Receptor Channel TRPC6 Is a Key Mediator of Notch-Driven Glioblastoma Growth and Invasiveness

Glioblastoma multiforme (GBM) is the most malignant primary brain tumor (1), Despite aggressive treatment approaches, median survival times remain less than 1 year (1). Invasion of GBM cells is the major reason for the failure of lasting success with surgical therapy and for tumor recurrence. A considerable effort has been focused on defining the mechanisms controlling GBM invasiveness and on developing therapeutic strategies aimed at reducing tumor growth and improving survival. Reduced oxygen availability (hypoxia) in the surrounding brain tissue is a major driving force behind GBM growth and aggressiveness (2). The molecular signal(s) that links tissue hypoxia to tumor aggressiveness is poorly understood.

Activation of the Notch pathway may contribute to these phenotypic changes of GBM (3). Notch plays an essential role in regulating cell fate proliferation and migration during normal development of many tissues and cell types (4). The Notch pathway consists of a family of transmembrane receptors and their ligands and Notch target transcription factors (4, 5). Binding of the ligand renders the Notch receptor susceptible to γ-secretase–mediated proteolytic cleavage, which in turn results in the release of the Notch intracellular domain (NICD) from the plasma membrane and its subsequent translocation into the nucleus. NICD interacts with the DNA binding protein CSL (CBF1, Suppressor of Hairless, Lag-1), also known as RBP-Jκ, to regulate the expression of the genes downstream of the Notch signaling, which include Hes1, Hes5, and Herp2 (4–6). In the absence of Notch signaling, CSL represses transcription of Notch target genes, and following activation by Notch, CSL is converted into a transcriptional activator and activates transcription of the same genes. The Notch signaling pathway can maintain cells in an undifferentiated state and have therefore been associated with a growing list of cancers (7). Deregulated expression of Notch receptors, ligands, and targets has been observed in many solid tumors (7–9). High-level expression of Notch1 and Jagged is associated with tumor growth and poor prognosis (7). Several members of the Notch family were found to be differentially expressed in GBMs depending on the degree of malignancy (10).

Although Notch has been associated with an oncogenic role in diverse malignancies, the Notch-regulated gene targets that are critical for the development of the aggressive phenotype remain poorly characterized. Here, we showed that inhibition of the Notch pathway in GBM blocks the hypoxia-induced upregulation of transient receptor potential 6 (TRPC6), a member cation channel of the transient receptor potential (TRPC) subfamily (11). Induction of TRPC6 promotes the aggressive phenotype by promoting a sustained elevation of intracellular Ca²⁺ level, which is critical for glioma proliferation and migration. Clinically, expression of Notch and TRPC6 was elevated in GBM biopsies in comparison with normal brain tissues. Collectively, these data enhance our understanding of the role of Notch and TRPC6 in human malignancies and reveal a specific molecular target that can provide the basis for developing the much needed therapies to treat malignant gliomas.

Culture medium, serum, growth factors, antibiotics, Trizol, SuperScript II RNaseH reverse transcriptase, 4′,6-diamidino-2-phenylindole (DAPI), and Oligofectamine were obtained from Invitrogen. Other reagents include N-[N-(3,5-difluorophenacetyl-l-alanyl)]-S-phenylglycine t-butyl ester (DAPT) and hypoxia-inducible transcription factor-1 (Hif-1) inhibitor {3-[2-(4-adamantan-1-yl-phenoxy)-acetylamino]- 4-hydroxybenzoic acid methyl ester; Calbiochem}; oleoyl-2-acetyl-sn-glycerol (OAG) and bromodeoxyuridine (BrdUrd; Sigma); and SK&F96365 (Tocris).

The human U373MG and HMEC-1 cell lines were obtained from American Type Culture Collection and Centers for Disease Control, respectively. Primary GBM was prepared by dissociation of human brain tumor patient specimens in accordance with a Florida Hospital Institutional Review Board–approved protocol.

