TRPC1 Protein Channel Is Major Regulator of Epidermal Growth Factor Receptor Signaling*

Background: EGFR regulates cell proliferation and therefore constitute a major target in cancer therapy. Results: EGF triggers a calcium entry through TRPC1 ion channels, which is crucial to allow complete activation of EGFR and its downstream pathways. Conclusion: Calcium entry through TRPC1 constitutes an amplification loop in EGFR activation. Significance: TRPC1 is a potential therapeutic target in cancers resistant to tyrosine kinase inhibitors. TRP channels have been associated with cell proliferation and aggressiveness in several cancers. In particular, TRPC1 regulates cell proliferation and motility, two processes underlying cancer progression. We and others have described the mechanisms of TRPC1-dependent cell migration. However, the involvement of TRPC1 in cell proliferation remains unexplained. In this study, we show that siRNA-mediated TRPC1 depletion in non small cell lung carcinoma cell lines induced G0/G1 cell cycle arrest resulting in dramatic decrease in cell growth. The expression of cyclins D1 and D3 was reduced after TRPC1 knockdown, pointing out the role of TRPC1 in G1/S transition. This was associated with a decreased phosphorylation and activation of EGFR and with a subsequent disruption of PI3K/Akt and MAPK downstream pathways. Stimulation of EGFR by its natural ligand, EGF, induced Ca2+ release from the endoplasmic reticulum and Ca2+ entry through TRPC1. Ca2+ entry through TRPC1 conversely activated EGFR, suggesting that TRPC1 is a component of a Ca2+-dependent amplification of EGF-dependent cell proliferation.

TRP channels constitute a large family of proteins that are expressed almost ubiquitously. The family was designated TRP because of a spontaneously occurring mutation in Drosophila, the photoreceptors of which lacked TRP and responded to a continuous light with a transient receptor potential. The homologous proteins in mammalian cells seem to mediate responses to agonists, pheromones, odorant ligands, temperature, pH, osmolarity, and oxidative stress (1). However, in contrast with voltage-operated and ligand-gated channels that have been studied in detail, the activation and regulation mechanisms of TRP channels are largely unknown and diverse. Structurally, TRP seems constituted of four subunits having each six transmembrane domains. This is similar to voltage-dependent channels, except that they are not positively charged and usually voltage-insensitive. The TRP channel family is divided in six subfamilies, among which the most important are the canonical TRP subfamily (TRPC1-7), the melastatin-related TRP subfamily (TRPM1-8) and the vanilloid-receptorrelated TRP (TRPV1-6) (2). Some of these have been involved in cell proliferation and cancer progression (3). In particular, TRPC1 has been shown to enhance cell proliferation and to regulate cell migration, two processes involved in cancer aggressiveness (4 -6).
TRPC1 is a non-selective cation channel (P Ca /P Na ϳ 1). It is involved in store-operated calcium entry (also named capacitative entry) in cooperation with Orai1 and activated by STIM1, the sensor of endoplasmic reticulum Ca 2ϩ content (7)(8)(9)(10). However, several evidences suggest that TRPC1 can be activated independently of store depletion, but the gating mechanisms are still unknown (11). We previously investigated the mechanisms underlying the role of TRPC1 in cell migration and showed the involvement of calcium-induced activation of calpains and proteolysis of myristoylated alanine-rich protein kinase C substrate, an actin-binding protein possibly involved in focal adhesion (12). Moreover, it has been shown that TRPC1 is a crucial determinant of directionality of migration in response to chemotactic agents (13). Besides, stimulation with epidermal growth factor (EGF) results in TRPC1 channel localization to the leading edge of migrating glioma cells and chemotaxis toward EGF was lost when TRPC1 channel was inhibited (14).
The mechanism underlying the effect of TRPC1 modulation on proliferation remains elusive. Activation of the calcium-sensing receptor by high external calcium ([Ca 2ϩ ] o ) increases cell proliferation and TRPC1 expression (15). Moreover, Ca 2ϩ entry through TRPC1 seems to be involved in the phosphorylation of ERK1/2 upon activation of the Ca 2ϩ -sensing receptor (16). Other reports have shown that TRPC1 depletion induced cell growth arrest by blocking the cell cycle in G 0 /G 1 phase in endothelial progenitor cells or by causing incomplete cytokinesis in gliomas (4,6).
