Abstract
In the past decade, T-type Ca2+ channels (TTCC) have been unveiled as key regulators of cancer cell biology and thus have been proposed as chemotherapeutic targets. Indeed, in vitro and in vivo studies indicate that TTCC pharmacologic blockers have a negative impact on the viability of cancer cells and reduce tumor size, respectively. Consequently mibefradil, a TTCC blocker approved in 1997 as an antihypertensive agent but withdrawn in 1998 because of drug–drug interactions, was granted 10 years later the orphan drug status by the FDA to investigate its efficacy against brain, ovary, and pancreatic cancer. However, the existence of different channel isoforms with distinct physiologic roles, together with the lack of selective pharmacologic agents, has hindered a conclusive chemotherapeutic evaluation. Here, we review the available evidence on TTCC expression, value as prognostic markers, and effectiveness of their pharmacologic blockade on cancer cells in vitro and in preclinical models. We additionally summarize the status of clinical trials using mibefradil against glioblastoma multiforme. Finally, we discuss the future perspectives and the importance of further development of multidisciplinary research efforts on the consideration of TTCCs as biomarkers or targetable molecules in cancer. Cancer Res; 78(3); 603–9. ©2018 AACR.
Introduction
T-type Ca2+ channels (TTCC) were first reported by Hagiwara and colleagues in the 1970s in voltage-clamped starfish eggs (1). These channels could be distinguished from other voltage-gated Ca2+ channels (VGCC) on the basis of their fast activation and inactivation at negative potentials and slow deactivation kinetics. Because of these peculiarities and small single-channel conductance, they were later coined as T-type Ca2+ channels (T for transient and tiny, TTCC) or low voltage–activated channels, in contrast with the high voltage–activated channels (HVA) comprising the rest of VGCC families. It took quite a few years before the molecular cloning of the first member of the TTCC family, α1G (2), termed Cav3.1 by the nomenclature established in 2000 (3). This was shortly followed by the molecular identification of the two other members of TTCC, α1H (Cav3.2; ref. 4) and α1I (Cav3.3; ref. 5).
TTCCs were initially linked to membrane excitability, such as cardiac pacemaker potentials (6) and neuron oscillatory firing (7, 8). TTCCs are present in central and peripheral neurons, and abnormalities in their expression or function have been linked to a range of neurologic diseases, including absence seizures, epilepsies, and neuropathic pain (9, 10). Nonetheless, TTCCs provide a key pathway for Ca2+ entry in nonexcitable cells, and eventually, their expression was found to be enhanced during the G1–S transition in proliferating cells (11, 12).
G1 to S-phase transition requires Ca2+ influx through multiple Ca2+ channels at the plasma membrane (13). As cells reenter the cell cycle in early G1, Ca2+ elevations promote the activation of AP-1 (Fos/Jun), cAMP-responsive element binding, and nuclear factor of activated T-cell (NFAT) transcription factors, which control the expression, assembly, or stability of cyclin/cyclin-dependent kinase (CDK) complexes essential for progression to the S-phase (Fig. 1; ref. 14).
TTCCs form functional complexes with K(Ca2+) channels in the plasma membrane and trigger signaling pathways that may favor cell-cycle progression or differentiation, depending on cell type and context. The figure combines a selection of relationships identified in different studies for TTCC-mediated signaling (green arrows) and effects of TTCC inhibition (red crosses/arrows), including activation of Cyclin/CDK complexes and CaM-dependent phosphorylation/dephosphorylation events. Arrow tips indicate whether the modulation is positive (pointed) or negative (round). Dashed lines, ion fluxes/transport across the plasma membrane.
