Voltage-gated sodium channel as a target for metastatic risk reduction with re-purposed drugs

Introduction

The capacity to metastasize is one of the hallmarks of cancer¹ and usually death due to cancer is not caused by the primary tumor but rather by the metastatic spread². The lack of an effective therapy in prevention of metastasis results in a high mortality rate in oncology. So it seems reasonable that if the risk of metastasis can be reduced, the outlook of cancer patients may significantly improve survival and quality of life. Solving the metastasis problem is solving the cancer problem³.

Many natural products, like genistein⁴, resveratrol⁵ and curcumin^6,7 have shown interesting anti-metastasis activity. The same effect has been observed with older pharmaceuticals like aspirin⁸, not-as-old pharmaceuticals such as celecoxib^6,9; new pharmaceuticals like ticagrelor¹⁰, as well as with more sophisticated molecules like dasatinib and ponatinib⁶ or ultrasophisticated drugs, like polymeric plerixafor¹¹.

Many other compounds have also been identified as possessing anti-metastatic effects, including increases in NO¹², cimetidine, doxycycline, heparin and low molecular heparins, and metapristone¹³.

High creativity has been employed in the search for anti-metastatic compounds. For example, Ardiani et al. developed a vaccine-based immunotherapy to enhance CD4 and CD8 T lymphocyte activity against Twist¹⁴. Twist is a transcription factor involved in invasion and metastasis.

Many known pharmaceuticals that are, or were, in use for other purposes than cancer treatment are demonstrating anti-metastatic activity. This is the case for thiobendazole, which is an antifungal, anti-parasitic drug that has been used in medical practice for over 40 years, but which also shows anti-migratory and apoptosis-inducing activity¹⁵. The introduction of Food and Drug Administration (FDA)-approved products which are used for a purpose different for which it was originally approved is called repurposing of a drug.

Many new drugs are being introduced in the area of anti-metastatic activity. One such example, zoledronic acid¹⁶ is a biphosphonate that decreases bone metastasis. Denosumab¹⁷ is another example. It is a monoclonal antibody directed against the receptor activator of nuclear factor kappa B ligand (RANKL) that diminishes the number of circulating cancer cells and prevents bone metastasis. It is in Phase II clinical trials and has the advantage of subcutaneous administration, while zoledronic acid requires intravenous route (for further information on these compounds, see clinical trials NCT01952054, NCT01951586, NCT02129699)¹⁸.

Metastasis is a multi-step development. The different steps in the metastatic cascade can be targeted with a combination of drugs against each step. Migration and invasion are necessary steps for the metastatic cascade. There is no metastasis without prior migration of malignant cells, so that if migration and invasion are blocked, metastasis should not occur.

Invasion is the first step in metastasis, and in a very simplified view, it can be divided into three stages (Shown schematically in Figure 1):

Figure 1. Repurposed drugs acting at different levels of the metastatic cascade.

(uPA: urinary plasminogen activator).

Voltage-gated sodium channels

Neurons and muscle cells (and excitable tissues in general) express voltage-gated sodium channel (VGSC) proteins; tumor cells may also express these proteins. VGSCs are important players in migration and invasion as it will be described in this manuscript.

Sodium channels were first described by Hodgkin and Huxley in 1952 and knowledge about structure and physiology of VGSCs are mainly the result of seminal investigations developed by William Catterall¹⁹.

Sodium channels are glycosylated transmembrane proteins that form passages in the cell membrane for the penetration of sodium into the intracellular space according to their electrical gradients. Voltage-gated sodium channels (also known as VGSCs or ‘NaV’ channels) refers to the mechanism that triggers these proteins to allow sodium movement across the membrane.

There are nine known VGSCs (NaV1.1 to Nav1.9) that are members of the superfamily of VGCSs. NaV1.1, 1.2, 1.3 and 1.6 are found in the central nervous system. NaV1.4 is found in muscle and NaV1.5 in cardiac muscle²⁰.

VGSC is formed by a large subunit (α) and other smaller subunits (β). The α subunit is the core of the channel and is fully functional by itself, even without the presence of β subunits^19–21.

