100ϫ in the medium. After the appropriate incubation time, the cells were washed twice with ice-cold PBS prior to being lysed.
Immunoprecipitation-The cells were lysed with 1 ml of ice-cold lysis buffer (1.25% Nonidet P-40, 1.25% sodium deoxycholate, 2 mM EDTA, 12.5 mM sodium phosphate, pH 7.2, 1 g/ml of soybean trypsin inhibitor, 5 g/ml of leupeptin, 100 M phenylmethylsulfonyl fluoride) supplemented with a phosphatase inhibitor mixture for 30 min on ice with gentle agitation. They were then scraped from the surface of the Petri dish and centrifuged at 15,000 ϫ g for 15 min at 4°C. The supernatant was collected and immunoprecipitated with 50 l of protein A-Sepharose beads (50% slurry) and anti-HA mouse antibody (1:1000) for 2 h at 4°C. Samples were then centrifuged for 1 min at 4°C at 800 ϫ g and washed twice with 500 l of ice-cold lysis buffer. Immunoprecipitated proteins were dissolved in 40 l of 2ϫ Laemmli buffer and boiled for 5 min before being separated on 7% SDS-polyacrylamide gels. The gels were then either dried and exposed to a film for autoradiography, or the protein bands were transferred to a 0.2-m nitrocellulose membrane (400 mA for 2 h or 100 mA overnight in 150 mM glycine, 20 mM Tris-base, and 20% methanol) for immunoblotting.
Immunoblots-The nitrocellulose membranes to which the whole cell lysates and immunoprecipitated proteins had been transferred were stained with Ponceau S (0.1% in 5% acetic acid) to visualize the marker proteins, destained in TBST (20 mM Tris-HCl, pH 7.5, 137 mM NaCl, 0.1% Tween 20) and blocked in TBST containing 5% (w/v) nonfat dry milk for either 1 h at room temperature or overnight at 4°C. The membranes were then washed and incubated in TBST for either 2.5 h at room temperature or overnight at 4°C with specific primary antibodies (rabbit anti-HA (1:1000) or mouse antiactin (1:10 000)). After three washes with TBST, the membranes were incubated for 1.5 h at room temperature in TBST containing peroxidase-conjugated donkey anti-rabbit-IgG (1:30,000) or peroxidase-conjugated sheep anti-mouse-IgG (1:10,000). The blots were washed three times with TBST and the immune complexes were detected with the Western Lightning Chemiluminescence Reagent Plus kit using the manufacturer's protocol.
Measurement of [Ca 2ϩ ] i -We used the method described by Zhu et al. (21) to measure [Ca 2ϩ ] i . Briefly, A7r5, T6.11, or transfected HEK293T cells grown on poly-L-lysine-treated coverslips were washed twice with HBSS (120 mM NaCl, 5.3 mM KCl, 0.8 mM MgSO 4 , 10 mM glucose, 1.8 mM CaCl 2 , 20 mM Hepes, pH 7.4) and loaded with Fura-2/AM (1.5 M in HBSS) for 20 min at room temperature in the dark. After washing and de-esterifying in fresh HBSS for 20 min at room temperature, the coverslips were inserted in a circular openbottom chamber and placed on the stage of a Zeiss Axovert microscope fitted with an Attofluor Digital Imaging and Photometry System (Attofluor Inc., Rockville, MD). The [Ca 2ϩ ] i in selected Fura-2-loaded cells was measured by fluorescence videomicroscopy at room temperature using alternating excitation wavelengths of 334 and 380 nm (10 nm bandpass filters), and emitted fluorescence was monitored through a 510 nm dichroic mirror and a 520 nm long pass filter set. Free [Ca 2ϩ ] i was calculated from the 334/380 fluorescence ratios using the method of Grynkiewicz et al. (22). Reagents were diluted to their final concentrations in HBSS and applied to the cells by surface perfusion. Ca 2ϩ -free HBSS was supplemented with 0.5 mM EGTA to chelate any remaining extracellular Ca 2ϩ . For the Ba 2ϩ entry assays, 1 mM Ba 2ϩ was added to Ca 2ϩ -free, EGTA-free HBSS. For the transient transfections, the cells were co-transfected with cDNA encoding the M5 muscarinic receptor, and only those responding to carbachol (CCh) were analyzed.