Hypoxic conditions were obtained by incubating cells in 100 μmol/L CoCl₂ (12). Cobalt has been widely used as a hypoxia mimetic in cell culture, and it is known to activate hypoxic signaling by stabilizing Hif-1α (12).

The expression levels of TRPCs, Hes1, and Hes5 were detected by quantitative real-time PCR (qRT-PCR) using the iCycler iQ (Bio-Rad) as described (13) using the primer sequences listed in Supplementary Table S1. Each sample was run in triplicate for the target gene and the internal control gene [glyceraldehyde-3-phosphate dehydrogenase (GAPDH)].

Cells were transfected with small interfering RNA (siRNA) targeting human TRPC6 (5′-GGGCAAGGCCUUGCAGCUCdTdT-3′; siRNA-TRPC6) or human Notch1 (5′-UGGCGGGAAGUGUGUG-AAGCG-dTdT-3′; siRNA-Notch1) using Oligofectamine according to the manufacturer's instructions. The target sequences were chosen based on previous experiments testing the gene-silencing effectiveness of three siRNA duplexes (Invitrogen). A nonsilencing sequence was included as a control. Gene silencing effect was evaluated by qRT-PCR and immunoblotting as described (13).

These methods are described previously (13). The following primary antibodies were used: Hif-1 (Abcam), TRPC6 (Chemicon), Notch1 intracellular domain (Val1774-NICD; Abcam), Jagged-1 (Cell Signaling), and nuclear factor of activated T cells (NFAT; Santa Cruz). Equal loading was confirmed by stripping and reprobing the membranes with β-actin (Sigma) or GAPDH (Sigma) antibodies. Densitometry quantitation was determined using the Image J software (NIH). Nuclei in immunostained specimens were visualized with DAPI (Molecular Probes).

Intracellular Ca²⁺ concentration ([Ca²⁺]_i) was measured by Fluo-4 epifluorescence with excitation at 480 nm and emission at 520 nm using a PolarStar plate reader per manufacturer's instructions (Fluo-4 NW Calcium Assay Kit, Molecular Probes).

Cell growth was measured by using MTT cell proliferation assay (14) and by BrdUrd, which incorporates in the newly synthesized DNA and is subsequently detected by immunocytochemistry using a BrdUrd antibody (DAKO), as described previously (15).

Anchorage-independent growth was assessed by colony formation in soft agar as described previously (14). Colonies were counted in a blinded manner using a 10× objective on a Nikon inverted microscope. Each condition was analyzed in triplicate, and all experiments were repeated thrice. Colonies with a diameter larger than 20 μm were scored.

The invasive ability of gliomas with or without treatments was examined by membrane transwell culture system as described previously (14).

The tube formation assay was done as previously described (14). Briefly, HMEC-1 cells were harvested and suspended in conditioned medium collected from U373 cells that were treated with CoCl₂ or siRNA alone or in combination. U373MG cultures left untreated were used as control.

Human GBM (grade 4) surgical biopsy specimens were obtained from the Preston Robert Tisch Brain Tumor Center; Duke University Medical Center and processed in accordance with the Duke University Medical Center Institutional Review Board–approved protocols. Serial sections of 11 specimens (TB# HP0308, HP0323, HP0549, HP0578, HP0591, 0430, 0444, 0445, 0456, 0457, and 0195-3691) were stained with the TRPC6 antibody as described previously (13, 16). The entire tumor was assessed by microscopy and a minimum of three separate fields were used for image collection. Images were acquired by using a Nikon Eclipse E600 fluorescence microscope and processed by using SPOT advance software (Diagnostic Instruments).

The statistical significance of differences between the means of two groups was evaluated by unpaired Student's t test. One-way ANOVA was used to test for differences between two or more independent groups. All statistical tests were two-sided and the level of significance was set at P < 0.05. Calculations were done using GraphPad Prism version 5 for Windows.