Because TRPC1 is involved in EGF-induced cell migration, we hypothesized that its effect on cell proliferation might be mediated by alteration of EGFR 3 signaling. EGFR has been implicated strongly in the biology of human epithelial malignancies, with therapeutic applications in cancers of the colon, head and neck, lung, and pancreas (17). In particular, overexpression or activating mutations of EGFR are found in 40 to 80% of non-small cell lung carcinoma (NSCLC) and associated with poor prognosis, rendering it an attractive therapeutic target (18,19). Inhibitors of EGFR tyrosine kinase activity are used largely in the treatment of advanced NSCLC (20,21). However, the efficiency of this therapeutic strategy is limited to a subset of patients who ineluctably develop resistance against currently available EGFR inhibitors such as erlotinib and gefitinib (22). It seems therefore crucial to better understand modulation of EGFR signaling pathways.
In the present study, we evaluated the role of TRPC1 channel in the modulation of EGFR activity in two different cellular models of NSCLC. We show that EGF-induced calcium influx through TRPC1 is an essential event in the triggering of EGFR activity and EGF-induced signaling pathways. Consequently, TRPC1 depletion inhibits cell proliferation. This study provides novel insights in the function of TRP channels in the regulation of signaling pathways involved in cell proliferation.
siRNA Transfection-Depletion of TRPC1 was achieved by using a pool of four siRNAs (called as siTRPC1) targeting four different sequences of human TRPC1 mRNA (5Ј-GGACUACG-GUUGUCAGAAA-3Ј; 5Ј-CGACAAGGGUGACUAUUAU-3Ј; 5Ј-GUAAGUGGAUUUGCUCUCA-3Ј; and 5Ј-GACGCAAGC-CCACCUGUAA-3Ј). siTRPC1 as well as the non-silencing control pool of siRNAs (siUNR) were purchased from Thermo Fisher Scientific (Lafayette, CO). A549 and H1299 cells were transfected using DharmaFECT reagent according to he manufacturer's instructions (Thermo Fisher Scientific). Twenty four hours later, cells were plated on six-well plates or on 10-cm diameter Petri dishes. Cells were analyzed 72 to 96 h after transfection. siRNA used in Fig. 1D was purchased from Invitrogen and targeted the 5Ј-AUAUUUAGAAGUCCGAAAGCCAAG-U-3Ј TRPC1 mRNA sequence. It was delivered into A549 cells using RNAiMAX transfection agent according to the manufacturer's instructions.
Quantitative RT-PCR-A549 and H1299 mRNAs were extracted with Ribopure kit (Ambion, Applied Biosystems, Lennik, Belgium) and reversed-transcribed using SuperScript II RNase H (Invitrogen). Gene-specific PCR primers were designed using Primer3. To avoid amplification of genomic DNA, primers were chosen in different exons. The following primers were purchased from Eurogentec (Seraing, Belgium): 5Ј-ACTGTGTAGG-CATCTTCTGTGAACA-3Ј (sense) and 5Ј-GGAGAAAATATA-CCAGAACAAAGCAAA-3Ј (antisense). The ␤2-microglobulin housekeeping gene and TRPC1 cDNAs were amplified in parallel. Real-time RT-PCR was performed using 5 l of cDNA, 12.5 l of SYBRGreen Mix (Bio-Rad) and 300 nM of each primer in a total reaction volume of 25 l. The reaction was initiated at 95°C for 3 min, followed by 40 cycles of denaturation at 95°C for 10 s, annealing at 60°C for 1 min, and extension at 72°C for 10 s. Data were recorded on a MyiQ real-time PCR detection system (Bio-Rad), and cycle threshold (C t ) values for each reaction were determined using analytical software from the same manufacturer.
