Ion channels and tranporters in cancer. 2. Ion channels and the control of cancer cell migration

Abstract

A hallmark of high-grade cancers is the ability of malignant cells to invade unaffected tissue and spread disease. This is particularly apparent in gliomas, the most common and lethal type of primary brain cancer affecting adults. Migrating cells encounter restricted spaces and appear able to adjust their shape to accommodate to narrow extracellular spaces. A growing body of work suggests that cell migration/invasion is facilitated by ion channels and transporters. The emerging concept is that K⁺ and Cl⁻ function as osmotically active ions, which cross the plasma membrane in concert with obligated water thereby adjusting a cell's shape and volume. In glioma cells Na⁺-K⁺-Cl⁻ cotransporters (NKCC1) actively accumulate K⁺ and Cl⁻, establishing a gradient for KCl efflux. Ca²⁺-activated K⁺ channels and voltage-gated Cl⁻ channels are largely responsible for effluxing KCl promoting hydrodynamic volume changes. In other cancers, different K⁺ or even Na⁺ channels may function in concert with a variety of Cl⁻ channels to support similar volume changes. Channels involved in migration are frequently regulated by Ca²⁺ signaling, most likely coupling extracellular stimuli to cell migration. Importantly, the inhibition of ion channels and transporters appears to be clinically relevant for the treatment of cancer. Recent preclinical data indicates that inhibition of NKCC1 with an FDA-approved drug decreases neoplastic migration. Additionally, ongoing clinical trials demonstrate that an inhibitor of chloride channels may be a therapy for the treatment of gliomas. Data reviewed here strongly indicate that ion channels are a promising target for the development of novel therapeutics to combat cancer.

according to the national cancer institute, cancer is the second most common cause of death in the United States. Given that 40 years have passed since President Richard Nixon declared a “War on Cancer,” this statistic is especially concerning. The failure to improve disease outcome can be largely attributed to a lack of specific therapies. The current standard of care for most cancers has changed little in the past 40 years and still includes surgery, radiation therapy, and chemotherapy. While these treatments decrease mortality and enhance the quality of life for many, it is inadequate for those suffering from more aggressive cancers. New targets for therapeutic intervention must be identified to combat this widespread disease.

A Role For Ion Channels in Cancer Cell Migration

A growing body of evidence indicates that ion channels and transporters play integral roles in cancer biology and may be promising novel targets for clinical intervention. Ion channels have been implicated in many aspects of cancer pathology, including uncontrolled growth, decreased apoptosis, disorganized angiogenesis, and aggressive migration, invasion, and metastasis (59). In this article, we will review accumulating evidence demonstrating that malignant cells hijack physiological mechanisms for cell migration, especially the use of ion channels to promote motility.

Cell migration plays an integral role in several normal physiological processes, including neural crest cell migration, leukocyte extravasation from the vasculature, and fibroblast migration during wound healing. Cell migration is also critical to cancer metastasis and malignant progression. Despite the heterogeneity in cell types, many of the underlying mechanisms facilitating migration are shared or identical. Migrating cells are polarized and move along a front-to-back axis (53). The cell's leading edge is characterized by a flat and mobile lamellipodium, which pulls the cell forward via rapid actin polymerization (77). Through a hypothesized “treadmilling model” actin monomers are added onto actin filaments directly abutting the plasma membrane of the leading edge. Thus the continuously growing edges of actin filaments push the membrane forward and extend the lamellipodium (87). The leading edge is additionally extended forward through the “lipid flow model,” involving the endocytosis of plasma membrane from the posterior of a migrating cell and subsequent insertion at the leading edge. This endocytic recycling of membrane brings integrins (57) and ion channels to the anterior of the cell, facilitating migration. Integrins serve as the point of attachment between migrating cells and the substratum, regulating adhesiveness and migration speed (55) and providing points of traction for directional movement. Beyond the leading edge, contraction of myosin II in the posterior of migrating cells propels the cell forward (14). While descriptions of cell migration have historically focused on the cytoskeleton, a growing body of evidence now indicates that ion channels are also a necessary component of the cellular migratory machinery.