The contribution of TTCCs to cell-cycle control is based on their distinct biophysical properties. From a closed-deactivated state, TTCCs activate by weak membrane depolarization, which is rapidly followed by adoption of a closed-inactivated (refractory) state. At the steady state, overlap between activation and inactivation potential ranges leaves a small fraction of TTCC open, enabling sustained inward Ca2+ currents (termed window currents) that may regulate Ca2+-sensitive processes (15, 16). Membrane potential (Vm) is a key regulator of cell cycle and is subject to fluctuations in proliferating cells (17). Of note, Vm is hyperpolarized in the G1 and S-phases, corresponding to the activation or increased expression of different K+ channels, including Ca2+-activated K+ (K(Ca2+)) channels, which form functional tandems with TTCCs (7, 15). G1–S hyperpolarization could lead to increased Ca2+ window currents, or TTCC availability, especially in cells with more depolarized mean Vm, such as stem or cancer cells. The involvement of TTCCs in cell-cycle progression promoted by growth factors was modeled by Gray and colleagues (12): Production of inositol triphosphate triggers Ca2+ release from the endoplasmic reticulum (ER), activating K(Ca2+) channels. The ensuing membrane hyperpolarization removes TTCC inactivation, facilitating a Ca2+ influx that, upon Ca2+ binding to S100 proteins, inhibits the p53/p21 pathway to pass the G1–S restriction point. However, the signaling mechanisms that control cell cycle downstream of TTCC-mediated Ca2+ entry do not appear to be limited to p53 inactivation. Proven transducers of TTCC activity are calmodulin (CaM) and downstream effector calmodulin kinase II (CaMKII; ref. 18). Activation of Cav3.1 heterologously expressed in HEK293 cells has been shown to transiently activate the Ras/MEK/ERK pathway (19), whereas TTCC blockade inhibits the PKB/Akt pathway in glioblastoma multiforme (GBM), both signaling routes being involved in G1–S progression (20). It is also known that Cav3.2 regulates the calcineurin (CaN)/NFAT pathway through both Ca2+ entry and direct binding to CaN, to induce cardiac hypertrophy (21). In addition, Ca2+ influx via Cav3.2 regulates the expression of the Sox9 transcription factor by CaN/NFAT activation during tracheal chondrogenesis (22). Furthermore, Cav3.1-deficient Th cells showed a reduced nuclear translocation of NFAT, in turn leading to a decreased secretion of granulocyte macrophage colony-stimulating factor and unveiling a role for TTCCs in lymphocyte differentiation (23). Thus and notwithstanding the proven role of TTCCs in G1–S progression, the expression of TTCCs can also be associated to cell-cycle exit (Fig. 1).
TTCCs as Prognostic Markers in Cancer
Increased basal Ca2+ influx and remodeled Ca2+ signaling pathways may contribute to tumor progression by enhancing proliferation, promoting invasiveness and conferring chemotherapeutic resistance (24, 25). Hence, important questions are whether TTCCs are differentially expressed in cancer cells, and whether the TTCC signature has prognostic value.
Available data indicate that TTCC expression levels depend on cancer type, stage, and TTCC isoform (Table 1). IHC staining showed that both Cav3.1 and Cav3.2 expression were increased in tumoral versus normal ovarian tissue (26). According to Human Protein Atlas, 82% of GBM biopsies expressed Cav3.1 and 27% expressed Cav3.2. The same database indicates that all prostate cancer samples expressed Cav3.2, while 75% expressed Cav3.1. The expression of Cav3.3 was not determined (27). Maiques and colleagues performed IHC against TTCC comparing normal skin, melanocytic nevi, and different types of melanoma (28). TTCC immunoexpression increased gradually from normal skin to common nevi, dysplastic nevi, and melanoma samples, with differences in the distribution of isoforms. Particularly, Cav3.2 was highly expressed in metastatic compared with primary melanoma. Positive correlations were found between Cav3.2, proliferative and hypoxia markers, and between Cav3.1, autophagy markers and the BRAFV600E mutation. Furthermore, Cav3.2 transcripts and protein were highly expressed in a subset of GBM tumors enriched in glioma stem cells (GSC; ref. 20), consistent with a previously described role for this isoform in stemness (29). However, the expression of the Cav3.1 isoform, which was previously shown at the mRNA level in a vast array of GSC lines (30), was not investigated.