[Ca 2ϩ ] i values were recorded every 3 s.

RESULTS
We first investigated whether the inhibition of PKC could modulate CCh-induced Ca 2ϩ mobilization in HEK 293 cells stably transfected with mTRPC6 (T6.11 cells) (23). As shown in Fig. 1A, in the presence of extracellular Ca 2ϩ , 50 M CCh caused a robust increase in [Ca 2ϩ ] i from a basal value of ϳ70 nM to a peak value of ϳ210 nM. The [Ca 2ϩ ] i then declined slowly and remained above the basal level for several minutes, which constituted a plateau phase. When the cells were pretreated with 100 nM GF1, a highly selective PKC inhibitor, the CCh-induced increase in [Ca 2ϩ ] i reached a peak value similar to that of untreated cells. However, the [Ca 2ϩ ] i remained elevated at a plateau of ϳ220 nM. In control HEK293 cells expressing angiotensin II receptor type 1 instead of TRPC6, the CCh-induced increase in [Ca 2ϩ ] i was similar to that of the T6.11 cells, but the [Ca 2ϩ ] i declined quickly to a plateau level of 130 nM within 45 s after the stimulation (Fig. 1B). Under the same conditions, the [Ca 2ϩ ] i in the T6.11 cells after 45 s was 165 nM (Fig. 1A). In HEK293 cells expressing the angiotensin II receptor type 1, GF1 slightly increased the peak [Ca 2ϩ ] i, but Ca 2ϩ levels quickly returned to values similar to those of untreated cells (Fig. 1B).
To discriminate between CCh-induced Ca 2ϩ release and CCh-induced Ca 2ϩ entry, we used a Ca 2ϩ depletion/Ca 2ϩ readdition protocol. Cells were incubated for 30 s in Ca 2ϩfree medium containing 0.5 mM EGTA before depleting the intracellular Ca 2ϩ stores with 5 M CCh. Once the [Ca 2ϩ ] i had returned to the basal level (2 min after the addition of CCh), the extracellular medium was replaced with medium containing 1.8 mM CaCl 2 . As shown in Fig. 2, in the absence of extracellular Ca 2ϩ , CCh-induced Ca 2ϩ release was similar in TRPC6-and mock-transfected cells. When 1.8 mM CaCl 2 was added to TRPC6-transfected cells, Ca 2ϩ entry raised the [Ca 2ϩ ] i to a plateau level of ϳ190 nM ( Fig. 2A). In the mocktransfected cells, the Ca 2ϩ entry raised the [Ca 2ϩ ] i to a lower plateau level of ϳ160 nM. Pretreating the cells with GF1 increased CCh-induced net Ca 2ϩ entry by 52.4 Ϯ 13.1 nM and 21.9 Ϯ 10.0 nM in TRPC6-and mock-transfected cells, respectively (Fig. 2B). Net Ca 2ϩ entry was calculated by subtracting the basal [Ca 2ϩ ] i from the maximal [Ca 2ϩ ] i recorded once the Ca 2ϩ had been restored. These results suggested that the inhibition of PKC causes increased TRPC6 activity and that PKC exerts an inhibitory effect on TRPC6. We next investigated how the activation of PKC with PMA could influence TRPC6mediated Ca 2ϩ entry. Because PMA greatly decreased CChinduced IP 3 production in the T6.11 cells (data not shown), we performed the Ca 2ϩ depletion/Ca 2ϩ readdition protocol by stimulating the cells with 100 M ATP. As shown in Fig.  2C, ATP caused a net Ca 2ϩ entry of 38.4 Ϯ 16.5 nM while a pretreatment of the cells with GF1 increased ATP-induced net Ca 2ϩ entry to 78.7 Ϯ 28.7 nM. The presence of PMA before the stimulation with ATP reduced Ca 2ϩ entry by 75% to a net Ca 2ϩ entry of 9.2 Ϯ 3.7 nM. These results further suggested that PKC decreases the activity of TRPC6.