When a cell expresses VGSC α subunits, this means that it is capable of conducting sodium into the cell. The structure of VGSC can be seen in Figure 2 and Figure 3. VGSCs modulate the exchange of Na+ across the cell membrane and the inflow of this electrolyte spikes the action potential in excitable tissues²².

Figure 2. An idealized drawing of α and β units of VGSCs.

Figure 3. Surface and side view of VGSC.

It is well known that expression of VGSCs appears in cancer cells where it is not expressed in their normal counterparts, and plays a significant role in disease progression. Table 1 shows examples of the cancer tissues in which dysregulated expression of VGSCs were identified and the role they play.

Targeting these channels may represent a legitimate way of reducing or blocking the metastatic process.

The role of sodium channel in invasion, metastasis and carcinogenesis is insufficiently known.

Sodium channel proteins and cancer

In 1995, Grimes et al.³³ investigated the differential electrophysiological characteristics of VGSCs in two different rodent prostate cancer cell lines: the Mat-Ly-Lu cell line, which is a highly metastatic line (more than 90% of metastasis to lung and lymph nodes under experimental conditions) and the AT-2 cell line with a much lower metastatic potential (less than 10% chance of developing metastasis in experimental conditions). They found fundamental differences in electrophysiological features between these two cell lines which displayed a direct relationship with in vitro invasiveness. Sodium inward currents were detected only in the Mat-Ly-Lu cell line and inhibition of VGSC protein with Tetrodotoxin (TTX; a powerful inhibitor of VGSCs) significantly reduced the capacity for invasion (mean reduction 33%). On the other hand, TTX showed no effect on invasion of AT-2 cell lines. The TTX-induced reduction of invasion showed a direct correlation with the amount of cells expressing VGSC in the culture.

No fundamental differences in the potassium channels were found between the two cell lines, except for a lower density of potassium channels in the Mat-Ly-Lu cell line. The authors concluded that ion channels may be involved in malignant cell behavior and that VGSCs could play a role in the metastatic process.

In 1997, Laniado et al.³⁴ investigated the presence of VGSC in human prostate cell lines. As in the Grimes research they used two different cell lines: one with a low metastatic potential: the LN-Cap cell line which is androgen dependent and expresses prostate-specific antigen, and the PC-3 line which is more malignant, does not express prostate-specific antigen and exhibits a high rate of metastatic potential.

As in the work by Grimes et al., they found that PC-3, the more malignant cell line, expressed VGSC protein and that inhibition of this channel protein with TTX reduced invasion in a significant way. LN-Cap cells did not express VGSC.

One of the conclusions reached by the authors was that cancer cells expressing functional VGSC had a selective advantage regarding migration and distant metastasis. In the case of both humans and rodents, not all cells in the highly malignant cell cultures showed the presence of the VGSC protein. For example, in PC-3 cell culture only 10% of cells expressed a functional VGSC protein. This is the reason why the authors consider these cells as a clonal evolution that gives pro-tumor and pro-invasive advantages.

The correlation between VGSC protein expression and invasiveness in human and rat prostate cancer cells was confirmed by Smith et al.³⁵ by comparing seven lines of rat prostate carcinoma cells with different metastatic ability, and nine human prostate carcinoma cell lines. In general, invading capacity of the basement membrane and metastatic ability showed a positive correlation with the percentage of cells expressing VGSC. But this positive correlation between percentage of cells expressing VGSCs and the percentage of cells being invasive occurred only up to 27% of the cells being invasive in the rat series and up to 12% of cells being invasive in the human series. Authors suggested that these discrepancies may be due to the necessity of other factors for invasive capability besides VGSC presence; i.e. this protein may represent a prerequisite for the invasive phenotype but other requirements must also be achieved for a full-blown invasive phenotype. Fraser et al.³⁶ determined the key role played by VGSCs in prostate cancer cells in invasion and motility and showed that TTX and phenytoin (PHEN) that are known VGSC blockers, decreased motility and invasiveness while channel openers increased motility. However, the increased invasion capacity in VGSC-expressing cancer cells is not limited to prostate cancers. The same features were found in breast cancer cell lines MCF-7 (estrogen receptor positive), MDA-MB-231 and MDA-MB-468 (both estrogen receptor negative).