Each cDNA was amplified in duplicate, and C t values were averaged for each duplicate. The average C t value for ␤2-microglobulin was subtracted from the average C t value for the gene of interest. This ⌬C t value obtained in siRNA-TRPC1 or shRNA-TRPC1 silenced myoblasts, or at different stages of differentiation, was then subtracted from the ⌬C t value obtained in control conditions (siRNA-or shRNA-treated cells, or at day 0 for the time course) giving a ⌬⌬C t value. As amplification efficiencies of the genes of interest and ␤2-microglobulin were comparable, the amount of mRNA, normalized to ␤2-microglobulin, was given by the relation 2 Ϫ⌬⌬Ct .
Flow Cytometry-Cells were counted by flow cytometry. Their volume of distribution was determined by concomitantly detecting fluorescent beads at a known concentration. Cell cycle analysis was performed by permeabilizing fixed cells with 0.01% Triton X-100 and subsequent staining by 50 g/ml propidium iodide (PI). Cytofluorometric analyses were performed on a FACSCalibur equipped with CellQuest Pro software (Becton Dickinson).
Double Thymidine Synchronization-Synchronization of A549 and H1299 cells at the G 1 /S border was performed by adding thymidine (2 mM) to the medium during two consecutive nights.  After the first and second overnight culture, the medium was removed, and cells were washed with PBS (3ϫ) and cultured in normal medium in the absence of thymidine. Cell cycle analysis was performed starting after the second removal of thymidine.
Assessment of EGFR Internalization-Assessment of EGFR internalization has been described by Duan et al. (23). Briefly, cells were stained with 100 ng/ml of Alexa Fluor 488-conjugated EGF at 4°C for 30 min and then rinsed and incubated for 2, 10, or 30 min at 37°C to allow internalization. Finally, they were treated with 0.2 M acetic acid for 5 min to remove noninternalized EGF. Cells were then fixed and analyzed by flow cytometry. Mean fluorescence intensity of cells after EGF binding but without the acid wash was set to 100%. Percentage internalization was calculated after subtracting background (fluorescence cells subjected to acid wash without allowing internalization).
Cytosolic Free Ca 2ϩ Measurements-A549 cells were plated on 22-mm round glass coverslips. Twenty-four hours after FCS removal, cells were incubated with 1 M Fura-2/AM (Calbiochem, Camarillo, CA) in Krebs-HEPES buffer (10 mM HEPES, 135 mM NaCl, 6 mM KCl, 2 mM CaCl 2 , 1.2 mM MgCl 2 , 10 mM D-glucose, pH 7.4) for 60 min at room temperature. Coverslips were then mounted in a heated (37°C) microscope chamber. Cells were alternately excited (1 Hz) at 340 and 380 nm using a Lambda DG-4 Ultra High Speed Wavelength Switcher (Sutter Instrument, Novato, CA) coupled to a Zeiss Axiovert 200 M inverted microscope (ϫ20 fluorescence objective) (Zeiss Belgium, Zaventem, BE). Images were acquired with a Zeiss Axiocam camera coupled to a 510-nm emission filter and analyzed with Axiovision software. Calcium concentration was evaluated from the ratio of fluorescence emission intensities excited at the two wavelengths using the Grynkiewicz equation (24).
Statistical Analysis-Data are presented as means Ϯ S.D. Student's t test and analysis of variance were used to determine statistical significance when appropriate.

RESULTS
TRPC1 Depletion Inhibits Cell Proliferation-We verified by quantitative RT-PCR that TRPC1 channel was by far the most expressed TRPC isoform in A549 and H1299 cells, two models of NSCLC. We used a pool of siRNAs targeting four different sequences of TRPC1 mRNA (referred to as siTRPC1 hereafter) to decrease TRPC1 expression. Immunodetection with an anti-TRPC1 antibody (Epitomics) revealed three bands at Ͼ170 kDa, ϳ130 kDa, and ϳ70 kDa. As suggested in previous studies, the three bands possibly correspond to multimeric and monomeric forms (25). All of these were decreased significantly 72 and 96 h after siTRPC1 transfection (Fig. 1A). Similar results were obtained with another antibody (provided by Alomone; data not shown). Due to the controversy about the specificity of commercially available TRPC antibodies (26,27), we confirmed TRPC1 depletion by quantitative RT-PCR. We observed that the content of TRPC1 decreased by 70 Ϯ 11% (n ϭ 3, p Ͻ 0.05) after siTRPC1 transfection in comparison with cells transfected with an unrelated siRNA (referred to as siUNR hereafter) (Fig.  1B). Treating A549 cells with siTRPC1 significantly inhibited cell proliferation with a doubling time of 31.08 Ϯ 6.5 h in TRPC1 depleted cells versus 23.07 Ϯ 3.5 h in control cells (n ϭ 3, p Ͻ 0.05, values calculated on cell quantification at 48 and 72 h) (Fig. 1C). Similar results were obtained with a siRNA targeted to another TRPC1 mRNA sequence (Fig. 1D).