The finding that ion channels play a role in the migration of malignant and nonmalignant cells is intriguing. While still poorly understood, it is hypothesized that ion channels facilitate hydrodynamic changes in migrating cells, which are 70% water (41). Specifically, ion channels facilitate migration by fluxing osmotically active ions, leading to the osmotic movement of water. Thus ion channels are able to efficiently regulate global cell volume or create local osmotic gradients facilitating the swelling or shrinking of cellular processes. Several models have sought to explain how ions acting as osmolytes can facilitate migration. In two-dimensional migration, channel-mediated influx of ions can facilitate osmotic water entry through aquaporins, protruding the leading edge (56, 73). Similarly, channel-mediated release of ions at the lagging edge allows osmotic water release and cell shrinkage (73). Models of three-dimensional migration additionally account for spatial constraints encountered by cells moving through narrow spaces. This has been especially well-characterized in gliomas, which are the most common and most lethal type of primary brain cancer affecting adults (79). Therefore, we will focus most of our attention in this review on gliomas and extrapolate salient features to other migratory cells. Glioma cells actively move from a central mass to adjacent parenchyma by migrating through the narrow extracellular space. This is facilitated by overexpression of ion channels involved in cytoplasmic volume regulation. In the following sections, we will review the mechanisms by which malignant cells flux ions acting as osmolytes to enhance migration, thereby hijacking physiological mechanisms for cell migration. We will also discuss the regulation of ion channels involved in migration and the transporters that accumulate ions to create a driving force for directional ion and water movement.

Ion Transporters Create a Driving Force For Directional Ion Movement

For nonmalignant and malignant cells to flux osmotically active ions through channels, there must be a driving force for ionic movement. This ionic gradient is created by transporters, which accumulate ions. NKCC1 is an active cotransporter that brings Na⁺, K⁺, and 2Cl⁻ into the cell and plays a major role in Cl⁻ accumulation. By accumulating Cl⁻ intracellularly, migrating cells can use the electrochemical driving force for Cl⁻ efflux to osmotically release cytoplasmic water, thereby modulating cellular volume and migration. A recent study demonstrated that glioma cells, probably via NKCC cotransport, accumulate intracellular Cl⁻ concentration ([Cl⁻]_i) to about 100 mM compared with 10 and 40 mM for mature neurons and astrocytes, respectively (23). The [Cl⁻]_i in glioma cells was determined by transfecting cells with the GABA_C channel, obtaining a gramicidin-perforated patch clamp to maintain [Cl⁻]_i, then stimulating with GABA to determine the reversal potential for Cl⁻ (E_Cl⁻) (23). E_Cl⁻ under gramicidin-perforated patch clamp in human glioma cells was measured to be −7 mV (23), which is positive to the resting membrane potential. If the high resting Cl⁻ gradient was dissipated in glioma cells, there was a 33% decrease in volume and 40% inhibition of migration (23). Since volume changes are mediated by the flux of Cl⁻ along with the obligated osmotic flux of water, the intracellular accumulation of Cl⁻ provides a gradient for efficient cell volume regulation and migration.

Human glioma cells strongly express NKCC1, which also directly plays an integral role in migration through confined spaces (22). NKCC1 localizes to the leading edge of migrating glioma cells, and genetic knockdown of NKCC1 inhibits in vitro and in vivo migration (22). Interestingly, NKKC1's potentiation of migration may be due to its role in regaining cellular volume after a hypertonic challenge (22). NKCC1 also plays a role in the migration of nonmalignant cells. For example, NKCC1 regulates the migration of neuroblasts along the rostral migratory stream (52).

If Cl⁻ channels play an important role in volume regulation and migration, there must be a cation that also concomitantly permeates the cell to maintain electroneutrality. K⁺ channels enhance the migration of malignant cells by allowing K⁺, which is intracellularly accumulated by the Na⁺-K⁺-ATPase, to exit cells and enhance volume regulation. Thus K⁺ channels and Cl⁻ channels may work in concert to allow the efflux of K⁺ and Cl⁻ ions, leading to osmotic water release and cytoplasmic condensation in migrating malignant cells.

Ion Channels Facilitate Malignant Cell Migration

Malignant cells can commandeer physiological mechanisms for migration to increase disease spread. Specifically, malignant cells overexpress a variety of K⁺, Cl⁻, and Na⁺ channels that use the ionic gradients created by NKCC1 and the Na⁺-K⁺-ATPase to facilitate motility and invasion. Several preliminary studies demonstrating that pharmacological inhibition of ion channels decreases malignant motility led investigators to hypothesize that ion channels facilitate the migration of transformed cells. The molecular identities of many of these channels were subsequently discerned through knockdown studies. Here we highlight several ion channels whose activity is correlated with migration.