Baciotglu et al.³⁸ when experimenting on a rat model of induced breast cancer showed the importance of inhibiting VGSCs in order to inhibit antioxidant response. They observed a survival improvement in rats treated with a VGSC blocker.

An important location of VGSCs in cancer cells is in a cellular region directly involved in migration and invasion: the invadopodia. Invadopodias are protrusions of the plasma membrane, rich in actin that are strongly related to degradation of the extracellular matrix (ECM). Figure 4 and Figure 5 summarize how invadopodia works and the relation between VGSC and invadopodia.

Figure 4. Proton extrusion through VGSC and NHE-1 (sodium hydrogen exchanger-1) produces acidification of extracellular matrix surrounding the cellular membrane (yellow area in the figure).

(Brisson 2013³⁹; Gillet 2009⁴⁰). Acidification activates cathepsine degradation of the extracellular matrix.

Figure 5. VGSCs.

The second mechanism of action is through activation of Src which increases MMP-2 and MMP-9 secretion and activity through phosphorylation of Cortactin. It is postulated that there is a feedback loop starting with MMPs products which induces the development of new invadopodia (Red circle around I). (This figure has been constructed based on references Gillet 2009⁴⁰, Brisson 2013¹³³, Clark 2007⁴² and Mader 2011⁴¹).

According to Brisson et al.³⁹, NaV 1.5 Na+ channels regulate the NHE-1 exchanger protein that increases proton extrusion with extracellular matrix acidification that promotes invasion and migration through activity of cystein cathepsines and degradation of extracellular matrix⁴⁰.

A second mechanism of invasion promotion was described by Mader et al.⁴¹ through the EGFR-Scr-cortactin pathway. Src is activated by VGSCs and Src phosphorylates cortactin. Cortactin is involved in MMP-9 and MMP-2 upregulation and secretion as can be seen in Figure 5. These events lead to matrix degradation, an integral step in cancer cell invasion⁴².

There are nine different VGSC α subunits and four different β subunits. The expression of these subunits may vary in the different tumor cells²¹. For example, NaV 1.5 is overexpressed in astrocytoma, breast and colon cancer. NaV 1.7 is found in breast, prostate and non small-cell lung cancer (NSCLC) and NaV 1.6 in cervical and prostate cancer. This suggests that the α subunits seem to be tissue specific.

The main players in the invadopodia complex, besides the VGSCs are Src kinase, cortactin and Rho-A GTPase. The exact relation between these players is not fully known and needs further research. (For further reading on invadopodia and cortactin, see references 43,44).

One possible relation between invadopodia-Src-VGSCs is described in Figure 6.

Figure 6. Tarone et al.⁴⁶ in 1985 reported the relation between the oncogenic Src and promotion of invadopodia.

Berdeaux et al.⁴⁷ reported that the small molecule GTPase Rho A activity is under control of oncogenic Src and localizes in the invadopodia complex and Durlong et al.⁴⁸ (2013) showed that Rho-A regulates the expression and activity of NaV1.5 and found a positive feedback between NaV1.5 and Rho A in breast cancer cells. According to Timpson et al.⁴⁹, cooperation between mutant p53 and oncogenic Ras activates Rho-A.

Figure 7. Activities of alpha and beta-1 subunits of VGSCs in cancer^131–133.

Onganer and Djamgoz⁴⁵ proposed the hypothesis that VGSC upregulation enhances the metastatic phenotype by enhancing endocytic membrane activity in SCLC.

Andrikopoulus et al.⁵⁰ have demonstrated that VGSCs have pro-angiogenic functions by significantly increasing vascular endothelial growth factor (VEGF) signaling in endothelial cells. Endothelial cells express NaV1.5 and NaV1.7. TTX blocks, and NaV1.5 RNAi decreases endothelial cell proliferation and tubular differentiation that are essential steps in the angiogenesis process.