TTCC signature and value as prognostic markers in cancer
Nonetheless, both up- and downregulation of TTCCs can become cancer's molecular signature. Phan and colleagues performed a bioinformatics analysis on the expression of TTCC transcripts in >4,000 cancer tissue samples by accessing Oncomine, a microarray database (31). The three TTCC isoforms exhibited variable levels in several cancer subtypes, including over- and underexpression when compared with normal tissues (Table 1).
The prognostic relevance of the TTCC gene signature in cancer is a crucial question. An in silico genomic analysis of The Cancer Genome Atlas (TCGA) database revealed that disease-free and overall survival correlated inversely with expression of Cav3.1 and Cav3.2 in melanoma (28). Another study using the TCGA database revealed alterations of the Cav3.2 gene in 15 of 136 GBM samples, and these cases presented a trend toward shorter overall survival (20).
Querying the Kaplan–Meier plotter database, Fornaro and colleagues performed a correlation study between TTCC expression in solid tumors and patient survival. In gastric cancer patients, the expression of Cav3.1 was associated to an extended overall survival, whereas Cav3.2 (best single predictor) and Cav3.3 were associated with poorer outcomes. This trend was repeated for lung cancer, whereas in ovarian cancer, patients Cav3.1 and Cav3.2 swapped the sign of their correlations with overall survival (32).
Thus, overexpression of specific TTCC isoforms appears to have a protective effect on specific cancer types and stages. Conversely, DNA aberrant methylation of the CACNA1G gene (encoding Cav3.1 channels) has been found in 18% to 35% of different human primary tumors, including pancreatic (33), hepatic (34), gastric, colorectal, and acute myelogenous leukemia (35, 36). The promoter region of CACNA1G is a target for the CpG island methylator phenotype, which implies the inactivation of multiple tumor suppressor genes (35, 36). The inactivation of CACNA1G may play a role in cancer development by favoring proliferation and/or avoiding apoptotic or autophagic pathways, but few studies have addressed these questions (Table 1).
Effects of TTCC Blockade/Gene Silencing on Cancer Cells: In Vitro Studies
The expression of TTCC in cancer cells was first reported in retinoblastoma Y79 cells shortly after their molecular identification (37). The notion that it was possible to halt cancer cell proliferation or induce cancer cell death by inhibiting TTCC was built up in the following years through in vitro studies on a wide range of cancer cells.
Reduced proliferation
Numerous studies have shown that TTCC pharmacologic blockade or gene silencing reduces the proliferation of cancer cells (12, 38). However, only a few provided insights on the pathways triggered by TTCC blockade/silencing. Lu and colleagues reported that pharmacologic inhibition of TTCC with mibefradil reduced cell proliferation via p53-dependent upregulation of CDK inhibitor p21 (39), halting G1–S progression (Fig. 1). Furthermore, mibefradil-induced activation of the p38–MAPK pathway caused p53 accumulation, resulting in cell-cycle arrest (and death) of colon cancer cells (40).
Inhibition of neuroendocrine differentiation
Unlike most cancer types, neuroblastoma can undergo a spontaneous complete regression through neuronal-like differentiation (41). Chemin and colleagues demonstrated that either pharmacologic blockade of TTCC or interfering with the expression of Cav3.2 (using antisense nucleotides) prevented cAMP-induced neuritogenesis of neuroblastoma–glioma NG108-15 cells and HVA Ca2+ channel expression, indicating a dual role of TTCC in promoting morphologic changes and membrane excitability at early stages of neuronal differentiation (42). Moreover, H2S-induced differentiation of NG108-15 cells involved the phosphorylation of Src kinase and was abolished by ascorbic acid (a proven inhibitor of Cav3.2) and by mibefradil (43). Other than Src activation, Chemin and colleagues found that Ca2+ influx through TTCC promotes neuroblastoma differentiation by an autocrine mechanism (44).