Cancer cells express several types of K⁺ channels contributing to migration. This has been characterized in human glioma cells, which express Ca²⁺-activated K⁺ channels contributing to motility through K⁺ efflux. Pharmacological inhibition of large conductance calcium-activated potassium channels (BK), a type of Ca²⁺-activated K⁺ channel, inhibits glioma cell migration by 60–80% (88). Glioma cells overexpress BK channels, and the tumor grade of human glioma tissue specimens directly correlates with BK channel expression (39). These BK channels produce a strongly outwardly rectifying K⁺ current that is inhibited by monovalvent cation substitution or drugs, such as iberiotoxin, charybdotoxin, quinine, tetrandrine, and tetraethylammonium ion (TEA) (64). Short hairpin RNA (shRNA) knockdown reduces the expression of BK channels in glioma cells and eliminates iberiotoxin-sensitive currents by 70% (88). The role of BK channels in cell motility may extend to breast cancer cells where the KCNMA1 gene, which forms the pore subunit of the BK channel, is upregulated (28). Genetic or pharmacological inhibition of the BK channel decreases breast cancer cell migration and invasion (28). Other Ca²⁺-activated K⁺ channels also play a role in cancer cell migration. Breast cancer cells express functional SK channels with currents sensitive to apamin, 4-aminopyridine (4-AP), and TEA inhibition, and genetic knockdown of SK channels decreases cell migration (26, 58). SK channels also promote the enhanced migration of melanoma cells (11). Intermediate-conductance calcium-activated potassium channels (IK) enhance glioma cell migration induced by chemokines such as CXCL12 (74). These data indicate that Ca²⁺-activated K⁺ channels facilitate neoplastic migration. Of note, these channels are regulated by intracellular Ca²⁺, which as discussed below, may be the coordinating signal that orchestrates channel activity during migration.

If K⁺ channels enhance the migration of malignant cells via K⁺ efflux, then these malignant cells must also express channels to release anions to maintain electroneutrality; Cl⁻ has been strongly implicated as this anion. Indeed, glioma cells express Cl⁻ channels on lipid rafts (51), colocalizing with BK channels on the invadapodia of migrating cells (51, 89). Whole cell patch-clamp of glioma cells in acute human glioma tissue exhibits voltage-activated chloride currents (85). These chloride currents are sensitive to inhibition by chlorotoxin, a 36-amino acid peptide purified from the venom of the deathstalker scorpion (Leiurus quinquestriatus) (85). Chlorotoxin, as well as substitution of Cl⁻ with impermeant anions, inhibits glioma cell migration and invasion (81). The molecular candidate for the Cl⁻ channel in glioma cells probably belongs to the CLC family of ion channels. Glioma cells express mRNA for ClC-2, -3, -4, -5, -6, and -7, but protein was only detected for ClC-2, -3, and -5 (42, 54). However, only ClC-2 and ClC-3 produce Cl⁻ currents that are discernable by pharmacological and genetic inhibition, suggesting the exclusive intracellular localization of ClC-5 (54). ClC-3 produces a slightly outwardly rectifying current that exhibits time- and voltage-dependent inactivation and is sensitive to DIDS, 5-nitro-2(3-phenylpropylamino)benzoic acid (NPPB), and Cl⁻ replacement by gluconate or glutamate (54). ClC-2 and -3 localize to lamellipodia on the leading edges of glioma cells (54), placing them in a prime location to flux Cl⁻ and facilitate migration. Indeed, ClC-3 knockdown significantly inhibits migration through Transwell barriers (15) and invasion through Matrigel (42). This appears to correlate with disease severity, as ClC-3 protein expression is enhanced in Grade IV tissue (15). In nasopharyngeal carcinoma cells, NPPB and impermeant anions such as gluconate also block Cl⁻ currents and migration (47). Thus ClC-3 endogenously expressed by human glioma cells and nasopharyngeal carcinoma cells facilitates migration and invasion (42, 46).

By use of pharmacological and genetic inhibitors, the integral role of ion channels in the migration of malignant cells has been elucidated. In the following section we explore how these channels, including Ca²⁺-activated K⁺ channels and ClC-3, mechanistically contribute to cell migration. As in nonmalignant cells, ion channels and transporters promote the migration of neoplastic cells by facilitating shape and volume changes, allowing cells to navigate through narrow and tortuous extracellular spaces.

Volume Regulation in Migrating Malignant Cells is Facilitated by Ion Channel Activity

Ca²⁺-activated K⁺ channels and ClC-3, along with several other ion channels and transporters, are involved in neoplastic cell migration. As depicted in Figs. 1 and 2, we hypothesize these channels facilitate migration via modulation of cellular volume, which is necessary when cells migrate through narrow spaces. Cl⁻ channels and K⁺ channels colocalize in caveolar lipid rafts on the invadapodia of migrating glioma cells (51). Concentration of ion channels on the leading edge of migrating cells allows K⁺ and Cl⁻ efflux, enabling invading cellular processes to collapse volume and migrate into confined spaces.