TRPC1 Depletion Induces G 0 /G 1 Cell Cycle Arrest-To grossly decipher the mechanism of siTRPC1-induced cell growth inhibition, we analyzed cell cycle by DNA staining and flow cytometry measurements. TRPC1 depletion induced a cell cycle arrest in G 0 /G 1 phase in a non-synchronized cell population (Fig. 2, A and C). This effect was made dramatically more visible after double thymidine block (Fig. 2, B and D). Indeed, 6 h after removal of thymidine, the proportion of siTRPC1transfected cells in G 0 /G 1 phase was three times times larger FIGURE 6. TRPC1 depletion disrupts EGF-dependent signaling pathways. Immunoblot analysis of pPDK1, total PDK1, phospho-Ser-473 Akt, total (tot) Akt, p-p44/p42, and total p44/p42 after transfection of A549 cells with siUNR or siTRPC1, cultured for 24 h in the absence of FCS and then treated with 100 ng/ml EGF for the indicated times. Data are representative of at least three independent experiments.

Role of TRPC1 in EGFR Signaling
than in siUNR-transfected cells. Importantly, the effect of siTRPC1 on the cell cycle could be extrapolated to H1299, another NSCLC cell line (Fig. 3).
On the basis of the effects of TRPC1 depletion on the cell cycle, we analyzed the expression of cyclins involved in G 1 /S transition. Expression of cyclins D1 and D3 was reduced 72 or 96 h after siTRPC1 transfection, consistent with G 0 /G 1 blockade induced by TRPC1 depletion (Fig. 4).
TRPC1 Mediates EGFR Phosphorylation and Activates EGFinduced Signaling Pathways-As TRPC1 and EGFR cooperate in cell migration, we measured EGFR phosphorylation after TRPC1 depletion in NSCLC cell lines cultured in complete medium (i.e. supplemented with FCS). We observed that phosphorylation on Tyr-1068 was reduced dramatically 72 and 96 h after transfection with siTRPC1 (Fig. 5A). The total amount of EGFR was not modified as shown in immunoblot (Fig. 5A) and quantitative RT-PCR (data not shown). Using FACS analysis of A549 stained with an Alexa Fluor-coupled EGF, we showed that the expression of EGFR at the membrane in basal conditions was unchanged and that kinetics of EGFR internalization was similar in siUNR-and siTRPC1-transfected cells (Fig. 5B). The decreased amount of phosphorylated EGFR protein can therefore be attributed to a decreased activity of the receptor.
We then analyzed the time course of EGFR autophosphorylation in A549 on three different tyrosine residues in response to stimulation with its natural ligand EGF in control conditions and after TRPC1 depletion. For these experiments, FCS was removed 24 h before EGF stimulation. EGF-induced phosphorylation on Tyr-1068 and Tyr-992, two sites involved in downstream signaling pathways, was largely decreased in TRPC1 depleted cells (Fig. 5, C and D). Similarly, we observed that phospho-Tyr-1068 was decreased in H1299 cells transfected with siTRPC1 and treated by EGF (Fig. 5E). In contrast, the ubiquitination triggering site Tyr-1045 was phosphorylated similarly by EGF stimulation in siUNR-and siTRPC1-transfected cells. This is in line with the observation that EGFR expression is not altered by TRPC1 depletion.