Cav3.2 channels also proved to be relevant for neuroendocrine differentiation of human prostate cancer cells (45–47). During this process, prostate cancer cells develop neurite-like processes and secrete diverse mitogenic factors with paracrine or autocrine actions (47). Opposite to neuroblastoma, differentiation of prostate epithelial cells is associated with an increased aggressiveness of prostate tumors, adoption of an androgen-refractory phenotype, and poor prognosis (48). Two recent works confirmed the mediation of Cav3.2 channels in the differentiation of prostate cancer LNCaP cells subject to physiologically relevant stimuli: sodium butyrate increased the expression of Cav3.2 channels at the mRNA and protein level, and their pharmacologic blockade decreased the number and length of neurite-like processes and cell viability (46). The same research group showed that IL6 upregulated Cav3.2 channels by a posttranscriptional mechanism. Again, pharmacologic blockade of TTCCs limited neurite number and extension (47).
Decreased cell survival
The year 2013 was a turning point in the consideration of the physiologic roles of TTCC in cancer cells. Dzigielewska and colleagues reported a dual effect of mibefradil on reducing proliferation and inducing caspase-dependent apoptosis of colon cancer cells, by p38-MAPK activation and p53 upregulation (40). Our research group proved that the structurally unrelated TTCC blockers mibefradil and pimozide halt melanoma cell proliferation at the G1–S transition and induce the intrinsic apoptosis pathway with activation of caspases-3 and -9 (49). Death occurred after induction of the unfolded protein response to ER stress, followed by subsequent blockade of constitutive autophagy. siRNA-mediated silencing of Cav3.1 and Cav3.2 isoforms exerted similar effects, demonstrating that TTCC play a role in Ca2+ homeostasis maintenance and in sustaining basal macroautophagy. The sequence of events linking TTCC blockade/silencing and autophagy deregulation has only been partially elucidated. Huang and colleagues showed that mibefradil and its derivative NNC-55-0396 exerted a dual role on leukemia cell viability, by promoting both G1–S arrest and apoptosis, which was preceded by ER Ca2+ release and depolarization of the mitochondrial membrane (50). Earlier this year, Niklasson and colleagues performed a drug screening assay on GSCs and found that among different disruptors of Ca2+ signaling network, TTCC and K(Ca2+) channel blockers decreased their viability (30). Transcriptomics and proteomics analyses revealed that upon channel blockade, the GSC plasma membrane depolarized, intracellular Na+ increased, and Na+-dependent transport was reduced, leading to nutrient starvation and cell death.
Nutrient starvation is a stimulus for macroautophagy by activating AMP-activated protein kinase and/or inhibiting the mTOR complex (51). Accordingly, KYS05090, a dihydroquinazoline with TTCC-blocking properties, induced autophagy and apoptosis in lung carcinoma A549 cells through reactive oxygen species generation and subsequent inhibition of glucose uptake (52). Nevertheless, autophagy deregulation by TTCC inhibition appears to be double edged. It is known that Ca2+ is required for phagosome–lysosome function (53); in this scenario, it is tempting to speculate that TTCC inhibition could prevent the influx of Ca2+ necessary for autophagolysosomal formation, de facto blocking macroautophagy at a late step, as observed for the effects of mibefradil in melanoma cells (49).
TTCC blockade/knockdown can also lead to apoptosis by inactivating Ca2+ signaling pathways relevant for cell survival. Valerie and colleagues showed that in addition of inhibiting proliferation, mibefradil or siRNA-mediated Cav3.1/Cav3.2 gene silencing induced the apoptotic death of GBM cells through dephosphorylation of prosurvival Akt and cleavage of caspase-3 and -7 (54). This pathway can also be triggered in ovarian cancer cells, resulting in reduced levels of antiapoptotic survivin (55). Remarkably, these works demonstrated that TTCC inhibition also sensitizes cancer cells to the chemotherapeutic of choice (temozolomide for GBM and carboplatin for ovarian cancer), thus establishing the grounds for the use of TTCC blockers in combined therapies.
Similarly, Zhang and colleagues have recently shown that mibefradil treatment reduced the viability of GSCs, partly due to inhibition of prosurvival Akt/mTOR pathways and upregulation of p27 and BAX proteins (20). A further RNA-seq transcriptomic analyses on GSCs found that mibefradil attenuated the expression of several oncogenes and promoted the expression of different tumor suppressor genes, proving that the signaling pathways stemming from TTCC blockade can be complex and intricate.