Fig. 1.

Ion channels facilitate neoplastic cell migration by modulating cell volume. In migrating malignant cells, Ca²⁺ influx from Ca²⁺-permeable channels increases intracellular Ca²⁺ concentration ([Ca²⁺]_i) (1). This may then lead to activation of Ca²⁺-activated K⁺ channels, like the BK channel, which is directly sensitive to increases in [Ca²⁺]_i (2). ClC-3, a voltage-gated Cl⁻ channel, is activated via phosphorylation by CaMKII, a Ca²⁺-sensitive kinase. Coordinated K⁺ and Cl⁻ efflux leads to osmotic water release from the cytoplasm, decreasing the volume of the migrating cell (3). Volume condensation then facilitates migration through narrow extracellular spaces (4).

Fig. 2.

A variety of ion channels and transporters enable malignant cells to invade through narrow spaces. Transporters and channels involved in regulatory volume decrease (RVD), such as ClC-3 and the K⁺-Cl⁻-cotransporter (KCC), can extrude osmotically active ions to collapse the cytosolic volume (1). This may facilitate cancer cell invasion through confined spaces. Transporters and channels involved in regulatory volume increase (RVI), such as Na⁺-K⁺-Cl⁻-cotransporters (NKCC) and epithelial sodium channels (ENaC), can take in osmotically active ions to increase the cytosolic volume (2). This may allow cancer cells to regain cytoplasmic volume after squeezing through a spatial constraint. Additionally, proteins like Na⁺/H⁺ exchangers (NHE) extrude protons at the leading edge of migrating cells, acidifying the extracellular space to facilitate degradation of matrix components. AQP, Aquaporin.

The Cl⁻ channels mediating voltage-activated chloride currents contributing to glioma cell migration may be the same channels mediating volume-sensitive chloride currents. In human glioma cells, hypotonic solutions activate an outwardly rectifying current that reverses near E_Cl⁻ and is time and voltage inactivating (63). This current is sensitive to Cl⁻ channel inhibitors such as DIDS, tamoxifen, Zn²⁺, and NPPB (63). Importantly, glioma cells have a basal Cl⁻ current at rest, and antagonists of volume-activated chloride currents significantly increase the membrane resistance of glioma cells (63). Glioma cells regulate [Cl⁻]_i so that E_Cl⁻ is positive to the resting membrane potential, leading to net Cl⁻ efflux at rest (23). Thus, given that there is a driving force for Cl⁻ efflux at rest, Cl⁻ channels are able to play an important role in glioma cell migration. Chemotactic transwell migration of glioma cells is inhibited by 30 μM NPPB, suggesting that volume-sensitive Cl⁻ currents are active during cell migration (63). This was confirmed when NPPB- and volume-sensitive Cl⁻ currents were detected after whole cell patch clamping of migrating glioma cells in the absence of an osmotic challenge (63). Additionally, 5 μM chlorotoxin inhibits glioma cell migration and invasion by decreasing regulatory volume decrease (RVD), the process allowing a cell to decrease in cellular volume after swelling. Chlorotoxin also decreases membrane permeability to Cl⁻, as indicated by the Cl⁻-sensitive dye MQE (81). NPPB and Cd²⁺, Cl⁻ channel blockers that inhibit Cl⁻ currents, also inhibit RVD in glioma cells by 60–70% (17). These data support the hypothesis that volume-sensitive Cl⁻ channels contribute to dynamic volume regulation necessary for cells migrating through narrow extracellular spaces.

This association of volume-activated Cl⁻ currents and cell migration extends to several other neoplastic cell types. NPPB blocks volume-activated Cl⁻ currents, RVD, and migration of nasopharyngeal carcinoma cells (47, 48). Hypotonic challenges increase volume-activated Cl⁻ currents and migration, whereas hypertonic challenges decrease volume-activated Cl⁻ currents and migration (47). Ion replacement studies also demonstrate a correlation between volume-activated Cl⁻ currents and cell migration, because replacement of Cl⁻ with impermeant anions such as gluconate impairs cell migration and volume-activated Cl⁻ current (47). In addition, Cl⁻ channel blockers such as NPPB, DIDS, niflumic acid, and tamoxifen inhibit RVD and invasion of human ovarian cancer cells through Matrigel (38). Thus there is a tight correlation between volume-activated Cl⁻ currents, RVD, and migration in malignant human cells.