Clinically, hypoxia contributes to the development of aggressive phenotype and resistance to radiation and chemotherapy and is predictive of a poor outcome in numerous tumor types including GBM (2). The expression of Notch1 and it ligands, Delta-like and Jagged-1, in human glioma cell lines and primary GBM cultures has previously been shown (16). Because Notch plays a role in the development and progression of various tumors, we assessed Notch activity in gliomas under hypoxia. To this end, we exposed the human glioma cell line U373MG to the hypoxia-mimicking compound CoCl₂ and measured by immunoblotting the level of NICD, which constitutes the activated form of Notch (4). NICD protein level was low under normoxic conditions but was rapidly upregulated by hypoxic treatment (Fig. 1A). Activation of the Notch pathway was also detected when U373MG cultures were subjected to hypoxia (Supplementary Fig. S1A). As control for the hypoxic effect, Hif-1α protein levels were increased in glioma cells (Fig. 1A; Supplementary Fig. S1A). Consistent with the elevated Notch response, the nuclear accumulation of NICD (Fig. 1A) and the expression of the Notch downstream gene Hes (Fig. 1B) were readily detected in glioma cell lines after the hypoxic switch. Specifically, Hes1 and Hes5 levels were markedly elevated in U373MG under hypoxia (Fig. 1B). The protein level of the Notch ligand Jagged-1 was also increased in hypoxic U373MG (Supplementary Fig. S1B), suggesting that ligand-dependent stimulation of the Notch receptor represents a potential mechanism by which activation of the Notch pathway is sustained in hypoxic gliomas. The Notch downstream response was also detected in primary GBM cultures following exposure to CoCl₂ (Supplementary Fig. S1C). These results indicated that Notch signaling is activated in gliomas by hypoxia.

Several recent studies report the involvement of TRPC channels in tumor development and malignant growth (17, 18). To explore the expression of TRPC channels in gliomas, we subjected U373MG to hypoxia and quantified TRPC transcripts by quantitative real-time PCR. TRPC6 mRNA level was markedly increased under hypoxic compared with normoxic conditions (Fig. 1C; Supplementary Fig. S1A). TRPC3 mRNA was also transiently elevated. Expression of other TRPCs seemed less affected by hypoxia. Similar results were obtained when total mRNA was extracted from primary GBM cultures subjected to hypoxia (Supplementary Fig. S1B and C). Immunoblotting and immunofluorescence staining confirmed the induction of TRPC6 protein in U373MG following the hypoxic switch (Fig. 1D). No staining was observed when the primary antibody was omitted or when the antibody was blocked with the TRPC6 peptide (data not shown).

To determine the involvement of Notch in hypoxia-induced TRPC6 expression in gliomas, we used the small-molecule γ-secretase inhibitor DAPT to pharmacologically inhibit the proteolytic processing of Notch to NICD (19). The level of inhibition of Notch activity was assessed by immunoblotting for NICD. Pretreatment with DAPT markedly reduced the amount of NICD (Fig. 2A and B) and substantially inhibited Hes1 mRNA expression (data not shown) in U373MG, which confirms the efficacy of DAPT in inhibiting the hypoxia-induced activation of the Notch pathway. Immunoblotting (Fig. 2A and B) and immunolabeling (Fig. 2C) indicated that pretreatment with DAPT greatly abrogated hypoxia-induced TRPC6 expression. No staining was observed when the primary antibody was omitted or when the antibody was blocked with the TRPC6 peptide (data not shown). To further establish that the Notch pathway is responsible for hypoxia-induced TRPC6 expression, we used siRNA to inhibit endogenous Notch1, which is known to be highly expressed in gliomas (3). Inhibition of Notch1 significantly inhibited TRPC6 protein expression in U373MG following the hypoxic switch (Fig. 2D), confirming that hypoxia-induced TRPC6 protein expression requires activation of the Notch1 pathway.