Together, these works indicate that the role of TTCCs spans beyond the control of cancer cell cycle into the regulation of cancer cell homeostasis, such that their pharmacologic blockade or gene silencing deregulates Ca2+-dependent physiologic processes pivotal for cell survival (Fig. 2).
TTCC-mediated signaling (green arrows) and effects of TTCC inhibition (red crosses/arrows) on cell survival/apoptosis. The figure combines a selection of relationships identified in different studies. Arrow tips indicate whether the modulation is positive (pointed) or negative (round). TTCC inhibition may induce apoptosis by PKB/Akt dephosphorylation or by activation of the p38–MAPK–p53 axis. TTCC/K(Ca2+) blockade may also cause plasma membrane depolarization and compromise Na+-dependent nutrient transport, in turn inducing the unfolded protein response (UPR) and the amino acid response (AAR), which may convey into apoptosis or autophagy. A question mark “?” indicates that the contribution of TTCC to these Ca2+-dependent processes is speculative Lys, lysosome.
Effects of TTCC Blockade/Gene Silencing on Tumor Growth: In Vivo Studies
Although TTCC blockade/gene silencing has shown to decrease the viability of cancer cells in vitro, albeit with notable exceptions regarding Cav3.1 knockdown, these strategies needed validation in preclinical models before consideration of TTCC targeting in clinical assays. Three-dimensional tumor growth implies a hypoxic microenvironment and altered focal/cell–cell adhesions that shape tumor progression, invasiveness, and sensitivity to therapeutic agents (56, 57). Chronic hypoxia triggers the transcriptional upregulation of Cav3.2 channels in several cell types, mediated by hypoxia-inducible factors (HIF; refs. 58, 59). Thus, the TTCC gene signature in cancer cells is likely to depend on O2 availability and HIF activity. Intriguingly, the connection between TTCC and HIFs appears to be bidirectional: Exposure to mibefradil reduced hypoxia-induced HIF1α and HIF2 in GSCs (20). Furthermore, TTCC blockade using different pharmacologic agents, or Cav3.2 silencing, resulted in a reduced stability of HIF1α protein and expression of VEGF, ultimately inhibiting GBM-induced angiogenesis (60).
Pioneer studies by Jung and colleagues evaluated the antitumor activity of KYS05090 in a mouse lung adenocarcinoma A549 xenograft, which slowed down tumor growth upon intravenous (61) or oral administration (62). Another 3,4-dihydroquinazoline able to block TTCC, KYS05047, demonstrated antitumor efficacy in the same xenograft model when administered orally (63).
Other groups studied the effect of mibefradil or NNC-55-096 against solid tumors growing in vivo. In a xenograft model of ovarian cancer, HO8910 cells developed smaller tumors when coinjected with NNC-55-096 (26). A similar approach was performed on a U87 GBM xenograft model (60). In consonance with in vitro results, intraperitoneal injection of NNC-55-0396 delayed tumor growth by inhibiting angiogenesis with a concomitant reduction of angiogenetic regulators (such as HIF-1α, VEGF and platelet–endothelial cell adhesion molecule).
More elaborated protocols of mibefradil administration have been performed against diverse subcutaneous and intracranial GBM xenografts. Keir and colleagues designed a chemotherapeutic strategy in which mibefradil was first administered to synchronize GBM cell cycle at the G1–S boundary, then withdrawn followed by administration of alkylating agent temozolomide (64). The rationale behind this approach, termed interlaced therapy, was that mibefradil exposure would reduce the time for DNA repair systems to act against temozolomide-induced damage. Indeed, this combined therapy enhanced the efficacy of best single treatment (temozolomide), increasing overall survival by 18% to 68% depending on tumor types, implant location, and treatment schedule.
In addition, TTCC pharmacologic blockade might synergize with radiotherapy, a common therapeutic tool for GBM. A study on rats carrying intracranial C6 glioma implants showed that intraperitoneal injection of mibefradil and simultaneous radiosurgery slowed tumor growth and extended median survival from 35 (radiosurgery alone) to 43 days. The benefits of initiating mibefradil treatment 1 week prior to radiotherapy were even stronger, achieving 52 days of median survival (65). These results suggest that the response to mibefradil in conjunction with ionizing radiation is also schedule dependent.