TRPC1 Depletion Alters EGFR Downstream Signaling-TRPC1 depletion dramatically decreased activation of two major downstream signaling pathways involved in cell proliferation. Phosphorylation of PDK1, the major transducer of PI3K action on growth factor-stimulated AGC kinases group to which Akt belongs (28), was decreased after siTRPC1 transfection (Fig. 6). In cells cultured for 24 h in the absence of FCS, Akt, another downstream target of PI3K, was much less phosphorylated in TRPC1depleted cells in comparison with control cells. However, the acute EGF stimulation seems to overwhelm the effect of siTRPC1 on Akt phosphorylation (Fig. 6). Nevertheless, in proliferating cells cultured in complete medium, siTRPC1 did reduce the amount of phospho-Ser-473 Akt (see Fig. 9). TRPC1 depletion also altered MAPK pathway as demonstrated by the decreased phosphorylation of p44/p42 (Fig. 6).
Calcium Influx through TRPC1 Is Essential for EGFR Activation-As observed previously, EGF induced an oscillatory Ca 2ϩ response in ϳ50% of A549 cells cultured 24 h without FCS (Fig. 7A). This response was composed of two phases. The first response was dependent on Ca 2ϩ release from internal stores because it was conserved when cells were stimulated with EGF in the absence of external Ca 2ϩ . The second phase consisted in repetitive peaks of Ca 2ϩ , which were dependent on extracellular Ca 2ϩ (Fig. 7, A and B). Importantly, the second phase was completely abolished in TRPC1 depleted cells (Fig. 7C). To assess that EGF-  triggered Ca 2ϩ entry occurred through TRPC1, we use Alomone blocking antibody to inhibit TRPC1 (29). We observed that second phase of the Ca 2ϩ response to EGF was almost abolished (Fig. 7D). Pretreatment of A549 cells with the specific of IP3 receptor Xestospongin B suppressed both Ca 2ϩ release from the endoplasmic reticulum (ER) and late Ca 2ϩ entry from the external medium (Fig.  7E), suggesting that emptying the ER is an obligatory step to allow Ca 2ϩ oscillations observed in the second phase. The release of Ca 2ϩ from the ER could be achieved by thapsigargin, a well known inhibitor of sarco-endoplasmic reticulum Ca 2ϩ ATPase (SERCA) pumps, and triggered a typical store-dependent entry of Ca 2ϩ (Fig.  7, F and G). The latter was reduced significantly by TRPC1 depletion. As expected, removal of external Ca 2ϩ and addition of 200 M EGTA dramatically reduced the activation of PI3K/Akt and MAPK pathways (Fig. 7H).
However, pharmacological inhibition of calmodulin by 100 M W13 blocked the activation of EGFR and the phosphoryla-tion of one of its downstream target, Akt, demonstrating that EGFR activation was mediated by Ca 2ϩ /calmodulin (Fig. 8A).
In agreement, treatment with W13 blocked cell cycle in G 0 /G 1 phase, mimicking the situation observed after TRPC1 depletion (Fig. 8B).

PI3K/Akt but Not MAPK Is Involved in Cell Cycle
Arrest Induced by TRPC1 Depletion-To further decipher the molecular pathways contributing to the cell growth arrest induced by TRPC1 depletion, we analyzed the involvement of EGFR downstream targets in A549 proliferation. As expected, inhibiting PI3K with LY294002 and MEK1 with PD98059 respectively inhibited phosphorylation of Akt and p44/p42 MAPK (Fig. 9, A  and D). LY294002 induced cell cycle arrest in G 0 /G 1 phase and reduced cyclin D3 and, to a lesser extent, cyclin D1 expression (Fig. 9, B and C). Importantly, siTRPC1 inhibited Akt phosphorylation independently of cell cycle progression (Fig. 9G). In contrast, PD98059 was almost inefficient in altering cell cycle pro- gression and cyclins D1 and D3 expression (Fig. 9, E and F). The effect of TRPC1 depletion on phosphorylation of p44/p42 was finally studied on synchronized cells. We observed that it induced a delay of phosphorylation (maximum observed 6 h after removal of thymidine block instead of 4 h in control cells). This is most likely a consequence of a reduced rate of cell cycle progression. Indeed, variation in phosphorylation status of p44/p42 during the cell cycle has been documented previously (30).