Next, we determined whether the hypoxia-induced TRPC6 expression was accompanied by a sustained increase in steady-state [Ca²⁺]_i by using Fluo-4 fluorescence spectrophotometry. Basal [Ca²⁺]_i was significantly elevated 8 hours after the hypoxia switch (Fig. 3A). To determine whether the expressed TRPC6 functionally augmented Ca²⁺ entry in U373MG, we applied the membrane-permeant diacylglycerol analogue OAG (100 μmol/L), which induces Ca²⁺ entry through the receptor-operated subtypes TRPC3, TRPC6, and TRPC7 (20, 21). Application of OAG evoked Ca²⁺ transients that were significantly greater in hypoxic compared with normoxic U373MG (Fig. 3A), suggesting that the expressed TRPC6 protein assembled into functional channels. Application of 10 μmol/L SK&F96365, a blocker of TRPCs (22, 23), blocked the OAG-evoked Ca²⁺ transients (Fig. 3A).

To verify that TRPC6 was mainly responsible for the increase in [Ca²⁺]_i and OAG-induced Ca²⁺ transients in hypoxic U373MG, we selectively knocked down TRPC6 expression. Transfection of U373MG with siRNA-TRPC6, but not siRNA-con, substantially blocked the hypoxia-induced expression of TRPC6 mRNA (Fig. 3B) and protein (Fig. 3C). Knockdown of TRPC6 did not alter the expression of other TRPC mRNAs (data not shown). TRPC6 knockdown inhibited the hypoxia-induced elevation of [Ca²⁺]_i and the OAG-stimulated Ca²⁺ entry (Fig. 3A), indicating that TRPC6, but not TRPC3 and TRPC7, is mainly responsible for the enhanced basal Ca²⁺ entry in hypoxic U373MG. Pretreatment with DAPT markedly inhibited the elevation of basal [Ca²⁺]_i and the OAG-evoked Ca²⁺ transients (Fig. 3A), consistent with the notion that Notch signaling mediates TRPC6-dependent Ca²⁺ entry in hypoxic U373MG. As expected, application of siRNA-TRPC6 and DAPT had no significant effect on basal (Fig. 3A) and OAG-evoked increase of [Ca²⁺]_i (data not shown) in U373MG maintained under normoxia.

Ca²⁺ signaling regulates cell growth and proliferation (24, 25). Because TRPC6-mediated Ca²⁺ entry is coupled to the activation of NFAT (26, 27), a Ca²⁺-dependent transcription factor implicated in cell proliferation and hypertrophy-associated gene expression (26–28), we validated the role of TRPC6 in NFAT signaling in gliomas. Hypoxia increased NFAT activation, as evidenced by increased nuclear localization of NFAT (Fig. 4A). Because NFAT activation requires sustained elevation of [Ca²⁺]_i (28), we determined whether the TRPC6-mediated Ca²⁺ entry is important for NFAT activation in gliomas. Knockdown of TRPC6 expression significantly inhibited the accumulation of NFAT in the nucleus (Fig. 4A), indicating that TRPC6 is required for NFAT activation in hypoxic U373MG. Pretreatment of U373MG with FK506, an inhibitor of the Ca²⁺-dependent calcineurin that dephosphorylates and activates NFAT (29–32), significantly attenuated the hypoxia-induced nuclear translocation of NFAT (Supplementary Fig. S2A), further establishing the role of TRPC6 in mediating the Ca²⁺ dependency of NFAT activation in hypoxic U373MG.

Next, we determined whether TRPC6 promotes NFAT-dependent cell proliferation following the hypoxic switch. BrdUrd incorporation was markedly inhibited in CoCl₂-treated cultures in the presence of siRNA-TRPC6 (Fig. 4B). MTT assay confirmed the antiproliferative effect of TRPC6 knockdown (Supplementary Fig. S2B). Similarly to TRPC6 knockdown, treatment with FK506 markedly decreased cell proliferation (Supplementary Fig. S2B). The decrease in cell proliferation in TRPC6 knockdown cells was not linked to cell death because no appreciable cell death was detected 72 hours after treatment with siRNA-TRPC6 (data not shown).