The important implications of VGSCs in cancer progression and invasion led Litan and Langhans⁵¹ to express that cancer is a channelopathy (For further reading on structure and functions of VGSC see reference 52).

Material and methods

A search was performed in the medical literature to find pharmaceuticals already in use for other purposes than cancer, that as an off-target effect could inhibit VSGCs and to determine if these pharmaceuticals can actually decrease migration, invasion and metastatic potential of cancer. A Pubmed advanced search retrieved 50519 articles under the search condition “voltage-gated sodium channel blocker” during the period of 1981–2015.

The articles that considered drugs that were not in clinical use or FDA-approved were not included in this study, with the exception of resveratrol and natural polyphenols.

The following drugs fulfilling these criteria were found: phenytoin, carbamazepine, lamotrigine valproate, ranolazine, resveratrol, ropivacaine, lidocaine, mexiletine, flunarizine, and riluzole.

A new search was performed in Pubmed for each of the above listed pharmaceuticals with two search criteria: 1) the drug and 2) the term cancer. The period considered was from 1962 to the current time.

Those VGSC blocking drugs that exhibited anti-cancer activity based mainly by other mechanisms are only briefly mentioned; valproic acid and lamotrigine probably act against cancer by histone deacetylase inhibition and riluzole’s anti-cancer mechanism is probably related to the glutamatergic pathway.

Tetrodotoxin is also analyzed in spite of the fact that it is not in clinical use, because it is the traditional model molecule of VGSC blocking, against which other drugs are comparatively tested in the experimental setting.

Results

Discussion

Functionally active VGSCs are expressed in many metastatic cancer cells. This functional expression is an integral element of the metastatic process in many different solid tumors.

The essential role of this protein in invadopodia has been established, so VGSCs became a legitimate target to decrease migration, invasion and metastasis. Repurposed drugs like anticonvulsants, (phenytoin in particular) have shown interesting anti-invasion effects.

Carbamazepine’s ability to induce Her2 protein degradation should be considered an interesting association to trastuzumab.

Targeting VGSCs may act in synergy with anti-angiogenic treatments and with other chemotherapeutic drugs like vinblastine.

On the other hand in the review by Besson et al.¹²⁷ there is a very important remark: VGSC is also present in macrophages and cells related with the immunologic system, so that disrupting VGSC`s activity may deteriorate also anti-tumor immunologic mechanisms.

Flunarizine represents a particularly interesting molecule because it may attack cancer from four different angles: invasion and migration through VGSC blocking, WNT pathway down-regulation, decreased lymphangiogenesis and better oxygenation of hypoxic areas which permits a better arrival of chemotherapeutic drugs and increased sensitivity to radiation. It has never been tested in clinical trials for cancer treatment.

Future directions

New VGSCs blockers are under research. Sikes et al.⁶⁷ developed new blockers based on the phenytoin binding site to VGSC. They found compounds with enhanced activity in VGSC blocking and antitumor activity against human prostate cancer cells.

The association of two or more VGSC blockers may show synergistic enhanced anti-metastatic activity. Nerve growth factor (NGF) increases the number of VGSCs¹²⁹; tanezumab, a new NGF inhibitor diminishes the amount of VGSCs¹²⁸, so we may assume that tanezumab may develop synergistic activity with VGSC blockers. Tanezumab has not been tested in cancer and we think it deserves more research because NGF is also an anti-apoptotic protein¹³⁰.

Stettner et al.¹³⁴ found that men over 50 years of age may be benefited with the use of anticonvulsivants, regarding prostate cancer prevention because they observed lower PSA levels compared with control groups. They also described a ranking of this preventive activity: valproic acid>levetiracetam>carbamazepine/oxcarbazepine>lamotrigine. The authors also observed synergy between these drugs.

Conclusions

Repurposed VGSC blocker drugs, particularly phenytoin, flunarizine and polyphenols, deserve clinical trials as complementary treatment to decrease the metastatic risk.