To confirm that the observed effect of PKC was due to a direct regulation of TRPC6, we exploited two characteristics that are distinctive of TRPC6. First, it has been shown that CCh generates a marked increase in receptor-operated Ca 2ϩ entry (ROCE) in addition to store-operated Ca 2ϩ entry (SOCE) in HEK293 cells overexpressing TRPC6 (23). The depletion of the intracellular Ca 2ϩ store was induced with 1 M thapsigargin, an irreversible inhibitor of the SERCA that causes a rapid leak of Ca 2ϩ from the endoplasmic reticulum and triggers store-operated channels. Fig. 3A shows that after an incubation of T6.11 cells with thapsigargin for 5 min, which completely depleted the intracellular Ca 2ϩ store (24,25), 50 M CCh caused an increase in [Ca 2ϩ ] i to a peak value of 245 nM, which slowly declined to a value of 140 nM 2 min after the addition of CCh. In the presence of GF1, CCh still caused an increase in [Ca 2ϩ ] i to a peak value similar to that of   (Fig. 3B).
Another distinctive characteristic of TRPC6 is that it can be activated by high concentrations of OAG, a non-metabolizable DAG analog (26 -28). Fig. 4A shows that 100 M OAG caused a slow elevation of [Ca 2ϩ ] i that reached a value of 122 nM in 1 min in T6.11 cells. In the presence of 100 nM PMA, OAG-induced Ca 2ϩ entry decreased to 33.3 Ϯ 11.0% of that of control cells (Fig. 4B). In the presence of 100 nM GF1, OAG-induced Ca 2ϩ entry increased to 144 nM (Fig. 4A), which corresponds to a relative elevation of 162.3 Ϯ 24.1% of that of control cells (Fig. 4B). These results further suggested that PKC down-regulates the activity of TRPC6.
To determine whether PKC phosphorylates TRPC6 in cellulo, HEK293 cells expressing TRPC6 were metabolically labeled using [ 32 P]orthophosphate. PKC was then activated through the stimulation of the muscarinic receptor with 50 M CCh. Fig. 5A shows that the immunoprecipitated TRPC6 was weakly phosphorylated under basal conditions. CCh increased the phosphorylation level of TRPC6 by 1.34 Ϯ 0.17fold, compared with the basal level (Fig. 5B). To further determine whether CCh-induced TRPC6 phosphorylation was due to PKC, cells were pre-incubated for 5 min with GF1 before the stimulation with CCh. GF1 did not modify the basal level of phosphorylation of TRPC6 (0.97 Ϯ 0.05 times the basal level), but it completely inhibited the effect of CCh on TRPC6 phosphorylation (0.98 Ϯ 0.13 times the basal level) (Fig. 5, A  and B). These results suggested that CCh-induced TRPC6 phosphorylation occurs through a PKC-dependent pathway.
It has previously been shown that the phosphorylation of Ser 712 in TRPC3 by PKC abolishes TRPC3 activity (13). Because Ser 712 in TRPC3 is part of a highly conserved five-residue motif (PS 712 PKS) common to all TRPC members, we first substituted the equivalent residue (Ser 768 ) of TRPC6 for alanine. Because Ba 2ϩ is a poor substrate for ATPase pumps (29), and its influx is weak through endogenous channels in HEK293 cells, we measured the activity of TRPC6 and TRPC6 S768A using a Ca 2ϩ depletion/Ba 2ϩ readdition protocol. Cells were incubated for 30 s in Ca 2ϩ -free medium containing 0.5 mM EGTA before depleting the intracellular Ca 2ϩ stores with 5 M CCh. Once the [Ca 2ϩ ] i had returned to the basal level (2 min after the addition of CCh), 1.0 mM Ba 2ϩ was added extracellularly. After CCh-induced Ca 2ϩ store depletion, mock-transfected cells displayed very weak Ba 2ϩ entry that was not modified in the presence of GF1 (Fig. 6A, upper  panel). In the case of TRPC6-expressing cells, CCh induced steady Ba 2ϩ entry that was potentiated by GF1 (Fig. 6A, middle panel). Unexpectedly, in the case of TRPC6 S768A -expressing cells, CCh induced steady Ba 2ϩ entry that was also potentiated by GF1 (Fig. 6A, lower panel). Fig. 6B shows the net fluorescence (⌬F340) increase recorded 2 min after the addition of Ba 2ϩ . These results indicated that Ser 768 is not involved in the PKC-mediated inhibition of TRPC6.  Because the amino acid sequence of TRPC6 contains 11 other intracellular consensus motifs ((S/T)X(R/K)) for PKC phosphorylation (supplemental Table S1), we next individually substituted all the serine and threonine residues for alanine and evaluated the functionality of the mutant channels and their sensitivity to PKC. Mutants TRPC6 T629A and TRPC6 S928A were poorly expressed and their activity could not be evaluated. Fig. 7 shows that GF1 could potentiate Ba 2ϩ entry through wild-type TRPC6 as well as all the TRPC6 mutants. These results suggested that none of the conventional PKC phosphorylation motifs is involved in the modulation of TRPC6 by PKC.