The molecular identity of this volume- and voltage-activated Cl⁻ channel in malignant cells contributing to volume regulation and migration is ClC-3. ClC-3 knockdown reduces RVD and volume-activated chloride currents in nasopharyngeal carcinoma cells, correlating with an inhibition of Transwell migration (46). Thus ClC-3 may regulate the migration of malignant cells by modulating cell volume. Likewise, in human glioma cells, ClC-3 knockdown significantly inhibits migration and invasion (15, 42).

Although several groups find that ClC-3 does not mediate volume-sensitive Cl⁻ currents using ClC-3 knockout animals (1, 21, 82), a recent study using an inducible ClC-3 knockout mouse demonstrates that ClC-3 deletion eliminates volume-sensitive Cl⁻ currents (91). These studies suggest that ClC-3 knockout animals may upregulate compensatory Cl⁻ channels to mediate volume-sensitive Cl⁻ currents, and inducible ClC-3 knockout allows measurement of volume-sensitive Cl⁻ currents before upregulation of compensatory Cl⁻ channels. Nevertheless, in migrating malignant cells, Cl⁻ channels like ClC-3 facilitate migration through efficient volume regulation.

Preliminary data from our laboratory indicate that IK channels play a role in the volume regulation of glioma cells. Therefore, Ca²⁺-activated K⁺ channels, including BK, IK, and SK channels, may contribute to cancer cell migration through modulation of cell volume. This is illustrated in Fig. 1, where Ca²⁺ influx leads to K⁺ efflux through BK channels, contributing to volume condensation. However, despite the data implicating BK channels in cancer cell migration, at least two reports have concluded that BK channels inhibit glioma cell migration (6, 34). In both studies, however, cell migration was assayed in 2D, without any requirement for the cells to modulate volume while migrating through confined spaces, as is expected in vivo. Therefore, K⁺ efflux through the BK channel may allow for cancer cell migration via volume modulation, and future studies on cancer cell migration should mirror the spatial constraints of the extracellular space found in vivo.

Remarkably, cancer cells seem to be recapitulating mechanisms for migration utilized by nontransformed cells. As described by Schwab et al. (73), K⁺ flux through IK potentiates migration through modulation of the cellular volume. Interestingly, cell migration decreases only when IK is inhibited at the posterior of the cell, where IK is predominantly active (72). Using atomic force microscopy to measure cell volume, activation and inhibition of IK leads to cellular shrinking and swelling, respectively, in the cell's posterior, indicating that IK activity is polarized (70). The polarized activity of IK seems to be secondary to intracellular Ca²⁺ gradients in migrating cells. Intracellular Ca²⁺ concentration ([Ca²⁺]_i) is higher in the soma compared with the lamellipodium, leading to preferential IK activation in the cell body, since IK is a Ca²⁺-sensitive channel (71). Thus IK, by mediating K⁺ efflux from the posterior of the cell secondary to high [Ca²⁺]_i, facilitates migration by modulating cell volume.

Several Other Channels and Transporters Facilitate Volume Regulation and Migration in Malignant Cells

In principle, any K⁺ or Cl⁻ channel may function to coordinate KCl flux, and different cell types have been shown to use a variety of channels and transporters in this vein. Figure 2 illustrates several other ion channels and transporters that facilitate neoplastic cell migration. NHE1, a Na⁺/H⁺ antiporter, is responsible for chemokine-induced cell swelling and migration in human polymorphonuclear leucocytes (65). NHE1 localizes to the lamellipodium along with AE2, a Cl⁻/HCO₃⁻ exchanger, where it facilitates volume increase at the cell anterior (29). Malignant cells also use NHE1 to promote migration. NHE1 localizes to actin-rich pseudopodia in migrating cancer cells (35), and inhibition of NHE1 decreases cell movement (10, 35). This may be due to the ability of NHE1 to localize to the leading edge and create a proton gradient across the surface of migrating cells, which is necessary for a migratory phenotype (84) (Fig. 2). Inhibition of NHE1 prevents the formation of lamellipodia important in cell invasion (83). NHE1 may also enhance migration by stabilizing integrin α2β1 interactions with the extracellular matrix through proton extrusion (83).

KCC, a cotransporter that releases K⁺ and Cl⁻ to the extracellular space, also plays a role in the migration of transformed cells (12). KCC expressed by human cervical cancer cells enhances dihydroindenyl-oxy-alkanoic acid (DIOA)-sensitive RVD (75), potentiating invasiveness (76).

KCC4, which localizes to lipid rafts and lamellipodia, plays a role in invasiveness and contributes to Cl⁻ efflux through RVD (13). In glioma cells, KCC contributes to about 40% of RVD, whereas Cl⁻ channels account for the other 60% (17).