Recently, administration of mibefradil inhibited the growth of GSC-derived intracranially implanted GBM murine xenografts and sensitized tumors to temozolomide treatment (20). In this study, two cycles of mibefradil (oral) and/or temozolomide (intraperitoneal) were concurrently administered. IHC revealed that proliferation marker Ki67 and stem cell marker SOX2 decreased, whereas astrocyte marker GFAP and caspase-3 increased in mibefradil-treated tumors. Data also showed that single treatments inhibited tumor growth by a similar magnitude and that the combined treatment inhibited tumor growth in an additive fashion. Consistently, both mibefradil and temozolomide alone significantly prolonged animal survival, which was further extended with the combined treatment.
Clinical Trials
Early results attained in murine xenografts encouraged the enrollment of high-grade GBM patients in clinical trials in which TTCC are pharmacologically targeted with mibefradil, a drug with a well-known pharmacokinetic and toxicity profile (66). Cavion Pharma LLC (formerly Tau Therapeutics LLC), a pharmaceutical company focused on drug development and on the repurposing of mibefradil for oncology and neurologic disease, performed in 2012 a dose escalation study to assess the safety of mibefradil dihydrochloride in 30 healthy patients, which rendered only mild and self-limited adverse effects (NCT01550458). This was followed by the launch of a second phase I study, in conjunction with the NCI (Rockville, MD), to assess the efficiency and optimal dosage of mibefradil sequentially administered in combination with temozolomide in patients with recurrent GBM (NCT01480050). The results for this trial indicate that the therapy was generally well tolerated (67). A third trial sponsored by the same company in collaboration with Yale University (New Haven, CT) has also been conducted (NCT02202993, 2014–2017). This was a dose escalation study to determine the safety and the maximum tolerated dose of mibefradil combined with hypofractionated radiation in patients with recurrent GBM, with no results published to date.
Conclusion
Pharmacologic blockade of TTCC reduces the viability of cancer cells in vitro and tumor growth in vivo. Preclinical results spearheaded the first clinical trials employing mibefradil in combined therapies against GBM. Yet, a compelling evaluation of TTCC as prognosis markers and/or targetable proteins in cancer will require a comprehensive characterization of the TTCC molecular signature, and a deeper knowledge of the cell signaling pathways stemming from TTCC activation/inhibition. Individual TTCC isoforms play different roles in cancer pathophysiology, but this notion is hampered by the absence of selective pharmacologic modulators. The expression of Cav3.1, which showed a positive correlation with autophagy markers, is predominant in some cancer types, while epigenetically silenced in others. The expression of Cav3.2, which increases in hypoxic conditions, has been associated to cancer stemness, aggressiveness, and metastasis. The expression of Cav3.3 across cancer tissues remains largely unexplored, in spite of current evidences for a negative correlation with survival of gastric, lung, and ovarian cancer patients. In addition to cancer, TTCCs are currently under the scope of different biomedical fields, including neurologic and cardiovascular disease. Multidisciplinary research efforts are bound to facilitate the development of isoform-specific tools and will hopefully galvanize fine-tuned approaches for different cancer types and stages.
Disclosure of Potential Conflicts of Interest
No potential conflicts of interest were disclosed.
Acknowledgements
We apologize to all colleagues whose contributions are not cited here due to space limitations. Marta C. Sallán and A. Visa were UdL predoctoral fellows. S. Shaikh was funded by the Marie Curie Cofund programme. Mireia Nàger was funded by IRB Lleida-Diputació de Lleida. Research in our laboratory is funded by Instituto de Salud Carlos III/FEDER “Una manera de hacer Europa” (grant PI13/01980 to J. Herreros).
- Received October 6, 2017.
- Revision received October 24, 2017.
- Accepted November 14, 2017.
- ©2018 American Association for Cancer Research.
References
Volume 78, Issue 3