GPS2.1 computational software (30) predicts that PKC␦ could phosphorylate TRPC6 on four additional putative phosphorylation sites (supplemental Table S2). We mutated two of these, Thr 69 and Ser 448 , into Ala and investigated the impact on GF1-potentiated TRPC6 activity. GF1 potentiated CChinduced Ba 2ϩ entry into cells expressing TRPC6 T69A (data not shown). Interestingly, however, CCh-induced Ba 2ϩ entry was not potentiated by GF1 in cells expressing TRPC6 S448A (Fig.  8A). The activity of TRPC6 S448A was also investigated by measuring CCh-induced Ca 2ϩ entry, which was higher in cells expressing TRPC6 S448A than in cells expressing TRPC6 (Fig.  8B). PKC (activated by CCh) likely did not exert its inhibitory effect on mutant TRPC6 S448A . In support of this interpretation, Fig. 8B also shows that GF1 did not modify CCh-induced Ca 2ϩ entry into cells expressing TRPC6 S448A whereas it potentiated CCh-induced Ca 2ϩ entry into cells expressing TRPC6. As shown in Fig. 2B, GF1 only slightly enhanced CCh-induced Ca 2ϩ entry in mock-transfected cells whereas in cells expressing TRPC6, it potentiated CCh-induced Ca 2ϩ entry 2.65 Ϯ 1.57-fold compared with mock-transfected cells (Fig. 8C). However, in cells expressing TRPC6 S448A , GF1 barely increased CCh-induced Ca 2ϩ entry compared with mock-transfected cells (0.37 Ϯ 0.09-fold) (Fig. 8C). These results demonstrated that Ser 448 mutation abolishes the sensitivity of TRPC6 to PKC.
To investigate whether TRPC6 S448A is a substrate for PKC, HEK293 cells were metabolically labeled using [ 32 P]orthophosphate. Under these conditions, PMA increased the phosphorylation level of TRPC6 (Fig. 9A) 1.63 Ϯ 0.20-fold over the basal phosphorylation level. In addition, the increase was prevented in the presence of GF1 (1.07 Ϯ 0.22-fold over the basal level) (Fig. 9B). Under the same conditions, the basal  HEK293T cells transiently transfected with pcDNA3 (mock), TRPC6 (T6WT), or mutant TRPC6 were stimulated with 5 M CCh, and Ba 2ϩ entry was monitored as described in the legend of Fig. 6. Net Ba 2ϩ entry was calculated by subtracting the basal F340 value (average of two values before adding the Ba 2ϩ ) from the average of three values taken 114 -120 s after adding the Ba 2ϩ . GF1-potentiated Ba 2ϩ entry was calculated by subtracting the net Ba 2ϩ entry value under control conditions from the net Ba 2ϩ entry value after the pretreatment with GF1. The histograms represent the potentiating effect of GF1 on CCh-induced Ba 2ϩ entry. The values are the average Ϯ S.E. from two to four independent experiments for the mutants and fourteen independent experiments for the mock and TRPC6 conditions.