In addition to K⁺ and Cl⁻ transporters, several studies have demonstrated a correlation between voltage-gated Na⁺ channel activity and the invasion of malignant cells. For example, colon cancer cells express Na_v1.5, which forms functional voltage-gated Na⁺ channels sensitive to TTX, and genetic inhibition of Na_v1.5 decreases colon cancer cell migration (24). Human ovarian cancer cells express Na_v1.5 at significantly higher levels than normal ovarian tissue, and pharmacological inhibition with TTX reduces migration by about 50% (19). Likewise, human lung, prostate, and breast cancer cells express voltage-gated Na⁺ channels, and migration/invasion is inhibited by application of TTX and/or knockdown of voltage-gated Na⁺ channel expression (2, 7, 36, 67). Inhibition of a TTX-sensitive inward Na⁺ current is associated with the inhibition of migration in rat prostate cancer cells, human mesothelioma cells, and breast cancer cells (8, 18, 66). Beyond voltage-gated Na⁺ channels, the epithelial sodium channel (ENaC)/Degenerin family of channels, including ENaC and acid-sensing ion channels (ASIC) channels, contributes to glioma cell migration. These heterotrimeric channels are sensitive to amiloride inhibition and permeable to monovalent cations, especially Na⁺ (3, 9). Human glioma cells express significantly more αENaC, γENaC, and ASIC subunits than nonmalignant glia, and these subunits can combine to form ENaC/ASIC hybrid channels (27). Knockdown of ASIC1, αENaC, and γENaC leads to a loss of amiloride-sensitive currents and inhibits glioma cell migration (27). In glioma cells ENaC/ASIC channels also promote regulatory volume increase (RVI), referring to a cell's ability to increase cellular volume after shrinking (68). Mechanistically determining whether amiloride-sensitive RVI couples to migration could further implicate Na⁺ channels in the migration of malignant cells.

For ion channels to modulate cellular volume through the movement of osmotically active ions, water must also efficiently permeate the plasma membrane. Aquaporins mediate water influx at the lamellipodium in several nonmalignant cell types (40, 69). Aquaporin activity increases when focal intracellular hypertonicity is induced by an accumulation of depolymerized actin and osmotically active ions at the lamellipodium. This leads to subsequent osmotic water influx through aquaporins causing swelling. The increase in cytoplasmic water content can then lead to greater hydrostatic pressure, causing membrane protrusion at the leading edge (56). Thus aquaporins facilitate shape changes at the leading edge and propel the cell forward through increases in intracellular hydrostatic pressure. Aquaporins and ion channels work in concert to modulate cellular volume, move the leading edge forward, and enhance dynamic shape changes. While these mechanisms are conserved across a variety of cell types to aid in homeostatic processes, malignant cells can co-opt the use of aquaporins to facilitate cancer spreading and disease progression. Human glioma tissue expresses AQP1 and AQP4, and expression of AQP1 and AQP4 in glioma cell lines increases water permeability and doubles glioma cell migration (49). AQP4 localizes to the leading edge of migrating glioma cells where it is negatively regulated by PKC, which inhibits glioma migration (50). Therefore aquaporins, along with ion channels and transporters, play an important role in the migration of malignant cells.

Ca²⁺ Signaling Regulates Ion Channels Facilitating Volume Regulation and Migration

Several ion channels that potentiate migration though volume regulation are directly or indirectly sensitive to increases in [Ca²⁺]_i. This has been best characterized in immature granule neurons migrating in the cerebellum (33), where Ca²⁺ signaling appears to be a critical prerequisite for migration. Ca²⁺ influx through N-type Ca²⁺ channels (30) or N-methyl-d-aspartate (NMDA) channels (31) regulates neuronal migration. These immature neurons move during oscillating Ca²⁺ increases, and raising the amplitude and frequency of Ca²⁺ spikes leads to faster migration (32).

Similarly, Ca²⁺ signaling appears to be a critical modulator of ion channel activity and migration in malignant cells. ClC-3 activation in human glioma cells occurs through Ca²⁺/calmodulin-dependent protein kinase II (CaMKII) phosphorylation (Fig. 1) (15). ClC-3 and CaMKII coimmunoprecipitate, and infusion of autoactivated CaMKII into the patch pipette of voltage-clamped cells activates Cl⁻ currents in glioma cells (15). However, if ClC-3 expression is inhibited, CaMKII does not activate Cl⁻ currents, indicating that CaMKII specifically activates a chloride conductance through ClC-3 (15). Interestingly, ClC-3 and CaMKII also coimmunoprecipitate from human Grade IV glioblastoma tissue (15). These data indicate that ClC-3 facilitates glioma migration through a CaMKII-dependent phosphorylation mechanism.