Next, we determined whether TRPC6-mediated Ca²⁺ entry is involved in the malignant growth of glioma cells under hypoxia. Treatment of U373MG with siRNA-TRPC6 decreased the number of colonies (Fig. 5A) and the average colony size (Fig. 5B), indicating that hypoxia-induced TRPC6 promotes in vitro tumorigenesis of glioma cells.

Because Ca²⁺ signals have also been associated with cell polarization and locomotion (29, 33), we next sought to determine whether TRPC6 affects cell migration in a Matrigel-based invasion assay. Under hypoxia, treatment with DAPT substantially reduced the percentage of cells that migrated through the inserts (Fig. 5C), indicating that Notch inhibition impaired glioma cell migration. The anti-invasion activity associated with DAPT was mimicked by TRPC6 knockdown (Fig. 5C), indicating that TRPC6 modulates the migratory and invasive activity of gliomas. NFAT inhibition by FK506 markedly attenuated, but did not abolish, the Ca²⁺ dependency of glioma migration (Fig. 5C), suggesting that Ca²⁺ dependent activation of NFAT contributes, in part, to the hypoxia-induced increase in the metastatic potential of glioma cells.

Hypoxia is associated with tumor growth through the formation of new blood vessels, a process called angiogenesis. Glioblastomas are among the most angiogenic of all human tumors, and the level of angiogenesis in glioblastomas is closely correlated with the degree of malignancy and patient prognosis (1). The calcineurin-NFAT pathway has been implicated in angiogenesis (30, 31), but the underlying mechanism is not clear. To determine whether the TRPC6-calcineurin-NFAT pathway plays a role in angiogenesis, we measured the effect of TRPC6 knockdown or NFAT inhibition by FK506 on the ability of hypoxic U373MG to induce endothelial cell tube formation in vitro. Inhibition of the hypoxia-induced TRPC6 expression and NFAT activation markedly reduced the number of branch points (Fig. 6A), indicating that TRPC6 is essential for the angiogenic potential of glioma cells.

To determine whether the Notch-induced TRPC6 expression is also seen in actual human GBMs, we performed TRPC6 immunohistochemistry on GBM specimens (grade 4) and normal brain tissues. We detected marked TRPC6 expression in GBM specimens by immunohistochemistry (Fig. 6B). By contrast, TRPC6 protein expression was low in the corresponding brain regions of age-matched normal subjects. No specific immunoreactivity was detected when the primary antibody was omitted or when the TRPC6 antibody was adsorbed with the TRPC6 peptide (Fig. 6B).

GBM is the most common and most malignant primary brain tumor in humans. Invasion of GBM cells is the major reason for the lack of lasting success with surgical therapy and for tumor recurrence. Hence, defining the mechanism controlling invasion is essential to improving cancer survival. The low oxygen environment in the brain is positively related to GBM aggressiveness and poor prognosis (32). The role of Hif-1α in tumor growth and invasion is well established (34). Hif-1α protein was undetectable or low in U373 cells under normoxic conditions but increased markedly under hypoxia. Similarly, Notch1 activity was low in glioma cell lines but was elevated after the hypoxic switch (Fig. 1A; Supplementary Fig. S1A). In addition to Notch1, other components of the Notch pathway were increased in glioma cells after the hypoxic switch. Specifically, the levels of Jagged-1 protein were increased under hypoxia (Supplementary Fig. S1B). Consistent with these findings, several members of the Notch receptor family were found to be differentially expressed in gliomas depending on the degree of malignancy (10). Although the findings presented in this study were from the U373 cell line, we observed similar results with the U118 cell line.