PKC-dependent Phosphorylation of TRPC6 on Ser 448
DECEMBER 24, 2010 • VOLUME 285 • NUMBER 52 phosphorylation level of TRPC6 S448A was similar to that of TRPC6 (1.01 Ϯ 0.19), while PMA had no effect (1.18 Ϯ 0.23) (Fig. 9B). As expected, GF1 did not modify the phosphorylation level of TRPC6 S448A in the presence of PMA (1.16 Ϯ 0.27). These results suggested that Ser 448 of TRPC6 is phosphorylated by PKC and is responsible for the PKC-mediated inhibition of TRPC6.
It has previously been shown that the A7r5 cells express high levels of TRPC6 and that knockdown of TRPC6 decreases AVP-induced Ca 2ϩ entry by 50 -70% (31)(32)(33). We investigated the influence of PKC on AVP-induced Ca 2ϩ entry in A7r5 cells. Using the Ca 2ϩ depletion/Ca 2ϩ readdition protocol, we showed that 100 nM AVP caused a net Ca 2ϩ entry of 82.6 Ϯ 26.3 nM 2 min after extracellular Ca 2ϩ was restored. PMA (100 nM) did not alter the Ca 2ϩ release phase (Fig. 10A). However, net AVP-induced Ca 2ϩ entry decreased to 21.2 Ϯ 12.2 nM in cells pretreated with PMA (Fig. 10C). To evaluate the implication of novel PKC isoforms in the inhibition of AVP-induced Ca 2ϩ entry, we used the two structurally similar PKC inhibitors, Gö 6983 and Gö 6976. Gö 6983 can inhibit conventional PKC and novel PKC with similar potencies whereas Gö 6976 is inactive on novel PKC (34,35).   duced Ca 2ϩ entry in A7r5 cells. The specific role of PKC␦ was assessed by transfecting A7r5 cells with PKC␦ siRNA. Treatment of cells with 50 nM PKC␦ siRNA significantly reduced the expression of PKC␦ to 47.0 Ϯ 4.6% (Fig. 11, A and B). Nonspecific PKC␦ siRNA showed no effect. The consequences of knocking down the expression of PKC␦ on AVPinduced Ca 2ϩ entry were evaluated using the Ca 2ϩ depletion/ Ca 2ϩ readdition protocol. Fig. 11C shows that AVP-induced Ca 2ϩ entry was higher in cells knocked down for PKC␦ compared with control cells. In cells treated with siCtl, AVP-induced net Ca 2ϩ entry was 122.8 Ϯ 8.4 nM, whereas in cells treated with siPKC␦, AVP-induced net Ca 2ϩ entry was  215.5 Ϯ 30.8 nM (Fig. 11, C and D). These results demonstrated that the activity of endogenously expressed TRPC6 in A7r5 cells is modulated by PKC␦ as with recombinant TRPC6 expressed in HEK293 cells.

DISCUSSION
PKC plays an important role in cellular functions by regulating many different signaling pathways, including Ca 2ϩ entry (36 -38). In the present study, we showed that PKC also inhibits the activity of TRPC6. These results are in agreement with those of Shi et al. (39), who showed that ionic currents in HEK293 cells transfected with TRPC6 are inhibited by PKC following the stimulation of the muscarinic receptor. We also observed an inhibitory effect of PKC on TRPC6 activity in cells endogenously expressing TRPC6. Stimulating mesenteric artery myocytes with angiotensin II activates two different types of current (40). One of these currents, which is inhibited by an intracellular application of an anti-TRPC6 antibody, is down-regulated by PKC. In PC12D cells, CCh-induced Ba 2ϩ entry is abolished after an siRNA knockdown of TRPC6 (41). CCh-induced Ba 2ϩ entry is also inhibited by PKC in these cells. All evidence to date indicates that the activity of TRPC6 is down-regulated by PKC.
In this study, we also showed that inhibition of PKC slightly increases the CCh-induced Ca 2ϩ entry in HEK293 cells. HEK293 cells endogenously express TRPC1 and TRPC3 (21,42). Previous studies have shown that PKC enhances the activity of TRPC1 (40,43,44) and inhibits the activity of TRPC3 (13,45,46), as well as TRPC4 (45), and TRPC5 (45,47). Also, PKC negatively regulates the activity of Orai1 (48). Therefore, the effects observed with the mock-transfected cells could result from an assortment of PKC-induced inhibition and activation of channels activity.