BK channels expressed by gliomas are also strongly activated by increases in [Ca²⁺]_i (Fig. 1). Gliomas express a splice variant of the human Slo gene (encoding the BK channel) known as gBK, which has an unusually high Ca²⁺ sensitivity (39, 62). Shifting [Ca²⁺]_i from 0 to 2.1 μM shifted the half-maximal voltage for BK channel activation from +138 to −14 mV (62). Therefore, BK expressed by glioma cells can be activated by physiological ligands that increase [Ca²⁺]_i. Using the amphotericin-perforated patch technique to maintain [Ca²⁺]_i, 1 μM bradykinin activated BK currents by increasing [Ca²⁺]_i (64). Similarly, cholinergic stimulation of glioma cells led to dose-dependent increases in [Ca²⁺]_i and subsequent activation of BK (6). Recent data also indicate that menthol may increase glioma cell migration by activating BK channels via a TRPM8-mediated increase in [Ca²⁺]_i (90). These data suggest that Ca²⁺ is the major regulator orchestrating volume changes mediated by ion channels.

Given that increases in [Ca²⁺]_i can lead to activation of BK and ClC-3 (Fig. 1), identifying sources of Ca²⁺ in migrating malignant cells may lead to new mechanistic insights. Members of the transient receptor potential (TRP) family of ion channels and Ca²⁺-permeable 3-hydroxy-5-methyl-4-isoxazole propionic acid (AMPA) receptors seem to be especially important in promoting the migration of malignant cells through Ca²⁺-dependent mechanisms. TRP channels are nonspecific cation channels that can increase [Ca²⁺]_i, leading to the possible activation of other K⁺ and Cl⁻ channels via membrane depolarization and/or Ca²⁺-dependent activation as seen in Fig. 1 (61). A role for TRP channels in the progression of malignant cell proliferation and metastasis has been hypothesized, and several studies support this prediction (20, 60). Glioma cells and human Grade IV glioma tissue express several members of the transient receptor potential canonical (TRPC) family of channels, including TRPC-1, -3, and -5 (4). These glioma channels form functional TRPC channels with currents sensitive to SKF96365, gadolinium chloride, and 2-aminoethoxydiphenyl borate (2-APB) inhibition (4). TRPC1 expressed endogenously in human glioma cells localizes to caveolar lipid raft domains (5), along with BK K⁺ channels and ClC-3 Cl⁻ channels (51). This localization may place Ca²⁺ influx through TRPC1 in proximity to Ca²⁺-activated BK channels and CaMKII-activated ClC-3 channels. Disruption of lipid rafts in glioma cells leads to a loss of SKF96365-sensitive current and store-operated Ca²⁺ entry attributed to TRPC1 (5). TRPC1 localizes to the leading edge of migrating cells and enhances migration in glioma cells after stimulation with epidermal growth factor (5). This migration can partly be attributed to larger currents ascribed to TRPC1 (5). Similarly, in human pancreatic cancer cells, TRPC1 partially mediates TGF-β-induced Ca²⁺ responses associated with migration, and knockdown of TRPC1 inhibits TGF-β induced Transwell migration (16). In nonmalignant cells, stretch-activated calcium channels, presumably belonging to the TRP family of channels, provide a Ca²⁺ influx pathway in migrating cells (37), potentially leading to posterior actomyosin contraction to move the cell forward. TRP channels also appear necessary for guidance of growth cones and netrin-1-induced Ca²⁺ signaling (86). These data demonstrate that TRP channels play a role in the migration of transformed cells, recapitulating mechanisms of migration in nonmalignant cells.