DISCUSSION
EGFR is a transmembrane tyrosine kinase that belongs to the human epidermal growth factor receptor/ErbB protein family. Its mechanism of activation relies on receptor dimerization and autophosphorylation (31). The Grb2 and Gab1 adaptor proteins bind EGFR at phospho-Tyr-1068 and phosphor-Tyr-1086 and lead to the activation of the MAPK and PI3K/ Akt pathways (32,33). PI3K stimulates synthesis of PIP 3 from PIP 2 . PIP 3 activates Akt directly or indirectly via PDK1. Phospholipase C␥ binds at phospho-Tyr-992, resulting in its activation (34). Receptor ubiquitination and degradation are the consequences of EGFR phosphorylation at Tyr-1045 (35).
Pioneer studies showed that addition of EGF causes an increase in cytoplasmic free calcium concentration ([Ca 2ϩ ] i ), which completely depends on extracellular Ca 2ϩ (36,37). More recently however, several studies showed that EGF induced complex oscillatory changes in [Ca 2ϩ ] i due to both a release of Ca 2ϩ from the endoplasmic reticulum and a Ca 2ϩ influx from the outer medium (for review, see Ref. 38). Conversely, Ca 2ϩcalmodulin complex regulates EGFR activity either directly or via calmodulin-dependent kinases. Indeed, binding of the calcium-calmodulin complex to EGFR allows its activation by EGF, whereas phosphorylation of EGFR on Ser-1046 and Ser-1047 by calmodulin-dependent kinases decreases its tyrosine kinase activity and increases its rate of internalization.
The major novelty of our study is to point out the TRPC1mediated Ca 2ϩ regulation of EGFR. We clearly demonstrate that Ca 2ϩ entry through TRPC1 is an obligatory step to allow complete activation of EGFR and its downstream targets in A549 and H1299, two NSCLC cell lines. This effect is mediated by calmodulin because its pharmacological inhibition abolishes EGFR activation. It results in a strong inhibition of G 1 to S transition. Interestingly, the MAPK pathway seems to have a marginal effect on A549 cell cycle progression. In contrast, pharmacological inhibition of the PI3K/Akt pathway strongly blocks cell cycle. This effect is associated with a decreased expression of cyclins D1 and D3. This is compatible with the fact that pAkt increases cyclin D1 and D3 expression, including in NSCLC (39,40). We therefore suggest that the previously reported anti-proliferative effect of TRPC1 blockade might due to EGFR signaling disruption (Fig. 10).
Our results also show that EGF stimulation induces a first [Ca 2ϩ ] i transient due to Ca 2ϩ release from the ER, followed by an oscillatory [Ca 2ϩ ] i response dependent on external Ca 2ϩ .
The first phase can be attributed to phospholipase C-mediated synthesis of IP 3 because xestospongin B completely abolishes Ca 2ϩ transients in response to EGF. The second phase is mediated by TRPC1 activation because it is abolished after TRPC1 depletion. The exact mechanism of EGF-mediated TRPC1 activation remains elusive but could be, at least partially, of capacitative nature. Indeed, it is dependent on IP 3 -mediated Ca 2ϩ release, and TRPC1 depletion inhibits capacitative Ca 2ϩ entry evoked by thapsigargin. We cannot exclude that ER depletion exerts a permissive effect on the process and that EGFR directly or indirectly activates TRPC1. Indeed, TRPC1 has been reported to be directly activated by PIP 3 and by phosphorylation by PKC (41,42). TRPC1 is also able to cluster with stimulated growth factor receptors (43). Altogether, our data suggest that Ca 2ϩ entry through TRPC1 constitutes an amplification loop: it is triggered by EGFR stimulation, and conversely, it enhances EGFR autophosphorylation and activity.
Recently, several TRP channels, including TRPC1, have been associated with proliferative phenotype in breast cancer (44). This correlation could be explained by our results about the effect of TRPC1 expression on EGFR signaling.
In conclusion, our study points out TRPC1 as a major regulator of EGFR signaling and makes TRPC1 an interesting target in the management of NSCLC patients.