The Notch-regulated transcriptional targets that are responsible for the development of the aggressive and malignant phenotypes in GBM remain poorly characterized. In this study, we showed that TRPC6 is markedly upregulated under hypoxia in a manner dependent on Notch activation. Basal expression of TRPC6 is low or undetectable in U373MG (Fig. 1D). Pharmacologic inhibition of Notch blocked the hypoxia-induced upregulation of TRPC6 in U373MG (Fig. 2). The induction of TRPC6 expression was subtype specific because other members of TRPC subfamily were unaffected (Fig. 1C), suggesting that TRPC6 is the major determinant of the increase in receptor-operated Ca²⁺ entry in hypoxic U373MG. Functionally, TRPC6 increases steady-state [Ca²⁺]_I, which was blocked by treatment with DAPT and siRNA-TRPC6 (Fig. 3A). Ca²⁺ entry in hypoxic U373MG was induced by OAG and inhibited by the TRPC blocker SK&F96365, further establishing that TRPC6 is responsible for the sustained increase in steady-state [Ca²⁺]_i in hypoxic U373MG.

Previous in vitro and in vivo studies have suggested that Ca²⁺ channels are important in growth control (35, 36). Specifically, TRPC6 channels have been implicated in cell proliferation and hypertrophic gene expression through the activation of the calcineurin-NFAT pathway in normal (26, 27) and malignant (37, 38) cells. Because glioma cells lack the expression of voltage-gated calcium channels (36) and because Ca²⁺ signaling promotes G₁-S phase transition and cell cycle progression in a variety of cell types (24, 26), the TRPC6-mediated sustained elevation of [Ca²⁺]_i and activation of the calcineurin-NFAT pathway is vital for the proliferation and malignant growth of gliomas under hypoxia (Fig. 4). Consistent with this notion, inhibition of the hypoxia-induced TRPC6 expression causes a dramatic decrease in the activation of NFAT, a transcription factor that is critical for glioma cell proliferation (39, 40).

In addition to cell growth and proliferation, Ca²⁺ signaling also plays a central regulatory role in migration (29, 33). Previous studies have shown that Notch signaling mediates hypoxia-induced tumor migration and invasion under hypoxic environment (41). Here, we showed that suppression of TRPC6 also greatly inhibited glioma cell migration and invasion in response to hypoxia (Fig. 5C). The molecular machinery that is responsible for cell movement is the actin cytoskeleton, which controls cell shape by assembling and disassembling itself, allowing the cell to move along the surface. Calcium-sensitive actin-binding proteins regulate the structure and dynamic behavior of the cytoskeleton. A role for TRPC6 in Rho activation and actin cytoskeleton rearrangements has been suggested (42). The TRPC6-mediated Ca²⁺ entry may contribute to invasion by promoting actin-myosin interactions and the formation and disassembly of cell-substratum adhesions that are important for glioma migration (41, 43). Recent evidence indicates that the activity of several TRPCs including TRPC6 is required for vascular endothelial growth factor–dependent angiogenesis (44) and is increased with epidermal growth factor receptor stimulation (45), suggesting that TRPCs may link growth factor response to tumor growth and invasiveness. Although Notch signaling is critical for TRPC6 upregulation, it remains to be determined whether the Notch pathway directly or indirectly, through cross talk with other transcription factors (46, 47), regulates TRPC6 transcription.

Expression of TRPC6 was higher in GBM biopsies compared with normal brain tissue (Fig. 6B), suggesting that TRPC6 plays a role in the malignant growth of gliomas in vivo. Although we examined only one type of malignant tumor in this study, our findings may also be applicable to other tumors in light of the mounting evidence for an oncogenic role of Notch in multiple types of cancers (7) including stem-like cells (48, 49).

No potential conflicts of interest were disclosed.

This work was supported in part by the James and Esther King New Investigator research grant (S.L. Chan).

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.