We showed that TRPC6 expressed in HEK 293 cells is slightly phosphorylated under basal conditions and that the level of phosphorylation is increased by CCh. The CCh-induced TRPC6 phosphorylation was prevented by GF1, a specific PKC inhibitor. It has previously been shown that Ser 712 in human TRPC3 is phosphorylated by PKC and that phosphorylation causes a loss of channel activity (13). For mouse TRPC6, the equivalent residue is Ser 768 . We showed that the activity of TRPC6 S768A is similar to that of wild-type TRPC6, and that activity increased following the inhibition of PKC. Mutations of nine putative consensus phosphorylation sites for PKC ((S/T)X(R/K)) in TRPC6 revealed that none of them is involved in PKC-mediated inhibition of TRPC6. The phosphorylation prediction software, GPS 2.1, identifies four putative phosphorylation sites for PKC␦. One of the potential phosphorylation sites, Ser 448 , was in a sequence that was favorable for phosphorylation by PKC␦. An oriented peptide library was used to show that the optimal sequence for phosphorylation by PKC␦ contains a Phe at position p ϩ 1 and hydrophobic residues at p ϩ 4 and p ϩ 5. In addition, basic residues at p ϩ 2, p ϩ 3, and p ϩ 4 are not favorable for phosphorylation (49,50). The sequence surrounding Ser 448 of TRPC6 contains Phe at p ϩ 1, Thr at p ϩ 2, Ile at p ϩ 3, Phe at p ϩ 4, and Leu at p ϩ 5, which corresponds to the criterion for PKC␦ phosphorylation. As expected, after being stimu-lated with CCh, mutant TRPC6 S448A displayed greater activity than wild-type TRPC6. Moreover, GF1 did not potentiate CCh-induced Ba 2ϩ or Ca 2ϩ entry into cells expressing TRPC6 S448A . More importantly, PMA did not increase the level of phosphorylation of TRPC6 S448A . All these results suggested that Ser 448 is phosphorylated by a non-conventional PKC. Further studies are needed to determine the involvement of PKC␦ or other isoforms of PKC in the phosphorylation of TRPC6. Kim and Saffen (51) previously showed that, in PC12D cells, PKC phosphorylates TRPC6 on Ser 768 . In PC12D cells, TRPC6 forms a multiprotein complex that includes PKC␣. However, the functional consequence of phosphorylating Ser 768 on TRPC6 by PKC had not yet been investigated. The results of all these studies suggest that TRPC6 can be phosphorylated on at least two distinct sites, depending on which isoform of PKC is activated and on the cellular context.
In A7r5 cells, TRPC6 is a major component in the AVPinduced cation current (33,52,53). We showed that, in A7r5 cells, Gö 6983 strongly potentiated AVP-induced Ca 2ϩ entry, whereas Gö 6976, which is inactive on novel PKC isoforms, did not potentiate AVP-induced Ca 2ϩ entry. Furthermore, PMA nearly abolishes Ca 2ϩ entry. This effect of PMA has also been reported by Soboloff et al. (32) who used OAG to activate TRPC6. Also, it has been shown that in smooth muscle cell from rat intrapulmonary arteries, sphingosylphosphorylcholine potentiated the contractile responses induced by prostaglandin F 2␣ and U436619. This potentiation, which is attributed to enhancement of Ca 2ϩ entry, was abolished by the broad spectrum PKC inhibitor Ro-31-8220, but not by Gö 6976 (54). Finally, it was observed that in PKC␦-deficient mast cells, a cell type expressing TRPC6 (55,56), the intracellular Ca 2ϩ concentration is elevated (57). Thus, all these results, which were obtained with recombinant cell models and cell models endogenously expressing TRPC6, suggest that PKC␦ plays a major role in regulating Ca 2ϩ entry.
In summary, our study demonstrated that PKC inhibits TRPC6 activity by phosphorylation of Ser 448 . Because both PKC and TRPC6 are activated following GqPCR signaling, PKC likely operates a negative feedback mechanism on TRPC6 activity to weaken Ca 2ϩ entry.