Beyond the TRP family of ion channels, Ca²⁺-permeable AMPA receptors play an important role in the invasiveness of human glioma cells. Glioma cells release large quantities of glutamate and can increase the extracellular glutamate concentration to 100 μM in a space 1,000-fold larger than the cellular volume within hours (92). Unlike astrocytes, gliomas cells have an impaired ability to uptake extracellular glutamate, leading to excitotoxic neuronal death (78, 92). The predominant source of this glutamate is system X_C⁻, a Na⁺-independent cystine/glutamate exchanger. Glutamate release from system X_C⁻ promotes glioma cell migration by activating Ca²⁺-permeable AMPA receptors, which do not contain the GluR2 subunit making the pores Ca²⁺ permeable (43). Glutamate activation of Ca²⁺-permeable AMPA receptors induces Ca²⁺ oscillations associated with migrating glioma cells (43). This translates to murine in vivo studies, where inhibition of system X_C⁻ leads to smaller and less invasive tumors (43). The Ca²⁺ permeability of AMPA receptors is critical for the invasive phenotype of glioma cells. If AMPA receptors are made Ca²⁺ impermeable by adenovirus-mediated transfer of GluR2 cDNA, glioma cell locomotion is inhibited and in vivo tumor invasion is suppressed (25). Pharmacological inhibition of Ca²⁺-permeable AMPA receptors changes the bipolar morphology of migratory cells, leading to a loss of processes and flattening of enlarged somata (25). In contrast, overexpression of Ca²⁺-permeable AMPA-R facilitates invasion and elongated cellular processes, typical of migratory cells (25). It remains to be determined whether Ca²⁺ influx through Ca²⁺-permeable AMPA receptors or TRP channels can lead to K⁺ and Cl⁻ channel activation, as seen in Fig. 1.

Ion Channels Play an Important Role in Cancer Cell Migration: Summary and Clinical Applications

Data presented in this review demonstrate that ion channels and transporters, while facilitating motility in nonmalignant cells, are also intimately involved in the migration of neoplastic cells by dynamic volume regulation. As illustrated in Fig. 1, it is hypothesized that cancer cells, as epitomized by glioma cells, express K⁺ channels and Cl⁻ channels that concomitantly release K⁺ and Cl⁻ ions, leading to the osmotic release of water. This in turn leads to cytoplasmic volume contraction, allowing cells to squeeze through narrow extracellular spaces and create a diffuse tumor mass. Several ion channels and transporters allow the release of K⁺ and Cl⁻ ions to facilitate migration (Fig. 2). Proteins such as ClC-3 and KCC are integral mediators of RVD. As depicted, it is hypothesized that these proteins decrease cytoplasmic volume, for example, before cells invade narrow spaces. Other proteins such as NKCC and ENaC are important in regaining cytoplasmic volume after a cell emerges from a confined space. Additionally, exchangers like NHE facilitate malignant cell invasion by acidifying the extracellular space to digest the extracellular matrix. By decreasing the ability of cancer cells to invade adjacent normal tissue, disease spread and clinical burden can be reduced leading to better patient outcomes.

TRP channels and Ca²⁺-permeable AMPA receptors increase [Ca²⁺]_i in malignant cells. These increases in [Ca²⁺]_i may in turn activate Ca²⁺-activated K⁺ channels, including BK, IK, and SK channels, as well as a Cl⁻ conductance via CaMKII activation of ClC-3 (Fig. 1). Although the mechanism by which voltage-gated Na⁺ channels regulate the migration of malignant cells is unclear, members of the ENaC family may facilitate malignant migration by controlling cell volume, probably through RVI (Fig. 2). ENaC may work in concert with NKCC, allowing cells to regain volume after squeezing through a confined space. Thus ion channels and transporters work in concert to modulate cell volume and promote migration.

Given that the many ion channels and transporters play a vital role in neoplastic cell migration, specific inhibition of these proteins should improve clinical outcomes. Indeed, this has proven to be the case in several preclinical and clinical studies. For example, bumetanide is an FDA-approved diuretic for the treatment of high blood pressure and heart failure. It acts by inhibiting the NKCC transporter, which is also expressed by gliomas where it accumulates intracellular Cl⁻ to enhance volume regulation and migration. When human glioma cells were intracranially implanted into SCID mice, inhibition of NKCC through injection of bumentanide significantly inhibited glioma cell invasion (22). Thus further investigation into the use of bumetanide to inhibit NKCC and reduce malignant migration certainly warrants further study.

Chlorotoxin is another promising therapeutic that inhibits ion channel activity to reduce cancer progression. It is a small peptide isolated from the venom of the deathstalker scorpion (Leiurus quinquestriatus) that specifically binds to human glioma cells (44, 80). Chlorotoxin inhibits Cl⁻ currents in human glioma cells (51) and has completed Phase I (45) and II clinical trials for the treatment of high-grade gliomas. Intracavitary administration of chlorotoxin was well tolerated and is a promising therapeutic for the treatment of gliomas. These clinical studies underscore the importance of targeting ion channels and transporters for the development of novel anti-cancer therapies as we continue Nixon's “War on Cancer” into the 21^st century.

GRANTS

The authors are grateful for the following grants from the National Institutes of Health: RO1 NS-31234, RO1 NS-52634, RO1 NS-36692, and P50-CA97247 to H. Sontheimer and F31 NS-73181 to V. A. Cuddapah.

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the author(s).